Agilent Pesticides App Compend

Solutions that meet your demands for
faster and more condent analysis
Agilent Pesticides Applications Compendium
Emerging technologies can help your lab adjust to changing industry requirements. However, implementing these advancements can be hampered by a shortage of time and resources. That is why, for over four decades, Agilent has taken an active role in developing instruments, methods, and applications that continually set new global standards for applications such as pesticides analysis. Agilent continually partners with the best, most experienced global food and environmental organizations, laboratories, government agencies, and universities to solve problems and bring new applications to light.
Table of Contents
The table of contents on the right has been linked to the individual sections in this compendium. Click on the text to jump to a specic section. If you prefer to search for words or phrases used in any of the enclosed Applications Notes, click on the search button below to open Acrobats search window.
This compendium is a collection of Agilent Application Notes and resources for pesticides analysis. You can browse the applications by technique using the links below.
GC/MS and LC/MS Analyzers and Application Kits GC/MS and GC/MS/MS Application Notes LC/MS and LC/MS/MS Application Notes Productivity Tools Sample Preparation Application Notes Index
Search
GC/MS and LC/MS

Analyzers and Application Kits
A quick, cost-effective path through your toughest problems

With any new technology, getting started is the biggest challenge. But Analyzers and Application Kits help you spend less time on developing your methods of analysis, and more time generating the highest quality results from new technologies. Ready-to-use packaged workow solutions: Factory-tested Analyzers Agilents pre-congured, GC/MS Analyzers, factory tested for pesticide analysis, make it easy to implement the latest technologies that would otherwise require hours or days to apply. So you can start producing consistent, highquality data from day one. Application Kits These easy-to-use application packages for Agilent LC/MS instruments include all the tools and components you need to get started quickly with your application.
> Return to Table of Contents > Search entire document
Agilent Analyzers and Application Kits
Triple Quadrupole LC/MS Pesticide Application Kit

Quickly and efciently implement target screening methods
Minimize the need for tedious manual method development with pretested methods and a Dynamic MRM database
Analyzing pesticide residues in food and environmental samples is challenging. Large numbers of compounds, many at low concentrations, must be monitored and quantitated. To complicate matters, method development is laborintensive and time-consuming, due to compound-dependent parameters that must be optimized. A faster, easier way to develop customized screening methods Agilents Pesticide Screening Dynamic MRM Application Kit is truly unique, because much of the development work has already been completed. The kit features easy-to-use examples that show you how to set up screening methods and quickly adapt them to your specic needs. It also includes: A more than 750-Pesticides Dynamic MRM database that includes compound names, MRM transitions, fragmentor voltages, collision energies, and retention times for each database compound for reliable pesticide screening. Pretested analysis methods, using the Dynamic MRM database, are provided for target screening of pesticides that are routinely monitored around the world.
The following components are included saving you time and money:
Pesticide Dynamic MRM database Positive and negative ion pesticide test mixes Agilent ZORBAX Eclipse Plus C18 column (2.1 mm 50 mm x 1.8 m) Free-trial QuEChERS sample preparation kit Quick-start guide and Application Note that show you how to run the test mixes and create dynamic MRM methods CD-ROM with examples of screening methods, data les, and reports
Quickly establish screening methods for complex matrices using these leading-edge technologies
A more than 750-compound Pesticide Dynamic MRM database and Agilent MassHunter Data Acquisition and Analysis software let you quickly generate acquisition and analysis methods, which can be easily modied to meet your future needs. The Agilent 1200 Series SL Rapid Resolution LC, interfaced with Agilents 6400 Series Triple Quadrupole LC/MS System delivers fast, high-resolution LC/MS/MS analysis. Agilents Jet Stream Electrospray lowers detection levels of pesticides in complex matrices.
Pre-developed examples help you start screening for pesticides in a fraction of time
10 6 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.6 0.9 1.4 1 1.9 2.2 2 2.7 3 3.2 3.8 4.1 4 4.7 5 6 6.8 7 7.2 17.0 8 9 10 11 12 13 14 15 16 17 18.2 18.7 18 19 19.7 3.6 5.4 7.1 6.2 4.5 5.0 5.9 6.5 7.4 8.0 8.4 8.9 9.4 5.2 9.1 11.7
10.9 1 12 23 34
15.5
16.6
12.4 7.9
10.2
13.0 12.1 14.0 14.7 15.1 11.4 13.4 19.0 14.3 17.6 16.1
5.7
9.3
10.5
9.9
Counts vs. acquisition time (min)
Faster identication and quantication of pesticides monitored worldwide: Here, a 300-compound pesticide mixture was analyzed using the Agilent 1200 SL Series RRLC and 6460 Triple Quad with Jet Stream Technology.
Dynamic MRM maximizes the Quadrupoles compound capacity without sacricing sensitivity when hundreds of residues are being analyzed at femtomole concentrations. Agilent SampliQ QuEChERS sample preparation kits make it quick and easy to extract pesticide residues from complex food matrices.
With over 750 entries, Agilents Pesticide Dynamic MRM database ensures fast, customized method development.
Ordering information: Agilents LC/MS Pesticide Application Kit (G1733AA)

The following are required but not included with the G1733AA kit:
Put your lab on the productivity fast track.

Contact your local Agilent Representative or Agilent Authorized Distributor Call 800-227-9770 (in the U.S. or Canada) or visit www.agilent.com/chem/appkits
This information is subject to change without notice. Agilent Technologies, Inc. 2010 Printed in U.S.A., July 7, 2010 5990-5309EN
Agilent 1200 Series Rapid Resolution LC System or Agilent 1290 Innity LC Agilent 6400 Series Triple Quadrupole LC/MS system Agilent MassHunter Acquisition Software 3.01 or higher
READY TO GO FOR PESTICIDES AND ENVIRONMENTAL POLLUTANTS

Triple Quadrupole GC/MS Multiresidue Analyzer
Agilent Analyzers and Applications Kits
Fully congured and factory chemically-tested analyzers make it easy to achieve the highest performance for multiresidue analyses in complex matrices
Based on Agilents 7890A GC and 7000 Series Triple Quadrupole GC/MS, our Multiresidue Analyzers feature a comprehensive MRM database with more than 1000 pesticides and environmental pollutants. The database is optimized with an average of 8 MRM transitions, plus relative intensities for each compound, helping you minimize matrix interference. It also includes a user-friendly tool that lets you use your own compound list to customize the analyzer acquisition method for your specic needs in just 5 minutes.
The following components are included saving you time and maximizing performance:
NEW 1000+ compound MRM database, with 8000+ optimized transitions DVD that shows you how to build an optimized MRM acquisition method based on your compound list Retention Time Locked HP-5MS Ultra Inert column for reliable peak identication and quantitation Checkout samples of 17 representative analytes Capillary Flow Technology and backush for reduced maintenance Quick-start guide that allows you to quickly utilize Agilent advanced technologies and get high performance analytical results on day one CD-ROM with system-specic checkout data les and reports
Perform highly sensitive, multiresidue target analysis in complex matrices using these built-in features:
Flexible, comprehensive MRM Database contains over 1000 pesticides and pollutants, and is optimized with an average of 8 MRM transitions to minimize matrix interference. It also includes a tool that makes it easy to build methods based on your own list. Integrated Capillary Flow Technology (CFT) backush promotes shorter analysis times, lower chemical background, longer column life, and less frequent ion source cleaning to improve uptime.
Boost your multiresidue screening productivity with the Agilent GC/MS/MS pesticide and environmental pollutant analyzer
Top two transitions of Methamidophos
141.0 95.0
Orange
Pear
95.0
79.0
medium matrix interference (overlapping peak gives incorrect ion ratio)

Orange
strong matrix interference (overlapping peak gives inaccurate quantitation result)

Pear
Two alternative transitions of Methamidophos
Multimode inlet (MMI) with large-volume injection enhances trace-level detection and adds exibility by including hot or cold split/ splitless capabilities.
95.0
64.0
141.0
80.0
minimum matrix interference

x104 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
minimum matrix interference
Ratio=23.3
11.8
11.9 12 12.1 12.2 Aquisition Time (min)
12.3
Superior GC/MS/MS selectivity and sensitivity eliminates false results, and simplies data review for improved productivity.
The G9250AA database has on average 8 transitions for each compound. This allows the user to choose alternative transitions to minimize matrix interferences and improve quantitation results.
Counts
Productivity tools to help you make the most out of every analysis: Agilent autotune, batch-at-a-glance data review, and parameter-less integrator streamline your data review and processing.
Ordering information: Choose one of the following options when you order an Agilent 7000 Series Triple Quadrupole GC/MS with an Agilent 7890A GC analyzer system:
The exibility-driven setup. G3445A Option 411: GC/MS/MS Pesticide and Environmental Pollutant Analyzer with constant pressure, post-column backush method The performance-driven setup. G3445A Option 412: GC/MS/MS Pesticide and Environmental Pollutant Analyzer with constant ow, mid-column backush method For customers who already own an Agilent GC/QQQ system: G9250AA: Pesticides and Environmental Pollutants MRM database
Agilent also offers Single Quadrupole GC/MSD pesticide analyzers for broad screening at 5 to 100 ppb. Each is equipped with CFT Backush, Deconvolution Reporting Software, and RTL Pesticide and Endocrine Disruptor library. Go to agilent.com/chem/library, brochure 5990-5310EN, for details.

This information is subject to change without notice. Agilent Technologies, Inc. 2011 Printed in U.S.A., December 9, 2011 5990-6045EN
GC/MSD Pesticide Analyzer

Simplify and accelerate your pesticide screening
Eliminate the need for tedious manual method development with a fully configured and factory-tested Analyzer.
Based on Agilents 5975C Series GC/MSD and 7890A GC System, our user-friendly GC/MSD Pesticide Analyzer quickly screens and quantitates large numbers of pesticides and endocrine disruptors in a single analysis. Its screening methods conform to the latest worldwide pesticide testing requirements. With inlet, column, capillary flow device, and software tools all installed and configured in the factory, and the Analyzer pre-tested for pesticide analysis you can save weeks of method development. Screen more pesticides... in less time Agilents GC/MSD Pesticide Analyzer makes use of productivity-boosting GC/MS technologies that allow you to: Increase the number of targets screened Reduce the analysis time required per sample Perform a complete screening and quantitation in 2-3 minutes Produce consistent, high-quality results from day one
The following components are included with Agilents Pesticide Analyzer saving you time and money.
Pesticide checkout samples Retention Time locked HP-5MS column ensuring reliable database matching Free-trial QuEChERS sample preparation kit Video training tutorials for easy learning of more advanced Analyzer features Quick-start guide and Application Note that show you how to run the screening method provided with the Analyzer CD-ROM with analysis methods, data files, and report
These built-in features make it faster and easier to screen large numbers of target compounds in complex matrices
Multimode Inlet (MMI) with large-volume injection enhances trace-level detection and adds flexibility by including standard split/splitless capabilities.
The Pesticide Analyzer will start your lab on the fast track to better broad-range screening
Retention Time Locking (RTL) for consistent retention times after column maintenance and easy matching with the 927-compound Pesticides and Endocrine Disruptors database.
11.82 11.84 11.86 11.88
Running the pesticides checkout sample: A quick-start guide included with Agilents GC/MSD Pesticide Analyzer shows you how to load the pesticide method, relock the method to get consistent retention times, and run the checkout sample.
Deconvolution peaks and spectra

Matrix
Interference
Target
Deconvolution Reporting Software (DRS) for fast data review in 2-3 minutes per sample, with screening and quantitation in one run. Capillary Flow Technology (CFT) and backflush promote shorter run times, low chemical background, longer column life, and less frequent source cleaning to improve uptime. Agilent SampliQ QuEChERS sample preparation kits let you quickly and easily extract pesticide residues from complex food matrices.
A DRS Checkout Report: You can analyze your results quickly using Deconvolution Reporting Software (DRS) and the RTL Pesticides database.
Ordering information: Order an Agilent 5975C Series GC/MSD along with an Agilent 7890A GC system with one of the following options:
SP1 7890-0456: Pesticide DRS Screening GC/MSD Analyzer

This information is subject to change without notice. Agilent Technologies, Inc. 2010 Printed in U.S.A., January 31, 2010 5990-5310EN
SP1 7890-0457: Japanese Positive List DRS Screening GC/MSD Analyzer
TOF and QTOF LC/MS Pesticide Application Kit

Simplify your broad-range screening for non-target compounds
Screen samples quickly and easily using pretested methods and a Pesticides Accurate Mass database
Conventional multi-target pesticide screening methods are based upon triple quadrupole technology. However, these methods are limited to target compounds, and do not allow a retrospective analysis of collected data. Packaging sensitive, high-performance (Q)TOF instruments with Screening Application Kits overcomes this limitation. Each user-friendly Application Kit combines pretested analysis methods with powerful software tools that, together, simplify the setup of screening applications. This enables even high-volume labs to perform truly comprehensive screening for large numbers of both target and non-target compounds. Capture all the data, all the time Using (Q)TOF technology for your pesticides screening allows you to retain all spectral data, not just your original range of interest. That means you can refer back to your data anytime without reruns to investigate samples further.
The following components are included saving you time and money:
Agilents 1600-pesticide Personal Compound Database and Library software (PCDL) Positive and negative ion pesticide test mixes Agilent ZORBAX Eclipse Plus C18 column (2.1 mm 50 mm x 1.8 m) Free-trial QuEChERS sample preparation kit Quick-start guide and Application Note that show you how to run the test mixes and create screening methods CD-ROM with examples of easy-to-use screening methods, data files, and reports that demonstrate method setup and adaptation
Leading-edge technologies deliver speed, sensitivity, and powerful data mining tools that make broad screening accessible to your lab
A 1600-pesticide Personal Compound database and Agilent MassHunter Data Acquisition and Analysis software let you quickly generate screening methods, which can be modified to meet your future needs. Industry-leading Rapid Resolution Liquid Chromatography (RRLC) and Ultra High Performance Liquid Chromatography (UHPLC) separations that enhance your results and productivity. Agilents Jet Stream Electrospray Ion Source lowers detection levels of pesticides in complex matrices. Best-in-class MS and MS/MS mass accuracy: our (Q)TOF Full Scan capability lets you access all the data, all the time, opening up endless possibilities for screening multiple analytes and non-target compounds.
Pre-developed examples help you start screening for pesticides in a fraction of the time
To demonstrate its functionality, Agilents (Q)TOF Pesticide Application Kit includes a pesticide test mix for both positive and negative ion modes. A general pesticide screening method example is also provided, along with a representative spinach extraction using the Agilent SampliQ extraction and dispersive SPE kits for complete food analysis.
x10 6 Cpd 323: Tribufos: +ESI ECC Scan Frag=150.0V SeqR_16-r003.d
1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 Counts vs. Acquisition Time (min) 11 12 13 14 15 16 15.212 Cpd 323: Tribufos
Here you see an extracted chromatogram of 200 pesticides using the Agilent 1200 Series SL LC with the Agilent 6230 TOF.
Agilent SampliQ QuEChERS sample preparation kits make it quick and easy to extract pesticide residues from complex food matrices.
Agilents Personal Compound database ensures fast, customized method development. The screen above gives an example of batch summary results with retention times updated.
Ordering information: Agilents (Q)TOF LC/MS Pesticide Application Kit (G6854AA)

The following are required but not included with the G6854AA kit: Agilent 1200 Series Rapid Resolution LC System or Agilent 1290 Infinity LC Agilent 6200 Series TOF or 6500 Series QTOF LC/MS system Agilent MassHunter Qualitative Analysis Software 3.01 or higher
This information is subject to change without notice. Agilent Technologies, Inc. 2010 Printed in U.S.A., April 12, 2010 5990-5642EN

Agilent Analyzers and Application Kits:
The
fast track
to greater productivity
Instruments alone do not solve business problems. Knowledge does. That is why, for over four decades, Agilent has taken an active role in developing methods and applications many of which have evolved into global standards for industries such as hydrocarbon processing, environmental, food safety, and forensics/toxicology. Now, Agilent brings this knowledge directly to your lab with our NEW industry-specific Analyzers and Application Kits.
A quick, cost-effective path through your toughest problems

With any new technology, getting started is the biggest challenge. But Analyzers and Application Kits help you spend less time on set-up and configuration, and more time generating the highest quality results from new technologies. Our Analyzer and Application Kit portfolio incorporates the latest GC, LC, GC/MS, and LC/MS productivity techniques for pesticide screening, forensic toxicology, semi-volatiles analysis, biodiesel, refinery gas analysis, and many more valve-based analyzers. And we are continually expanding our line of kits for critical applications.
Actual kit contents may vary based on application.
Simplify your startup and reduce the time to technical mastery

Every Agilent Analyzer and Application Kit is backed by decades of experience, customer feedback, and collaborations with key laboratories around the world. Analyzers and Application Kits include: Applications methodology, column, supplies, and a checkout standards mixture so you wont waste time or money purchasing individual components
Lower your method development costs by up to 90%

Developing new methods is time-consuming and costly. Your most experienced chemists are busy learning and investigating new productivity-enhancing technologies and you cannot yet run real samples on your new instrument. That is why Agilent Analyzers and Application Kits include a method of analysis and checkout standard, which decrease the time and hassle of developing new methods and costly test mixtures.
Analyzers also include:

Shipped fully assembled for your application from our factory including any hardware or software components required to optimize the application, minimizing the downtime and disruption of on-site installation Factory configured and tested a checkout report at shipping, and a duplicate on-site checkout, make sure that your application is ready to go Concise, application-focused video training tools facilitate smooth operation right out of the box And unlike other providers, Agilent Analyzers and Application Kits can be custom-tailored to suit your unique needs, ensuring excellent results throughout the 10-year life of your instrument.
A unique combination of application-specific tools and our expertise can reduce the number of steps between start-up and results.
Example: Pesticides GC/MS Analyzer Factory All checkout on the instrument as ordered by the customer Instrument setup and leak tested Application appropriate column, liner and septa installed Retention time locked based on customer instrument Capillary flow device setup based on application DRS, AMDIS database setup according to application Factory run checkout method using application standard mix Signal/Noise check Delivery Quick start guide to run the checkout method All method and data files on DVD for easy out of the box operation Advanced features video tools Consumables included, no separate ordering required Trial SPE QuEChERS kit Easy consumables re-ordering information Installation Focus is on the startup of the application Duplicate factory checkout with standard mix on site Review video tutorials on how to use advanced features
Implement advanced technologies quickly and confidently

New technologies can help your lab adjust to changing business metrics and industry requirements. However, implementing these technologies to their fullest potential can be hampered by a shortage of time and resources. An Agilent Analyzer solves this problem by ensuring that your application is appropriately configured for optimal performance. So you can apply new capabilities with minimal disruption, shorten your time-to-results (or time-to-market), and realize the full benefit of the latest technologies. Moreover, Analyzers and Application Kits include current advances for increasing analysis speed, improving sensitivity in complex matrices, simplifying data review, and promoting ease of use.
Employ the very latest technologies without worrying about your first results
Concern about questionable results can be a significant obstacle to investigating newer, more productive methods and technologies. But an Agilent Analyzer or Application Kit can help you produce consistent, high-quality data right from day one. The products, supplies, and optimization services in Agilent Analyzers and Application Kits can help you achieve the lowest possible detection limits while reducing or eliminating false positives and negatives. In addition, our LC/MS kits are equipped with databases that can increase your ability to automatically find compounds in matrices such as fruits, vegetables, and body fluids.
You choose the level of service that best fits your lab
Each Analyzer and Application Kit encompasses the latest productivity-enhancing techniques, such as GC Capillary Flow Technology, Deconvolution Reporting Software for fast data review and reporting, or LC/MS and GC/MS mass databases for efficient screening of large compound groups. Individual components and services may include: Method of analysis, column, supplies, and checkout samples DVD with method parameters Video training tools* Factory checkout* Sample preparation pack* Application Notes Quick-start guide Optimization of the analysis Testing and verification with an appropriate suitability standard
*For select systems.
Capillary Flow Technology (CFT) Deconvolution Reporting SW (DRS) App. specific supplies Sample Prep QuEChERS
Retention Time Locking (RTL) RRLC/UHPLC Multimode Inlet Jet Stream Technology
RTL databases for 927 pesticide compounds
MassHunter SW
Synchronous SIM/SCAN App. specific supplies
Dynamic MRM PCD or DMRM database
Together with a comprehensive database, a preconfigured, pretested system can help you achieve reliable sample screening and confirmation, while helping you take full advantage of the productivity gains that new technologies offer.
Sample Prep QuEChERS
Start your lab on the fast track to success.

Contact your local Agilent Representative or Agilent Authorized Distributor at www.agilent.com/chem/contactus Or call 800-227-9770 (in the U.S. or Canada) Visit www.agilent.com/chem/appkits for a description of available kits and their contents.
3
Beyond the box: Advice and support customized for your samples
As a supplement to our new Analyzers and Application Kits, Agilent also delivers local, on-site assistance at the following levels: Training courses and application consulting that teach you the best practices for running your samples Extensions to your standard warranty coverage Preventive maintenance for your complete system On-site duplicate factory checkout We also provide customized solutions in conjunction with our Agilent Channel Partners. Whether you need to improve data integrity, reduce matrix interference, or decrease your consumption of columns and supplies, Agilents extensive industry knowledge and technical expertise can help you meet your most challenging analytical demands and timeframes.
For more information

Learn more about individual kits and their contents: www.agilent.com/chem/appkits Find an Agilent customer center in your country: www.agilent.com/chem/contactus U.S. and Canada 1-800-227-9770 agilent_inquiries@agilent.com Europe info_agilent@agilent.com Asia Pacific adinquiry_aplsca@agilent.com
Contact your Agilent Representative today to discuss your specific needs.
This information is subject to change without notice. Agilent Technologies, Inc. 2009 Printed in U.S.A., August 28, 2009 5990-4116EN
GC/MS and GC/MS/MS

Application Notes
The GC/MS/MS Analyzer and the Pesticides and Environmental Pollutants MRM Database
Application Note
Food Safety and Environmental
Author
Chin-Kai Meng Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
Based on an Agilent 7890A GC System and an Agilent 7000 Series Triple Quadrupole GC/MS, these GC/MS/MS Multiresidue Analyzers were developed to simplify a labs startup process. The database has an average of eight MRM transitions with relative intensities for each compound to provide alternative measurements for minimizing matrix interferences. Easy-to-use tools, as well as tutorial videos are included in the database to build an MRM acquisition method based on your list of compound CAS numbers in less than 5 minutes.
Introduction
Pesticide residue analysis is a complex task requiring the analyst to search for dozens, or even hundreds, of compounds in a wide variety of crop or environmental matrices. Triple Quadrupole GC/MS (GC/MS/MS) provides excellent sensitivity and selectivity in analyzing complex matrices. Agilent Technologies offers several preconfigured and pretested GC/MS/MS analyzers to simplify a labs startup process. The analytical capability of a lab, however, is largely determined by the completeness of the MS/MS MRM transitions in an acquisition method. Agilent Technologies developed an MS/MS MRM Pesticides and Environmental Pollutants Database (G9250AA) of over 1,070 compounds to address many of the limitations most labs are facing. This database is in the form of a spreadsheet for ease-of-use. Significant efforts were invested to acquire multiple transitions (eight on average, with relative intensities) of each compound in the database to work around matrix interferences.
Key features of the MS/MS MRM Database include: Allows users to build acquisition methods without buying all compound standards, thus saves time and money Includes retention times for either constant flow or constant pressure methods; corresponding retention index (RI) values are also included for easy migration to other GC oven programs Has multiple MS/MS transitions, eight on average, for each compound which provides alternatives to work around matrix interferences Shows relative intensity of each MS/MS transition within a compound that facilitates transition selection Allows a quick sort according to compound category (phthalates, PAHs, organophosphorus, fungicides, or semivolatile pollutants and so forth, see Appendix A), CAS number, molecular formula, or molecular weight and so forth. Using the tools in the database, an MRM acquisition method, based on a list of compound CAS numbers, can be easily created from a subset of MS/MS transitions in the database in less than 5 minutes.
Agilents QuEChERS extraction kits and dispersive SPE cleanup kits have demonstrated excellent recoveries for the frequently used pesticides in different food matrices [2, 3]. The sample extracts used in this study were prepared using the QuEChERS technique as described in an application note by Zhao, L. et al [4].
GC/MS/MS Analyzer
The Multiresidue GC/MS/MS Analyzer is configured with Agilents proprietary Capillary Flow Technology (CFT), enabling rugged, reliable GC column backflushing. Backflushing the GC column shortens run times, extends column life, reduces chemical background, provides consistent retention times and spectra, and keeps the MS ion source clean. A Multi-Mode Inlet (MMI) provides the flexibility to inject samples in cold, hot, or solvent-vent modes. Each analyzer system is tested with a 17-compound mixture and retention-time locked at the factory. Two hardware configurations are available to meet different lab needs (see Figure 1): G3445A option 411: This configuration is based upon constant pressure mode method with post-column backflushing. It provides the flexibility to add GC detectors and can be easily scaled for shorter run times. G3445A option 412: This configuration is based upon constant flow mode method with mid-column backflushing. This method provides ultimate performance and shorter cycle times with reduced carrier gas consumption. Both configurations (option 411 and option 412) are interchangeable by changing the column and adding or removing a capillary flow-restrictor. A Quick Start Guide for each analyzer discusses the retention time locking, checkout method results and report for your specific system, a list of supplies, and some troubleshooting tips.
Experimental Conditions
Sample Preparation
Without proper extraction and cleanup procedures, it is difficult to detect trace levels of analytes in complex matrices. The QuEChERS sample preparation technique was first introduced for pesticide analysis in foods by USDA scientists in 2003 [1]. It has been rapidly accepted worldwide for multiresidue pesticide analysis due to its special features known as Quick, Easy, Cheap, Effective, Rugged, and Safe. The QuEChERS extracts can be analyzed by LC and GC combined with MS to determine a wide range of pesticide residues.
G3445A option 411 Setup

EPC Inlet 30 m HP-5ms UI
Purged ultimate union
Constant pressure Post-column backflush MS/MS
1070+ compounds MRM Database
Flexible to add GC detectors and easily scaled for shorter run times
G3445A option 412 Setup
Constant flow Mid-column backflush EPC
1070+ compounds MRM Database
Inlet 15 m HP-5ms UI
Purged ultimate union
MS/MS 15 m HP-5ms UI
Provides ultimate performance and the shortest cycle time

Figure 1. System configurations of Agilent GC/MS/MS Multiresidue Analyzers.
Methods
There are three sets of method parameters included with the database. Some of the method highlights are shown in Table 1.
Table 1.
Method Parameters Included with the Databases

Method 1 Method 2 41.867 min Constant pressure mode MS/MS Analyzer G3445A option # 411 Method 3 19.75 min Constant flow mode MS/MS Analyzer G3445A option # 412
Run time Column flow Feature
40.5 min Constant flow mode Allowing many more transitions in an analysis than Method 3 Agilent J&W HP-5ms UI 0.25 mm 15 m, 0.25 m (two each) Initial at 60 C, hold for 1 min 40 C/min to 120 C, hold for 0 min 5 C/min to 310 C, hold for 0 min
Column(s)
Agilent J&W HP-5ms UI 0.25 mm 30 m, 0.25 m (one each) Initial at 70 C, hold for 2 min 25 C/min to 150 C, hold for 0 min 3 C/min to 200 C, hold for 0 min 8 C/min to 280 C, hold for 10 min Chlorpyrifos-methyl locked to 16.593 min 300 C Q1 = Q2 = 180 C Post-column, post-run
Agilent J&W HP-5ms UI 0.25 mm 15 m, 0.25 m (two each) Initial at 60 C, hold for 1 min 40 C/min to 170 C, hold for 0 min 10 C/min to 310 C, hold for 2 min
Oven program
Locking compound and RT MS source temperature Quad temperature Backflush
Chlorpyrifos-methyl locked to 18.111 min 300 C Q1 = Q2 = 180 C Mid-column, post-run
Chlorpyrifos-methyl locked to 9.143 min 300 C Q1 = Q2 = 180 C Mid-column, post-run
Additional method details for each method are listed on individual pages (tabs) in the database. The retention times (RTs) and retention indexes (RIs) corresponding to these three sets of methods are included for all compounds in the database. Therefore, you can either use one of the above prescribed methods (use RTs in the database) or your existing lab method (convert database RIs to expected RTs for your method). The database RIs were calculated using retention times of straight-chain hydrocarbons from C-8 to C-35. An RI_to_RT conversion tool is included with the database, so you can calculate expected retention times of your analytes based on the RIs in the database and RTs of hydrocarbon markers (C-8 to C-35) from your existing GC method. If your existing method uses HP-5ms UI column of the same phase ratio as a 0.25 mm, 0.25 m column, you will see the smallest difference between the expected and actual retention times for your analytes.
Database Overview
The G9250AA MRM Database is in Microsoft Excel format for easy searching and filtering. Compounds are separated by color bands for clarity. The following basic compound information is included for each compound: Common name Molecular formula Molecular weight (averaged) Molecular weight (mono-isotopic) CAS number, without dashes for easy sorting Classification 1 (see Appendix A) Classification 2 (see Appendix A) Retention times (RT) and retention indexes (RI) for constant flow and constant pressure methods (total three methods) Relative intensity of each transition within a compound Chinese name and Japanese name where available In addition, information is included for building MassHunter MRM acquisition methods: CAS number, standard format with dashes Method RT Common name ISTD (true or false) Precursor ion MS1 resolution Product ion MS2 resolution Dwell time Collision energy (Voltage) Retention time window (used in the MassHunter Compound List Assistant tool) Figures 2 and 3 give an overview of the database layout. Using the Excel filtering tool, it is easy to display the table array according to the criteria chosen in any column. Figure 4 shows the database after using Excel filtering in column AE to hide all transitions except the top two (Q0 and Q1). This flexibility allows the user to build methods according to compound categories, (for example, PAHs, Phthalates, or PCBs), or regulatory methods and so forth. Two groups of compound classifications in the Database are listed in Appendix A as a reference.
Backflush
Food or environmental extracts after cleanup are usually still very complex containing various matrix residues such as highboiling compounds. The extracts used in GC/MS analyses can cause contamination and deterioration of the analytical column and MS ion source, affecting data quality due to poor peak shape and loss of responses of active analytes. It also leads to a shorter lifetime of the analytical columns and frequent MS maintenance. Therefore, it is necessary to use best techniques and supplies to achieve reliable results and to protect the analytical column and MS ion source. Column backflushing can be beneficial for the analysis of complex extracts because it significantly reduces analysis time and reduces both column head trimming and MSD ion source cleaning frequency [5]. Agilent CFT makes column backflushing routine [6, 7].
Database has RTs (and RIs) to be used with three GC methods (CF-40 min, CP-40 min, and CF-20 min
Average and exact Molecular Weight
Each compound is classified in two ways

Figure 2. Layout of the Database 1: molecular weights, classifications, and three RTs and RIs.
MassHunter format for building acquisition methods
The scale and relative intensities of transitions

(Color Scale): Orange denotes strong intensity and blue denotes weak intensity among ALL transitions.
One Quant (Q0) and several Qualification ions for each compound
Figure 3. Layout of the Database 2: MassHunter format for building acquisition methods, multiple transitions, and relative intensities.
Use the filtering function to quickly select a Quant (Q0) and a Qualifier ion to build an acquisition method
Figure 4.
Using Excel filter to hide all transitions except the top two of each compound.
Table 2 gives a breakdown of the compounds included in the database.

Table 2. Compounds Included in the Database
Total number Pesticides (fungicides, herbicides, insecticides, rodenticides, and others) Breakdown products Deuterated compounds Polybrominated Diphenyl Ether (PBDE) Polybrominated Biphenyl (PBB) Polychlorinated Biphenyl (PCB) Polycyclic Aromatic Hydrocarbon (PAH) Phthalates Additional semivolatile pollutants 675
A complete list of compounds can be found on the DB Compound List tab in the database. There are three videos included with the database to help the user learn the database: An overview of the content and layout of the database. & Each individual column and tab is explained in the video. A tutorial of building an MRM acquisition method based on your list of compound CAS numbers. & An MRM acquisition method, based on your list of CAS numbers, can be easily and quickly created from a subset of MS/MS transitions in the database. A tutorial showing how to add new compounds to the database.
42 6 4 1 209 26 17 94
The database ReadMe file shows a few Excel shortcuts to use with the database and a few additional ways of using the database. For example, it shows how to find all nitrogen containing compounds in the database, or how to select all PCB congeners, or the 14 most toxic PCB planar congeners in the database.
Results and Discussion

Chemical Background from the Matrix
Figure 5 shows four MRM total ion chromatograms (TICs) of pepper, spinach, orange, and pear extracts acquired using method 3 described on page 3. Thirty-five analytes at 10 ppb each were spiked in each matrix. Seven transitions of each analyte were used in the acquisition method. The TICs showed that the chemical background from the four matrices was quite different and sizeable. In this study, pear extract showed the highest background response in terms of number of peaks and intensity. The TIC from the orange extract was the cleanest among the four chromatograms. These different and high background responses all came from the matrix. To understand the matrix effect, we need to evaluate the chemical background in each individual transition.
10 5 3 2 1 0 10 5 3 2 1 0 10 5 3 2 1 0 10 5 3 2 1 0
Pepper
Spinach
Orange
Pear
10
11
12
13
14
15
16
17
18
Figure 5.
Total ion chromatograms (TICs) of pepper, spinach, orange, and pear extracts including 35 analytes spiked at 10 ppb each.
Spinach
Orange
Pear
141 & 95
95 & 79 Small matrix interference Medium matrix interference

(overlapping peak gives incorrect ion ratio)
Strong matrix interference

(overlapping peak gives inaccurate quantitation result)
Figure 6.
Top two transitions of methamidophos (at 10 pg) in three matrices.
Orange
95 & 64
Pear
141 & 80 Minimum matrix interference Minimum matrix interference
Figure 7.
Two alternative methamidophos transitions with minimum matrix interference.
A typical MRM database or acquisition method has two MRM transitions for each analyte. Figure 6 shows extracted ion chromatograms (EICs) of the top two transitions of methamidophos (at 10 pg) in three matrices. The retention time of methamidophos is about 4.6 minutes. The transitions are arranged in the descending order of responses with the larger one on top. Figure 6 shows the obvious issues of getting inaccurate quantitation results due to medium or strong matrix interference. For orange matrix, an overlapping peak in the second transition marked by a blue arrow, affected integration results and the qualifier ion ratio. For pear matrix, an overlapping peak in the first transition, marked by a green arrow, which is typically used for quantitation, gave higher and inaccurate quantitation results.
If a user only has two MRM transitions available for each analyte, it is difficult to work around the matrix effect as seen in Figure 6. The G9250AA database has an average of eight transitions for each compound. This allows the user to choose alternative transitions easily when matrix interference affects peak shape and integration results. Figure 7 shows EICs of two alternative methamidophos transitions in the database. Both transitions showed minimum matrix interferences in orange and pear matrices. In fact, the EICs of these two transitions showed minimum matrix interference in all four matrices. Although these two transitions do not provide the highest responses, they are suited for a universal or screening MRM method. It is always best to evaluate the chemical background of an analytes multiple transitions in different matrices before selecting the most appropriate transitions in a particular matrix.
Signal-to-Noise Ratio Comparison for MRM Transitions in Spinach and Orange

18000 3000
Dichlorvos
16000 Dash Lines are Absolute Responses (area) 14000 12000 10000
Acephate
Atrazine
Lindane
2500 Solid Lines are Signal/Noise
S/N
Responses (area) in spinach and orange
2000
1500 8000 6000 4000 500 2000 0 200 & 122 181 & 145 200 & 94 215 & 200 173 & 138 219 & 183 145 & 109 183 & 147 217 & 181 136 & 94 136 & 42 142 & 96 125 & 47 185 & 93 187 & 93 109 & 79 215 & 58 0 1000
50 ppb Spinach, resp.
50 ppb Orange, resp.
50 ppb Spinach, S/N
50 ppb Orange, S/N
Figure 8.
Comparison of area counts and signal-to-noise ratios of four analytes MRM transitions in spinach and orange matrices.
Signal-to-Noise Ratios
Evaluating the signal-to-noise ratios (S/N) of MRM transitions is another way to identify matrix effects. Some pesticides showed consistent MRM responses in different matrices, but many pesticides had different MRM responses in different matrices due to either matrix enhancement or matrix suppression. Figure 8 shows responses, or area counts, and S/Ns of several MRM transitions for four analytes in spinach and orange matrices. The orange dashed line and dark green dotted line represent area counts from four or five MRM transitions of each analyte in these two matrices. The solid blue and green lines represent the S/N from the same MRM transitions in these two matrices. The dashed and dotted lines, signifying area counts, superposed tightly. However, the solid lines, S/N, showed significant variations for some transitions within each analyte. Using atrazine as an example, the area count for transition 215&58 was about the same (approximately 7,000) for both matrices, but the S/N for this transition
in spinach was about 40% higher than the S/N in orange matrix. In contrast, for transition 200&122, the S/N in orange matrix was almost double the S/N in spinach, even though the area counts in both matrices were about the same (approximately 4,000). This matrix effect was not unique to atrazine. The S/N variations from some of the MRM transitions of dichlorvos and lindane were more pronounced even though the area counts were comparable in both matrices. Again, if the number of MRM transitions available for each analyte is limited to two or three, it is difficult to select optimal MRM transitions suited for the matrix analyzed. The multiple transitions available in the G9250AA database allow users to choose several selective transitions to achieve accurate confirmation and quantitation results. This study showed that MRM transitions should be chosen according to matrix to achieve optimal and reliable quantitation results. It is important to use matrix-matched calibrations and low background transitions to achieve accurate quantitation results.
Conclusion
Based on the Agilent 7890A GC and 7000 Series Triple Quadrupole GC/MS, the GC/MS/MS Multiresidue Analyzers were developed to simplify a labs startup process. A special feature of the analyzer is the comprehensive and flexible MRM database of over 1,070 pesticides and environmental pollutants. The analyzer also includes CFT backflush for superior system robustness during routine operations. Matrix can cause quantitation interference, lower responses, or poor peak shape. Each matrix has a different matrix effect. Therefore, it is critical to choose the most selective transitions for a particular matrix and use matrix-matched calibration curves to achieve accurate and reliable quantitation results. The G9250AA MRM Database has an average of eight MRM transitions with relative intensities for each compound to provide alternative measurements to minimize matrix interference. Easy-to-use tools as well as tutorial videos are also included in the database to build an MRM acquisition method based on your list of compound CAS numbers in less than 5 minutes.
References
1. M. Anastassiades and S.J. Lehotay, Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and Dispersive Solid-Phase Extraction for the Determination of Pesticide Residues in Produce, J. AOAC Int., 2003, 86, 412- 431. 2. L. Zhao, D. Schultz, and J. Stevens, Analysis of Pesticide Residues in Apple Using Agilent SampliQ QuEChERS AOAC Kits by GC/MS, Agilent Technologies publication 5990-4068EN. 3. L. Zhao and J. Stevens, Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS AOAC Kits by GC/MS, Agilent Technologies publication 5990-4305EN. 4. L. Zhao and C.-K. Meng, Quantitative and Repeatability Analysis of Trace Level Pesticides in Plantation Food by GC/MS/MS, Agilent Technologies publication 59909317EN. 5. M.J. Szelewski and B. Quimby, New Tools for Rapid Pesticide Analysis in High Matrix Samples, Agilent Technologies publication 5989-1716EN. 6. C-K. Meng, Improving Productivity and Extending Column Life with Backflush, Agilent Technologies publication 5989-6018EN. 7. P.L. Wylie and C-K. Meng, A Method for the Trace Analysis of 175 Pesticides Using the Agilent Triple Quadrupole GC/MS/MS, Agilent Technologies publication 5990-3578EN.
For More Information

For more information on our products and services, visit our Web site at www.agilent.com/chem.
10
Appendix A
List of First Compound Classification in the Database

algicide bird repellent breakdown defoliant, plant growth regulator deuterated fragrance fumigant fungicide fungicide, insecticide fungicide, microbiocide fungicide, plant growth regulator herbicide herbicide safener herbicide, algaecide herbicide, microbiocide herbicide, plant growth regulator insect attractants insect growth regulator insect repellent, synergist insecticide insecticide, fungicide insecticide, insect repellent insecticide, molluscicide insecticide, nematicide insecticide, plant growth regulator microbiocide microbiocide, fungicide molluscicide nematicide plant growth regulator plant growth regulator, herbicide pollutant rodenticide synergist wood preservative, microbiocide
List of secondary compound classification in the Database

1,3-Indandione 2,6-Dinitroaniline Amide Anilide Anilinopyrimidine Aromatic Aryl phenyl ketone Arylalanine Aryloxyphenoxy propionic acid Auxins Benzamide Benzimidazole benzofuranyl alkylsulfonate Benzoic acid Benzothiazole Botanical Bridged diphenyl Carbamate Carbanilate Carbofuran Carboxamide Chitin synthesis inhibitor Chlorinated phenol Chloroacetanilide Dinitrophenol derivative Diphenyl ether Dithiocarbamate Folpet Formamidine Formamidine Fumigant Halogenated organic Hydrobenzonitrile Hydroxybenzonitrile Imidazolinone Juvenile hormone mimics Keto-enol Mercaptobenzothiazole Morphactins Morpholine Naphthalene acetic acid derivative Neonicotinoid Nitrophenyl ether N-Methyl carbamate Organochlorine Organophosphorus Oxadiazolone PAH Phthalate Phthalic acid Picolinic acid Pyrazole Pyrethroid Pyrethroid ester Pyridazine Pyridazinone Pyridine Pyridinecarboxylic acid Pyrimidinamine Pyrimidine Pyrimidine organothiophosphate Pyrimidinyloxybenzoic acid Pyrrole Quinoline Quinone Quinoxaline SemiVOC Strobin Substituted benzene Sulfite ester Thiadiazole Thiocarbamate Continued
11
List of secondary compound classification in the Database

Chlorophenoxy acid or ester Conazole Coumarin Cyclic dithiocarbamate Cyclodiene Cytokinins Defoliant Deuterated PAH Deuterated semiVOC Dicarboximide PBB PBDE PCB Phenol Phenoxyacetic Phenoxybutyric Phenoxypropionic Phenylsulfamide Phosphoramidate Phosphorodiamide Thiophthalimide Triadimefon Triazine Triazinone Triazole Triazolone Uracil Urea Xylylalanine
www.agilent.com/chem
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice.
Agilent Technologies, Inc., 2011 Printed in the USA November 10, 2011 5990-9453EN
Organophosphorus Pesticides Analysis Using an Agilent J&W DB-5ms Ultra Inert Capillary GC Column
Application Note
Environmental
Authors
Doris Smith and Kenneth Lynam Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
Agilent Technologies Inc. has implemented new testing procedures to more effectively evaluate GC column inertness performance. This new testing procedure employs deliberately aggressive probes to thoroughly investigate and verify column inertness and quality. In challenging separations, knowing that the GC column has been thoroughly investigated for column inertness gives analysts higher confidence in the accuracy of their results. Trace- and ultra trace-level pesticide analyses are important tools for accessing food supply and environmental quality worldwide. In this application note, trace-level organophosphorus pesticide analysis is demonstrated using electron impact single quadrupole scanning mass spectrometry. Agilent's J&W DB-5ms Ultra Inert capillary GC column provides excellent peak shape for even the most problematic pesticides.
Introduction
Pesticides are commonly used in agricultural and residential applications throughout the world. Organophosphorus pesticides make up approximately 70 percent of the insecticides currently in use. Unfortunately, these highly toxic materials have three main routes of human exposure: inhalation, ingestion, and skin penetration. Sources of these exposures include consumption of foodstuff containing pesticide residues, aerosol inhalation, and dermal contact during pesticide application. [1] Organophosphorus pesticides use the same mechanism of action as deadly nerve agents such as sarin, soman, and VX. These pesticides affect the nervous system of insects, mammals, and wildlife by inhibiting the enzyme cholinesterase, important in helping regulate nerve impulses. Inactivation of cholinesterase leads to the accumulation of the neurotransmitter acetylcholine in the central and peripheral nervous system, which leads to depressed motor function and respiratory depression. Human toxicities for this class of molecules have shown acute as well as chronic effects from pesticide poisoning. [2,3] Organophosphorus pesticides tend to be difficult to quantify due to poor peak shape, as evidenced by broad, asymmetrical peaks. An EPA Method 525.2 standard containing organophosphorus pesticides along with a custom pesticide mix acquired from Ultra Scientific (North Kingstown, RI) were analyzed to highlight the value of using a 30-m Agilent J&W DB5ms Ultra Inert capillary GC column for difficult pesticide analysis. Many pesticides are sensitive to chromatographic system activity and will readily breakdown. The Ultra Scientific custom mix contains several types of these pesticides, which are useful in quickly evaluating system performance with particularly challenging pesticide analytes. Capillary GC column activity as a potential source of result uncertainty has been virtually eliminated with the Ultra Inert series of columns. [4]
Table 1A.
Chromatographic Conditions for EPA Method 525.2 Calibration Standards
GC Sampler
Agilent 6890N/5975B MSD Agilent 7683B, 5.0-L syringe (Agilent p/n 5181-1273) 1.0-L splitless injection Helium 44 cm/sec, 1.5 mL/min constant flow Pulsed splitless; 250 C, 40 psi until 0.75 min, purge flow 50 mL/min at 1.0 min Deactivated dual taper direct connect (Agilent p/n G1544-80700) Agilent J&W DB-5ms Ultra Inert 30 m 0.25 mm 0.25 m (Agilent p/n 122-5532UI) 40 C (1 min) to 110 C (50 C/min), 7 C/min to 190 C, 12 C/min to 285 C, hold 2 min. MSD source at 250 C, quadrupole at 150 C, transfer line at 280 C, EI mode, scan range 45450 amu
Chromatographic Conditions for Ultra Scientific Calibration Standards
Carrier Inlet Inlet liner Column Oven Detection
Table 1B.
GC Sampler Carrier Inlet Inlet liner Column Oven Detection
Agilent 6890N/5975B MSD Agilent 7683B, 5.0-L syringe (Agilent p/n 5181-1273) 1.0-L splitless injection Helium 52 cm/s, constant flow Pulsed splitless; 250 C, 40 psi until 0.75 min, purge flow 50 mL/min at 1.0 min Deactivated dual taper direct connect (Agilent p/n G1544-80700) Agilent J&W DB-5ms Ultra Inert 30 m 0.25 mm 0.25 m (Agilent p/n 122-5532UI) 75 C to 175 C (15 C/min), 10 C/min to 275 C (1 min) MSD source at 250 C, quadrupole at 150 C, transfer line at 280 C, EI mode, scan range 45450 amu
Flow Path Supplies
Table 2.
Vials Vial caps Vial inserts Syringe Septum Inlet liners Ferrules 20x magnifier
Amber crimp-top glass vials (Agilent p/n 5183-4496) Crimp caps with 11-mm septa (Agilent p/n 5181-1210) 100-L glass/polymer feet (Agilent p/n 5181-8872) 5 L (Agilent p/n 5181-1273) Advanced Green (Agilent p/n 5183-4759) Deactivated dual taper direct connect (Agilent p/n G1544-80700) 0.4 mm id short; 85/15 Vespel/graphite (Agilent p/n 5181-3323) 20x magnifier loupe (Agilent p/n 430-1020)
Experimental
An Agilent 6890N GC/5975B MSD equipped with a 7683B autosampler was used for this series of experiments. Table 1 lists the chromatographic conditions used for these analyses. Table 2 lists flow path consumable supplies used in these experiments.
Sample Preparation A six-component EPA Method 525.2 pesticide standard mix and internal/surrogate standard mix were purchased from Accu-Standard (New Haven, CT) and used to prepare a sixlevel calibration standard set. The stock pesticide solution as delivered had a nominal concentration of 1,000 g/mL. The internal/surrogate solution as delivered had a nominal concentration of 500 g/mL. The calibration standards were prepared with component concentrations of 10, 5, 2, 1, 0.5, and 0.1 g/mL and a constant level of 5 g/mL of internal/surrogate standard as per EPA Method 525.2. All solutions were prepared in acetone using class A volumetric pipettes and flasks. Acetone used was JT Baker Ultra Resi Grade purchased thorough VWR International (West Chester, PA). Acetone was used as a reagent blank and syringe wash solvent. An 11-component pesticide standard mix was purchased from Ultra Scientific and used to prepare a seven-level calibration standard set. The stock pesticide solution as delivered had a nominal concentration of 1,000 g/mL. The calibration standards were prepared with component concentrations of 10, 5, 2.5, 1, 0.5, 0.25, and 0.1 g/mL. All solutions were prepared in 2,2,4-trimethylpentane using class A volumetric pipettes and flasks. The 2,2,4-trimethylpentane used was JT Baker Ultra Resi Grade purchased thorough VWR International (West Chester, PA). 2,2,4-Trimethylpentane was used as a reagent blank and syringe wash solvent.
sites, particularly at trace level, like the organophosphorus pesticides in this application example. A more detailed description of the test mix and additional application examples can be found in references 6 through 8. Organophosphorus Pesticide Analysis In this application note, a multilevel pesticide calibration curve set was evaluated over the concentration range of 0.1 to 10 g/mL on an Agilent J&W Ultra Inert DB-5 ms 30 m 0.25 mm 0.25 m (Agilent p/n 122-5532UI). Separate calibration curves were developed for both the EPA 525.2 organophosphorus and Ultra Scientific standards. The standard levels used for the 525.2 calibration were 0.1, 0.5, 1, 2, 5, and 10 g/mL, while the Ultra Scientific calibration levels were 0.1, 0.25, 0.5, 1, 2.5, 5, and 10 g/mL. The custom pesticide standard from Ultra Scientific was used to determine system performance by analyzing difficult pesticides, such as endrin and p,p'-DDT, which are prone to analyte breakdown. No tailing was observed for any of the organophosphorus pesticide peaks across the range studied in either standard set. Sharp, symmetrical peak shapes were noted for all the organophosphorus pesticides analyzed. Good resolution was obtained for each of the pesticides investigated. Linearity for the 525.2 standard components was excellent across the range studied, giving R2 values of 0.997 or greater in all cases but fenamiphos, which had an R2 value of 0.978. This value increases to 0.991 at the midlevel concentrations as suggested by EPA Method 525.2 Sec. 13.2.3.3. Figure 5 indicates the correlation coefficients for each of the individual pesticides and shows an example linear regression plot for disulfoton. Linearity for the Ultra Scientific standard components was also quite good across the range studied. R2 values of 0.990 or greater were obtained for the organophosphorus pesticides. Figure 6 indicates the correlation coefficients for each of the individual pesticides and shows an example linear regression plot for mevinphos.

Baseline Inertness Profile for Ultra Inert Columns The basic approach for inertness verification for the Agilent J&W Ultra Inert series of capillary GC columns is testing with aggressive active probes at low concentration and low temperature. [5] This is a rigorous approach that establishes consistent baseline inertness profiles for each column in the Agilent J&W Ultra Inert GC column series. The baseline inertness profile then serves as a predictor for successful analysis of chemically active species that tend to adsorb onto active
1. 2. 3. 4. 5. 6.
Abundance 70000 65000 60000 55000 50000 45000 40000 35000 30000 25000 20000 15000 10000 5000
1,3-Dimethyl-2-nitrobenzene (SS) Acenapthene-d10 (IS) Cycloate Prometron Phenanthrene-d10 (IS) Disulfoton
7. 8. 9. 10. 11. 12.
Ametryn Fenamiphos Tribufos (DEF) Triphenylphosphate (SS) Chrysene-d12 (IS) Perylene-d12 (SS)
3 4
5 6
10
11
12
6.00 Figure 1.
8.00
10.00
12.00
14.00 Time
16.00
18.00
20.00
22.00
Total ion chromatogram (scan mode) of the 1-ng on-column EPA Method 525.2 standard solution loading on an Agilent J&W DB-5ms Ultra Inert 30 m 0.25 mm 0.25 m capillary GC column (p/n 122-5532UI). Chromatographic conditions are listed in Table 1A.
Abundance 40000
35000
30000 25000
1. Disulfoton 2. Fenamiphos 3. Tribufos (DEF)
20000
15000
10000
5000
13.00 Figure 2.
13.50
14.00
14.50
15.00
15.50
16.00 Time
16.50
17.00
17.50
18.00
18.50
19.00
Enlarged section of the total ion chromatogram (scan mode) for a 1-L injection of 1.0 g/mL EPA Method 525.2 standard pesticide mix. The peaks noted in the figure are the three organophosphorus pesticides of interest. Chromatographic conditions are listed in Table 1A.
Figure 3.
Total ion chromatogram (scan mode) of the 0.1-ng on-column Ultra Scientific standard solution loading on an Agilent J&W DB-5ms Ultra Inert 30 m 0.25 mm 0.25 m capillary GC column (p/n 122-5532UI). Chromatographic conditions are listed in Table 1B.
Abundance 1800 1600 1400 1200 1000 800 600 400 200
S/N ratio = 5.33
Peak Height = 1600 abundance units
Noise = 300 abundance units
5.00
5.20
5.40
5.60
5.80
6.00 Time
6.20
6.40
6.60
6.80
7.00
7.20
Figure 4.
Enlarged section of the total ion chromatogram (scan mode) for a 1-L injection of 0.1 g/mL Ultra Scientific standard pesticide mix on an Agilent J&W DB-5ms Ultra Inert 30 m 0.25 mm 0.25 m capillary GC column (p/n 122-5532UI). The peak in the figure is mevinphos, an organophosphorus pesticide of interest. This injection represents an on-column loading of 0.1 ng per component. Chromatographic conditions are listed in Table 1B.
Component Cycloate Prometon Disulfoton Ametryn Fenamiphos Tribufos (DEF)
R2 1.000 0.999 0.999 0.999 0.978 0.997
1.0000 0.8000 0.6000 0.4000 0.2000 0.0000 0.00
Disulfoton
5.00
10.00
15.00
Figure 5.
Correlation coefficients for the EPA Method 525.2 pesticide components over the 0.1 to 10 g/mL range of this study and an example linear regression plot for disulfoton.
Component 1,3-Dimethyl-2-nitrobenzene Mevinphos Pentachlorophenol Terbufos Bromacil Aldrin Carboxin Endrin p,p'-DDT Triphenylphosphate DEHP
Figure 6.
R2 0.999 0.990 0.989 0.996 0.988 0.999 0.996 0.998 0.996 0.997 0.996
14000000 12000000 10000000 8000000 6000000 4000000 2000000 0 0.00
Mevinphos
5.00
10.00
15.00
Correlation coefficients for the Ultra Scientific pesticide components over the 0.1 to 10 g/mL range of this study and an example linear regression plot for mevinphos.
Conclusions
This application successfully demonstrates the use of an Agilent J&W DB-5ms Ultra Inert capillary GC column for trace-level organophosphorus pesticides. Linearity was excellent for all organophosphorus pesticides studied, yielding 0.99 or greater R2 values down to a 0.1-ng on-column loading of each component. One of the reasons for excellent linearity and high R2 values is the highly inert surface of the column. The lack of chemically active sites makes these columns an excellent choice for trace-level applications.
This study was done using scan mode on an Agilent 6890/5975B GC/MSD equipped with an inert electron impact source. The signal-to-noise ratio for a 0.1-ng on-column loading of mevinphos was greater than 5 to 1 with this system. This result shows clearly the power of using an Agilent J&W DB-5ms Ultra Inert column for trace-level organophosphorus pesticides analysis. Lower limits of quantification are expected when using one of Agilent's latest GC/MS offerings, such as the 7890/5975C GC/MSD Triple-Axis Detector coupled with an Agilent J&W DB-5ms Ultra Inert GC capillary column.
References
1. J. Routt Reigart and James R. Roberts, Recognition and Management of Pesticide Poisoning, 5th Ed, 1999, pp 3447 [Online] Avaliable: www.epa.gov/oppfead1/safety/healthcare/handbook/ handbook.htm Kenneth D. Katz, et al, "Organophosphate Toxicity," [Online]. Available: www.emedicine.com/med/TOPIC1677.HTM (Article last updated May 30, 2008) K. Steenland, "Chronic Neurological Effects of Organophosphate Pesticides," BMJ, May 25, 1996; 312(7042): 1312 1313. J. W. Eichelberger, et al., US EPA Method 525.2 Revision 2.0, Determination of Organic Compounds in Drinking Water by Liquid-Solid Extration and Capillary Column Gas Chromatography/Mass Spectrometry, National Exposure Research Laboratory, USEPA, Cincinnati, Ohio, 1995. Mitch Hastings, Allen K. Vickers, and Cameron George, "Inertness Comparison of Sample of 5% Phenyldimethylpolysiloxane Columns," Poster Presentation, 54th Annual Pittsburg Conference, Orlando, FL, March 2003 "Agilent J&W Ultra Inert GC Columns: A New Tool to Battle Challenging Active Analytes," Agilent Technologies publication 5989-8685EN, May 29, 2008 Kenneth Lynam, "Semivolatile Analysis Using an Inertness Performance Tested Agilent J&W Ultra Inert DB-5ms Column," Agilent Technologies publication 5989-8616EN, May 13, 2008 Kenneth Lynam and Doris Smith "Polycyclic Aromatic Hydrocarbon (PAH) Analysis Using an Agilent J&W DB5ms Ultra Inert Capillary GC Column," Agilent Technologies publication 5989-9181EN, in press
2.
3.
4.
5.
6.
7.
8.

Agilent Technologies, Inc., 2010 Published in the USA March 30, 2010 5989-9879EN
Highly Sensitive and Rugged GC/MS/MS Tool

For Pesticide Multiresidue Analysis in Food Samples
Agilent 7000 Series Triple Quadrupole GC/MS. The worlds first MS/MS designed specifically for GC Analysis
Introduction
Multi-residue methods are efficient and cost-effective for analysis of pesticide residues. For methods with a very wide scope, generic sample preparation procedures are usually employed. Inherent to this approach is that clean up of extracts is only possible to a limited extent1. When applying such methods to complex matrices like baby food, herbs, spices and tobacco, enhanced selectivity in detection is required to make up for the low selectivity in sample preparation. The analytical challenge is to maximize the number of pesticides, minimize the variety of methods, keep run times short and achieve limits of detection (LODs) at or below the maximum residue limits (MRLs) which are specified for pesticides under EU legislation. As regulations in the European Union require very low MRLs for pesticide residues, the latest challenge has been to reach part-per-billion level concentrations for hundreds of pesticides in complex matrices, which in turn has required greater sensitivity and efficiency in pesticide screening. Quantitation and confirmation of identity of trace level compounds can be complicated by the matrix, resulting in qualifier ion ratios out of range, or target ions buried in the high chemical background noise. With single quadrupole mass spectrometry, selected ion monitoring (SIM) is often used to improve the detection limit and quantitative reproducibility. In SIM mode, only a few ions are monitored for each target within the retention time (RT) range that the target elutes from the column. SIM may not work well for trace levels in matrix as the interferences in SIM are the same as in full scan mode.
Your Challenges
Our Integrated Approach
Triple quadrupole mass spectrometry allows for drastic reduction or elimination of matrix interferences that limit the accuracy and detection limits of SIM methods. This process, referred to as Multiple Reaction Monitoring (MRM), has two fundamental advantages over SIM. First, detection is based secondary product ion produced by the collisional dissociation of an analyte precursor ion. The analyte precursor ion (isolated in Q1 by a SIM mechanism) has the same selectivity as SIM, but there is a high probability that at least one of the resultant product ions will be unique to the precursor and not the interference. The increase selectivity of MRM is often apparent by the reduced offset of the baseline as compared to SIM. Secondly, during the mass filtering process in Q1, all lower m/z ions from the sample are eliminated. The unique product ions from the collisional dissociation are measured in this zero noise region of the spectrum. The combination of a unique product ions (more selectivity) and the elimination of background noise results in consistently low limits of detection even for complex matrices. This application brief describes the analysis of pesticides in fruit and vegetable extracts using the Agilent 7000 Series Triple Quadrupole GC/MS system in MRM mode and in combination with Retention Time Locking2 and Agilent Capillary Flow Technology to provide backflushing of high-boiling materials.3
Column backflushing is essential for the analysis of complex samples such as food extracts4 because they usually contain high-boiling indigenous compounds. In just a few runs, these materials can collect on the head of the column, causing peak tailing, retention time shifts and increased chemical noise. Over time, they can migrate from the column to the ionization source, which would eventually have to be cleaned. Agilent's proprietary capillary flow technology makes column backflushing routine and easy to setup for non experts. The method robustness is drastically increased and the analysis cycles are shortened5. In conclusion the system up-time is maximized allowing significant productivity gains. The need for maintenance is reduced by keeping the chromatographic system and MS ion source cleaner between each injection.
Experimental
Samples were prepared using the QuEChERS6 method. QuEChERS stands for Quick Easy, Cheap, Effective, Rugged and Safe and is a food sample preparation for multi-class, multi-residue pesticide analysis. See more at www.agilent.com/chem/Quechers
Instrumentation
The Triple quadrupole GC/MS system used for these experiments are described in Table 1 and shown in Figure 1.
Instrument conditions
Figure 1 Agilent 7890A/7000A Triple Quadrupole GC/MS system with the new high capacity 7693 ALS. Agilent 7000A Agilent 7890A PTV, in splitless mode, 1L Injection, Multi-baffle liner 80 C for 0.5 min, then 500 C/min to 280 C for 2 min Capillary flow technology device: 3-way splitter with analytical column in and restrictor out to the triple quadrupole helium pressure provided by Aux EPC at 1 psi Column: Agilent J&W HP-5ms Ultra Inert 30 m x 0.25 mm ID, 0.25 m HP-5MSUI Restrictor: 80 cm x 0.180 mm deactivated fused silica Carrier gas: Helium 30.883 psi (constant pressure mode) Oven temperature: 70 C (1 min), 25 C/min to 150 C (0 min), 3 C/min to 200 C (0 min), 8 C/min to 280 C (10 min) Backflush: Time 5 min, inlet press. 1 psi, Aux EPC 80 psi, oven temp. 280 C Retention time locking: Chlorpyrifos-methyl locked to 16.53 min Collision cell gases: N2 2.60 psi and He 6.25 psi Inert source temperature: 260 C Quadrupole temperature: 150C Table 1 Instrument conditions. Instrumentation GC/MS Triple Quadrupole: GC: Inlet:
Method Development
In order to maximize the response of the instrument for each residue the choice of precursor ion, product ion and collision energy were optimized. The spectra of a typical pesticide in full scan mode (50-500 m/z) e.g. Dicloran (MW = 206) is displayed in Figure 2. The product ion scan spectra of 206 m/z at different collision energy are displayed in Figure 3.
Results
x103 6 5 4 3 2 1 0 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 Counts vs. Mass-to-Charge [m/z]
Figure 2 Full scan spectrum of Dicloran under EI ionization mode.
Results
Figure 3 Product ion scan of Dicloran for the Precursor (206 m/z ) at different collision energies (5-40 V).
The optimum collision energy for the 206>176 transition product ion was found to be at 10V and the resulting MRM chromatogram is shown in Figure 4.
Results
x106 3 2.5 2 1.5 1 0.5 12.35 12.45 12.55 12.65 Counts vs. Acquisition Time [min] 12.75 12.85
Results and discussion

Figure 5 shows the TIC chromatogram acquired in MRM mode for 360 pesticides. Each MRM segment is indicated by a grey marker line. An enhanced view on a selected part of the analysis with an overlay of all MRM for the compounds in this part is shown Figure 6. The MRM mode allows for accurate quantification of many coeluting analytes as shown between 13.6 and 14.2 minutes. Figure 7 demonstrates the identity confirmation of Diclobenil in a peppermint extract at 10 pg on column using two MRM transitions. The dashed lines indicate the allowed range of the ion ratio as specified in the method.
Figure 4 MRM chromatogram of Dicloran at 10ppb.
Results
x104 2.0 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Counts vs. Acquisition Time [min] 32 34 36 38 40
Figure 5 Total ion chromatogram of the vegetable extract by GC/MS/MS.
Results
x103 2.4 2.0 1.6 1,2 0.8 0.4 0 11.5 11.7 11.9 12.1 12.3 12.5 12.7 12.9 13.1 13.3 13.5 13.7 13.9 14.1 14.3 Counts vs. Acquisition Time [min]
Figure 6 Overlay of extracted MRM transitions.
Results
171.0 -> 136.0, 171.0 -> 100.0 x102 Ratio=80.7 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 -0.2 6 6.4 6.8 7.2 7.6 Acquisition Time [min]
Results
Dichlobenil 7 Levels, 7 Levels Used, 7 Points, 7 Points Used, 0 QCs y = 953.8898 * x - 58.7038 R2 = 0.99989052 Chlormefos Levels, 7 Levels Used, 7 Points, 7 Points Used, 0 QCs y = 554.7783 * x - 1176.5985 R2 = 0.99902029
x105 1.6 1.3 1 0.6 0.2 0
Counts
x105 1 0.8 0.6 0.4 0.2 0
0 20 40 60 80 100 120140160180200 Concentration (ng/ml)
0 20 40 60 80 100 120 140160180200 Concentration (ng/ml)
Figure 8 Calibration curves showing excellent linearity over the concentration from 1 ppb to 200 ppb range R2= 0.999 respectively for Dichlobenil and Chlormefos.
Figure 7 Two transitions identifying Diclobenil in a peppermint extract at 10 pg on column. The dashed lines indicate the allowed range of the ion ratio.
Results
6,150 6,130 6,110 RT [min] 6,090 6,070 6,050 6,030 6,010 5,990 5,970 5,950 1 11 21 31 41 51 61 Injection Number 71 81 91
Linearity was also tested with five levels between 1 and 200 ppb for 360 pesticides and Figure 8 shows the calibration curves for Diclobenil and Chlormefos. The correlation coefficients of the external standard calibration curve were 0.99 on average. The LOD was estimated based on the calculated S/N of the 10 pg standard. For the majority of pesticides, LODs were below 2 pg on column (based on S/N >3:1 Peak to Peak). Retention time reproducibility was also tested to demonstrate the robustness of the analytical method. Figure 9 shows the outstanding retention time stability of one representative compound: Trifluralin. Calculated %RSD is 0.0306 at 6.073 min for one hundred consecutive injections of lettuce extract into the GC/MS/MS system. Only 3 minute Backflush was necessary to remove all high boiling matrix compounds, the total cycle time for this stability test was 21 hours.
Min Max Mean Std Devn % RSD
6.070 6.075 6.073 0.00186 0.0306
Figure 9 Exceptional retention time stability of Trifluralin with 100 injections of lettuce extract thanks to column backflushing.
Conclusion
Agilents 7000 Series Triple Quadrupole GC/MS in combination with the 7890 GC is a sensitive and rugged tool for target pesticide analysis in complex matrices. The single multi-residue method we developed also meets the performance and identity confirmation criteria defined by the stringent EU regulations. Excellent selectivity has been achieved to allow unambiguous confirmation of identity for these 360 pesticides even in very complex food matrices and generic sample clean up. Agilent SampliQ QuEChERS kits enable you to prepare food samples for multiresidue, multi-class pesticide analysis with just a few simple steps. For this new GC/MS/MS method, the Agilent Retention Time Locking (RTL) database was used to calibrate the retention times of all pesticides. Therefore, the presented GC/MS/MS method can be easily transferred to other Agilent 7000 Series Triple Quadrupole systems with minimum effort and time.
References
1. M. Mezcua, M. A. Martnez-Uroz, P. L. Wylie, and A. R. Fernndez-Alba, Simultaneous screening and target analytical approach by GC-q-MS for pesticide residues in fruits and vegetables, Accepted for publication by J. AOAC Int. 2. Bruce D. Quimby, Leonid M. Blumberg, Matthew S. Klee, and Philip L. Wylie, Precise Time-Scaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking, Agilent Technologies publication 5967-5820E. 3. Capillary Flow Technology, more information at http://www.agilent.com/chem/cft 4. Chin-Kai Meng Improving Productivity and Extending Column Life with Backflush, Agilent Technologies publication 5989-6018EN.
5. Chris Sandy Improving GC-MS Method Robustness and Cycle Times Using Capillary Flow Technology and Backflushing Agilent Technologies publication 5990-3367EN. 6. Limian Zhao, Joan Stevens Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS AOAC Kits by GC/MS Agilent Technologies publication
www.agilent.com/chem/7000B
Agilent Technologies, Inc., 2009 Published November 1, 2009 Publication Number 5990-5044EN
A Method for the Trace Analysis of 175 Pesticides Using the Agilent Triple Quadrupole GC/MS/MS
Application Note
Food Safety
Authors
Philip L. Wylie and Chin-Kai Meng Agilent Technologies, Inc. 2850 Centerville Rd. Wilmington, DE 19808 USA
Abstract
A GC/MS/MS multiple reaction monitoring (MRM) method has been developed on the Agilent 7890A/7000A GC triple quadrupole mass spectrometer system (GC/QQQ) for 175 commonly analyzed pesticides. Numerous fruit and vegetable extracts were analyzed by this method and by a GC single quadrupole mass spectrometer (GC/Q) for comparison. The GC/Q was operated in the selected ion monitoring (SIM) mode and in the scan mode. Scan results were evaluated using Agilent's Deconvolution Reporting Software (DRS) with the RTL Pesticide and Endocrine Disruptor Database. The GC/Q instrument was equipped with a Multimode inlet and injections of 5 L were made in the cold splitless mode. These were compared to 1 L injections of the same extracts on the GC/QQQ. The GC/QQQ was found to be far more sensitive and selective than either GC/Q approach, primarily because there is much less interference from co-extracted matrix. There is, however, still a need for GC/Q methods that use DRS to screen for more than 900 pesticides and other contaminants since the GC/QQQ in the MRM mode is only for target compound analysis.
Introduction
Pesticide residue analysis is a complex task requiring the analyst to search for dozens, or even hundreds, of compounds in a wide variety of crop matrices. Extraction techniques, such as the QuEChERS method [13] leave large amounts of indigenous materials in the extract. The use of more extensive cleanup steps risks removing pesticide residues in addition to the matrix. As required detection limits for many pesticides fall to 10 g/Kg (10 ppb) or lower, more sophisticated analytical tools are needed. For GC-amenable pesticides, many laboratories are using two complementary techniques for screening and confirmation purposes. For broad screening at the 5 to 100 ppb level, GC/single quadrupole (GC/Q) is employed with Deconvolution Reporting Software (DRS) and the RTL Pesticide and Endocrine Disruptor library from Agilent Technologies [46]. This is a scan method to screen for 927 GC-amenable pesticides and endocrine disruptors in a single GC/MS run. Detection limits for most pesticides vary from approximately 5 to 100 ppb, depending upon the matrix and the injection volume [4]. For target pesticide analysis in the most complex matrices, the Agilent 7890A/ 7000A GC/triple quadrupole (GC/QQQ) is unmatched. This paper compares three mass spectral techniques for the analysis of pesticide residues in a variety of crop matrices. Spiked and unspiked samples were analyzed by GC/Q in the selected ion monitoring (SIM) mode and in the scan mode with DRS analysis. The same samples were also analyzed by GC/QQQ using a multiple reaction monitoring (MRM) method for 175 pesticides. The objective was to compare the ability of these GC/Q and GC/QQQ methods to detect low levels of pesticides in several different crop matrices.
Table 1.
Instrumentation and Analytical Conditions for the GC/Q System Agilent 7890A Series Agilent 7693A Injector and sample tray Multimode inlet Helium 18.420 psi (constant pressure mode) during run 2.0 psi (during backflush)
GC Autosampler Inlet Carrier gas Inlet pressure
Splitless Mode Inlet Parameters Temperature 250 C Inlet liner Helix double taper, deactivated (P/N 5188-5398) Injection volume 1 L Purge flow to split vent 30 mL/min at 0.75 min Cold Splitless Mode Inlet Parameters Temperature program 60 C (0.01 min), 700 C/min to 280 C (hold) Inlet liner Helix double taper, deactivated (P/N 5188-5398) Injection volume 5 L Purge flow to split vent 30 mL/min at 1.25 min Oven temperature program 70 C (1 min), 50 C/min to 150 C (0 min), 6 C/min to 200 C (0 min), 16 C/min to 280 C (5 min) 2-way splitter with one port capped used for backflushing the analytical column and retention gap Helium plumbed to 2-way splitter 4.0 psi during run, 60.0 psi during backflush Agilent J&W HP-5ms UI 15 m 0.25 mm 0.25 m (P/N 19091S-431UI) Between retention gap and 2-way splitter 2.0 m 0.25 mm Siltek deactivated fused silica tubing (Restek, Bellefonte, PA) Between inlet and analytical column using an Ultimate Union (P/N G3182-61580) to couple the retention gap to the column 80 cm 0.15 mm deactivated fused silica tubing (Agilent) Between the 2-way splitter and the MSD 2.705 mL/min (nominal) Chlorpyrifos-methyl locked to 8.298 min Agilent 5795C Series with performance turbo pump Electron impact 280 C 230 C 150 C 100 A/D = 4 1 Variable from 4 to 25 ms On 2.5 min
Capillary flow technology
Pneumatic Control Module (PCM) PCM pressure Analytical column Connections Retention gap Connections
Restrictor Connections Initial column flow rate Retention time locking Mass selective detector Mode Transfer line temperature Source temperature Quadrupole temperature Threshold Sampling rate Gain factor SIM dwell times Trace ion detection Solvent delay Backflushing Conditions Timing Oven temperature
Experimental
Samples Spiked and unspiked extracts of fresh produce were provided by the U.S. Food and Drug Administration (U.S. FDA, CFSAN, College Park, MD) and the U.S. Department of Agriculture (USDA ARS, ERRC, Wyndmoor, PA). Samples from the FDA were prepared using the QuEChERS [13] method modified to include the use of activated carbon as an additional sorbent. The resultant toluene solution contained 4.5 grams of produce per milliliter of extract. Samples from the USDA were extracted using the published QuEChERS method and contained 1 gram of produce per milliliter of acetonitrile solvent. Instrumentation The GC/Q and GC/QQQ systems used for these experiments are described in Tables 1 and 2.
2
5 min duration during post-run 280 C

(Continued)
Instrumentation and Analytical Conditions for the GC/Q System (continued) Aux EPC pressure 60 psi Inlet pressure 2 psi Software GC/MSD
Table 1.
Agilent GC/MS ChemStation control and data analysis software (P/N G1701EA E.02.00 SP1)
Triple Quadrupole Mass Spectrometer Mode Transfer line temperature Solvent delay Source temperature Quadrupole temperature MRM Mode Conditions MS1 resolution MS2 resolution Collision gas flows Backflushing Conditions Timing Oven temperature Aux EPC pressure Inlet pressure Software Data acquisition Qualitative analysis Quantitative analysis
Agilent 7000A Series Electron impact 280 C 2.3 min 300 C Q1 and Q2 = 150 C
Deconvolution Reporting Software Library Searching Software
Agilent P/N G1716AA (Ver. A.04.00)
1.2 u 1.2 u Nitrogen at 1.5 mL/min, Helium at 2.35 mL/min
NIST MS Search (Ver 2.0d) (comes with NIST mass spectral library Agilent P/N G1033A) Automated Mass Spectral Deconvolution and Identification Software (AMDIS_32 version 2.62 or greater; comes with NIST mass spectral library Agilent P/N G1033A) NIST 08 mass spectral library (Agilent P/N G1033A) Pesticide and Endocrine Disruptor Database (Agilent P/N G1672AA)
Deconvolution software
3 min duration during post-run 280 C 80 psi 1 psi
MS Libraries
Agilent MassHunter Data Acquisition Software (Ver. B.04.00) MassHunter Workstation Software for Qualitative Analysis (Ver. B.03.01) MassHunter Workstation Software for Quantitative Analysis (Ver. B.03.01)
Table 2.
Instrumentation and Analytical Conditions for the GC/QQQ system Agilent 7890A Series Agilent 7683A Injector and sample tray Split/splitless Helix double taper, deactivated (P/N 5188-5398) Helium 18.350 psi (constant pressure mode) during run 1.0 psi (during backflush) 250 C 1 L 30 mL/min at 0.75 min On (20 mL/min at 2.0 min) 70 C (1 min), 50 C/min to 150 C (0 min), 6 C/min to 200 C (0 min), 16 C/min to 280 C (5.5 min) Purged Ultimate Union (P/N G3186B) used for backflushing the analytical column and retention gap Helium plumbed to Purged Ultimate Union 4.0 psi during run, 80.0 psi during backflush Agilent J&W HP-5ms UI 15 m 0.25 mm 0.25 m (P/N 19091S-431UI) Between retention gap and Purged Ultimate Union (P/N G3186B) 2.0 m 0.25 mm Siltek deactivated fused silica tubing (Restek, Bellefonte, PA) Between inlet and analytical column using ultimate union (P/N G3182-61580) to couple the retention gap to the column 65 cm 0.15 mm deactivated fused silica tubing (Agilent) Between the Purged Ultimate Union (P/N G3186B) and the MSD 2.688 mL/min (nominal) Chlorpyrifos-methyl locked to 8.298 min
GC Autosampler Inlet Inlet liner Carrier gas Inlet pressure Inlet temperature Injection volume Purge flow to split vent Gas saver Oven temperature program

GC Configuration Both GC systems used a 15-m 0.25 mm 0.25 m Agilent J&W HP-5ms UI column and were running the standard Agilent pesticide method [7] at 2X speed. This method uses an initial oven temperature of 70 C, which works for most GC solvents without using a retention gap. However, 1-L injections of samples in toluene lead to poor peak shape, so a 2-m deactivated retention gap was coupled to the front of the column. This improved the peak shapes. Column backflushing is essential for the analysis of food extracts [4, 89] because they usually contain high-boiling indigenous compounds. In just a few runs, these materials can collect on the head of the column (or retention gap), causing peak tailing and retention time shifts. Over time, they can migrate from the column to the mass spec source, which would eventually have to be cleaned. Agilent's capillary flow technology makes column backflushing routine (4, 89) and several different capillary flow devices can be used for the purpose. The GC/QQQ system used a Purged Ultimate Union, while the GC/Q system used a twoway splitter (with one port capped). In both cases, the analytical column was connected to the capillary flow device. A short restrictor was used to couple the capillary flow device to the mass spectrometer. Figure 1 shows the configuration of each instrument.
3
Aux EPC gas Aux EPC pressure Analytical column Connections Retention gap Connections
Restrictor Connections Initial column flow rate Retention time locking
7693 Injector
A
7693 Sample Tray
a b c g d
PCM
5975C MSD
e
Figure 1A. The GC/MSD used for scan and SIM analyses was configured with a) Multimode inlet, b) 2 m 0.25 mm deactivated retention gap, c) Ultimate Union, d) 15 m 0.25 X 0.25 m Agilent J&W HP-5ms UI column, e) two-way purged splitter with one port capped, f) helium purge flow controlled by a pneumatic control module (PCM), and g) 80 cm 0.15 mm deactivated restrictor.
7683 Injector
7683 Sample Tray
a b c g
Aux EPC
7000A QQQ
f d
Figure 1B. The GC/QQQ used for MRM analyses was configured with a) split/splitless inlet, b) 2 m 0.25 mm deactivated retention gap, c) Ultimate Union d) 15 m 0.25 0.25 m Agilent J&W HP-5ms UI column, e) Purged Ultimate Union, f) helium purge flow, and g) 65 cm 0.15 mm deactivated restrictor.
MRM Method A method was developed for the analysis of 175 commonly analyzed pesticides. Two transitions were determined for each compound and the collision energy was optimized for each. Since the method was locked to the Agilent Pesticide method (running at twice the original speed), the retention times correspond to those recorded in Agilent's RTL Pesticide
Table 3. Target and Qualifier Transitions for 175 Pesticides
and Endocrine Disruptor Database (P/N G1672AA) divided by two. There are small differences in RT between the database and values shown here because this method used a retention gap, capillary flow device, and a restrictor. Table 3 lists the pesticides in alphabetical order with their retention times, quant and qual transitions, and the collision energies for each.
Compound name Acrinathrin Akton Alachlor Aldrin Allethrin Atrazine Azamethidaphos (Azamethiphos) Azinphos-methyl Benfluralin BHC, BHC, BHC, Bifenthrin Bromacil Bromophos Bromophos-ethyl Bromopropylate Captan Carbophenothion Chlordane, cisChlordane, transChlordene, Chlordene, Chlordene, Chlorfenvinphos, Chlorobenzilate Chloroneb Chlorothalonil Chlorpyrifos Chlorpyrifos-methyl Chlorthiophos Coumaphos Cyanazine Cyanophos Cyfluthrin 1 Cyfluthrin 2
RT (min) 15.371 11.403 8.507 9.247 10.908 6.581 13.248 14.835 5.842 6.025 6.595 7.266 14.428 9.186 10.020 11.261 14.320 10.617 13.316 11.410 11.010 8.562 9.376 9.314 10.779 12.706 4.323 7.395 9.606 8.284 13.051 15.859 9.694 6.887 16.144 16.212
Quant transition Precursor ion Product ion 181.1 282.9 188.1 262.9 123.1 200.1 215.0 160.1 292.1 181.0 181.0 181.0 181.1 205.0 330.9 358.9 183.0 79.1 157.0 372.9 372.9 230.0 230.0 230.0 267.0 139.0 191.0 265.9 196.9 286.0 268.9 362.0 212.1 243.0 163.0 163.0 152.1 219.0 130.1 192.9 81.1 122.1 171.0 77.1 264.0 145.0 145.0 145.0 165.1 132.0 315.9 302.9 155.0 77.1 121.0 265.9 265.9 160.0 160.0 160.0 159.0 111.0 113.0 133.0 168.9 93.0 205.0 109.0 123.1 109.0 127.1 91.1
CE 25 10 40 40 10 10 15 20 10 15 15 15 30 30 20 15 15 10 25 40 20 40 35 40 20 15 15 40 15 25 15 15 20 10 5 15
Qual transition Precursor ion Product ion 181.1 282.9 188.1 262.9 123.1 200.1 215.0 160.1 292.1 181.0 181.0 181.0 181.1 205.0 330.9 358.9 183.0 79.1 157.0 372.9 372.9 230.0 230.0 230.0 267.0 139.0 191.0 265.9 196.9 286.0 268.9 362.0 212.1 243.0 163.0 163.0 127.1 184.0 160.1 190.9 79.1 104.0 128.0 132.1 160.1 109.0 109.0 109.0 166.1 187.9 285.9 284.8 76.0 51.1 75.1 263.9 263.9 195.0 195.0 195.0 81.0 75.0 141.0 230.9 107.0 270.9 177.0 81.0 151.1 79.0 91.1 127.1
CE 30 25 10 40 20 20 30 0 20 30 30 30 15 20 35 35 35 25 40 30 25 25 25 25 40 30 10 20 40 20 25 40 10 30 15 5
(Continued)
Compound name Cyfluthrin 3 Cyfluthrin 4 Cyhalothrin, Cypermethrin 1 Cypermethrin 2 Cypermethrin 3 Cypermethrin 4 Dacthal (DCPA) (Chlorthal-Dimethyl) DDD, o,p'DDD, p,p'DDE, o,p'DDE, p,p'DDT, o,p'DDT, p,p'DEF (Tribufos) Deltamethrin Demeton-S Demeton-S-methyl Dialifos Diallate 1 Diallate 2 Diazinon Dicapthon Dichlofenthion Dichlofluanid Dichlorobenzophenone, 4,4'Dichlorvos Diclobenil Dicloran Dieldrin Dimethachlor Dioxathion Disulfoton Ditalimfos Edifenphos Endosulfan ether Endosulfan I Endosulfan II Endosulfan sulfate Endrin Endrin aldehyde Endrin ketone EPN Ethalfluralin Ethion Ethoprop (Ethoprophos) Etridazole
RT (min) 16.273 16.307 15.208 16.381 16.463 16.531 16.558 9.721 12.170 12.841 11.241 12.007 12.882 13.492 12.054 18.016 6.303 5.230 15.432 5.957 6.127 7.226 9.694 8.067 9.199 9.593 2.905 3.367 6.269 11.926 8.080 15.934 7.260 11.586 13.377 7.660 11.308 12.570 13.377 12.366 12.956 14.116 14.333 5.632 12.997 5.357 3.963
Quant transition Precursor ion Product ion 163.0 127.1 163.0 181.1 181.1 181.1 181.1 181.1 300.9 235.0 235.0 246.0 246.0 235.0 235.0 169.0 181.1 88.1 88.1 208.0 234.1 234.1 179.1 262.0 279.0 123.0 139.0 109.0 171.0 206.0 262.9 134.1 125.0 88.1 130.0 173.0 240.9 240.9 195.0 271.9 262.9 249.9 316.9 157.0 276.1 231.0 158.0 183.0 127.1 152.1 152.1 152.1 152.1 152.1 222.9 165.1 165.1 176.1 176.1 165.1 165.1 57.1 152.1 60.0 60.0 102.1 150.0 150.0 121.1 216.0 223.0 77.1 111.0 79.0 100.0 176.0 192.9 105.1 97.0 60.0 102.1 109.0 205.9 205.9 125.0 236.9 193.0 214.9 101.0 77.1 105.1 128.9 97.0 139.9
CE 5 5 30 25 30 25 25 25 30 25 40 40 30 30 5 25 5 5 40 20 20 40 15 15 20 15 5 25 10 40 15 5 5 15 15 20 15 25 20 35 35 20 25 35 25 15 20
Qual transition Precursor ion Product ion 163.0 91.1 163.0 181.1 181.1 181.1 181.1 181.1 300.9 235.0 235.0 246.0 246.0 235.0 235.0 169.0 181.1 88.1 88.1 208.0 234.1 234.1 179.1 262.0 279.0 123.0 139.0 109.0 171.0 206.0 262.9 134.1 125.0 88.1 130.0 173.0 240.9 240.9 195.0 271.9 262.9 249.9 316.9 157.0 276.1 231.0 158.0 183.0 91.1 127.1 127.1 127.1 127.1 127.1 166.9 199.1 199.1 211.0 175.1 199.1 199.1 112.9 127.1 59.0 59.0 89.0 192.0 192.0 137.2 123.0 205.0 51.1 75.1 47.0 136.0 124.0 190.9 77.1 65.0 59.0 75.0 65.1 203.9 136.0 159.0 116.9 190.9 141.9 245.0 110.0 202.0 174.9 114.0 108.0
CE 15 15 35 35 35 30 30 40 15 20 20 40 20 20 5 25 20 15 40 10 10 20 40 30 40 30 15 15 30 35 30 25 25 30 40 20 40 10 40 35 40 20 15 20 10 5 40
(Continued)
Compound name Famphur Fenamiphos (Phenamiphos) Fenarimol Fenchlorphos (Ronnel) Fenitrothion Fenpropathrin Fensulfothion Fenthion Fenvalerate 1 Fenvalerate 2 Fluchloralin Flucythrinate 1 Flucythrinate 2 Fluridone Fluvalinate - 1 Fluvalinate - 2 Folpet Fonophos Heptachlor Heptachlor exo-epoxide isomer A Heptachlor exo-epoxide isomer B Hexachlorobenzene Hexazinone Iprobenfos (IBP) Iprodione Isazophos Isofenfos Jodfenphos (Iodofenphos) Leptophos Lindane (-BHC) Malathion Methidathion Methoxychlor, o,p'Methoxychlor, p,p'Metolachlor Mevinphos Mirex Nonachlor, cisNonachlor, transOxadiazon Parathion Parathion methyl Pentachloroaniline Pentachlorobenzene Pentachlorobenzonitrile Pentachlorophenyl methyl ester Pentachlorothioanisole
RT (min) 13.329 11.803 15.222 8.650 9.030 14.503 12.780 9.552 17.202 17.412 7.321 16.571 16.741 16.944 17.412 17.480 10.807 6.934 8.379 10.474 10.352 6.168 13.702 7.660 14.211 7.517 10.813 11.776 14.876 6.710 9.396 11.146 13.730 14.442 9.450 3.782 14.923 12.848 11.539 12.210 9.633 8.284 7.761 4.459 6.866 6.283 9.016
Quant transition Precursor ion Product ion 218.0 109.0 303.1 139.0 284.9 277.0 181.1 292.0 278.0 167.1 167.1 306.1 199.1 199.1 328.1 250.1 250.1 147.1 246.1 271.9 183.0 352.9 283.9 171.1 204.0 187.0 161.1 213.1 376.9 171.0 181.0 173.1 145.0 227.1 227.1 162.1 127.0 271.9 408.8 408.8 175.0 291.1 263.0 264.9 249.9 274.9 264.9 295.9 80.0 111.0 269.9 109.0 152.1 156.0 109.0 125.0 125.0 264.1 107.1 107.1 259.0 55.1 55.1 103.1 109.0 236.8 118.9 262.8 213.9 71.1 91.1 124.0 119.0 121.0 361.9 77.1 145.0 99.0 85.1 121.1 141.1 133.1 109.0 236.9 109.0 299.8 112.0 109.0 109.0 193.9 214.9 239.9 236.9 245.8
CE 15 40 15 15 20 30 25 20 15 10 5 30 25 30 15 15 5 15 25 30 25 35 15 10 25 10 20 20 25 15 15 5 15 40 15 10 15 20 25 15 10 10 30 25 20 10 40
Qual transition Precursor ion Product ion 218.0 79.0 303.1 139.0 284.9 277.0 181.1 292.0 278.0 167.1 167.1 306.1 199.1 199.1 328.1 250.1 250.1 147.1 246.1 271.9 183.0 352.9 283.9 171.1 204.0 187.0 161.1 213.1 376.9 171.0 181.0 173.1 145.0 227.1 227.1 162.1 127.0 271.9 408.8 408.8 175.0 291.1 263.0 264.9 249.9 274.9 264.9 295.9 154.0 75.0 239.9 260.0 127.1 109.0 125.0 89.1 89.1 206.0 157.1 157.1 189.1 200.1 200.1 76.0 137.0 116.9 154.9 281.9 248.8 85.1 121.0 159.0 146.0 185.0 93.0 124.1 109.0 117.0 58.1 91.1 169.1 132.1 95.0 116.9 299.9 301.8 76.1 81.0 79.0 155.9 142.0 204.9 142.9 262.9
CE 30 20 35 35 5 35 20 20 40 35 15 10 5 40 20 25 30 5 40 15 20 25 15 40 15 5 5 35 10 30 10 15 35 30 25 15 40 25 30 40 40 35 30 40 35 40 15
(Continued)
Compound name Permethrin, cisPermethrin, transPhenanthrene-d10 Phenothrin Phenthoate Phorate Phosalone Phosmet Pirimiphos ethyl Pirimiphos methyl Procymidone Profenofos Propachlor Propargite Propazine Propetamphos Propyzamide Prothiophos Pyraclofos Pyrazophos Pyridaphenthion Quinalphos Quintozene Resmethrin Simazine Sulfotep-ethyl Sulprofos Tebupirimfos Tecnazene (TCNB) Tefluthrin Temephos Terbufos Terbuthylazine Tetrachloroaniline, 2,3,5,6Tetrachlorvinphos Tetramethrin I Tetramethrin II Thiometon Tolclofos methyl Tolyfluanid Triallate Triazophos Trifluralin Triphenyl phosphate Vinclozolin
RT (min) 15.703 15.798 6.863 14.713 10.861 5.961 14.855 14.259 10.332 9.138 10.983 11.953 5.164 13.858 6.676 6.948 6.975 11.878 15.439 15.351 14.272 10.827 6.832 13.994 6.473 5.902 13.180 7.687 5.110 7.524 20.525 6.890 6.907 5.293 11.478 14.299 14.421 6.161 8.392 10.623 7.470 13.241 5.808 13.865 8.311
Quant transition Precursor ion Product ion 183.1 153.1 183.1 188 183.1 274.0 231.0 182.0 160.0 318.1 290.1 283.0 207.9 120.1 135.1 214.1 138.0 173.0 162.0 360.0 221.1 340.1 146.1 236.9 123.1 201.1 322.0 322.0 261.1 202.9 177.1 125.0 231.0 214.1 230.9 329.0 164.1 164.1 125.0 265.0 137.0 268.0 161.0 306.1 326.1 212.0 155.1 160 153.1 121.0 128.9 111.0 77.1 166.1 125.0 96.1 63.1 77.1 107.1 172.0 110.0 145.0 63.1 96.9 193.1 199.1 118.1 118.9 81.1 173.1 146.0 97.0 137.1 83.0 127.1 47.0 128.9 104.0 158.0 109.0 107.1 107.1 47.0 250.0 91.1 183.9 134.1 264.0 169.1 145.0
CE 15 10 10 15 10 25 15 30 15 25 10 40 20 15 10 5 15 40 35 10 5 10 25 5 5 25 30 15 25 20 20 25 20 25 25 15 10 20 15 20 25 10 5 35 25
Qual transition Precursor ion Product ion 183.1 168.1 183.1 188 183.1 274.0 231.0 182.0 160.0 318.1 290.1 283.0 207.9 120.1 135.1 214.1 138.0 173.0 162.0 360.0 221.1 340.1 146.1 236.9 123.1 201.1 322.0 322.0 261.1 202.9 177.1 125.0 231.0 214.1 230.9 329.0 164.1 164.1 125.0 265.0 137.0 268.0 161.0 306.1 326.1 212.0 165.1 186 168.1 125.0 174.9 75.1 133.0 182.1 233.0 67.1 99.0 92.1 77.1 104.0 64.0 109.0 98.0 194.0 149.1 97.0 91.1 142.9 95.1 138.1 65.0 156.0 153.1 142.9 137.0 79.0 174.9 132.0 122.0 79.0 135.1 135.1 79.0 93.0 65.1 226.0 91.1 160.0 233.0 109.0
CE 15 10 10 15 20 10 40 15 15 10 40 25 5 30 20 15 35 20 15 15 40 30 30 5 10 40 5 20 20 20 10 10 10 40 35 10 5 10 25 35 15 20 30 10 40
Carrot Extract A carrot extract with incurred pesticide residues was analyzed in the scan and SIM modes with the GC/Q. In each case, 5-L injections were made using Agilent's new Multimode inlet operated in the cold splitless mode. Three SIM methods were used to monitor > 170 compuonds with about 60 pesticides in each method. Four ions were monitored for each compound. The scan data were analyzed automatically using Agilent's Deconvolution Reporting Software, together with the 927compound RTL Pesticide and Endocrine Disruptor Database. The same carrot sample was also analyzed on the 7890A/ 7000A GC/QQQ system using the MRM transitions listed in Table 3. An 11-point calibration curve was prepared in carrot matrix for 170 pesticides from 3.33 g/kg (ppb) to 6670 g/kg. Table 4 shows the results of these analyses.
Table 4. Results from the Analysis of a Carrot Extract with Incurred Pesticides by GC/MS in the Scan Mode with DRS Analysis, by GC/MS in the SIM Mode, and by GC/MS/MS in the MRM Mode (An X implies that the compound was found by that method.) GC/Q 5 L (Multimode inlet) Cold SL scan Cold SL SIM + DRS GC/QQQa 1 L Hot SL (ppb) 0.38b 0.75b 2.3b 0.53b 1.2b 24.7 3.7 X X X X X X Not in method X 240 9 Sum = 45 130 Not in method
The single quad methods were not quantitative, so Table 4 only indicates (with an X) if a pesticide was found, either by DRS or by manual examination of the SIM data. Since the triple quad method was calibrated, the amount of each pesticide could be determined. The amounts reported are those found in the extract. Because the extraction method concentrated this sample by a factor of 4.5:1 (4.5 g of carrot to 1.0 mL of final extract), the pesticide concentrations in the original carrot samples were actually lower by this factor. The scan method with DRS analysis has the capability to find any of the 927 compounds in the database, while the SIM and MRM methods are limited to the 175 target compounds listed in Table 3. DRS found fenazaquin, a pesticide that was not in the SIM or MRM methods. This demonstrates the advantage of using GC/MS with DRS for screening purposes in combination with GC/MS/MS for target compound analysis. In spite of the concentrated carrot matrix, the GC/QQQ was able to detect three pesticides below 1 ppb (1 g/kg) and three more below 5 ppb. The lowest level calibration standard was prepared at 3.33 ppb, so numbers reported below that level are extrapolated values. The optimal MRM transitions for p,p'-DDD and o,p'-DDT are the same and, since these two compounds were only partially resolved chromatographically, they are reported together. Figure 2A shows the extracted quant ion (m/z 246) for p,p'-DDE from the scan analysis of the carrot sample. Interferences in these chromatograms make it harder to do an accurate quantitative analysis without first deconvoluting the spectrum. After deconvolution (Figure 2B), ChemStation integration is trivial. Figure 2C shows the EIC (m/z 246) from the GC/MS SIM analysis of the same sample. Although the signal/noise ratio (S/N) is 10-fold better, there appear to be more interferences. It is easy to see the advantage of the GC/QQQ for target compound analysis. A 1-L injection of the carrot extract on this instrument gave a clean MRM chromatographic peak (Figure 2D) with better S/N (434) than was obtained for the 5-L GC/Q SIM analysis (S/N = 375)(Figure 2C).
Pesticide Diclobenil Pentachlorobenzene Trifluralin Tefluthrin 4,4'-Dichlorobenzophenone Chlorpyrifos o,p'-DDE p,p'-DDE o,p'-DDD p,p'-DDD o,p'-DDT p,p'-DDT Fenazaquin
a. The actual concentration of these compounds was lower in the original carrot sample by a factor of 4.5 since the extraction method results in 4.5 g of produce per mL of extract. b. The reported values fall below the lowest point on the calibration curve.
A S/N = 35
C S/N = 375
1.5 1.4 1.3 1.2 1.1 1.0 0.9 Counts 10 5 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 _0.1 11.8
246.0 & 176.1 246.0 & 175.1 Ratio=23.8
S/N = 434
11.9
12
12.1
12.2
12.3
Acquisition time (min)
Figure 2.
A) p,p-DDE quant ion (m/z 246) extracted from the scan chromatogram obtained from a 5-L cold splitless injection of a carrot extract with incurred pesticides. B) Same as in (A) but after deconvolution. C) p,p-DDE quant ion (m/z 246) extracted from a SIM chromatogram obtained from a 5-L cold splitless injection of the same sample. D) Quant and qualifier transitions (246.0 & 176.1 and 246.0 & 175.1, respectively) for the GC/MS/MS analysis of a 1-L hot splitless injection of the same carrot extract. Peak-to-peak signal/noise ratios for the extracted ions and the quant transition are shown. The ratio of the two transition ions (D) is 23.8, confirming the presence of p,p-DDE.
Comparing GC/MS SIM to GC/MS/MS MRM Various Matrices Figure 3 compares GC/MS SIM results to GC/MS/MS MRM results for p,p'-DDE spiked into various commodities at 10 ppb. On the left, the SIM EICs for the quant ion (m/z 246) show increasing amounts of matrix interference from the
apple, cabbage, ginseng, orange, and spinach samples. In contrast, the p,p'-DDE GC/MS/MS transitions shown on the right have no interferences from any of the extracts. The large S/N values shown for the quant transition (246.0 & 176.1) suggest that one should be able to detect p,p'-DDE at the sub-ppb level.
10
EIC (246)
1.2 1.0 Counts 10 4 0.8 0.6 0.4 0.2 0.0 -0.1 11.8
MRM
246.0 & 176.1 246.0 & 175.1 Ratio = 23.5
Apple
S/N = 448
11.9 12 12.1 12.2 Acquisition time (min) 246.0 & 176.1
12.3
Cabbage
8 7 6 5 4 3 2 1 0
246.0 & 175.1 Ratio = 22.9
Counts 10 3
S/N = 241
11.8
11.9 12 12.1 12.2 Acquisition time (min)
12.3
Ginseng
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
246.0 & 176.1 246.0 & 175.1 Ratio = 23.3
Counts 10 4
S/N = 446
11.8
11.9 12 12.1 12.2 Acquisition time (min) 246.0 & 176.1 246.0 & 175.1
12.3
Orange
8 7 6 5 4 3 2 1 0
Counts 10 3
Ratio = 23.6
S/N = 456
11.8 1.4 1.2 Counts 10 4 1.0 0.8 0.6 0.4 0.2 0.0 11.8
12.3
246.0 & 176.1 246.0 & 175.1 Ratio = 24.0
Spinach
S/N = 260
12.3
Figure 3.
A comparison of GC/MS SIM and GC/MS/MS MRM analysis for p,p-DDE spiked at 10 ppb (10 g/kg) into five different produce extracts. On the left, the EICs for the p,p-DDE quant ion (m/z 246) show increasing amounts of interference from the matrix. The transitions on the right (246.0 & 176.1 and 246.0 & 175.1) for p,p-DDE are clean, with peak-to-peak S/N values ranging from 241 to 448. All injections were 1 L.
11
Tomato Extract All three techniques being discussed were able to identify incurred chlorothalonil in a tomato extract, which was present at 1 ppm. However, only the GC/QQQ was able to identify pentachlorobenzonitrile, a chlorothalonil metabolite, which it measured at 9.3 ppb. Figure 4 shows the MRM transitions for pentachlorobenzonitrile and a calibration curve for the compound ranging from 3.33 ppb to 6670 ppb. Backflushing the Column The norm when analyzing dirty samples by GC/MS is to replace the inlet liner and clip the column frequently. Many labs do this daily. Otherwise, matrix accumulates in the liner and column, degrading the chromatography. Over time, these materials migrate through the GC column and contaminate the source, which then needs to be cleaned. This problem may be compounded with a GC/QQQ instrument because one does not see much evidence of the matrix and the temptation is to ignore maintenance until the source (and sometimes the first quadrupole) needs to be cleaned. The Agilent 7000A Series triple quad MS uses the same inert source and gold-plated quartz quadrupole that are found in the 5975C MSD. These can be heated up to 350 C and 200 C, respectively, which greatly minimizes the need for cleaning, even when high-boiling matrix compounds do reach the detector. The best way to prevent chromatographic degradation and reduce the need for source cleaning is to backflush the GC column during or after each run. With the configurations shown in Figure 1, backflushing is done for 3 to 5 minutes after the run by raising the pressure at the capillary flow device (two-way splitter or the Purged Ultimare Union) and A
2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 274.9 & 239.9 274.9 & 204.9 Ratio=62.1 Responses 10 6
lowering the inlet pressure. This reverses the flow through the column and purges high-boiling matrix components from the head of the column and out through the inlet's split vent. During the course of this work, approximately 100 1-L injections of concentrated food extracts were made into the GC/QQQ system with no evidence of column or MS performance problems. Nearly 300 L of these same extracts were injected into the GC/Q system before column and inlet maintenance was required. With the capillary flow device installed, you can do this maintenance without venting the mass spectrometer.
Conclusions
Agilent's 7890A/7000A triple quadrupole MS system is a sensitive and rugged tool for target pesticide analysis. There is far less interference from matrix than one sees in single quadrupole methods, making it much easier to quantify pesticides at the low ppb levels required by today's legislation. In many cases a 1-L injection into the GC/QQQ produced far better results than a 5-L injection into the GC/Q. Nevertheless, there is still a need for screening methods that look for hundreds of pesticides. For this, we recommend using large-volume injection with Agilent's new Multimode inlet, GC/Q analysis in the scan mode, and data analysis using Deconvolution Reporting Software with Agilent's Pesticide and Endocrine Disruptor Database. The combination of these two approaches is the best way to screen for more than 900 contaminants (by GC/Q with DRS) while performing ultra-trace analysis for a smaller list of target compounds (using GC/QQQ). Both approaches benefit from column backflushing, which is highly recommended when analyzing dirty samples, such as food extracts.
4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0
Pentachlorobenzonitrile - 11 Levels, 11 Levels Used, 11 Points, 11 Points Used, 0 QCs y = -0.0030 * x ^ 2 + 670.6288 * x - 343.7991 R^2 = 0.99970460
Counts 10 3
6.6
6.7 6.8 6.9 7 Acquisition time (min)
7.1
_ 500
500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 Concentration (ng/mL)
Figure 4.
A) MRM transitions identifying pentachlorobenzonitrile at 9.3 ppb in a tomato extract. B) A calibration curve for pentachlorobenzonitrile from 3.33 to 6,670 ppb with a quadratic curve fit > 0.999.
12
References
1. 2. 3. 4. M. Anastassiades, S. J. Lehotay, D. Stajnbaher, and F. J. Schenck, J AOAC Int, 86 (2003) 412. S.J. Lehotay, A. de Kok, M. Hiemstra, and P. Bodegraven, J AOAC Int, 88 (2005) 595. QuEChERS Web site, http://www.quechers.com M. Mezcua, M. A. Martnez-Uroz, P. L. Wylie, and A. R. Fernndez-Alba, "Simultaneous screening and target analytical approach by GC-q-MS for pesticide residues in fruits and vegetables," Accepted for publication by J. AOAC Int. Chin-Kai Meng and Mike Szelewski, "Replacing Multiple 50-minute FPD/ELCD/SIM Analyses with One 15-Minute Full-Scan Analysis for 10X Productivity Gain," Agilent Technologies publication 5989-7670EN. Philip L. Wylie, "Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library," Agilent Technologies publication 5989-5076EN. Bruce D. Quimby, Leonid M. Blumberg, Matthew S. Klee, and Philip L. Wylie, "Precise Time-Scaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking," Agilent Technologies publication 5967-5820E. Chin-Kai Meng, "Improving Productivity and Extending Column Life with Backflush," Agilent Technologies publication 5989-6018EN. Philip L. Wylie, "Direct Injection of Fish Oil for the GCECD Analysis of PCBs: Results Using a Deans Switch with Backflushing," Agilent Technologies publication 5989-6095EN.
5.
6.
7.
8.
9.
Acknowledgements
The authors wish to thank Dr. Jon Wong of the U.S. Food and Drug Administration (College Park, MD, USA) and Dr. Steven Lehotay of the U.S. Department of Agriculture (ARS, ERRC, Wyndmoor, PA, USA) for providing numerous standards and food extracts.

13
Agilent Technologies, Inc., 2009 Printed in the USA February 25, 2009 5990-3578EN
Analysis of Complex Samples by GC/MS/MS Pesticides in Marine Biota

Application Brief
Author
Chris Sandy Agilent Technologies, Inc. UK and Ireland Sales Headquarters 710 Wharfedale Road Winnersh Triangle Wokingham, Berkshire, RG41 5TP UK
Introduction
The analysis of organochlorine pesticide residues in marine biota samples (mussels) by GC/MS is extremely challenging. Despite accelerated solvent extraction (ASE), followed by size exclusion chromatography and alumina clean-up, the extracted samples still contain large amounts of matrix materials. On a single quadrupole GC/MS system, these materials not only interfere with the quantitative analysis in selected ion monitoring (SIM) mode, but also cause problems with both the liner and GC column. This results in the shifting of GC retention times and decreasing signal intensities. In addition the MS ion source is quickly contaminated. Avoiding retention time shifts is especially important when acquiring data in multiple reaction monitoring (MRM) mode with a GC/QQQ since complex multiresidue analysis requires careful placement of MRM segment times. This application brief describes the analysis of a marine biota sample using the Agilent 7000A Triple Quad GC/MS system in MRM mode in combination with Agilent Capillary Flow Technology to provide backflushing of high-boiling materials. System up-time is maximized and the need for maintenance is reduced by keeping the chromatographic system and MS ion source cleaner of high-boiling matrix materials between each injection.
The Analytical Challenge

The full-scan total ion chromatogram (TIC) (Figure 1) shows the chromatogram of a marine biota extract from 50 to 550 amu. After all the analytes have eluted from the column, large amounts of matrix material still remain in the chromatographic system.
Figure 1.
Full-scan TIC of a marine biota extract from 50 to 550 amu.
The SIM TICs of clean solvent injections on either side of two extract injections (Figure 2) shows the extent of the chemical background that builds up in the gas chromatograph.
Figure 3. Agilent GC Method Translation software.
Clean solvent after 2 extract injections

Autosampler 2-way Capillary flow splitter with makeup
AUX EPC 4.0 psig
Clean solvent before extract injections Figure 2. SIM TICs of clean and extract injections illustrating the buildup of the chemical background in the GC.
G7000A QQQ
15 m 0.25 mm id 0.25 um HP-5MS UI 0.80 m 0.15 mm id UDFS
Experimental
Agilent Method Translation software (Figure 3) was used to create a 2x speed method from the normal 42-minute analysis provided by the Agilent RTL Pesticide method. The column used was a 15 m 0.25 mm id 0.25 m HP-5MS Ultra Inert column (19091S-431 UI). The column head-pressure was converted to allow for the 4.0 psig AUX pressure applied at the end of the column at the capillary flow device and the system run in constant pressure mode (Figure 4). Injection volume was 2 L splitless.
Figure 4.
7890A
Agilent 7000A Triple Quad GC/MS system in MRM mode.
The 7000A mass spectrometer was operated in MRM mode. Table 1 gives the retention times, mass transitions, and collision voltages used for each analyte.
Table 1. Analyte Retention Times, Mass Transitions, and Collision Voltages
The collision cell was operated with nitrogen at 2.60 psi and helium quench gas at 6.25 psi. The TIC MRM is shown in Figure 5; each MRM segment is indicated by a grey marker line.
Figure 5.
TIC of the Agilent 7000A Triple Quad GC/MS system in MRM mode.
Calibration
Internal standard calibration curves were created over the range of 0.8 to 200 ppb. Example MRM quant transitions for low-concentration standards of dieldrin and endrin are shown in Figure 6.
Figure 6.
Example MRM quantitative transitions for low-concentration standards of dieldrin and endrin.
Quantitative Results
Table 2 shows quantitative results (ppb) from three replicate injections of the marine biota extract. Two of the chlorinated pesticides (dieldrin and beta hexachlorocyclohexane [HCH]) were identified and quantitated.
Table 2.
Quantitative Results from Three Replicate Injections of Marine Biota Extract
Conclusion
The combination of Agilent Capillary Flow Technology and backflush mode with the high selectivity and sensitivity of the Agilent 7000A Triple Quad GC/MS system operated in MRM mode provides a robust solution for the analysis of organochlorine pesticides at trace levels in marine biota extracts.

Agilent Technologies, Inc., 2008 Published in the USA December 19, 2008 5989-9727EN
November 2008
Comprehensive Screening, Confirmation, and Quantification of Organic Pesticides in Foods by GCMS and LCMS
The global focus on keeping food safe has intensified, resulting in major changes in the number of pesticides that are being regulated and monitored, as well as the allowable levels of those pesticides in food. These new regulations drive a need for a wide range of comprehensive and complementary analytical methods and instrumentation that possess the throughput, sensitivity, and breadth of pesticide compound coverage to meet the demand of testing laboratories all over the world. This article provides an overview of the instrument platforms, tools, and workflow for analyzing pesticides.
Philip Wylie, Jerry Zweigenbaum, Melissa Churley, Chin-Kai Meng, and Cao Zhe
ith the globalization of trade, food production and distribution have become truly international businesses. When we dine out, the fish might come from Japan, the rice from Australia, the spices from China, and the strawberries from Mexico. We take it for granted that the food we eat is safe and free from contamination that could make us seriously ill. One of the 20th-century advancements that made this cornucopia possible was the development of pesticides. Crops that were previously devastated by pest damage can now be produced at very high yields to meet worldwide demand. However, all pesticide compounds are potentially harmful to humans, and safeguards must be in place to assure that consumers are not exposed to dangerous levels of these pesticides. Many countries have had regulations regarding the use of pesticides, and their allowable levels in foods, for many years.
With the increasing globalization of the food industry, there is greater scrutiny on food safety, resulting in major changes in the number of pesticides that are being regulated and monitored, as well as the allowable levels of those pesticides in food.
A Global Effort to Keep Food Safe

The Food Quality Protection Act (FQPA) of 1996 updated federal regulation of pesticides in the United States to implement a single health-based standard for all pesticides in foods. It provides special protection for infants and children, expedites approval of safe pesticides, and requires periodic evaluation of pesticide registrations and tolerances to ensure that the scientific data supporting registrations will remain up to date in the future. Under this mandate, the US Food and Drug Administration (FDA) performs regulatory monitoring by sampling individual lots of domestically produced and imported foods
their products to assure compliance with FQPA, the EU regulation, and the Japan Positive List, if they hope to sell their products in the U.S., Europe, and Japan, which generate much of the worlds demand for imported food. These new regulations drive a global need for analytical methods that possess the throughput, sensitivity, and breadth of pesticide compound coverage to meet the demand of testing laboratories all over the world that must ensure compliance.
A Broad Range of Complementary Analytical Methods Is Needed

Food testing laboratories need the ability to detect and quantify hundreds of pesticides, in myriad foodstuffs, at very low levels of contamination. They might be screening for a set of hundreds of known target compounds, using a defined test method for each target in the set. Alternatively, a testing laboratory might wish to look for the presence of any pesticide (unknowns), without the need for defining a method for each possible compound. They are often called on to do both types of testing, and no single analytical approach can provide the flexibility required to meet this need. Wide variations in the chemical properties of pesticide contaminants and the necessity to detect a very large number of compounds require a range of chromatography and mass spectrometry (MS) systems. For pesticides that can be vaporized easily without degradation, gas chromatography (GC)MS (single quadrupole) is an ideal analytical tool, due to the availability of large libraries of pesticide spectra and deconvolution software. This approach can be used to screen, confirm, and quantify both target and unknown compounds. Many foodstuffs, such as garlic and ginger, are very complex, or dirty, due to the presence of a very large number of background compounds. GC triple quadrupole MS (QQQ) is very useful for screening, confirming, and quantifying trace-level target compounds in these complex matrices. Pesticides that are quite polar, not easily vaporized, thermally labile, or not easily derivatized are best analyzed using liquid chromatography (LC) methods. LCQQQ is particularly useful for analyzing sets of known target compounds.
Figure 1: GCMS analysis of p,p-DDE in spinach. Data were generated using a J&W DB-5ms column, a model 7890A GC system, and a model 5975C MSD mass spectrometer (Agilent Technologies).
and analyzes them for pesticide residues to enforce established tolerances (http://www.cfsan.fda.gov/~dms/pes0406.html). Domestic samples are collected at the point of production, and import samples are collected at the point of entry into U.S. commerce. Products typically are analyzed as the unwashed, whole, raw commodity. The FDA also performs a total diet study (TDS) on four regional market baskets per year, each basket comprising about 300 foods that represent the average U.S. consumers diet. The foods are prepared as they would be consumed (table-ready) and tested for regulated pesticides. The FDA also performs focused sampling that is generally used to follow up on suspected problem areas or acquire residue data on food commodities that are not usually covered during regulatory monitoring. There are more than 1000 registered pesticides in the U.S., and approximately 400 with tolerances established by the Environmental Protection Agency and enforced by FDA. The European Union (EU) has also recently updated a harmonized regulation of pesticide residues in food, with the new standard taking effect on September 8, 2008 (http://ec.europa.eu/food/plant/protection/pesticides/index_en.htm). This regulation sets maximum residue levels (MRL) for food and animal feed. The
MRL is the highest level of a pesticide residue that is legally tolerated. The regulation covers around 1100 pesticides currently or formerly used in agriculture in or outside the EU. For pesticides that are not specifically mentioned, a general default MRL of 0.01 mg/kg (10 ppb) applies. The EU designates the minimum number of samples to be taken by each member state, and EU reference laboratories coordinate, train staff, develop methods of analysis, and organize tests to evaluate the skills of the various national control laboratories. Japan also adopted a stricter pesticide regulation in 2006, referred to as the Positive List system, which establishes MRLs for about 500 pesticides (http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/introduction.html). Produce with pesticide residues exceeding these MRLs or the default tolerance of the uniform limit of 0.01 ppm (10 ppb) for those pesticides for which there is no Japanese MRL cannot be marketed in Japan. The regulation also lists 16 pesticides that must not be detected in food. Japan tested more than 200,000 food samples in 2007. The increased international focus on pesticide contamination in food impacts food producers in virtually any country, since the food market is a global one. Producers in Thailand, for example, must test
m/z
m/z
m/z
Figure 2: Confirmation of p,p-DDE in spinach using DRS and the Pesticide and Endocrine Disruptor spectral library (Agilent Technologies).
A unique fragment ion for each pesticide is routinely used for screening of each target compound a second precursor to product ion transition can then be used to confirm the identity of the compounds. The transition with the greatest response is typically used to quantify the pesticides. Laboratories may be faced with the challenge of determining all pesticides, if any, present in a food sample. LCtime of flight MS (TOF) screening methods provide extremely accurate molecular mass information (typically < 2 ppm) that can be searched against exact mass databases of thousands of compounds. Compounds found in this type of screen can be confirmed and quantified using LCMS-MS analysis on either QQQ or quadrupole time-of-flight (Q-TOF) instruments.
Analyzing for Target Pesticides

One of the most common activities in a food-testing laboratory is analysis for a set of target pesticides of specific interest. A finite set of pesticides might be used for production of a particular food crop, or a specific set might be especially
toxic to children, necessitating the routine analysis for those compounds. GCMS (single quadrupole) utilizing deconvolution software has broad applicability for this type of analysis. The Automated Mass spectral Deconvolution and Identification System (AMDIS) was developed by the National Institute of Standards and Technology, and is incorporated into deconvolution reporting software (DRS) available from Agilent Technologies. Deconvolution separates unresolved peaks at any given retention time into sets of cleaned component spectra that can be matched to library spectra. The analyst runs the method that was used to generate the Pesticide and Endocrine Disruptor Database Library and locks the method, so that retention times will match those stored in the database. DRS then uses deconvolution and locked retention times to identify pesticides from the database. This is particularly useful in the analysis of extracts from foods such as spinach (Figure 1). In this case, one of the pesticides of interest is p,p-
dichlorodiphenyldichloroethylene (p,pDDE), a breakdown product of DDT. The total ion chromatogram (TIC) does not show a discernible peak at the p,p-DDE retention time (24.07 min). Deconvolution analysis found 768 different components over the whole TIC. After a pesticide library search, 12 of these had a match, including p,p-DDE. The black ions in the bottom window of Figure 1 represent the raw scan, before deconvolution. The white ions are a deconvoluted component that matches the known spectrum of p,p-DDE. Plotting the extracted ion chromatograms for this component (246, 248, 316, and 318) in the top window of Figure 1 also helped to confirm the presence of the p,p-DDE peak in the TIC. Figure 2 illustrates the extraction of the p,pDDE spectrum from the raw spectrum, and its match to the library spectrum, resulting in a positive confirmation of the presence of this pesticide in the spinach sample. The use of powerful integrated deconvolution software such as DRS, in combination with specialized database libraries such as the Japanese Positive List Pesticide Library (1) or the more comprehensive Pesticide and Endocrine Disruptor Library (2) (Agilent Technologies, Palo Alto, California), can increase the efficiency of a food testing laboratory greatly. These GCMS methods can screen for hundreds of compounds and quantify a subset of them in as little as 2 min after the run using DRS. With traditional GCMS data processing, an analyst would need as much as 15 min per sample to review the data and confirm pesticide hits. LCMS methods are also very useful for analysis of target pesticides that are too polar or thermally labile for GCMS. LCQQQ is especially suited to very complex food matrices that can interfere with the analysis.
Abundance
Abundance
Abundance
Dealing with Dirty Sample Matrices

Sample preparation is a significant challenge with food analysis and makes up a considerable portion of the total pesticide analysis time. Robust analysis methods that can tolerate very complex sample matrices without loss of sensitivity or accuracy are a must to reach throughput goals.
y=208335.0516 x - 152717.0728 R2=0.99956076
Figure 3: Calibration curve for diazinon in pepper using a six-point curve from 0.1 to 100 ng/mL with a linear fit and no origin treatment. Data were generated using a ZORBAX SB-C-18 1.8-m LC column with a model G6410A QQQ system in the positive ESI mode (Agilent Technologies).
Area 1 Area 2 Area 3 Area 4 Area 5 Mean RSD %
1197 1280 1347 1186 1291 1260 5.4
Figure 4: Precision of GC-QQQ quantification of 500 fg of cyanophos spiked in garlic. Data were generated using an HP5ms UI (Ultra Inert) 30 m 0.25 mm, 0.25-m film thickness GC column and a model 7000A QQQ MS system (Agilent Technologies). The pesticide cyanophos was spiked into a garlic matrix at 0.5 ppb and replicate analyses performed at 500 fg on-column. The peaks are superimposed for the five runs, with the replicate peak areas, mean peak area, and RSD listed in the table to the left.
Multiple reaction monitoring (MRM) with a QQQ system provides detection of trace amounts of target compounds in complex matrices, providing screening, confirmation, and quantification in one analysis (although a replicate analysis is usually required in most laboratories for confirmation). The precursor ion of interest generated by the source is selected and isolated in the first (Q1) quadrupole. A hexapole collision cell (Q2) fragments the molecule, and the product ions are selected in Q3. By choosing product ions from this transition that are characteristic for the pesticides of
interest, chemical noise is separated from signal, providing very high sensitivity and selectivity, even in very dirty matrices. With careful selection of highly specific product ions for each pesticide, it is possible to create an MRM assay that can be used to identify and quantify hundreds of pesticides in a sample simultaneously. Efficient use of LCQQQ MRM methods requires optimization of a few parameters, including voltages on the fragmentor and the collision cell. Parameters must be optimized to generate the most intense product ions representative for each target compound, and
automated method optimization software simplifies this task enormously. LCQQQ is an excellent platform for screening, confirmation, and quantification of pesticides.A screening method uses one transition per compound to look for a large number of pesticides on the target list. Confirmation of pesticides detected in the screening is done using two transitions per compound. For example, 301 pesticides have been rapidly identified in one analysis with most having a limit of detection (LOD) of 0.01 mg/kg (10 ppb), which is the MRL for baby food in the U.S. and the general default MRL for foods in
the European Union. These levels were reached in a single analysis using positive ion electrospray with 99 transitions per time segment, and a quantifying and qualifying ion for each compound (3). Two types of vegetables (pepper and tomato) were evaluated, and a linearity of response over three orders of magnitude was shown for quantification (r2 > 0.99; Figure 3). GCQQQ also provides sensitive and
reproducible screening, confirmation, and quantification of pesticides that are not highly polar, particularly in very complex matrices such as garlic and ginger. Despite the high background of these matrices, pesticide compounds can be detected at levels as low as 0.5 ppb, with very high signal to noise and peak area precision as good as 6% RSD (Figure 4). For certain active pesticides such as acephate, the
Using oven backflush it took an additional 33 min at 320 C to remove those boilers
Run stopped at 42 min, backflushed at 290 C for 7 min Blank run after backflushed shows the column was clean
Time (min)
Figure 5: Backflushing the GC column for higher signal to noise and shorter cycle times. Comparison of backflushing using capillary flow (red chromatogram) versus column bake-out (blue chromatogram) utilizing a model 7890A GC system and an HP-5ms GC column (Agilent Technologies).
use of chemical protectants can reduce adsorption of analytes during the analysis, increasing sensitivity and improving peak shape. The result can be a marked improvement in accuracy of quantification in complex matrices (4). Backflushing the GC column, rather than using oven bakeout, also removes many high-boiling interfering substances in the sample (Figure 5) without altering retention times in subsequent runs or shortening the life of the column (5). Backflushing improves data quality while shortening cycle time and increasing column life. It also eliminates ghost peaks and reduces ion source contamination by keeping excess column bleed and heavy residues from being introduced into the mass spectrometer. One key to the successful rapid analysis of large numbers of pesticides is the dwell time of the QQQ mass spectrometer the time spent to measure the signal of any one precursor to product ion transition. The shorter the dwell time, the more transitions that can be analyzed per unit time, and the more pesticides that can be quantified per unit time.
Figure 6: Pesticide screening for unknowns using LCQTOF. The top panel shows the full spectrum total ion chromatogram (TIC) using a model 1200 LC system, a ZORBAX Eclipse XDB, 5-m column, and a model 6510 LCMS QTOF system (Agilent Technologies). The Molecular Feature Extractor found 510 compounds in the TIC, 15 of which had hits from the EXACT MASS database search (bottom panel). The three highlighted compounds were further confirmed by MSMS analysis.
Analyzing for Unknown Pesticides

Often, food-testing laboratories want to determine the entire spectrum of pesticides that might be present in a sample, rather than just a limited set of target compounds. While GCMS using deconvolution is an effective tool to screen for over 900 pesticides (2), LCTOF or LCQTOF can be used to screen for any number of LCMS-amenable compounds. TOF and QTOF are extremely accurate MS techniques, routinely achieving <2 ppm mass accuracy.Analysis software such as the Molecular Feature Extractor (MFE) from Agilent Technologies can find all ions representing real compounds in the sample. Noise and other extraneous ions are excluded. The resulting list of masses is then searched against a database of theoretical exact masses of compounds based on their molecular formulas and selected adducts, to identify pesticide compounds (Figure 6). This method has been used to screen food extracts for 600 pesticides and their degradates. The limit of detection (LOD) was determined for 100 of the 600 pesticides, and varied from <0.01 mg/kg (10 ppb) for 34% of the compounds, to <0.5 mg/kg (500 ppb) for 95% of the compounds (6). Using the MFE algorithm streamlines the data analysis for screening of hundreds of compounds at sensitive levels to minutes, compared with hours or even days using a manual approach. TOF can be used to identify an unlimited number of ionizable compounds, and the sensitivity is not impacted by the number of compounds screened.While LCTOF can provide screening and quantification of both unknown and targeted pesticides, LCQTOF or another MS-MS approach is required to confirm identification.
Conclusion
One need only look to recent events to appreciate the importance of pesticide test-
ing in food.A series of poisonings in Japan in December of 2007 and January of 2008 that sickened 10 people was linked to methamidophos pesticide contamination of gyoza dumplings imported from China. The impact on Japanese food buying habits was immediate, as importation of Chinese vegetables fell 40% shortly after the source of the poisoning was determined. In another example, pet food vegetable proteins contaminated with melamine were imported into the United States from China in early 2007, resulting in the deaths of several cats and dogs. More recently, the same contaminant has appeared in baby formula in China. Food-testing laboratories require a broad range of complementary MS-based analytical technologies and methods to screen, confirm, and quantify the presence of as many as 1000 pesticide compounds. Screening methods can be used to detect as many contaminants as possible in a single run, and these methods dont necessarily have to quantify all the compounds. The screening method can be followed by target compound analysis that is very selective, sensitive, and quantitative. For LCamenable pesticides, this means running LCTOF or LCQTOF followed by LCQQQ. For GC-amenable pesticides, this means running GCMS with DRS for screening followed by GCQQQ for target compound analysis at low levels. The method used by any given laboratory is determined by its specific pesticide analysis goals, existing instrumentation, the budget for acquisition of new instrumentation, and required LODs. When choosing new instrumentation, laboratories doing pesticide testing also should consider whether a particular manufacturer has a complete solution that includes all of the above platforms, and whether those platforms share common, interchangeable ion sources, a unified soft-
ware platform, and transferable collision voltages, thus providing a uniform workflow. The robustness of a manufacturers systems is a key factor, including the tolerance of the ion source to contamination and its required frequency of cleaning and maintenance. Utility of the software and its integration with comprehensive pesticide spectra libraries and exact mass databases are also keys to success. Finally, the manufacturers reputation for excellence in GC and LCMS instrumentation should be considered.
References
(1) P.L. Wylie, Screening for pesticides in food using the Japanese Positive List Pesticide Method, Agilent Application Note 59897436 (2007). (2) P.L. Wylie, Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library, Agilent Application Note 5989-5076 (2006). (3) J.A. Zweigenbaum, Multiresidue analysis of 301 pesticides in food samples by LC/Triple Quaduprole mass spectrometry, Agilent Application Note 5989-8614 (2008). (4) M. Anastassiades et al., J. Chromatogr., A 1015, 163184 (2003). (5) C.-K. Meng, Improving Productivity and Extending Column Life with Backflush, Agilent Application Note 5989-6018 (2006). (6) J.A. Zweigenbaum, Automated screening of 600 pesticides in Food by LC/TOF MS using a molecularfeature database search, Agilent Application Note 59895496 (2006).
Philip Wylie, Jerry Zweigenbaum, Melissa Churley, ChinKai Meng, and Cao Zhe are Senior
Applications Chemists, Agilent Technologies, Santa Clara, California.
Printed in U.S.A.
Reprinted from Current Trends In Mass Spectrometry, a Spectroscopy Supplement, November 2008
A Rapid GC/MS Solution for the Japanese Positive List Pesticide Method Application Brief
Philip L. Wylie
In 2006, the Japanese Ministry of Health, Labour and Welfare introduced a new system for the regulation of pesticides, feed additives, and veterinary drugs. This "Positive List System" stipulates that only compounds on the approved list can be used in food production and it provides the framework for regulation of these compounds. These new regulations have increased the need for analytical methods capable of detecting residues of these chemicals in a wide variety of food products. The majority of the chemicals under regulation are pesticides, and their residues are most often measured by gas or liquid chromatography with mass spectral detection (GC/MS or LC/MS). To address the need for rapid and comprehensive analysis of food samples in the Japanese market, Agilent has introduced a new Japanese Positive List Pesticide Database for use with its Deconvolution Reporting Software (DRS). With this new database and DRS, analysts can screen their GC/MS data files for the 430 GC-amenable pesticides that are being regulated by the Japanese government. The process is fully automated and takes about two minutes per sample. Food extracts for pesticide residue analysis are usually very dirty because one risks removing pesticides along with endogenous compounds if elaborate cleanup steps are used. So, the challenge is to detect traces of pesticides in samples that contain a lot of interferences. DRS helps the analyst "see through" the interferences and identify target compounds pesticides in this case. The heart of DRS is software from the National Institute of Standards and Technology (NIST) called Automated Mass Spectral Deconvolution and Identification System (AMDIS). AMDIS profiles background noise, corrects for spectral skew, and interpolates between scans to find a more accurate retention time for every ion in every spectrum. It then looks for ions that rise and fall together and have the same retention time. These ions are grouped together into a "component" spectrum, which is searched against the Japanese Positive List Pesticide Library. In essence, AMDIS pulls clean, library-searchable spectra out of the background. To illustrate the power of deconvolution, a green pepper extract was spiked with 229 pesticides, of which approximately 148 can be analyzed by GC/MS. The total ion chromatogram for this extract is shown in Figure 1. Figure 2 shows the undeconvoluted spectrum at 17.840 minutes (top) and the deconvoluted spectrum (middle). The deconvoluted spectrum is an excellent match to the fenhexamid library spectrum (bottom).
Highlights
Deconvolution extracts clean, library-searchable spectra from coeluting peaks. Agilent's Japanese Positive List Pesticide Database and Deconvolution Reporting Software (DRS) enable screening of GC/MS data files for the 430 pesticides regulated by the Japanese government in a single run. Retention time locking eliminates almost all false positive hits. The fully automated process takes only 2 to 3 minutes per sample and is more reliable than conventional GC/MS methods. The DRS summary report quantifies calibrated compounds on the basis of their undeconvoluted and deconvoluted spectra.
Figure 1.
GC/MS total ion chromatogram (scan mode) of a green pepper extract spiked with 229 pesticides at 100 ppb each. The arrow shows where the pesticide fenhexamid elutes. The pesticide is buried under a much larger unresolved peak.
Figure 2.
Top: Undeconvoluted spectrum taken at 17.840 minutes for the green pepper extract shown in Figure 1. Middle: Deconvoluted spectrum at 17.840 min. Bottom: Japanese Positive List Pesticide Library spectrum of fenhexamid.
Figure 3 shows part of a DRS report for the spiked green pepper extract. Large portions of the table were removed since it is shown simply to illustrate the table's format and the information it provides. The first three columns give the retention time, CAS number, and name for each target analyte found by the DRS software. The next two columns show the calculated amount determined by the Agilent GC/MS ChemStation. The first value (column 4) is determined using the undeconvoluted target ion and represents the typical result one gets from a standard GC/MS calibration method. The amount under the "AMDIS" heading uses the deconvoluted target ion for its calculation. Note that the quant results shown in Figure 3 are only estimates based on an average response factor that was used for all pesticides in the calibration table. Compounds were not calibrated individually. Many laboratories prepare a multilevel calibration for a list of target pesticides and use average calibration factors for the remaining compounds.
MSD Deconvolution Report

Sample name: Data file: Date/time: Gr pepper C:\MSData\Sept 04_07 Lehotay samples using TID & Japanese method\Gr Pepper_TID_2uL.D 1:59:58 PM Monday, July 14, 2008 Adjacent peak subtraction = 2 Resolution = High Sensitivity = Very high Shape requirements = Low
The NIST library was searched for the components that were found in the AMDIS target library.
Amount (ng/uL) AMDIS NIST Chem- AMDIS Match R.T. diff Reverse sec. match Station 2.4 88 82 0.19 0.17 6.4699 10265926 Methamidophos 1.2 86 84 0.13 0.10 Dichlorvos 6.6154 62737 1.0 80 88 0.07 0.05 Carbofuan-7-phenol 7.2062 1563388 0.5 83 87 0.28 0.22 15.397 32809168 Procymidone 1.5 66 67 0.46 0.37 Thiabendazole 15.407 148798 0.6 90 97 1.02 0.15 Piperonyl butoxide 18.217 51036 0.3 86 0.35 0.16 18.354 106325080 Epoxiconazole 59 18.355 52669928 1-Butanone, 4-[4-(4-chlorophenyl)1,2,3,6-tetrahydro-1-pyridyl]-1-(4fluorophenyl)0.5 58 44 0.07 0.04 18.532 36734197 Iprodione 66 0.2 0.20 82 0.25 Pyridaphenthion 18.539 119120 0.1 52 46 0.03 0.05 18.573 135410207 Acetamiprid 90 0.7 96 1.06 1.39 18.696 82657043 Bifenthrin 90 0.7 92 0.84 1.04 20.674 96489713 Pyridaben 60 0.8 69 0.04 0.10 20.682 67747095 Prochloraz 86 1.4 88 0.15 0.16 24.042 110488705 Dimethomorph-(E) R.T. CAS # Compound name 12.578
Figure 3.
Hit num. 1 1 1 1 1 1 1
11 1 1 1 1 1 1
Phenanthrene-d10
10.00
Partial DRS report for a green pepper extract spiked with 229 pesticides at 100 ppb each. Some of these are not GC-amenable. Of the 229 spiked pesticides, 156 were found by this GC/MS method. Only a portion of the report is shown. The quantitative results shown in columns 4 and 5 are only approximations based on an average response factor. No individual compound calibrations were performed.
The column labeled AMDIS/Match shows the quality of the spectral match between the deconvoluted spectrum and the Japanese Positive List Pesticide Library spectrum. The reporting threshold can be set by the user. Matches below the threshold are not reported. The Japanese Positive List Pesticide Database contains retention times for each pesticide using the published Japanese positive list method. With retention time locking (RTL), users can easily reproduce the database retention times. AMDIS can be configured to reject a "hit" if the compound's retention time falls outside of a user-settable window (typically 10 sec). This feature greatly reduces false positive hits. The next column in Figure 3, labeled AMDIS/R.T. Diff sec, shows the difference between the retention time of the analyte in the chromatogram and its value stored in the Japanese Positive List Pesticide Library. Of the 156 pesticides identified by DRS, fewer than 10 had retention times more than 2 seconds away from their database values. In extremely dirty samples, peaks may shift a bit but the retention time window can be widened to accommodate them. Even then, the retention time requirement eliminates almost all false positives.
3
As a general rule, matches greater than 60 are very reliable, especially when there is also a good match to the NIST library. For the DRS report shown in Figure 3, the minimum match value was set to 40. Several pesticides were reported with AMDIS match factors between 40 and 60. For these, it is good to review the results in Qedit, which now (using DRS version A.04 or greater) includes AMDIS results combined with the ChemStation results. As a confirmation step DRS takes the deconvoluted spectrum for each "hit' and searches it against the entire NIST mass spectral library using the NIST library searching algorithm. If the hit is found in the NIST library, its match value and rank are reported. A "hit number" of 1 means that the compound identified by AMDIS was also the best match in the NIST library. Much more information about this method may be found in an earlier Agilent Application [1]. The benefits of DRS Version A.04 are discussed in a recent Agilent eSeparation Times article [2].
References
1. Philip L. Wylie, "Screening for Pesticides in Food Using the Japanese Positive List Pesticide Method: Benefits of Using GC/MS with Deconvolution Reporting Software and a Retention Time Locked Mass Spectral Database," Agilent Technologies publication 59897436EN, www.agilent.com/chem. 2. Mike Szelewski, "I'm interested in quantitation using deconvolved areas. Are the advantages of DRS Version A.04 great enough to justify learning this new software?" eSeparation Times, 21(2), www.agilent.com/chem.
Summary
This Applications Brief discusses a method to screen for the GC-amenable pesticides that are being regulated by the Japanese Ministry of Health, Labour and Welfare (MHLW) under the Positive List System. The new Japanese Positive List Pesticide Database contains mass spectra for 430 pesticides and their locked retention times. Deconvolution is used to extract clean, library-searchable spectra from coeluting peaks. Deconvolution Reporting Software (DRS) compares all of the deconvoluted spectra to the Japanese Positive List Pesticide Database to find any spectral matches. Hits must also fall within narrow retention time windows, made possible by the use of retention time locking. Screening takes 2 to 3 minutes per data file and is more reliable than conventional GC/MS methods. DRS produces a summary report showing compounds identified by the ChemStation, AMDIS, and NIST library searching. Calibrated compounds are quantified on the basis of their undeconvoluted and deconvoluted spectra. This method screens for all 430 pesticides of interest to the MHLW in a single run. Reference 1 contains a complete description of the method.
Acknowledgement
The author would like to thank Drs. Stephen Lehotay and Katerina Mastovska of the U.S. Department of Agriculture for providing the spiked green pepper extract.

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2008 Published in the USA September 18, 2008 5989-9320EN
Screening for 430 Pesticide Residues in Traditional Chinese Medicine Using GC/MS: From Sample Preparation to Report Generation in One Hour Application
Traditional Chinese Medicine
Authors
Wei Luan and Zhixiu Xu Agilent Technologies (Shanghai) Co., Ltd. 412 Ying Lun Road Waigaoqiao Free Trade Zone Shanghai 200131 China
Introduction
Traditional Chinese Medicine (TCM), which has been used in China for thousands of years, shows very good therapeutic effects in China as well as other countries. More TCM crops than ever are being cultivated in large-scale farming operations. The pesticides that are used on the farm need to be monitored to make sure that the amount of residues is kept within the safe range. TCM have very complex matrices, several hundred kinds of pesticides are often used, and the amount of residue is very small. All these facts make pesticide residue screening a very difficult task to accomplish in a short time. There are several steps to solve this problem, including sample pretreatment, instrument acquisition, data processing, and report generation. The basic requirements for using a gas chromatograph/mass spectrometer (GC/MS) are listed here: 1. An effective and easy sample preparation method. 2. Repeatable and reproducible retention times. 3. A database containing the information about retention time and mass spectra QuEChERS, a sample preparation method for pesticide analyses in foodstuff, is the acronym for Quick, Easy, Cheap, Effective, Rugged, and Safe developed by Dr. Steven J. Lehotay and his colleagues [1-5]. QuEChERS is selected here because it has many advantages over traditional techniques, such as high recovery for pesticides wide
Abstract
A total solution of screening for pesticide residues in Traditional Chinese Medicine (TCM) is described in this application, including sample preparation procedure, instrument acquisition parameter, advanced mass spectrum data processing, and automatic qualification and quantitation report generation. Several techniques are used in this screening method. QuEChERS is employed to shorten the sample pretreatment time, reduce its cost, and environmental contamination. A significant instrumentation innovation is the capillary flow technology of Agilent 7890A, which can perform the backflushing to remove chemical noise caused by dirty matrices and save cycle time. Agilent's Deconvolution Reporting Software (DRS) uses the AMDIS deconvolution algorithm, which can be used to screen for 430 GC-amenable pesticides with the new Japanese Positive List RTL database in only 2 or 3 minutes. Overall, this screening method, from preparing the sample to generating a report, can be accomplished in about 1 hour.
range of polarity and volatility, high sample throughput, and the use of a smaller amount of organic solvent and no chlorinated solvents [1-5]. QuEChERS meets the first basic requirement mentioned above. Retention time locking (RTL) [6] is the ability to very closely match chromatographic retention time in any Agilent 6890/7890 GC system to those in another 6890/7890 GC system with the same nominal column. It is remarkable that the repeatability of retention time in Agilent 7890GC is better than 0.01 minute. RTL can precisely match the retention time between different instruments, which meets the second basic requirement mentioned above. On November 29, 2005, the Japanese Ministry of Health, Labour, and Welfare (MHLW) published a Positive List system for the regulation of pesticides, feed additives, and veterinary drugs. This system is one of the most restrictive regulations on pesticide residues all over the world. Referring to this regulation, Agilent developed a new database called the Japanese Positive List RTL database, which includes all GC-amenable pesticides in the Japanese Positive List, together with other pesticides that are monitored by the Japanese Quarantine Stations. The information in the Agilent database comprises the retention time and spectra of each pesticide, which meets the last basic requirement mentioned above [7]. In addition, the advanced requirement for screening is to rapidly identify pesticides, with minimal false positives and negatives. Agilents Deconvolution Reporting Software (DRS) [7] is designed to automatically deconvolute the spectra from matrices and generate a qualification and quantitation report. DRS integrates information from three processes: MSD ChemStation, Automated Mass Spectral Deconvolution and Identification System (AMDIS), and National Institute of Standards and Technology (NIST) search. DRS increases the confidence in results for complex matrices; the typical data processing time is about 2 or 3 minutes, which perfectly meets the advanced requirement mentioned above.
This application describes how to use the techniques mentioned above to develop a whole process for screening pesticides in TCM matrices. The investigated targets include different officinal parts of TCM, such as Flos Lonicerae Japonicae for flower, Radix Astragali for root and stem, Folium Ginkgo for leaf, Radix et Rhizoma Ginseng for root, and Fructus Lycii for fruit. After about a 20-minute pretreatment by QuEChERS, the extract is injected into an Agilent 7890A/5975C. The system then runs the Japanese Positive List method with backflushing in about 37 minutes. At the end, after a 2- or 3-minute DRS processing with the Japanese Positive List RTL database, a qualification and quantitation report is automatically generated.
Experimental
TCM raw materials were purchased from a local medicine store, including Flos Lonicerae Japonicae, Radix Astragali, Folium Ginkgo, Radix et Rhizoma Ginseng, and Fructus Lycii. All materials are finely ground. Sample Preparation Weigh 1 g of ground TCM and add it into a 50-mL Teflon tube; add 6 mL distilled water to thoroughly soak the sample and store the sample at ambient temperature for 2 hours. Add 4 mL 0.1% HAc/acetonitrile, and shake vigorously for 1 minute by hand, then add 1.2 g MgSO4, 0.6 g NaAc, and 4 g NaCl, and shake vigorously for 1 minute by hand. Centrifuge for 3 minutes at 4,000 rpm. Add 500 mg PSA (p/n 5188-5364), 250 mg C18 (p/n 189-1302), and 250 mg MgSO4 in a 5-mL centrifuge tube. Transfer an aliquot (1.5 mL) of the above extracts upper layer into the tube and then centrifuge for 1 minute at 3,000 rpm. If the solution is dark after centrifugation, graphitized carbon black can be added to adsorb the pigment (for example, chlorophyll). Transfer an aliquot (1.0 mL) of the extract into the GC vials. For dedicated quantitation analysis after screening, analytical protectants and internal standard can be added to the final solution.
GC/MS The Japanese Positive List method was published by MHLW for pesticide residues on all Japanese agricultural products and imports to Japan. Here, the method was modified to handle the dirty TCM matrices by adding the backflushing. The detail is displayed in Table 1. System Software Requirement Agilent MSD ChemStation (p/n G1701EA, Ver. E.02.00); Agilent Deconvolution Reporting Software (p/n G1716AA, Ver. A.04.00); NIST05b Mass Spectral Library (p/n G1033A), includes NIST MS Search (Ver. 2.0d) and AMDIS (Ver. 2.65); Agilent Japanese Positive List Pesticide RTL Database (p/n G1675AA).

QuEChERS Method Optimization for Versatile TCM Matrices In Traditional Chinese Medicine formulation, different edible parts of TCM plants can be chosen to make each dose. Moreover, the TCM matrices are versatile in chemical constituents, such as saponins, polysaccharids, flavones, phenolic acid, and fatty acid. Therefore, it is a tough job to determine the pesticide residues in such kinds of complex matrices as TCM. In our study, Flos Lonicerae Japonicae was investigated at first, then we mixed different kinds of TCM together to be more representative in respect to chemical constituents. Flos Lonicerae Japonicae for flower, Radix Astragali for root and stem, Folium Ginkgo for leaf, Radix et Rhizoma Ginseng for root, and Fructus Lycii for fruit are the five selected TCM.
Table 1. GC
Gas Chromatography and Mass Spectrometers Conditions Agilent Technologies 7890A or 6890N Split/splitless Splitless 280 C Helix double taper deactivated (p/n 5188-5398) 17.446 psi 50 mL/min 0.75 min 54.51 mL/min On 20.0 mL/min 1.00 min Helium C/min 25 10 36.5 min 0.5 min J&W DB-5ms Ultra Inert (p/n 122-5532UI) 30 m 0.25 mm 0.25 m Constanr flow 17.446 psi (2 psi for post run) 1.51 mL/min (4.05 mL/min for post run) Front Inlet PCM (p/n G1530-63309) 2 psi (60 psi for Post run) Next C 50 125 300 300 Hold min 1 0 10 5 Front Injector Sample washes Sample pumps Injection volume Syringe size Preinj solv A Preinj solv B Postinj solv A Postinj solv B Viscosity delay Plunger speed Preinjection dwell Helium MSD Acquistion mode Solvent delay Low mass High mass Threshold Sampling Quad temp Source temp Transfer line temp Tune file Gain factor Trace ion detection RTL Backflush Device Restrictor Size
Front Inlet Injection type Inlet temp Inlet liner Pressure Purge flow Purge time Total flow Gas saver Saver flow Saver time Gas type Oven Oven ramp Initial time Ramp rate Ramp rate Post run Total run time Equilibration time Column Length Diameter Film thickness Mode Pressure Nominal initial flow Inlet Outlet Outlet pressure
0 6 2 L 10 L 0 0 6 6 1 second Fast 0 minutes 0 minutes Agilent Technologies 5975C with TripleAxis Detector Scan 3.30 min 35 450 100 (or setting according to noise level) 2 150 C 230 C 280 C Atune.u 10 On System retention time locked to Chlorpyrifos methyl at 13.443 min 3-way splitter (p/n G3183B) 0.706 m 0.18 mm id 3
Regarding the QuEChERS method, the amount of absorbent, such as PSA and C18, is critical to the recovery of pesticides. It is demonstrated that PSA is more effective in dealing with the acidic component in TCM extract. By screening for different doses of PSA, 500 mg is enough for the analysis. A typical chromatogram of mixed TCM sample after QuEChERS is shown in Figure 1. Backflushing Using Capillary Flow Technology for Complex TCM Matrices Backflushing is demonstrated to significantly remove chemical noise from run to run by reversing the column flow to push the higher boilers out of the inlet end of the column. Traditionally, the higher boilers stay near the front of the column until the oven temperature is high enough to move them through the column. It is well known that baking the column at high temperature is inclined to cause heavy column bleed and result in shorter column lifetime. For a dirty matrix, even baking for a long time cannot thoroughly remove the higher boilers, which may result in a retention time shift in the following injection. Backflushing is a perfect solution to avoid high-temperature baking and retention time shift from run to run. Backflushing minimizes the contamination that goes into the detector, which is especially valuable for the MSD because it reduces ion source cleaning. Another
positive attribute of backflushing is that it shortens cycle time and improves lab productivity. Details about backflushing, including a method development procedure, can be found in the Agilent application in reference 8. In our experiment, a three-way splitter with makeup gas was used to perform the backflushing. The device has extra makeup gas supply tubing and four plug-in spots: one for the analytical column and three for the restrictor tube connecting to three available detectors. Since only MSD is used as a detector in this application, the first two spots are plugged in the direction of makeup flow. The third spot is used for column in and the last spot is used for restrictor in. The main purpose of this configuration is to avoid peak broadening due to extra dead volume. The length and internal diameter of the restrictor tubing is 0.706 m and 0.18 mm, respectively. A photo of the device and the GC/MS system configuration schematics are displayed in Figure 2. Target Pesticides Identification in TCM Matrices by DRS Another challenge for pesticides analysis is the identification of trace-level target compounds in complex matrices, which are often disturbed by high-level chemical noise and result in poor library match factors. Background subtraction is the tradi-
Figure 1.
Typical Total Ion Chromatogram of mixed TCM sample after QuEChERS: Flos Lonicerae Japonicae, Radix Astragali, Folium Ginkgo, Radix et Rhizoma Ginseng, and Fructus Lycii.
0.706 m 180 m restrictor
7683 Auto-sampler
AUX EPC
3-way splitter with make gas 2 ports plugged
Restrictor out to MSD Column in
5975C MSD 7890A GC
Plugged Column Plugged
30 m 250 m 0.25 m DB-5MS UI
Aux EPC in
Figure 2. Schematics of the GC/MS system configuration and tubing connection for three-way splitter.
tional way to improve the match factor; however, it is both matrix- and operator-dependent and can yield inconsistent results. DRS, which Agilent introduced in 2004, is a software package that combines the information from three separate processes into one easy-to-read report. The primary benefit of DRS is significant time saving when interpreting results from complex matrices analyses. Moreover, a suitable database is critical to perform DRS successfully. The Japanese Positive List RTL database is a new DRSusable library for pesticide analysis, and includes the retention time and mass spectra for 430 GCamenable pesticides. Detailed information about the database can be found in the Agilent application in reference 7. Using DRS with the Japanese Positive List RTL database is highly recommended for data processing and report generation for pesticide residues analysis in TCM matrices. Figure 3 is part of a typical DRS report for a spiked mixed TCM sample. Figure 4 shows the benefit of using DRS in the analysis of pesticides in TCM matrices. Pirimiphosmethyl was spiked into mixed TCM sample at a level of 10 ppb. The compound was missed by MSD ChemStation but successfully picked up by DRS. The upper window is the Total Ion Chromatogram (TIC), the middle window is the raw or dirty spectrum in the scan No. 2199 (13.981 min), and the lower window is the comparison of the deconvoluted spectrum (the white plot) with the spectrum of pirimiphos-methyl in the Japanese Positive List RTL database (the black plot). After deconvolu-
tion, the spectrum of the scan No. 2199 is "clean," and the Pirimiphos-methyl can be easily identified in the mixed TCM sample.
Conclusions
The primary interference for trace-level pesticides analysis is chemical noise in complex TCM matrices. This application introduces several techniques to eliminate the interference. Well-developed sample preparation is the main approach to remove as much of the chemical noise as possible, QuEChERS was employed to simplify the sample preparation procedure and improve lab productivity, completing the process in only about 20 minutes. However, even with a pretreatment method such as QuEChERS, the chemical noise cannot be completely removed. Backflushing is an alternative approach to eliminate the chemical noise from run to run. Although backflushing is not a new concept, capillary flow technology can perform a backflushing much more effectively than traditional techniques can. In this application, a three-way splitter with makeup gas is used for backflushing, and the data-acquisition step with backflushing is almost 37 minutes. The last step related to chemical noise in data processing is the utilization of a mathematic algorithm called deconvolution. Agilent's DRS can shorten the data processing time to only 2 or 3 minutes, which is very helpful for pesticide analysis in TCM matrices with the new Japanese Positive List RTL database. Altogether, only 1 hour is needed to screen 430 pesticide residues in TCM matrices from sample pretreatment to screening report generation.
5
Figure 3.
Part of typical DRS report.
Figure 4.
AMDIS display showing the total ion chromatogram of a mixed TCM sample (top window), the spectrum where pirimiphos-methyl elutes (middle window), and the deconvoluted spectrum (in white) juxtaposed to the library spectrum for pirimiphos-methyl (in black) (bottom window).
References
1. M. Anastassiades, S. J. Lehotay, D. Stajhbaher, and F. J. Schenck, Fast and Easy Multiresidue Method Employing Acetonitrile Extraction/Partitioning and 'Dispersive Solid-Phase Extraction' for the Determination of Pesticide Residues in Produce, J. AOAC Int., 2003; 86(2): 412 2. S. J. Lehotay, K. Mastovska, and S. J. Yun, Evaluation of Two Fast and Easy Methods for Pesticide Residue Analysis in Fatty Food Matrixes, J. AOAC Int., 2005; 88(2): 630 3. S. J. Lehotay, K. Mastovska, and A. R. Lightfield, Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005; 88(2): 615 4. S. J. Lehotay, A. de Kok, M. Hiemstra, and P. van Bodegraven, Validation of a Fast and Easy Method for the Determination of 229 Pesticide Residues in Fruits and Vegetables Using Gas and Liquid Chromatography and Mass Spectrometric Detection, J. AOAC Int., 2005; 88(2): 595 5. M. Anastassiades, K. Mastovska, S. J. Lehotay, Evaluation of Analyte Protectants to Improve Gas Chromatographic Analysis of Pesticides, J. Chromatogr. A, 2003; 1015(1-2): 163 6. V. Giarocco, B. Quimby, and M. Klee, Retention Time Locking: Concepts and Applications, Agilent Technologies publication 5966-2496E 7. Philip L. Wylie, Screening for Pesticides in Food Using the Japanese Positive List Pesticide Method: Benefits of Using GC/MS with Deconvolution Reporting Software and a Retention Time Locked Mass Spectral Database, Agilent Technologies publication 5989-7436EN 8. Chin K. Meng, Improving Productivity and Extending Column Life with Backflush, Agilent Technologies publication 5989-6018EN

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2008 Printed in the USA August 18, 2008 5989-9341EN
Replacing Multiple 50-Minute GC and GC-MS/SIM Analyses with One 15-Minute Full-Scan GC-MS Analysis for Nontargeted Pesticides Screening and >10x Productivity Gain Application
Food Safety
Author
Chin-Kai Meng and Mike Szelewski Agilent Technologies 2850 Centerville Road Wilmington, DE 19808
Introduction
To have a plentiful food supply, most fruits and vegetables are treated with pesticides (insecticides, fungicides, herbicides, etc.) to protect primarily against insects, molds, and weeds. Therefore, in order to ensure food safety, the food supply is frequently monitored for pesticide residues. Nowadays, the pesticide monitoring is expanding beyond food, for example, to botanical dietary supplements. The analytical challenge to monitor (identify and quantify) trace multiresidues requires an effective and universal extraction and analysis method for maximum productivity and efficiency. Up to now, the trade-off in analysis has been between sensitivity and confirmation. Element-selective gas chromatograph (GC) detectors, such as the flame photometric (FPD), electron capture (ECD), electrolytic conductivity (ELCD), and halogen selective (XSD) detectors, provide excellent selectivity and sensitivity; however, they lack the capability to identify. On the other hand, mass spectrometry (MS) is capable of identifying an analyte by fullscan library match or multiple target and qualifier ion ratios from selected ion monitoring (SIM). However, MS sometimes lacks the selectivity to
Abstract
Pesticide analysis of fruits and vegetables requires finding trace-level residues in complex matrices. Up to now, the typical trade-off is between sensitivity and confirmation. Therefore, multiple injections are needed for screening and confirmation using gas chromatography with mass spectrometry (GC-MS) or GC-MS in combination with GC with element-selective detection. With the recent introduction of hardware and software tools, for example, capillary flow three-way splitter, trace ion detection, and deconvolution, a 15-minute fast analysis can match the results obtained from three injections of approximately 50 minutes each. A table comparing the results from the Food and Drug Administration/Center for Food Safety and Applied Nutrition (FDA/CFSAN) procedure using the traditional multi-instrument approach and the new Agilent single injection approach shows that Agilent's fast analysis is capable of finding all the target analytes in less than one-tenth of the current FDA/CFSAN total analysis time.
find target analytes in a complex matrix full of interferences and chemical background. Analyte spectra are sometimes overwhelmed by similiar ions contributed from the coextractives in the matrix that prevent the analyte of interest from being identified or confirmed. The compromise and the typical approach are to use selective GC detector(s) to flag potential target analytes and use MS SIM for confirmation. For instance, many laboratories screen food samples for semivolatile pesticides using the ECD or ELCD (or XSD) for organohalogen, FPD or pulsed FPD (PFPD) for organophosphorus, and NPD for nitrogen-containing targets [1 5]. Any found targets are further confirmed by GC-MS/SIM. In addition, other procedures have used GC-MS/SIM entirely for the screening of pesticides in foods [6 8]. In most of these procedures, multiple injections are needed to identify hundreds of compounds at the detection limit in the low parts-per-billion (ppb) levels. To improve the efficiency and increase the productivity of screening for all of these pesticides, the challenge is to reduce the GC-MS or the combination of GC and GC-MS analysis times. There are several hundred pesticides typically used in the world, and each country has its own pesticide tolerance levels for different agricultural commodities. This presents another analytical challenge in multiresidue monitoring: to develop a nontargeted procedure to identify pesticides at trace levels in different food matrices. These challenges are met by the recent introduction of hardware and software tools, including GCMS, capillary flow three-way splitter, trace ion detection, and deconvolution reporting software (DRS). The splitter allows multiple GC as well as MS signals to be acquired from a single injection for productivity gains (from three injections down to one). Trace ion detection minimizes noise on the signal and DRS separates target analyte ions from matrix background ions. Several sample extracts were analyzed by the current Food and Drug Administration/Center for Food Safety and Applied Nutrition (FDA/CFSAN ) multiple injection process and this new Agilent pesticide system. With DRS, the demand for chromatographic resolution is minimum; therefore, the Agilent system was running the analysis at a 3x faster speed (one-third the analysis time) to further increase productivity. A table comparing the
results from the current FDA/CFSAN multi-instrument approach and the new Agilent single-injection approach shows that not only is Agilents fast analysis capable of finding all the target analytes, but it is also accomplished in just one-tenth of the current FDA/CFSAN total analysis time.
Experimental
Sample Preparation Sample extracts of fresh produce were prepared by FDA based on modifications of the QuEChERS protocols [9, 10]: Homogenize 1 to 2 kg of sample 15 g sample + 15 mL 1% AcOH/ACN, homogenized Add 6 g MgSO4 and 2.5 g NaOAc, shake vigorously for 1 minute, and centrifuge Transfer ~15 mL + 0.5 g C-18 + 1.2 g MgSO4, shake, and centrifuge for 5 min at 3,000 rpm Transfer ~12 mL + 0.4 g PSA + 0.2 g GCB + 1.2 g MgSO4, vortex Add 4 mL toluene, shake, and centrifuge Transfer 6 to 8 mL, evaporate and bring to volume with toluene, add I.S. Add MgSO4, vortex and centrifuge, transfer to ALS vials GC and GC-MS analysis Sample preparation of dried ginseng powder is similar to that used for fresh produce, but smaller sample sizes (2 g) were used [11]. Capillary Flow Three-Way Splitter One of the capillary flow devices is a three-way splitter, which consists of two half plates bonded together (diffusion bonding) to form a plate with the etched flow channels inside. The splitter is only 6.5 cm tall and 3 cm wide and is mounted on the side of the oven wall (see Figure 1). The low thermal mass minimizes cold spots and peak broadening. All capillary flow devices use metal column ferrules, have extremely low dead volumes, are inert, and do not leak, even after many oven cycles.
To MSD
To FPD To ECD Column in
Aux EPC in
different inlets and outlets. Aux pressure can be either an inlet (for the splitter flow restrictors connected to different detectors) or an outlet (for the analytical column). A graphical user interface makes the configuration easy to set up. Once all the columns and restrictors are configured, the backflush can be executed easily. Backflush Traditional bakeout step for removing late eluters could be very time consuming, or even as long as the analysis time depending on the matrix. Backflush is a simple technique to remove high boilers from the column faster and at a lower column temperature to cut down analysis time and increase column lifetime. Capillary flow devices (in this case, a three-way splitter) also provide backflush [13, 14] capability. Backflush is a term used for the reversal of flow through a column such that sample components in the column are forced back out the inlet end of the column. By reversing column flow immediately after the last compound of interest has eluted, the long bake-out time for highly retained components can be eliminated. Therefore, the column bleed and ghost peaks are minimized, the column will last longer, and the MS ion source will require less frequent cleaning. The split vent trap may require replacement more frequently than usual. Figures 2 and 3 are two screen shots from the MSD ChemStation software, providing a summary of the backflush operation. In Figure 2, the column and three restrictor dimensions and respective detectors are shown (the setup came from the column configuration section). For MSD, the user can choose the vacuum pump installed on the system. This information will be used to calculate if the backflush is within the system flow limits. By clicking on the Evaluate button, the screen shown in Figure 3 appears, listing the maximum flow for each detector and the void volumes for a certain backflush time. In this example, Aux pressure is at 60 psi, inlet is at 1 psi, and oven is at 280 C. The backflushing flow is shown to be 8.66 mL/min, and the void time is shown to be 0.16 min. Therefore, backflushing for 2.5 minutes will send 15.6 void volumes through the column. This is useful for developing the backflush method. Figures 2 and 3 simplify the setup and development of a backflush method.
Etched flow channel inside two diffusion bonded plates Capillary tube connection via metal ferrule
Figure 1.
Flow diagram of the three-way splitter. The picture shows the splitter mounted on the oven wall.
The three-way splitter enhances productivity by splitting column effluent proportionally to multiple detectors: MSD, dual flame photometric detector (DFPD) and micro-electron capture detector (ECD). Therefore, two GC detector signals can be acquired together with the MS data (both SIM and scan signals if desired) from one injection [12]. The exit end of the analytical column is installed into one of the four ports on the splitter using a metal ferrule. The other three ports are connected to three detectors via restrictors (deactivated capillary tubing) of varying diameter and length to set the split ratio among the three detectors. Restrictors are sized for 1:1:0.1 split ratio in favor of MSD and DFPD (ECD has 1/10 of the flow to MSD), with similar hold-up times. The splitter uses auxiliary (Aux) electronic pneumatics control (EPC) for constant pressure makeup flow. The makeup gas (Aux pressure 6) at the splitter is fixed at 3.8 psi to maintain the split ratio throughout the run. This multisignal configuration provides full-scan data for library searching, SIM data for trace analysis, DFPD (phosphorus or sulfur mode), and ECD data for excellent selectivity and sensitivity from complex matrices. The trade-off is the decrease of analyte concentration in any detector due to the flow splitting and the additional makeup gas from the splitter. An analyte would have similar retention times in all three detectors. Therefore, the GC data can be used in two ways: first, to confirm the presence of target analytes found by the MSD deconvolution reporting software (DRS), and second, to highlight potential target compounds to be further confirmed by MSD. With the new 7890A GC software, up to six columns/ restrictors can be configured/assigned to
Figure 2.
Backflush setup in ChemStation.
Figure 3.
Automated backflush calculations in ChemStation.
Deconvolution
Another useful feature in Figure 3 is the warning, shown as highlighted yellow cells. In this example, setting the backflush pressure to 60 psi sends more than the allowable flow (60 mL/min) to the FPD. Therefore, the backflush pressure setting and the actual flow value to FPD are shown in yellow as warnings. Although the system will accept the setup, the high flow may cause consequences in the analysis, for example, flameout. Trace Ion Detection Trace ion detection [15] is a filtering algorithm to smooth peaks. This filtering is an advanced form of averaging used to remove the noise riding on the signal. The implications from TID are typically a slight loss in peak height and some peak broadening. The default setting in ChemStation for TID is off. It should be turned on for any analysis that uses deconvolution and has more than six sampling points across a peak. TID provides better signal-to-noise ratios and helps deconvolution to confirm target compounds as shown in the Results section. Deconvolution In GC/MS, deconvolution is a mathematical technique that separates overlapping mass spectra into deconvoluted spectra of the individual components. Figure 4 is a simplified illustration of this process. Here, the total ion chromatogram (TIC) and apex spectrum are shown on the left. In a complex matrix, a peak may be composed of multiple overlapping components and matrix background ions; therefore, the apex spectrum is actually a composite of these constituents. A mass spectral library search would give a poor match, at best, and certainly would not identify all of the individual components that make up the composite spectrum. The deconvolution process groups ions whose individual abundances rise and fall together within the spectrum. The deconvolution process first corrects for the spectral skew that is inherent in quadrupole mass spectra and determines a more accurate apex retention time of each chromatographic peak. As illustrated in Figure 4, deconvolution produces a cleaned spectrum for each overlapping component. These individual spectra can be library searched with a high expectation for a good match. Deconvolution significantly reduces chromatographic resolution requirements, allowing much shorter analysis times.
TIC and spectrum

TIC
Deconvoluted peaks and spectra

Component 1 extracted spectrum
Library search each component to identify
Figure 4.
Deconvolution process of three overlapped peaks.
Agilent Deconvolution Reporting Software (DRS) utilizes the AMDIS deconvolution program from the National Institute of Standards and Technology (NIST), originally developed for trace chemical weapons detection in complex samples [16]. DRS presents the analyst with three distinct levels of compound identification: (1) ChemStation, based on retention time and four ion agreement; (2) AMDIS, based on cleaned spectra full spectral matching and expected retention time window as a qualifier; and (3) NIST05 search using a >163,000compound library [17, 18]. In this application, both the ChemStation quantitation database and the AMDIS library have the same 926 entries. These entries include pesticides, numerous metabolites, endocrine disruptors, important PCBs and PAHs, certain dyes, synthetic musk compounds, and several organophosphorus fire retardants [18]. The AMDIS software, shipped with the NIST05 Library CD-ROM, is also capable of deconvoluting selected ion monitoring (SIM) data [19], while previous AMDIS revisions were not. Testing has shown that proper compound identification requires four ions per compound. All Agilent DRS databases are retention time locked and have both full-scan and SIM libraries for AMDIS. Instrument Method The system used for this study consists of an Agilent 7890A GC with split/splitless inlet, a threeway splitter, ECD, DFPD, and 5975 MSD. For a detailed description of SIM/scan and the splitter system configuration, please refer to the experimental section of reference [12]. See Table 1 for hardware detail and settings.
Table 1. Gas Chromatograph, Mass Spectrometer, and Three-Way Splitter Operating Parameters GC Injector Syringe size Injection volume Solvent A wash Solvent B wash Sample wash Sample pump Plunger speed Inlet Mode Inlet temperature Pressure Purge flow Purge time Septum purge flow Septum purge mode Gas saver Gas type Liner Oven Oven ramp Initial Ramp 1 Ramp 2 Ramp 3 Runtime Oven equilib time Post-run time Post-run temperature Column Length Diameter Film thickness Mode Nominal initial flow Outlet Outlet pressure Backflush (post-run) Oven Time Inlet Aux pressure 6 Front detector Temperature Const col + makeup Make gas type Data rate Back detector Temperature 6 Agilent Technologies 7890A with 240V fast oven option Agilent Technologies 7683 10 L 1 L 1 (pre), 3 (post) 1 (pre), 3 (post) 0 4 Fast EPC split/splitless Splitless 250 C ~24.4 psi (chlorpyrifos methyl RT locked to 5.531 min, 3x speed) constant pressure mode 50.0 mL/min 2 min 3 mL/min Switched Off Helium Helix double taper liner, deactivated, p/n 5188-5398 C /min 75 9 24 13.96 min 1.0 min 2.5 min 280 C Agilent Technologies HP-5MS, p/n 19091S-431 15.0 m 0.25 mm 0.25 m Constant pressure RT locked to chlorpyrifos methyl at 5.531 min, 3x analysis speed 3.5 mL/min Aux pressure 6 3.8 psi (Aux EPC pressure to three-way splitter), helium gas 280 C 2.5 min 1 psi 60 psi (column outlet) ECD 300 C 60.0 mL/min Nitrogen 20 Hz Dual FPD 250 C Final (C) 70 150 200 280 Hold (min) 0.67 0 0 3.33 Hydrogen flow Air flow Const Col + Makeup Make gas type Lit offset Data rate Transfer line AUX Thermal 1 AUX Pressure 6 Gas type Initial pressure Backflush pressure MSD Tune file Mode Solvent delay EM voltage Low mass High mass Threshold Samples Scans/sec Quad temp Source temp Three-way splitter Pressure Split ratio MSD restrictor DPFD restrictor ECD restrictor Flow to MSD Flow to DFPD Flow to ECD Makeup (Aux 6) Software GC/MSD ChemStation MS Libraries Agilent part number G1701EA (version E.01.00 or higher) NIST05a mass spectral library (Agilent part number G1033A) Agilent RTL Pesticide and Endocrine Disruptor Libraries (926 entries) in Agilent and AMDIS formats (part number G1672AA) Automated Mass Spectral Deconvolution and Identification Software (AMDIS_32 version 2.65 Build 116.66) NIST MS Search (version 2.0d or greater) (comes with NIST'05a mass spectral library Agilent part number G1033A) Agilent part number G1716AA (version A.03.00 or higher) 75.0 mL/min 100.0 mL/min 60.0 mL/min Nitrogen 2.00 20 Hz 250 C MSD transfer line, 280 C Three-way splitter Helium 3.8 psi 60 psi Agilent Technologies 5975C MSD Atune.u Scan 1.50 min Atune voltage 50 amu 550 amu 0 2 2.91 150 C 230 C Agilent 7890A Option 890, installed during factory assembly 3.8 psi (Aux pressure 6 setting) 1:1:0.1 MSD:DFPD:ECD 1.444 m 0.18-mm id deactivated fused silica tubing, p/n 160-2615-10 0.532 m 0.18-mm id deactivated fused silica tubing, p/n 160-2615-10 0.507 m 0.10-mm id deactivated fused silica tubing, p/n 160-2635-1 3.43 mL/min (at 70 C), 1.53 mL/min (at 280 C) 3.43 mL/min (at 70 C), 1.53 mL/min (at 280 C) 0.343 mL/min (at 70 C), 0.153 mL/min (at 280 C) 3.19 mL/min (at 70 C), 1.52 mL/min (at 280 C)
Deconvolution software Library searching software Deconvolution reporting software

Backflush Example Blank runs, made after separate milk analyses with different backflush (BF) times, are shown in Figure 5. The top TIC is a blank run after a milk extract analysis stopped at 42 minutes and the system backflushed for 1 minute. The next TIC is a blank run after another milk extract analysis stopped at 42 minutes and backflushed for 2 minutes and so on for the other five TICs. It is interesting to confirm graphically that the latest eluters disappeared from the TIC earliest in backflushing.
100000 50000 0 100000 50000 0 100000 50000 0 100000 50000 0 100000 50000 0 100000 50000 0 100000 50000 0
Trace Ion Detection Figure 6 compares the signals when TID is on and off. Visually, it is obvious that TID smoothes the noise riding on top of the signal. When TID was on, Atrazine was successfully identified by AMDIS. When TID was off, Atrazine was not found by AMDIS and resulted in a false negative. Figure 7 compares TID on and off for two different analytical conditions of the same ginseng extract. On the right, the fast (3x) analysis was a 1-L splitless injection with TID on. The analyte Diazinon was found by AMDIS with a peak width less than 5 seconds. On the left side, the normal (1x)
BF for 1 min
BF for 2 min
BF for 3 min
BF for 4 min
Note: late eluters were backflushed out first
BF for 5 min
BF for 6 min
BF for 7 min
10.00 20.00 30.00 40.00 50.00
Column is clean
60.00 70.00 80.00
Figure 5.
Differences in blank runs as the result of seven different backflush times. Atrazine found by AMDIS
TID on
TID off
Atrazine not found by AMDIS Figure 6. The power of deconvolution with TID for atrazine, from AMDIS. 7
analysis was a 5-L cold splitless injection using a PTV with TID off. The 9-second-wide Diazinon was not found by AMDIS, also a false negative. Both examples show that TID is a very useful feature for trace target analysis. Benefits of TID: Improves the signal-to-noise ratio AMDIS is more thorough in identifying components, resulting in fewer false positives Improves library match quality Improves area repeatability, resulting in more reliable quantitation DRS and Splitter Figures 8 and 9 are DRS reports for ginseng and peach extracts with pesticides highlighted (for a detailed explanation of the report, please refer to references [17] and [20]). Figures 10 and 11 show simultaneously collected GC and MS signals (RT locked) for the corresponding ginseng and peach extracts from a three-way splitter. The presence of the GC peaks from the ECD and FPD (P) helps confirm the targets reported by DRS. Each run is finished at 15 minutes using the 3x speed and a 240V oven. With deconvolution, less peak resolution is required for compound identification. A 4minute backflush is added after the run to make sure that the column is clean to maintain the next runs locked RTs for all peaks.
PTV, 1x speed, 5 L injection No TID, Diazinon not found (false negative)
Deconvolution Figures 12 through 15 show the results from AMDIS. There are three spectra for each target compound found by AMDIS. The top window shows the spectrum (scan) from the TIC. This is the only spectrum that would be available for library searching without deconvolution obviously quite useless. The middle window shows the deconvoluted spectrum and the bottom window is the target compounds spectrum in the library. The compound confirmation can be done easily and with confidence by visually comparing the bottom two spectra. The power of deconvolution is appreciated while comparing the top two spectra (the raw scan and the spectrum hidden in the raw scan). It is easy to further confirm the hits found by deconvolution. In Figure 9, four pesticides found by AMDIS in the peach extract have a match factor of about 80 or lower. The four pesticides are Cabaryl, Captan, Propiconazole, and Fenbuconazole. A SIM method of these compounds was set up to analyze the peach extract. By selecting the proper AMDIS library (full-scan or SIM), DRS can process full-scan as well as SIM data files [19]. Figure 16 is the DRS report of the peach SIM analysis. The high match factor (99 or higher) and the small RT difference of all targets found by AMDIS confirm the presence of all compounds.
Splitless, 3x speed, 1 L injection with TID, Diazinon found
Figure 7.
The power of deconvolution with TID for diazinon.
Figure 8.
DRS report for ginseng with pesticides highlighted.
Figure 9.
DRS report for peach with pesticides highlighted.
1e+07 8000000 6000000 4000000 2000000 2.00 9e+07 8e+07 7e+07 6e+07 5e+07 4e+07 3e+07 2e+07 1e+07 2.00 1.660 3.00 3.00 4.00 3.524 5.00 6.00 7.00 8.00 8.699 9.00 10.00 9.413
Ginseng TIC
11.00
12.00
13.00
Tetrachloro-m-xylene (ISTD) Chlorthaldimethyl

4.501 6.248 4.00 3.665 5.00 6.00 7.005 7.00 8.00 9.00 10.00 11.00 12.00 13.00 8.744 8.992 9.093 9.059 9.265 9.528 9.646 9.784
ECD Azoxystrobin
8.290
7.783
6000000 5000000 4000000 3000000 2000000 1000000
Tributylphosphate (ISTD)
9.413
Diazinon
4.831 9.313 7.641
FPD (P)
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
11.00
12.00
13.00
Figure 10. Simultaneous display of MSD and GC selective detector signals for ginseng.
2200000 1800000 1400000 1000000 600000 200000 2.00 1.838 3.5e+07 3e+07 2.5e+07 2e+07 1.5e+07 1e+07 5000000 2.00 5000000 4000000 3000000 2000000 1000000 2.00 3.00 4.00 9.358 4.773 5.00 6.00 7.00 8.00 9.00 10.00 9.826 9.910 2.703 4.696 3.932 3.00 4.00 3.641 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00
Peach TIC
11.00
12.00
13.00
Tetrachloro-m-xylene (ISTD)
3.519
Carbaryl
5.640 5.00
Endosulfan(alpha) 7.556 Phosmet 9.538 Captan

7.093 6.680 7.00 8.00 9.081 9.00 9.523 10.559 10.00 11.00
ECD
12.382 12.00 13.00
6.00
Tributylphosphate (ISTD)
Phosmet FPD (P)
11.066 11.00 12.00 13.00
Figure 11. Simultaneous display of MSD and GC selective detector signals for peach. 10
Scan at 12.299 min
Deconvoluted/extracted spectrum
Library spectrum Azoxystrobin
Figure 12. Raw (dirty) spectrum, deconvoluted (clean) spectrum, and library spectrum of azoxystrobin found in ginseng, from AMDIS.
11
Scan at 5.615 min
Library spectrum Carbaryl
Figure 13. Raw (dirty) spectrum, deconvoluted (clean) spectrum, and library spectrum of cabaryl found in peach, from AMDIS.
12
Scan at 10.776 min
Library spectrum Fenbuconazole
Figure 14 Raw (dirty) spectrum, deconvoluted (clean) spectrum, and library spectrum of fenbuconazole found in peach, from AMDIS.
13
Scan at 8.934 min
Library spectrum Endosulfan sulfate
Figure 15 Raw (dirty) spectrum, deconvoluted (clean) spectrum, and library spectrum of endosulfan sulfate found in tomato, from AMDIS.
14
Figure 16. DRS report from the SIM analysis of peach. Please refer to reference 18 for the explanation of the fictitious CAS number assigned to Propiconazole-II (999048032).
Comparison of Incurred Samples The current approach at FDA/CFSAN is to find a wide suite of organohalogen and organophosphorus pesticide residues. This requires four injections (GC-MS/SIM and GC-ELCD for organohalogen and GC-MS/SIM and GC-FPD for organophosphorus screening) of approximately 50-minutes runtime each (total runtime = 200 minutes). Table 2 shows that FDA found several target compounds in three extracts as well as quantitation
Table 2.
results from both GC and MS. In comparison, using the new tools (splitter, TID, and deconvolution) found as many target compounds and a few more in just one short (15-minute) full-scan analysis. The three-way splitter was used to get selective GC signals (ECD and FPD) for confirmation purposes. Due to column effluent splitting to three detectors (1:1:0.1), the MSD is getting less than half of the amount injected. FDA/CFSAN GC and GC/MS/SIM analyses for organohalogen monitor-
Comparison of the Agilent Pesticide System Results with the FDA Results Agilent DRS (full scan/TID) FDA (FPD, ELCD, SIM) Diazinon (FPD, SIM) GC-FPD or ELCD 25 3 ppb GC-MS/SIM 25 2 ppb
Ginseng
Diazinon Chlorthal-dimethyl Azoxystrobin Carbaryl Captan Endosulfan (alpha) Phosmet Propiconazole I and II Fenbuconazole Chlorothalonil Endosulfan (alpha) Endosulfan (beta) Endosulfan sulfate 1 15-min injection (splitter) found these
Peach
Phosmet (FPD, SIM)
320 37
230 23
Tomato
Chlorothalonil (ELCD, SIM) Endosulfan (alpha) (ELCD, SIM) Endosulfan (beta) (ELCD, SIM) Endosulfan sulfate (ELCD, SIM) 2 50-min injections found these
205 10 16 2 34 4 14 2
153 47 26 4 47 5 21 6
FDA quant results
15
ing found endosulfan sulfate at 14/21 ppb (pg/L) in tomato. Agilent MSD/DRS also found this compound in full-scan mode with less than half the amount reported by FDA/CFSAN available at the MSD due to the split. Several other target compounds that did not contain any halogens or the organophosphorus moeity at the low ppb concentrations were also identified by DRS, such as carbaryl (C12H11NO2) in peach and azoxystrobin (C22H17N3O5) in ginseng. The two FDA/CFSAN procedures for organohalogen and organophosphorus pesticides never would have been able to detect these additional nitrogen-containing pesticides. This shows that deconvolution of data acquired with TID is capable of identifying compounds below 10 pg on column in full-scan mode.
Department of Food and Agriculture, Fresenius J. Anal. Chem, 1991, 339, 376383 2. J. Cook, M. P. Beckett, B. Reliford, W. Hammack, and M. Engel, Multiresidue Analysis of Pesticides in Fresh Fruits and Vegetables Using Procedures Developed by the Florida Department of Agriculture and Consumer Services, J. AOAC Int, 1999, 82, 14191435 3. G. E. Mercer, Determination of 112 Halogenated Pesticides Using Gas Chromatography/ Mass Spectrometry with Selected Ion Monitoring, J. AOAC Int, 2005, 88, 14521462 4. G. E. Mercer and J. A. Hurlbut, A Multiresidue Pesticide Monitoring Procedure Using Gas Chromatography/Mass Spectrometry and Selected Ion Monitoring for the Determination of Pesticides Containing Nitrogen, Sulfur, and/or Oxygen in Fruits and Vegetables, J. AOAC Int, 2004, 87, 12241236 5. USDA Pesticide Data Program Analytical Methods: http://www.ams.usda.gov/science/pdp/ Methods.htm 6. J. Fillion, R. Hindle, M. Lacroix, and J. Selwyn, Multiresidue Determination of Pesticides in Fruit and Vegetables by Gas ChromatographyMass-Selective Detection and Liquid Chromatography with Fluorescence Detection, J AOAC Int, 1995, 78, 12521266 7. S. Nemoto, K. Sasaki, S. Eto, L. Saito, H. Sakai, T. Takahashi, Y. Tonogai, T. Nagayama, S. Hori, Y. Maekawa, and M. Toyoda, Multiresidue Determination of 110 Pesticides in Agricultural Products by GC/MS(SIM), J. Food Hyg. Soc, Japan, 2000, 41, 233241 8. G. F. Pang, C. L. Fan, Y. M. Liu, Y. Z. Cao, J. J. Zhang, X. M. Li, Z. Y. Li, Y. P. Wu, and T. T. Guo, Determination of Residues of 446 Pesticides in Fruits and Vegetables by Three-Cartridge SolidPhase Extraction Gas Chromatography-Mass Spectrometry and Liquid ChromatographyTandem Mass Spectrometry, J. AOAC Int, 2006, 89, 740771 9. M. Anastassiades, S. J. Lehotay, D. Stajnbaher, and F. J. Schenck, Fast and Easy Multiresidue Method Employing Acetonitrile Extraction/Partitioning and Dispersive Solid-Phase Extrac-
Conclusions
The trade-off in trace-level pesticide residue analysis is sensitivity versus confirmation. Therefore, the common practice is to use element-selective GC detectors to screen the extracts and use MS/SIM to confirm hits found by GCs. This can take as many as four injections to have a complete residue analysis from a sample extract. Recent introduction of hardware and software tools, which include the capillary flow three-way splitter, trace ion detection, and deconvolution reporting software, can increase productivity dramatically. With deconvolution the demand for chromatographic resolution is lowered; therefore, the Agilent system can run the analysis at a 3x faster speed to further increase productivity. A single-injection approach even at the 3x fast speed can replace the three-injection approach. A table comparing the results from the current FDA/CFSAN multi-instrument approach and the new Agilent single-injection approach shows that not only is Agilents fast analysis capable of finding all the target analytes, but it can also do it in just one-tenth of the current FDA/CFSAN total analysis time.
References
1. S. M. Lee, M. L. Papathakis, H. C. Feng, G. F. Hunter, and J. E. Carr, Multipesticide Residue Method for Fruits and Vegetables: California
16
tion for the Determination of Pesticide Residues in Produce, 2003, J. AOAC Int, 86:412431 10.S. J. Lehotay, K. Matovsk, and A.R. Lightfield, Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, 2005, J. AOAC Int, 88:615629 11.J. W. Wong, M. K. Hennessy, D. G. Hayward, A. J. Krynitsky, I. Cassias, and F. J. Schenck, Analysis of Organophosphorus Pesticides in Dried Ground Ginseng Root by Capillary Gas Chromatography-Mass Spectrometry and -Flame Photometric Detection, J. Agric. Food Chem, 2007, 55, 11171128 12.Chin-Kai Meng and Bruce Quimby, Identifying Pesticides with Full Scan, SIM, ECD, and FPD from a Single Injection, Agilent Technologies publication, 5989-3299, July 2005 13.Chin-Kai Meng, Improving Productivity and Extending Column Life with Backflush, Agilent Technologies publication, 5989-6018EN, December 2006 14.Matthew Klee, Simplified Backflush Using Agilent 6890 GC Post Run Command, Agilent Technologies publication, 5989-5111EN, June 2006 15.Randy Roushall and Harry Prest, The 5975C Series MSDs: Method Optimization and Trace Ion Detection, Agilent Technologies publication, 5989-6425EN, March 2007 16.http://chemdata.nist.gov/mass-spc/amdis/ explain.html
17. Philip L. Wylie, Michael J. Szelewski, Chin-Kai Meng, and Christopher P. Sandy, Comprehensive Pesticide Screening by GC/MSD Using Deconvolution Reporting Software, Agilent Technologies publication, 5989-1157EN, May 2004 18.Philip L. Wylie, Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library, Agilent Technologies publication, 5989-5076EN, April 2006 19.Mike Szelewski and Chin-Kai Meng, New Features of Deconvolution Reporting Software Revision A.02, Agilent Technologies publication, 5989-4159EN, November 2005 20.Bruce Quimby and Mike Szelewski, Screening for Hazardous Chemicals in Homeland Security and Environmental Samples Using a GC/MS/ ECD/FPD with a 731 Compound DRS Database, Agilent Technologies publication, 5989-4834EN, February 2006
Acknowledgements
The authors would like to thank Jon Wong, FDA/CFSAN, for the sample extracts and results used in this study as well as valuable feedback on this application.

17
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2007 Printed in the USA December 18, 2007 5989-7670EN
Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library Application Note
Food and Environmental
Authors
Philip L. Wylie Agilent Technologies, Inc. 2850 Centerville Rd. Wilmington, DE 19808-1610 USA
endocrine disrupters in about two minutes per sample. Deconvolution helps identify pesticides that are buried in the chromatogram by co-extracted materials. The new database was compared to the smaller one for the DRS analysis of 17 surface water samples. With the new database, DRS found 99 pesticides, metabolites, fire retardants, and related contaminants that were not contained in the original RTL Pesticide and Endocrine Disruptor Library.
Abstract
An updated and greatly expanded collection of mass spectral libraries has been introduced, replacing Agilents RTL Pesticide Library and DRS pesticide solution. The new library contains 926 pesticides, endocrine disruptors, and related compounds 359 more than the original library. Included are all compounds specified for GC/MS analysis in the new Japanese Positive List regulations. All compounds have locked retention times that can be accurately reproduced using an Agilent GC/MS system with the ChemStation's Retention Time Locking software. The new Database can be used as a standard GC/MS library for compound identification or with Agilent's Screener software for identifications based upon retention time and mass spectral matching. The greatest benefit accrues when these libraries are used with Agilents new version of Deconvolution Reporting Software (part number G1716AA version A.03.00). This solution allows one to screen GC/MS files for all 926 pesticides and
Introduction
Several years ago Agilent Technologies introduced Retention Time Locking (RTL) for gas chromatography (GC) and GC with mass spectral detection (GC/MS). RTL software makes it possible to reproduce retention times from run-to-run on any Agilent GC or GC/MS, in any laboratory in the world, so long as the same nominal method and GC column are used (1). Since any laboratory can reproduce retention times generated in another, it is possible to create mass spectral libraries that contain locked retention times. By locking their method to the published database, users can screen GC/MS files for all of the librarys compounds. Hits are required to have the correct retention time as well as the correct spectrum, which eliminates many false positives and gives more confidence in compound identifications (2).
More recently, Agilent introduced Deconvolution Reporting Software (DRS) that incorporates mass spectral deconvolution with conventional library searching and quantification. DRS results from a marriage of three different GC/MS software packages: 1) The Agilent GC/MS ChemStation, 2) The National Institute of Standards and Technology (NIST) Mass Spectral Search Program with the NIST 05 MS Library, and 3) The Automated Mass Spectral Deconvolution and Identification System (AMDIS) software, also from NIST. The original DRS software was intended to be a comprehensive solution for pesticide analysis and, therefore, included the mass spectra (in AMDIS format) and locked retention times for 567 pesticides and suspected endocrine disrupters (3). Recently, Agilent introduced an updated and greatly expanded Pesticide and Endocrine Disruptor Database (part number G1672AA) that now contains 926 entries. This represents the addition of 359 new compounds to the original library. At the same time, Agilent introduced a new version of the DRS software (part number G1716AA version A.03.00) that can be used with any Agilent-provided or user-developed DRS library. Pesticide and Endocrine Disruptor Database Contents The G1672AA Pesticide and Endocrine Disruptor Database contains virtually all GC-able pesticides, including those introduced very recently. In addition, the database includes numerous metabolites, more endocrine disruptors, important PCBs and PAHs, certain dyes (for example, Sudan Red), synthetic musk compounds, and several organophosphorus fire retardants. This new database includes: A conventional mass spectral library for use with Agilent GC/MS ChemStations
A screener database for use with Agilents powerful screener software that is integrated into the GC/MS ChemStation Locked Retention Times for all 926 compounds that any Agilent 5975 or 5973 GC/MS user can reproduce in their laboratory Files for use with Agilents G1716AA (A.03.00) Deconvolution Reporting Software An e-method that can be loaded into Agilents G1701DA (version D.02.00 SP1 or higher) with instrument parameters for acquiring GC/MS files and analyzing the data with DRS. These parameters are listed in Table 1. Example files Application notes On November 29, 2005, the Japanese Government published a Positive List system for the regulation of pesticides, feed additives, and veterinary drugs. Maximum Residue Limits (MRL) have been set for 758 chemicals while 65 others have been exempted from regulation. Fifteen substances must have no detectable residues. Other agricultural chemicals not mentioned have a uniform MRL of 0.01 ppm (4). This new regulation is scheduled to take effect on May 29, 2006. Of the pesticides in the Japanese Positive List, 265 are to be analyzed by GC/MS. The new G1672AA Pesticide library contains mass spectra and locked retention times for all of these compounds. Thus, a laboratory could screen for all 265 positive list compounds and several hundred more pesticides in just 13 minutes after the GC/MS run.
Experimental
Table 1 lists the instrumentation, software, and analytical parameters used by Agilent for pesticide analysis. Depending upon the desired injection volume, a PTV inlet or split/splitless inlet can be used.
Table 1.
Instrumentation and Conditions of Analysis Agilent 6890N Agilent 7683 Injector and AutoSampler Agilent PTV operated in the solvent vent mode or Split/Splitless Agilent 30 m 0.25 mm 0.25 m HP-5MSi (part number 19091S-433i) Helium in the constant pressure mode Chlorpyrifos-methyl locked to 16.596 min (nominal column head pressure = 17.1 psi) 70 C (2 min), 25 C/min to 150 C (0 min), 3 C /min to 200 C (0 min), 8 C /min to 280 C (1015 min) Temp program: 40 C (0.25 min), 1600 C/min to 250 C (2 min); Vent time: 0.2 min; Vent flow: 200 mL/min; Vent pressure: 0.0 psi; Purge flow: 60.0 mL/min; Purge time: 2.00 min 15 L (using a 50-L syringe) Agilent 5975 inert Atune.u Scan (or SIM with SIM DRS library) 50550 u 230, 150, and 280 C, respectively 4.00 min Autotune voltage
Gas Chromatograph Automatic Sampler Inlet Column Carrier gas Retention time locking Oven temperature program PTV inlet parameters Injection volume Mass Selective Detector Tune file Mode Scan range Source, quad, transfer line temperatures Solvent delay Multiplier voltage Software GC/MSD ChemStation Deconvolution Reporting Software Library searching software Deconvolution software MS Libraries
Agilent part number G1701DA (version D02.00 sp1 or higher) Agilent part number G1716AA (version A.03.00) Deconvolution Reporting Software NIST MS Search (version 2.0d or greater) (comes with NIST '05 mass spectral library Agilent part number G1033A) Automated Mass Spectral Deconvolution and Identification Software (AMDIS_32 version 2.62 or greater; comes with NIST '05 mass spectral library Agilent part number G1033A) NIST 05 mass spectral library (Agilent part number G1033A) Agilent RTL Pesticide and Endocrine Disruptor Libraries in Agilent and NIST formats (part number G1672AA)

DRS, which has been described in preceding papers (3,5,6), can be summarized as follows: Three separate, but complimentary, data analysis steps are combined into the DRS. First, the GC/MS ChemStation software performs a normal quantitative analysis for target pesticides using a target ion and up to three qualifiers. An amount is reported for all calibrated compounds that are detected. For other compounds in the database, an estimate of their concentration can be reported based upon an average pesticide response factor
that is supplied with the DRS software. The DRS then sends the data file to AMDIS, which deconvolutes the spectra and searches the Agilent RTL Pesticide Library using the deconvoluted full spectra. A filter can be set in AMDIS, which requires the analytes retention time to fall within a userspecified time window. Because RTL is used to reproduce the RTL database retention times with high precision, this window can be quite small typically 1020 seconds. Finally, the deconvoluted spectra for all of the targets found by AMDIS are searched against the 147,000-compound NIST mass spectral library for confirmation; for this step, there is no retention time requirement.
This approach was rapidly adopted by many laboratories because of its ability to identify pesticides in complex chromatograms containing high levels of co-extracted interferences. Indeed, the solution proved to be so useful that users began to create their own DRS libraries (7). Therefore, the DRS was unbundled from the pesticide database so that it could be used with any agilent-provided or user-created database. The original 567-compound RTL Pesticide Library (G1049A) included pesticides, a few metabolites, and most of the GC-amenable endocrine disruptors that were known at the time. The new version of the library includes many more pesticides, endocrine disruptors, and metabolites. This update also contains important compounds from other classes of contaminants that have been found in food and water supplies. Included are eighteen polychlorinated biphenyls (PCBs), four polybrominated biphenyls (PBBs), several polynuclear aromatic hydrocarbons (PAHs), several organophosphorus fire retardants, three important toxaphene congeners, and three Sudan dyes.
Advantages of Deconvolution Figure 1 shows a screen from AMDIS that illustrates the power of this deconvolution software. The white trace in Figure 1A is the total ion chromatogram while the other three are extracted ions of a deconvoluted peak (a component in AMDIS terminology). Note that the TIC and extracted ions are not scaled to each other and this component is actually obscured by co-eluting compounds. Figure 1B juxtaposes the deconvoluted component spectrum (white) with the complete undeconvoluted spectrum (black). Clearly, this component is buried under co-eluting peaks that would ordinarily obscure the analyte. Figure 1C shows that the deconvoluted peak (white spectrum) is a good library match for norflurazon (black spectrum). The locked retention time for norflurazon in the RTL Pesticide Database is 26.933 min, which is just 2.3 seconds away from its observed RT in this chromatogram. Confidence in peak identifications is greatly enhanced by the combination of spectral deconvolution and locked retention time filtering.
Figure 1.
AMDIS screen showing the identification of norflurazon. A) The total ion and extracted ion chromatograms where norflurazon elutes. B) The deconvoluted component spectrum (white) juxtaposed with the spectrum at 26.972 min (black). C) The deconvoluted component matched to the library spectrum of norflurazon.
Surface Water Analysis - Revisiting an Earlier Study In an earlier study, a comparison was made between Agilents DRS and conventional pesticide analysis (3). The California Department of Food and Agriculture (CDFA) provided data files for 17 surface water extracts that had been analyzed in their laboratory. Since the GC/MS chromatograms were locked to the Agilent pesticide method, it was possible to analyze these data files using DRS without having to re-run the samples. The original DRS analysis was made using the 567-compound RTL Pesticide Database. For comparison, these same data files were re-analyzed using the new 926-compound RTL Pesticide Database. The chromatogram (Figure 2) and the DRS report (Figure 3) from one of these samples are shown below.
Excluding phthalates, seven new compounds (shown with bold type in Figure 3) were identified using the 926-compound database: 4-chlorophenyl isocyanate (a phenylurea herbicide metabolite); 3,4-dichlorophenyl isocyanate (diuron metabolite); tris(2-chloroethyl) phosphate (a fire retardant); caffeine (a stimulant); Cyprodinil (a fungicide); desmethyl-norflurazon (a metabolite of norflurazon, an herbicide); and tris(2-butoxyethyl) phosphate (a fire retardant). Although caffeine is not generally considered to be dangerous, it is included in the database because it has been found frequently in sewage effluent and in numerous waterways together with a various pharmaceuticals and pesticides (8).
2800000 2600000 2400000 2200000 2000000 Abundance 1800000 1600000 1400000 1200000 1000000 800000 600000 400000 200000 5.00 10.00 15.00 20.00 Time 25.00 30.00 35.00 40.00 TIC: E02-557.d\data.ms
Figure 2.
Chromatogram of a surface water extract that was analyzed by DRS using the new RTL Pesticide and Endocrine Disrupter Database. The results of this analysis are shown in Figure 3.
MSD Deconvolution Report Sample Name: E02-557 Data File: C:\MSDChem\1\DATA\CDFA surface water data\E02-557.d Date/Time: 11:24 AM Tuesday, Apr 4 2006 The NIST library was searched for the components that were found in the AMDIS target library. Agilent ChemStation amount (ng) AMDIS match 62 84 99 84 93 67 63 RT Diff (sec.) 3.2 1.8 3.1 2.0 2.1 1.7 7.7 62 1.29 85 98 86 96 83 88 79 95 80 90 97 90 99 90 69 2.2 2.6 2.6 3.0 0.7 1.4 1.0 1.3 1.6 3.2 0.4 1.5 0.4 0.7 0.1 70 87 87 94 89 75 98 65 4.5 1.5 0.5 0.8 3.3 0.3 1.9 71 10 1 69 79 94 83 83 90 1 2 1 1 1 1 3 84 92 88 90 74 86 78 83 74 88 90 84 94 87 2 1 2 1 1 2 1 1 1 4 1 1 1 1 1 NIST reverse match 48 86 95 85 89 84 Hit number 1 2 1 1 2 2
RT 4.4689 4.4689 4.8840 6.3879 6.8357 7.6988 7.9342 8.1112 8.1112 8.941 9.7903 10.0019 10.7109 10.9684 11.6491 12.9326 13.4309 13.7478 15.4048 15.9474 16.5988 17.3653 18.4213 18.9214 20.5633 20.5633 26.4247 26.9700 26.9992 27.3984 28.0127 29.6537 33.9298 33.9298 13.739 Figure 3.
Cas # 106445 0000 104121 102363 759944 95761 131113 25013165 0000 29878317 134623 84662 119619 126738 1582098 122349 115968 1517222 58082 84695 5598130 7287196 84742 51218452 121552612 76470252 23576241 27314132 85687 51235042 78513 117817 84764 0000
Compound name 4-Methylphenol 3-Carbobenzyloxy-4-ketoproline 4-Chlorophenyl isocyanate Diuron Metabolite [3,4-Dichlorophenyl isocyanate] EPTC 3,4-Dichloroaniline Dimethylphthalate Butylated hydroxyanisole 7-Methoxy-2,2,4,8-tetramethyltricyclo [5.3.1.0(4,11)]undecane Tolyltriazole [1H-Benzotriazole, 4-meth-] N,N-Diethyl-m-toluamide Diethyl phthalate Benzophenone Tributyl phosphate Trifluralin Simazine Tris(2-chloroethyl) phosphate Phenanthrene-d10 Caffeine Diisobutyl phthalate Chlorpyrifos Methyl Prometryn Di-n-butylphthalate Metolachlor Cyprodinil 9,9-Dimethoxy-9-sila-9, 10-dihydroanthracene Norflurazon, DesmethylNorflurazon Butyl benzyl phthalate Hexazinone Tris(2-butoxyethyl) phosphate Bis(2-ethylhexyl)phthalate Di-n-nonyl phthalate Phthalic acid, 3,4-dichlorophenyl propyl ester Phenanthrene-d10
DRS report from the analysis of a surface water sample. The compounds shown in bold type were found by the new RTL Pesticide Database but not the original one because these compounds were not included.
For this sample, the ChemStation identified only tolyltriazole at 8.941 min, but AMDIS did not confirm this assignment, nor could it be confirmed manually. Butylated hydroxyanisole was tentatively identified by AMDIS with a low match value, but the retention time is off by 7.7 seconds which is considerably more than most other hits. This compound is not in the NIST library so it could not be confirmed. The ChemStation method used for this analysis required that all three qualifier ions fall within 20% (relative) which is a rigorous requirement for such a complex sample. This explains why so few compounds were found by the ChemStation. Cyprodinil (20.563 min) was identified by AMDIS but the NIST library search failed to confirm its presence. The next line shows that the best NIST library match is an anthracene derivative that is nothing like cyprodinil. This result was obtained when AMDIS was configured to use uncertain peaks as shown in Figure 4. When this feature is
turned off in DRS Compound Identification Configuration, the best NIST library hit for this spectrum is, indeed, cyprodinil. When a compound's identity is ambiguous, as with cyprodinil, it may be useful to perform the DRS search both ways and compare the results. In the comparison described earlier (3), DRS was able to identify all 37 pesticides found by the CDFA chemist. However, DRS completed the task for all 17 samples in about 20 minutes compared to ~8 hours for the manual procedure (Table 2). Moreover, DRS identified one false positive in the CDFA report and found 34 additional pesticides and related compounds. Using the new 926-compound Database, it took 32 minutes to analyze all of the samples and DRS was able to find an additional 99 pesticides, metabolites, fire retardants, and related compounds (Table 2).
Figure 4.
DRS configuration screen for the method called Tri_Pest. When the box labeled Use Uncertain Peaks is checked, AMDIS will use uncertain peaks for library searches. When unchecked, AMDIS ignores uncertain mass spectral peaks. Sometimes, this can affect the quality of a library match.
Table 2.
Comparison of the Results Obtained by Screening 17 Surface Water Extracts Using Traditional Methods (CDFA) and Using DRS With Two Different Databases the G1049A With 567 Compounds and the G1672AA With 926 Entries Agilent DRS (Original G1049A database) Same 37 +34 more 0 20 minutes Agilent DRS (G1672 AA database) Same 37 +99 more 0 32 min
CDFA Targets found (not counting ISTD) False positives Processing time 37 1 ~8 hrs (ChemStation only)
Handling Stereoisomers Many pesticides have multiple stereoisomers with virtually identical mass spectra. For example, cyfluthrin has four diastereomers arising from its three chiral centers. It is very difficult and sometimes impossible to determine the elution order of these isomers and most analysts report them as a sum of the isomer amounts. Agilents G1049A RTL Pesticide database arbitrarily assigned each isomer a Roman numeral with I for the earliest eluting isomer, II for the next, and so on. The same Chemical Abstracts Service number (CAS #) was assigned to all of the isomers. Generally, it was a CAS # for the compound with unstated stereochemistry. This caused some incompatibility with AMDIS as explained below. AMDIS software differentiates among compounds using a chemical identification number. The easiest and most consistent approach is to use each compound's CAS #. The default setting for AMDIS is to allow each CAS # to be used only once when analyzing a GC/MS data file. While this seems logical, it requires that each database entry have a different CAS #. It is possible to allow multiple hits per compound by checking the box in AMDIS found in the drop down menu under Analyze/ Settings/Identif. However, this allows multiple peaks to be assigned the same compound name.
In the new RTL Pesticide Database (G1672AA), the Roman numeral designations remain and the first isomer in the series is given its genuine CAS #. Subsequent isomers in the series are given unique, but fictitious CAS #s generated by Agilent. The compound's real CAS # appears in braces after the compound name. For example, the cyfluthrin isomers are entered into the database as shown in Table 3.
Table 3.
RT 32.218 32.359 32.477 32.536
Method for Listing Compounds with Multiple Stereoisomers in the New G1672AA RTL Pesticide Database Compound name* CAS #** Cyfluthrin I Cyfluthrin II {CAS # 68359-37-5} Cyfluthrin III {CAS # 68359-37-5} Cyfluthrin IV {CAS # 68359-37-5} 68359-37-5 999028-03-4 999029-03-7 999030-03-4
* In a series, the earliest eluting isomer is identified with I and is assigned its legitimate CAS #. Subsequent isomers are assigned unique, but fictitious CAS #s (see footnote **). Their actual CAS # is put in braces behind the compound name. **Cyfluthrin I has been given it's genuine CAS #. Cyfluthrin II-IV have been given unique numbers that can be distinguished from actual CAS numbers because they all have six digits before the first hyphen (9 total) and all begin with the series 999.
Figure 5 shows how permethrin was identified in a spinach sample using both databases with AMDIS configured to allow one hit per compound. Using the older 567-compound database (G1049A) only one permethrin isomer was identified because its CAS # could be used only once. With the new format used in the 926-compound RTL Pesticide Database (G1672AA), both isomers of permethrin were identified. Not surprisingly, the NIST library search found no hits with the same fictitious CAS # assigned to permethrin II. So, the software printed the best match on the following line. This compound, a cyclopropanecarboxylic acid derivative, is a permethrin isomer. So long as the NIST library search is turned on in DRS, it will always print another line after reporting a compound with a fictitious CAS #. Note that these fictitious CAS #s always contain 9 digits and begin with 999. A)
Agilent ChemStation amount (ng) AMDIS match 88 RT Diff (sec.) 3.9 NIST reverse match 91 Hit number 3
RT 31.6158
Cas # 52645531
Compound name Permethrin II
B)
Agilent ChemStation amount (ng) AMDIS match 78 65 RT Diff (sec.) 2.6 3.5 95 1 NIST reverse match 81 Hit number 3
RT 31.4127 31.6088 31.6088
Cas # 52645531 999046036 51877748
Compound name Permethrin I Permethrin II {CAS # 52645-53-1} Cyclopropanecarboxylic acid, 3-(2,2-dichlorovinyl)-2,2-dimethyl-, (3-phenoxyphenyl)methyl ester, (1R-trans)-
Figure 5.
A) A single isomer of permethrin was identified by DRS using the G1049A 567-compound database when AMDIS was not allowed to use multiple hits per compound. B) Two permethrin isomers are identified by DRS with the G1672AA 926-compound database under the same circumstances.
Conclusions
The new G1672AA RTL Pesticide and Endocrine Disruptor library contains substantially more target analytes than its predecessor. With the addition of 359 new compounds, it is the most comprehensive library of its type available today. Many new pesticides, metabolites, and endocrine disruptors were added along with important PCBs, PBBs, PAHs, synthetic musk compounds, Sudan dyes, and organophosphorus fire retardants. The database contains all of the analytes specified for GC/MS analysis in the new Japanese Positive List regulations. When combined with the complete DRS solution, one can screen GC/MS data files for all 926 compounds in about two minutes per sample. This is the fastest, most comprehensive, most accurate, and least tedious method for screening food and environmental samples for these compounds.
5. C. P. Sandy, A Blind Study of Pesticide Residues in Spiked and Unspiked Fruit Extracts using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1564EN, www.agilent.com/chem 6. C. Lesueur and M. Gartner, Routine Identification and Quantification of Pesticide Multiesidues in Fruit and Vegetable Samples with Full Scan, SIM, and Deconvolution Reporting Software, 2005 Ernhrung/Nutrition, 29 (11) 466471 7. X. Ping, C.-K. Meng, and M. Szelewski, Building Agilent GC/MSD Deconvolution Reporting Libraries for any Application, Agilent Technologies, publication 5989-2249EN, www.agilent.com/chem 8. Large-scale studies of the occurrence and distribution of new contaminants in the environment Reconnaissance studies, USGS, Contaminant Occurrence Studies, http://toxics.usgs.gov/topics/reconnaissance_ studies.html
References
1. V. Giarocco, B. Quimby, and M. Klee, Retention Time Locking: Concepts and Applications, Agilent Technologies, publication 5966-2469E, www.agilent.com/chem 2. H. Prest, P. Wylie, K. Weiner, and D. Agnew, Efficient Screening for Pesticides and Endocrine Disrupters Using the 6890/5973 GC/MSD System, Agilent Technologies, publication 5968-4884E, www.agilent.com/chem 3. P. L. Wylie, M. J. Szelewski, C.-K. Meng, C. P. Sandy, Comprehensive Pesticide Screening by GC/MSD using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1157EN, www.agilent.com/chem 4. Introduction of the Positive List System for Agricultural Chemical Residues in Foods Department of Food Safety, Ministry of Health, Labour and Welfare http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/introduction.html

Acknowledgments
The author wishes to thank Dr. G. Kempe of the Landesuntersuchungsanstalt Sachsen, Institut, Chemnitz, Germany for his help in acquiring much of the data for this library update. The author also thanks Dr. Mark Lee and Mr. Steve Siegel of the California Department of Food and Agriculture for providing the surface water extract data files.
10
Appendix A
Lists of Compounds in Databases

1,2,4-Trichlorobenzene 1,2-Dibromo-3-chloropropane 1,3,5-Tribromobenzene 1,3-Dichlorbenzene 17a-Ethynylestradiol 1-naphthalenol 2-(1-naphthyl)acetamide 2-(2-Butoxyethoxy)ethyl thiocyanate 2-(Octylthio)ethanol 2,3,4,5-Tertrachloronitrobenzene 2,3,4,5-Tetrachlorophenol 2,3,4,6-Tetrachlorophenol 2,3,5,6-Tetrachlorophenol 2,3,5,6-Tetrachloro-p-terphenyl 2,3,5-Trichlorophenol 2,3,5-Trimethacarb 2,3,6-Trichloroanisole 2,3,7,8-Tetrachlorodibenzofuran 2,3,7,8-Tetrachlorodibenzo-p-dioxin 2,4,5,6-Tetrachloro-m-xylene 2,4,5-T methyl ester 2,4,5-Trichloroaniline 2,4,5-Trichlorophenol 2,4,5-Trichloro-p-terphenyl 2,4,5-Trimethylaniline 2,4,6-Tribromoanisole 2,4,6-Tribromophenol 2,4,6-Trichloroanisole 2,4,6-Trichlorophenol 2,4-D methyl ester 2,4-D sec-butyl ester 2,4-DB methyl ester 2,4'-Dichlorobenzophenone (2,4'-Dicofol decomposition product) 2,4-Dichlorophenol 2,4-Dichlorophenyl benzenesulfonate 2,4-Dimethylaniline 2,4-Dimethylphenol 2,6-Dichlorobenzamide 2,6-Dichlorobenzonitrile 2,6-Dimethylaniline 2-[3-Chlorophenoxy]propionamide 2-Chlorophenol 2-Ethyl-1,3-hexanediol 2-ethyl-6-methylaniline 2-Hydroxyestradiol 2-Methyl-4,6-dinitrophenol 2-Methylphenol 2-Nitrophenol 2-Phenoxypropionic acid 3,4,5-Trimethacarb 3,4-Dichloroaniline 3,5-Dichloroaniline 3-Aminophenol 3-Chloro-4-fluoroaniline 3-Chloro-4-methoxyaniline 3-Chloroaniline 3-Hydroxycarbofuran 3-Indolylacetonitrile 3-Trifluormethylaniline 4,4'-Dichlorobenzophenone 4,4'-Oxydianiline 4,6-Dinitro-o-cresol (DNOC) 4-Aminodiphenyl 4-Bromoaniline 4-Chloro-2-methylaniline 4-Chloro-3-methylphenol 4-Chloroaniline 4-Chlorophenyl isocyanate 4-Isopropylaniline 4-Methylphenol 4-Nitrophenol 4-Nonylphenol 5,7-Dihydroxy-4'-methoxyisoflavone 9,10-Anthraquinone Acenaphthene Acenaphthylene Acephate Acequinocyl acetamiprid Acetochlor Acifluorfen methyl ester Aclonifen Acrinathrin Alachlor Aldrin Allidochlor Ametryn Amidithion Aminocarb Amitraz Amitraz metabolite [Methanimidamide, N(2,4-dimethylphenyl)-N'-methyl-] Ancymidol Anilazine Aniline Anilofos Anthracene Aramite I Aramite II {CAS # 140-57-8} Atraton Atrazine Atrazine-desethyl Azaconazole Azamethiphos Azibenzolar-S-methyl Azinphos-ethyl Azinphos-methyl Aziprotryn metabolite [2-Amino4-isopropylamino-6-methylthio1,3,5-triazine] Aziprotryne Azobenzene Azoxybenzene Azoxystrobin Barban Beflubutamid Benalaxyl Benazolin-ethyl Bendiocarb Benfluralin 11
Benfuracarb Benfuresate Benodanil Benoxacor Bentazone Bentazone methyl derivative Benthiocarb Benzene, 1,3-bis(bromomethyl)Benzenesulfonamide Benzidine Benzo(a)anthracene Benzo(a)pyrene Benzo[b]fluoranthene Benzo[g,h,i]perylene Benzo[k]fluoranthene Benzophenone Benzoximate metabolite Benzoylprop ethyl Benzyl benzoate b-Estradiol BHC alpha isomer BHC beta isomer BHC delta isomer BHC epsilon isomer Bifenazate metabolite (5-Phenyl-o-anisidine) Bifenox Bifenthrin Binapacryl Bioallethrin Bioallethrin S-cyclopentenyl isomer Bioresmethrin Biphenyl Bis(2,3,3,3-tetrachloropropyl) ether Bis(2-butoxyethyl) phthalate Bis(2-ethylhexyl)phthalate Bisphenol A Bitertanol I Bitertanol II {CAS # 55179-31-2} Boscalid (Nicobifen) Bromacil Bromfenvinphos-(E) Bromfenvinphos-(Z) Bromobutide Bromocyclen Bromophos 12
Bromophos-ethyl Bromopropylate Bromoxynil Bromoxynil octanoic acid ester Bromuconazole I Bromuconazole II {CAS # 116255-48-2} Bufencarb Bupirimate Buprofezin Butachlor Butafenacil Butamifos Butoxycarboxim Butralin Butyl benzyl phthalate Butylate Butylated hydroxyanisole Cadusafos Cafenstrole Caffeine Captafol Captan Carbaryl Carbetamide Carbofuran Carbofuran-3-keto Carbofuran-7-phenol Carbophenothion Carbosulfan Carboxin Carfentrazone-ethyl Carpropamid Carvone Cashmeran Cekafix Celestolide Chinomethionat Chloramben methyl ester Chloranocryl Chlorbenside Chlorbenside sulfone Chlorbicyclen Chlorbromuron Chlorbufam Chlordecone Chlordene, trans-
Chlordimeform Chlorethoxyfos Chlorfenapyr Chlorfenethol Chlorfenprop-methyl Chlorfenson Chlorfenvinphos Chlorfenvinphos, cisChlorfenvinphos, transChlorflurecol-methyl ester Chlormefos Chlornitrofen Chlorobenzilate Chloroneb Chloropropylate Chlorothalonil Chlorotoluron Chlorpropham Chlorpyrifos Chlorpyrifos Methyl Chlorthal-dimethyl Chlorthiamid Chlorthion Chlorthiophos Chlorthiophos sulfone Chlorthiophos sulfoxide Chlozolinate Chrysene Cinerin I Cinerin II Cinidon-ethyl cis-Chlordane Clodinafop-propargyl Clomazone Cloquintocet-mexyl Coumaphos Crimidine Crotoxyphos Crufomate Cyanazine Cyanofenphos Cyanophos Cyclafuramid Cycloate Cyclopentadecanone Cycluron
Cyflufenamid Cyfluthrin I Cyfluthrin II {CAS # 68359-37-5} Cyfluthrin III {CAS # 68359-37-5} Cyfluthrin IV {CAS # 68359-37-5} Cyhalofop-butyl Cyhalothrin I (lambda) Cyhalothrin (Gamma) Cymiazole Cymoxanil Cypermethrin I Cypermethrin II {CAS # 52315-07-8} Cypermethrin III {CAS # 52315-07-8} Cypermethrin IV {CAS # 52315-07-8} Cyphenothrin cisCyphenothrin trans- {CAS # 39515-40-7} Cyprazine Cyproconazole Cyprodinil Cyprofuram Cyromazine d-(cis-trans)-Phenothrin-I d-(cis-trans)-Phenothrin-II {CAS # 260002-80-2} Dazomet DDMU [1-Chloro-2,2-bis(4'-chlorophenyl)] Decachlorobiphenyl Deltamethrin Demephion Demeton-S Demeton-S-methylsulfon Desbromo-bromobutide Desmedipham Desmetryn Dialifos Di-allate I Di-allate II {CAS # 2303-16-4} Diamyl phthalate Diazinon Diazinon-oxon Dibenz[a,h]anthracene Dicamba Dicamba methyl ester Dicapthon Dichlofenthion Dichlofluanid
Dichlofluanid metabolite (DMSA) Dichlone Dichlormid Dichlorophen Dichlorprop Dichlorprop methyl ester Dichlorvos Diclobutrazol Diclocymet I Diclocymet II {CAS # 139920-32-4} Diclofop methyl Dicloran Dicrotophos Dicyclohexyl phthalate Dicyclopentadiene Dieldrin Diethatyl ethyl Diethofencarb Diethyl dithiobis(thionoformate) (EXD) Diethyl phthalate Diethylene glycol Diethylstilbestrol Difenoconazol I Difenoconazol II {CAS # 119446-68-3} Difenoxuron Diflufenican Diisobutyl phthalate Dimefox Dimepiperate Dimethachlor Dimethametryn Dimethenamid Dimethipin Dimethoate Dimethomorph-(E) Dimethomorph-(Z) {CAS # 110488-70-5} Dimethylphthalate Dimethylvinphos(z) Dimetilan Dimoxystrobin Di-n-butylphthalate Di-n-hexyl phthalate Diniconazole Dinitramine Di-n-nonyl phthalate Dinobuton
Dinocap I Dinocap II {CAS # 39300-45-3} Dinocap III {CAS # 39300-45-3} Dinocap IV {CAS # 39300-45-3} Di-n-octyl phthalate Dinoseb Dinoseb acetate Dinoseb methyl ether Dinoterb Dinoterb acetate Di-n-propyl phthalate Diofenolan I Diofenolan II {CAS # 63837-33-2} Dioxabenzofos Dioxacarb Dioxathion Diphacinone Diphenamid Diphenyl phthalate Diphenylamine Dipropetryn Dipropyl isocinchomeronate Disulfoton Disulfoton sulfone Ditalimfos Dithiopyr Diuron Diuron Metabolite [3,4-Dichlorophenyl isocyanate] Dodemorph I Dodemorph II {CAS # 1593-77-7} Drazoxolon Edifenphos Empenthrin I Empenthrin II {CAS # 54406-48-3} Empenthrin III {CAS # 54406-48-3} Empenthrin IV {CAS # 54406-48-3} Empenthrin V {CAS # 54406-48-3} Endosulfan (alpha isomer) Endosulfan (beta isomer) Endosulfan ether Endosulfan lactone Endosulfan sulfate Endrin Endrin aldehyde Endrin ketone 13
EPN Epoxiconazole EPTC Erbon Esfenvalerate Esprocarb Etaconazole Ethalfluralin Ethidimuron Ethiofencarb Ethiolate Ethion Ethofenprox Ethofumesate Ethofumesate, 2-Keto Ethoprophos Ethoxyfen-ethyl Ethoxyquin Ethylenethiourea Etoxazole Etridiazole Etridiazole, deschloro- (5-ethoxy3-dichloromethyl-1,2,4-thiadiazole) Etrimfos Eugenol Exaltolide [15-Pentadecanolide] Famoxadon Famphur Fenamidone Fenamiphos sulfoxide Fenamiphos-sulfone Fenarimol Fenazaflor Fenazaflor metabolite Fenazaquin Fenbuconazole Fenchlorazole-ethyl Fenchlorphos Fenchlorphos-oxon Fenclorim Fenfuram Fenhexamid Fenitrothion Fenitrothion-oxon Fenobucarb Fenoprop 14
Fenoprop methyl ester Fenothiocarb Fenoxanil Fenoxaprop-ethyl Fenoxycarb Fenpiclonil Fenpropathrin Fenpropidin Fenson Fensulfothion Fensulfothion-oxon Fensulfothion-oxon -sulfone fensulfothion-sulfone Fenthion Fenthion sulfoxide Fenthion-sulfone Fenuron Fenvalerate I Fenvalerate II {CAS # 51630-58-1} Fepropimorph Fipronil Fipronil, desulfinylFipronil-sulfide Fipronil-sulfone Flamprop-isopropyl Flamprop-methyl Fluacrypyrim Fluazifop-p-butyl Fluazinam Fluazolate Flubenzimine Fluchloralin Flucythrinate I Flucythrinate II {CAS # 70124-77-5} Fludioxonil Flufenacet Flumetralin Flumiclorac-pentyl Flumioxazin Fluometuron Fluoranthene Fluorene Fluorodifen Fluoroglycofen-ethyl Fluoroimide Fluotrimazole
Fluoxastrobin cisFluquinconazole Flurenol-butyl ester Flurenol-methylester Fluridone Flurochloridone I Flurochloridone II {CAS # 61213-25-0} Flurochloridone, deschloroFluroxypyr-1-methylheptyl ester Flurprimidol Flurtamone Flusilazole Fluthiacet-methyl Flutolanil Flutriafol Fluvalinate-tau-I Fluvalinate-tau-II {CAS # 102851-06-9} Folpet Fonofos Formothion Fosthiazate I Fosthiazate II {CAS # 98886-44-3} Fuberidazole Furalaxyl Furathiocarb Furilazole Furmecyclox Halfenprox Haloxyfop-methyl Heptachlor Heptachlor epoxide isomer A Heptachlor exo-epoxide isomer B Heptenophos Hexabromobenzene Hexachlorobenzene Hexachlorophene Hexaconazole Hexazinone Hexestrol Hydroprene Imazalil Imazamethabenz-methyl I Imazamethabenz-methyl II {CAS # 81405-85-8} Imibenconazole Imibenconazole-desbenzyl
Indeno[1,2,3-cd]pyrene Indoxacarb and Dioxacarb decomposition product [Phenol, 2-(1,3-dioxolan-2-yl)-] Ioxynil Ioxynil octanoate Ipconazole Iprobenfos Iprodione Iprovalicarb I Iprovalicarb II {CAS # 140923-25-7} Irgarol Isazophos Isobenzan Isobornyl thiocyanoacetate Isocarbamide Isocarbophos Isodrin Isofenphos Isofenphos-oxon Isomethiozin Isoprocarb Isopropalin Isoprothiolane Isoproturon Isoxaben Isoxadifen-ethyl Isoxaflutole Isoxathion Jasmolin I Jasmolin II Jodfenphos Kinoprene Kresoxim-methyl Lactofen Lenacil Leptophos Leptophos oxon Lindane Linuron Malathion Malathion-o-analog MCPA methyl ester MCPA-butoxyethyl ester MCPB methyl ester m-Cresol Mecarbam
Mecoprop methyl ester Mefenacet Mefenpyr-diethyl Mefluidide Menazon Mepanipyrim Mephosfolan Mepronil Metalaxyl Metamitron Metasystox thiol Metazachlor Metconazole I Metconazole II {CAS # 125116-23-6} Methabenzthiazuron [decomposition product] Methacrifos Methamidophos Methfuroxam Methidathion Methiocarb Methiocarb sulfone Methiocarb sulfoxide Methomyl Methoprene I Methoprene II {CAS # 40596-69-8} Methoprotryne Methoxychlor Methoxychlor olefin Methyl (2-naphthoxy)acetate Methyl paraoxon Methyl parathion Methyl-1-naphthalene acetate Methyldymron Metobromuron Metolachlor Metolcarb Metominostrobin (E) Metominostrobin (Z) {CAS # 133408-50-1} Metrafenone Metribuzin Mevinphos Mirex Molinate Monalide
Monocrotophos Monolinuron Musk amberette Musk Ketone Musk Moskene Musk Tibetene (Moschustibeten) Musk xylene Myclobutanil N,N-Diethyl-m-toluamide N-1-Naphthylacetamide Naled Naphthalene Naphthalic anhydride Naproanilide Napropamide Nicotine Nitralin Nitrapyrin Nitrofen Nitrothal-isopropyl N-Methyl-N-1-naphthyl acetamide Nonachlor, cisNonachlor, transNorflurazon Norflurazon, desmethylNuarimol o,p'-DDD o,p'-DDE o,p'-DDT Octachlorostyrene o-Dianisidine o-Dichlorobenzene Ofurace Omethoate o-Phenylphenol Orbencarb ortho-Aminoazotoluene Oryzalin Oxabetrinil Oxadiazon Oxadixyl Oxamyl Oxycarboxin Oxychlordane Oxydemeton-methyl Oxyfluorfen 15
p,p'-DDD p,p'-DDE p,p'-DDM [bis(4-chlorophenyl)methane] p,p'-DDT p,p'-Dibromobenzophenone p,p'-Dicofol Paclobutrazol Paraoxon Parathion PBB 52 Tetrabrombiphenyl PBB 101 PBB 15 PBB 169 Hexabrombiphenyl PCB 101 PCB 105 PCB 110 PCB 118 PCB 126 PCB 127 PCB 131 PCB 136 PCB 138 PCB 153 PCB 169 PCB 170 PCB 180 PCB 30 PCB 31 PCB 49 PCB 77 PCB 81 p-Dichlorobenzene Pebulate Penconazole Pendimethalin Pentachloroaniline Pentachloroanisole Pentachlorobenzene Pentachloronitrobenzene Pentachlorophenol Pentanochlor Permethrin I Permethrin II {CAS # 52645-53-1} Perthane Phantolide Phenamiphos 16
Phenanthrene Phenanthrene-d10 Phenkapton Phenol Phenothiazine Phenothrin I Phenothrin II Phenoxyacetic acid Phenthoate Phorate Phorate sulfone Phorate sulfoxide Phorate-oxon Phosalone Phosfolan Phosmet Phosphamidon I Phosphamidon II {CAS # 13171-21-6} Phthalide Phthalimide Picloram methyl ester Picolinafen Picoxystrobin Pindone Piperalin Piperonyl butoxide Piperophos Pirimicarb Pirimiphos-ethyl Pirimiphos-methyl Plifenat p-Nitrotoluene Potasan Prallethrin, cisPrallethrin, trans- {CAS # 23031-36-9} Pretilachlor Probenazole Prochloraz Procymidone Prodiamine Profenofos Profenofos metabolite (4-Bromo2-chlorophenol) Profluralin Prohydrojasmon I Prohydrojasmon II {CAS # 158474-72-7}
Promecarb Promecarb artifact [5-isopropyl3-methylphenol] Prometon Prometryn Propachlor Propamocarb Propanil Propaphos Propargite Propargite metabolite [Cyclohexanol, 2-(4-tert-butylphenoxy)] Propazine Propetamphos Propham Propiconazole-I Propiconazole-II {CAS # 60207-90-1} Propisochlor Propoxur Propyzamide Prosulfocarb Prothioconazole-desthio Prothiofos Prothoate Pyracarbolid Pyraclofos Pyraflufen-ethyl Pyrazon Pyrazophos Pyrazoxyfen Pyrene Pyrethrin I Pyrethrin II Pyributicarb Pyridaben Pyridaphenthion Pyridate Pyridinitril Pyrifenox I Pyrifenox II {CAS # 88283-41-4} Pyriftalid Pyrimethanil Pyrimidifen Pyriminobac-methyl (E) Pyriminobac-methyl (Z) {CAS # 136191-64-5}
Pyriproxyfen Pyroquilon Quinalphos Quinoclamine Quinoxyfen Quintozene metabolite (pentachlorophenyl methyl sulfide) Quizalofop-ethyl Rabenzazole Resmethrin Resmethrine I Resmethrine II {CAS # 10453-86-8} Rotenone S,S,S-Tributylphosphorotrithioate Schradan Sebuthylazine Sebuthylazine-desethyl Secbumeton Silafluofen Silthiopham Simazine Simeconazole Simetryn Spirodiclofen Spiromesifen Spiroxamine I Spiroxamine II {CAS # 118134-30-8} Spiroxamine metabolite (4-tert-butylcyclohexanone) Sudan I Sudan II Sudan Red Sulfallate Sulfanilamide Sulfentrazone Sulfotep Sulfur (S8) Sulprofos Swep Tamoxifen TCMTB Tebuconazole Tebufenpyrad Tebupirimifos Tebutam Tebuthiuron
Tecnazene Tefluthrin, cisTemephos Terbacil Terbucarb Terbufos Terbufos-oxon-sulfone Terbufos-sulfone Terbumeton Terbuthylazine Terbuthylazine-desethyl Terbutryne Tetrachlorvinphos Tetraconazole Tetradifon Tetraethylpyrophosphate (TEPP) Tetrahydrophthalimide, cis-1,2,3,6Tetramethrin I Tetramethrin II {CAS # 7696-12-0} Tetrapropyl thiodiphosphate Tetrasul Thenylchlor Theobromine Thiabendazole Thiazopyr Thifluzamide Thiofanox Thiometon Thionazin Thymol Tiocarbazil I Tiocarbazil II {CAS # 36756-79-3} Tolclofos-methyl Tolfenpyrad Tolylfluanid Tolylfluanid metabolite (DMST) Tolyltriazole [1H-Benzotriazole, 4-methyl-] Tolyltriazole [1H-Benzotriazole, 5-methyl-] Tonalide Toxaphene Parlar 26 Toxaphene Parlar 50 Toxaphene Parlar 62 trans-Chlordane Transfluthrin Traseolide Triadimefon
Triadimenol Tri-allate Triamiphos Triapenthenol Triazamate Triazophos Tributyl phosphate Tributyl phosphorotrithioite Trichlamide Trichlorfon Trichloronate Triclopyr methyl ester Triclosan Triclosan-methyl Tricresylphosphate, metaTricresylphosphate, orthoTricresylphosphate, para Tricyclazole Tridemorph, 4-tridecylTridiphane Trietazine Triethylphosphate Trifenmorph Trifloxystrobin Triflumizole Trifluralin Triphenyl phosphate Tris(2-butoxyethyl) phosphate Tris(2-chloroethyl) phosphate Tris(2-ethylhexyl) posphate Triticonazole Tryclopyrbutoxyethyl Tycor (SMY 1500) Uniconizole-P Vamidothion Vernolate Vinclozolin XMC (3,4-Dimethylphenyl N-methylcarbama XMC (3,5-Dimethylphenyl N-methylcarbama Zoxamide Zoxamide decomposition product
17
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2006 Printed in the USA April 18, 2006 5989-5076EN
Screening for Hazardous Chemicals in Homeland Security and Environmental Samples Using a GC/MS/ECD/FPD with a 731 Compound DRS Database Application Note
Homeland Security, Environmental
Authors
Bruce Quimby Mike Szelewski Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 UASA
interpretation of the MS data, especially in samples with high matrix contamination. The combination of selective GC detectors, SIM/Scan, and deconvolution makes a very powerful hazardous chemical analysis system that shows significant progress toward the above goals.
Introduction
In recent years, there has been increasing concern over the release of hazardous chemicals through either accidental or intentional acts. Both the homeland security and environmental communities recognize the need for preparing analytical laboratories that can respond quickly to such incidents. The terms toxic industrial chemicals/toxic industrial materials (TIC/TIM) are used in homeland security to refer to hazardous chemicals, while the environmental community uses different terminology like hazardous materials. In either case, the challenge is to develop laboratory methods with the capability of identifying any hazardous chemical(s) involved in an incident and to be able to measure its concentration in collected samples. There are several significant challenges to face when developing methods for this analysis. The methods must able to: Rapidly and accurately identify the specific toxic agents involved Measure concentration correctly at high levels of agent at the epicenter (high dynamic range) Measure concentration correctly at low levels of agent at perimeters and during decontamination (low detection limits)
Abstract
Response to homeland security or environmental incidents involving hazardous chemicals requires first, the rapid and accurate identification of the chemical agent(s) involved and second, the quantitative measurement of that agent in large numbers of samples to aid in managing the response. Given the unknown nature of the analytes and the complexity of matrices that could be encountered, developing analytical methods for this analysis is challenging. The approach described in this work uses a gas chromatography/mass spectrometry (GC/MS) system with a micro-fluidic splitter added to the end of the column. The splitter divides the column effluent between the MS and either a dual-wavelength flame photometric detector (DFPD) or a micro electron-capture detector (ECD) and a single-wavelength FPD. This approach allows the simultaneous collection of MS and two channels of selective GC detector data from a single injection. This multisignal configuration provides: full-scan MS data for library searching, selective ion monitoring (SIM) data for trace analysis, ECD and FPD data for excellent selectivity and sensitivity in complex matrices. The systems use retention time locking (RTL) to produce retention times (RTs) that precisely match those in a 731 compound database of hazardous chemicals. Deconvolution Reporting Software (DRS) is used to provide fast and accurate
Be highly selective over matrix interferences (wood smoke, fuels, burning tires, etc.) to minimize both false positives and false negatives Indentify as many toxic agents as possible Handle large numbers of samples It is clear that there is no single analytical technique that can be used for detecting all possible hazardous chemicals. However, one technique that is widely applicable for the identification and measurement of broad classes of hazardous chemicals is GC/MS. GC/MS is widely used in laboratories worldwide for the analysis of thousands of different chemicals. GC/MS methods are typically developed to analyze between 10 and 100 individual compounds. A target compound is deemed to be present if the target ion and two or three qualifier ions, with specific abundance ratios, fall within a defined RT window. The identity of the target may be further confirmed by comparison of the scan at the apex of the peak with a library reference spectrum. Matrix interferences are usually minimized by optimizing a combination of the sample preparation, GC, and MS parameters. Since most methods only deal with at most a few matrix types, the ions chosen for identification purposes can be selected such that they are minimized in the matrix. With the limited number of targets addressed by the method, recalibration of response factors, RTs, and qualifier ion abundance ratios can be accomplished with the injection of a few calibration mixtures. General screening methods for very large numbers of targets in widely varying and complex matrices offer a new set of challenges for the method developer. When screening for hundreds of targets, several factors must be addressed: Use of sample preparation to reduce matrix interferences is now significantly limited because rigorous cleanup steps may unintentionally remove targets. This reduced level of cleanup can result in significantly higher levels of matrix interferences to contend with. Recalibration of response factors, RTs, and qualifier abundance ratios is difficult or impossible because of the large number of targets. The methods may be deployed in laboratories without access to standards for all of the targets.
The time required for data review of hundreds of targets in complex matrices can become unmanageably large. Even with a very large database of targets, it is possible that hazardous chemicals not in the target list could be present in a sample. Recently, several techniques have become available to help address the above set of challenges. RTL produces RTs that precisely match from instrument-to-instrument and to those in a database [1]. This eliminates the need for recalibration of the individual RTs and timed events. The introduction of reliable and inert microfluidic splitters allows for the simultaneous collection of mass spectral data and, for example, phosphorus, sulfur, and/or electron capture data [2]. The selective detector chromatograms can highlight suspect compounds even if they are not in the MS target list. They can also offer an alternative means for quantitation of target analytes. The introduction of the synchronous SIM/Scan feature allows for the simultaneous acquisition of both full scan and SIM data from the same injection [2, 3]. The scan data can be used for screening the full list of targets in the database while the SIM data looks for a high priority subset of compounds down to very low levels. One of the most significant tools developed for dealing with complex matrices is Agilents Deconvolution Reporting Software (DRS) [4]. It uses advanced computational techniques to extract the spectra of targets from those of overlapped interference peaks. It then compares the extracted spectrum with a library to determine if the target is present. Any hits are confirmed by searching against the main NIST MS reference library. This process is automated and provides significant time savings in data interpretation. Since it deals with the entire spectrum instead of just four ions, DRS can often correctly identify a target in the presence of interferences where the typical approach would fail. The use of DRS substantially reduces the number of both false positives and false negatives. This application note describes the combination of the above techniques with a database of 731 hazardous chemicals, the Agilent Hazardous Chemical DBL (HCD), to be used for screening purposes. The compounds were chosen because of their significance in environmental or food safety analysis. The reasoning is that if the materials are manufactured in significant quantities and are toxic, they would be likely to appear in an
environmental method. The pesticides are included because many exhibit toxicity. The list is comprised of: Chlorinated Dioxins and Furans: EPA 8280A, 10 compounds Polychlorinated biphenyls: EPA 8082, 19 compounds Volatiles: EPA 502/524, 60 compounds Semivolatiles: EPA 8270C Appendix IX, 140 compounds Pesticides: Agilent RTL Pesticide Database (adapted), 567 compounds Total: 796 compounds, with 65 compounds in two groups, or 731 individual compounds The names of all the compounds in the database are listed in Appendix A at the end of this note. The above list by no means contains all of the hazardous chemicals that could be encountered. However, it does screen for a large number of known hazards and with the addition of selective detection can highlight other nontarget compounds that may be of interest.
and where the fuel components are not of interest. In the examples shown below the database with hydrocarbons removed was used, since fuels were used as prototype matrices.
System Configuration
The system configurations used are shown in Figure 1A and 1B.
A
Auto-sampler Phosphorus FPD
3-Way effluent splitter with makeup
AUX EPC 3.8 psig
ECD
Column 6890N GC
5975 Inert MSD
The chromatographic conditions chosen for development of the database are general in nature and are compatible with the analysis of other types of compounds beyond those in the table. For example, laboratories with access to calibration standards for chemical warfare agents (CWA) can add CWA data to the tables and screen for them as well. The RTs for compounds in the database were collected with the column outlet pressure at 3.8 psig using a microfluidic splitter. This was done to assure that the RTs observed during sample analysis would closely match those in the database when a microfluidic splitter or QuickSwap is used. The chromatographic conditions for the database were chosen to be compatible with the method translation technique. Constant pressure mode was used in the GC inlet so that method translation can be used to precisely time scale the methods for faster operation [5]. Provided with the Agilent Hazardous Chemicals DBL are the files to run the analysis precisely threefold (3X) and sevenfold (7X) faster than the primary database (1X). Also, each of the three-speed variations of the database are provided in two forms: one with the entire set of 731 compounds and one with the 36 aromatic hydrocarbons removed. The latter is provided for use with samples known to contain fuels
Auto-sampler Dual Flame Photometric Detector Sulfur Phosphorus 3-Way effluent splitter with makeup
AUX EPC 3.8 psig
Column 6890N GC
5975 Inert MSD + Performance electronics
Figure 1.
System configurations. A). GC/MS/ECD/FPD system used for 1X and 3X screening analyses. B). GC/MS/DFPD system used for 7X screening analyses.
Key components are: Fast Oven The primary 1X method only requires the 120V oven. With the 6890N 240V oven (option 002), the screening analysis method can be run precisely three times faster (14.33 min) using a 15-m HP-5MS column. If the 240V GC is further equipped with SP1 2310-0236 (puts MSD interface in back of oven under rear injection port) and
3
using the G2646-60500 oven-insert accessory, the speed can be increased to seven times faster (6.14 min) with a 5-m HP-5MS column. Note that use of the oven insert prevents use of the front inlet and detector positions. Only one detector is available for splitting. The DFPD is a good choice for this configuration, as it uses only one detector position but generates two signals. ECD The 6890N Option 231 is a ECD. The signal from the electron capture detector (ECD) is collected, stored, and processed by the MS ChemStation simultaneously with the MS data. ECDs are selective in nature and exhibit very sensitive response to halogenated compounds, with detection limits below 1 pg for polyhalogenates. They also respond to several other functional groups like nitro compounds. They do, however, also respond to some fairly low-priority compounds, like phthalate esters. The ECD data can be used in several ways. Nontarget halogenated or nitro compounds are highlighted. The presence of an electrophore at the RT of an identified compound can be used to support confirmation of identity. The response on the ECD can be used for quantitative analysis, but only after calibration with a standard, as the response factors are compound dependent and can vary significantly with compound class. Single FPD The 6890N Option 240 is a single FPD. It is used to selectively detect either sulfur or phosphorus. The detector is usually run in the phosphorus mode to highlight such compounds as organophosphorus pesticides and nerve agents. In the phosphorus mode, the detector is highly selective (>106) with a very low (~0.050 pg) detection limits for phosphorus. The ability of the FPD to uncover nontarget organophosphorus compounds like new pesticides or designer nerve agents is especially helpful. The presence of phosphorus at the RT of an identified compound can be used to support confirmation of identity. Because the response per unit weight of phosphorus is relatively consistent from compound to compound, the FPD can be used for semi-quantitative analysis in situations where no calibration standard is available for an identified analyte. Dual FPD The 6890N Option 241 is a DFPD with two optical detection channels that measures sulfur and phosphorus simultaneously. The DFPD sulfur response is also selective (>104) and sensitive (detection limits <10 pg) , although not as much as phosphorus.
4
The sulfur signal is also quadratic with respect to the amount of sulfur injected. It is often used to detect sulfur-mustard agents and for confirmation of sulfur-containing pesticides. The response per unit weight of sulfur is relatively consistent from compound to compound, but varies more than that of the phosphorus signal. Microfluidic Splitter The 6890N Option 890 (3-way splitter) or Option 889 (2-way splitter) uses diffusion-bonded plate technology combined with metal column ferrules to make an inert, easy-to-use, leak free, hightemperature column-effluent splitter. The splitter uses Auxiliary EPC for constant pressure makeup (6890N Option 301). The Auxiliary EPC makeup can be pressure programmed at the end of the run to higher pressure, while at the same time the inlet pressure is lowered to near ambient. This causes the flow in the column to reverse direction, backflushing heavy materials out the split vent of the inlet. Backflushing can greatly reduce analysis times for samples that contain high-boiling matrix components [6]. The Aux EPC also allows column changing and maintainance without venting the MSD. When the column fitting is removed from the splitter, helium from the makeup supply purges the fitting, preventing air from entering the MSD. If the column is attached to the splitter but removed from the inlet, helium flows backwards through the column and out the inlet end. Inlet maintainance or column headtrimming can be done without cooling and venting the MSD to prevent sucking air into a hot source. MSD System The 5975 inert MSD with performance turbo (G3243A) or 5973N inert MSD with performance electronics and performance turbo (G2579A), EI (electron impact ionization mode) MSD is used. These configurations provide faster full scan rates while maintaining sensitivity. The scan rates are compatible with the narrower peaks generated by fast chromatography. The performance turbo pump is required to handle the higher flows associated with the screening method. Synchronous SIM/Scan The D.02.00 (or higher) revision of the Agilent MSD ChemStation is used because it supplies the synchronous SIM/Scan feature. SIM/Scan operates by collecting SIM data every other cycle and scan data on alternate cycles throughout the entire chromatogram. The signal-to-noise performance of the collected SIM and scan data is virtually identical to that obtained with SIM-only and scan-only
methods. As with conventional SIM methods, not all 731 targets can be monitored in a single run due to the required time separation between SIM groups. In general, the acquisition of SIM data is set up to collect high-priority targets at very low levels. Examples would be the chlorinated dioxins and CWAs. DRS Software (G1716AA) Spectral deconvolution of the MS data enables identification of analytes in the presence of overlapped matrix peaks [4]. This significantly reduces chromatographic resolution requirements, which allows detection of targets in higher levels of matrix or can be used with fast chromatography to shorten analysis times. DRS uses the AMDIS deconvolution program from NIST, originally developed for trace chemical-weapons detection in complex samples. DRS presents the analyst with three distinct levels of compound identification: ChemStation, based on RT and four-ion agreement AMDIS, based on cleaned spectra full-ion matching and locked RT NIST05 search using a 163000 compound library Hazardous Chemical DBL (G1671AA) This supplies the mass spectral library, method, and DRS files for the 731 compound-screening method.
volatility from gases to large polynuclear aromatic hydrocarbons (PAHs). Splitless injections are usually incompatible with the lowest boiling volatiles due to problems with the solvent. For low matrix samples where semivolatiles are of interest, splitless injections can be used. For ambient headspace analysis [7], the conditions are listed separately at the bottom of Table 1. The liner used for ambient headspace was 1-mm id straight through (no glass wool) and Siltek coated (Restek, part number 20973-214.5). The auto injector parameters are critical in ambient headspace and are listed in Table 1. The volatiles samples run by ambient headspace were prepared as described in Reference 7. While the targets in the table cover a very broad range of boiling points, it is usually not practical to screen for all of them in one run. This is because an analysis for semivolatile compounds would be done with a solvent that would occlude the lowest boiling volatiles in the table. Conversely, a method for injecting the lowest boiling compounds would usually not be suitable for the highest boiling. The MSD solvent delays listed in Table 1 are based on isooctane as the solvent in a semivolatiles analysis. If a lower boiling solvent is used, it may be possible to reduce these delays accordingly. Some of the target compounds were found to have sufficiently high boiling points to require higher inlet and detector temperatures. These were the higher molecular weight PAHs, the polychlorinated dioxins, and the polychlorinated furans. For these compounds the inlet temperature, MS source, and transfer line were also raised to 300 C. Without this increase in temperature, the compounds would exhibit tailing and in some cases reduction in signal. The trade-off with temperature is that the performance of some thermally labile compounds is degraded at the higher temperatures. The MSD data acquisition sampling rates listed in Table 1 are for scan mode only. For volatiles analysis, the scan rate is increased one step. It is also increased one step when SIM/Scan is used. In SIM/Scan mode the SIM dwell time was set to 40 milliseconds for each ion monitored. The microfluidic splitter parameters are chosen to provide the desired flow ratio between detectors while meeting the flow requirements of the detectors used. A primary consideration is to make sure that the flow to the MSD does not exceed ~4 mL/min while collecting analyte data. It was also desired to split the effluent equally between the DFPD and MSD in the 2-way split configuration. In the 3-way configuration, the split to the ECD was reduced
5
Instrument Operating Parameters

The instrument operating parameters used (unless noted otherwise) are listed in Table 1. These are starting conditions and may have to be optimized. The split/splitless injection port was used for all work described here. It was chosen for its flexibility, allowing splitless injections for clean samples and split injections for dirty or high-concentration samples. It is also compatible with column backflushing. For all cases (except ambient headspace), the inlet liner used was the 4-mm id Siltek Cyclosplitter (Restek, part number 20706-214.1). This inlet liner was found to be of low activity, as it does not contain glass wool. Proper mixing for split injections is done by the internal liner geometry. Except as noted, split injections with a split ratio of 10:1 were used. For high matrix samples, this roughly matches the amount of matrix injected with the column capacity. If excess amounts of matrix are injected, the RTs of targets can shift. Split injection is also the easiest and most reliable way of screening samples for analytes ranging in
Table 1.
Gas Chromatograph and Mass Spectrometer Conditions Original 1X Method 3X Method 7X Method
GC Agilent Technologies 6890N 7683 Autoinjector and Tray Inlet Mode Injection type Injection volume (uL) Inlet temp ( C) Pressure, nominal (psig) RT Locking compound RT Locking time (min) Split ratio Gas saver Gas type Oven Voltage (VAC) Initial oven temp (C) Initial oven hold (min) Ramp rate (C/min) Final temp (C) Final hold (min) Total run time (min) Equilibration time (min) Column Type Agilent part number Length (m) Diameter (mm) Film thickness (um) Outlet pressure (AUX EPC, psig) FPD or DFPD Type Temperature (C) Hydrogen flow (mL/min) Air flow (mL/min) Mode: Constant makeup flow Nitrogen makeup flow (mL/min) Data rate (Hz) ECD Temperature (C) Nitrogen makeup flow (mL/min) Mode: Constant makeup flow Data rate (Hz) AUX EPC Pressure Pressure (psig) Gas type EPC Split/Splitless Constant pressure Split 1.0 250 31.17 Tripropyl phosphate 12.874 10:1 Off Helium EPC Split/Splitless Constant pressure Split 1.0 250 23.96 Tripropyl phosphate 4.291 10:1 Off Helium EPC Split/Splitless Constant pressure Split 1.0 250 8.84 Tripropyl phosphate 1.839 10:1 Off Helium
120 or 240 40 2 10 300 15 43.00 0.5
240 40 0.667 30 300 5 14.33 0.5
240 (and pillow) 40 0.286 70 300 2.143 6.14 0.5
HP 5-MS inert 19091S-433i 30 0.25 0.25 3.8
HP 5-MS 19091S-431 15 0.25 0.25 3.8
HP 5-MS Custom 5 0.25 0.25 3.8
Single, Phosphorus 250 75 100 60 5
Dual, S and P 250 75 100 60 10
30 0 60 5
300 60 10
N/A N/A N/A
3.8 Helium
3.8 Helium
3.8 Helium
Table 1.
Gas Chromatograph and Mass Spectrometer Conditions (Continued)
MSD Agilent Technologies Tune file Mode Solvent delay (min) EM voltage Low mass (amu) High mass (amu) Threshold Sampling Scans/s Quad temp (C) Source temp (C) Transfer line temp (C) Splitter Type 6890N option number Flow ratio [Deactivated fused silica tubing] MSD restrictor length (m) MSD restrictor id (mm) FPD/DFPD restrictor length (m) FPD/DFPD restrictor id (mm) ECD restrictor length (m) ECD restrictor id (mm) Ambient Headspace Inlet Mode Injection type Inlet temp ( C) Pressure, nominal (psig) RT locking compound RT locking time (min) Split ratio Gas saver Gas type Autoinjector Sample washes Sample pumps Injection volume (L) Syringe size (L) PreInj Solvent A washes PreInj Solvent B washes PostInj Solvent A washes PostInj Solvent B washes Viscosity delay (s) Plunger speed Pre-injection dwell (min) Post-injection dwell (min) Sampling depth (mm) [critical!]
5975 inert MSD Atune.U Scan 2.20 Atune voltage 35 565 0 1 5.23 150 230 280
5973 inert with Performance Electronics Atune.U Scan 0.40 Atune voltage 35 565 0 0 9.46 150 230 280
3 way 890 1:1:0.1 MSD:FPD:ECD 1.44 0.18 0.53 0.18 0.51 0.10
3 way 890 1:1:0.1 MSD:FPD:ECD 1.44 0.18 0.53 0.18 0.51 0.10
2 way 889 1:1 MSD:DFPD 1.44 0.18 0.53 0.18 N/A N/A
EPC Split/Splitless Constant pressure Split 200 31.17 Tripropyl phosphate 12.874 1:1 Off Helium
0 3 50 100 0 0 1 3 5 Fast 0 0 20
to 1/10th that going to the MSD and FPD because of the extreme sensitivity of the detector. The lengths and diameters of the detector restrictors were calculated using the spreadsheet calculator included with the splitter. The peak recognition windows used in the Agilent ChemStation were set to 0.2 min and in AMDIS to 12 s. these values were found to be sufficiently wide enough to compensate for some RT drift yet narrow enough to minimize the number of false positives. The minimum match factors setting in AMDIS was set to 45. This value seemed to give the least number of false positives and false negatives.
relatively non-polar volatiles in water. It is convenient for labs that need to screen samples for volatiles but do not have a dedicated headspace sampler. The conversion from liquid sampling to ambient headspace simply requires changing the inlet liner and the autosampler syringe. Figure 2 shows the chromatograms from a run using the system in Figure 1A. A mixture of 14 halogenated volatiles was spiked into water at 2 ppm. Fifty microliters of the approximately 1 mL of headspace in the vial was injected. With the exception of peaks 3 and 4, which coelute, the compounds are well separated. The ECD chromatogram is inverted for comparison with the MS total ion chromatogram from the full-scan data. All of the volatiles respond on the ECD, although the response to compounds 1, 2, and 8 is significantly lower than for the rest of the compounds. In general with an ECD the response to a compound increases dramatically with the number of halogens in the molecule. Since none of the compounds contain phosphorus, there is no response on the FPD. Figure 3 shows the DRS report for the sample. For each compound identified, the RT, Chemical Abstracts number (CAS#), and compound name are listed. A line is generated in the report if a compound is found by either the Agilent ChemStation, AMDIS, or both.
Results
Volatiles To evaluate the HCD method for volatiles analysis, headspace injection was chosen. Headspace injections are usually done with an automated heated sampler specifically designed for the purpose. Ambient headspace [7] is a variant of the technique that uses a gastight syringe in the liquid autosampler and injects the headspace from a 2-mL vial. It is unheated, and is thus limited to compounds that are volatile at room temperature. Ambient headspace works well for the analysis of
3,4 7 2 1 5 6 8 10 9 11 12
TIC: 2ppmMIX 3_Only_simscan.D\DATA.MS
14 13
TIC
ECD
FPD
2 4 6 8 10 12
Peak identities 1) 1,2-Dichloroethane 2) 1,1-Dichloropropylene 3) 1,2-Dichloropropane 4) Trichloroethylene
5) 6) 7) 8) 9)
cis-1,3-Dichloropropylene trans-1,3-Dichloropropylene 1,1,2-Trichloroethane 1,3-Dichloropropane 1,2-Dibromoethane
10) 11) 12) 13) 14)
1,1,1,2-Tetrachloroethane 1,1,2,2-Tetrachloroethane 1,2,3-Trichloropropane 3-Chloro-1,2-dibromopropane Hexachlorobutadiene
Figure 2. 8
Ambient headspace analysis of volatile organics in water, spiked at 2 ppm per component.
Figure 3.
DRS report for the analysis in Figure 2.
The report shows that a compound has been determined as present by the Agilent ChemStation if a value appears in the Agilent ChemStation Amount column. This means the identification criteria set in the DATA ANALYSIS section of the method have been met. Typically the criteria are that the target ion is present and all three qualifier ions are present in ratios that fall within the percent uncertainty values for that compound. The Agilent ChemStation Amount listed is a very rough approximation of the amount of the compound, in nanograms, reaching the MS. This is based on the response factor originally observed when the HCD table data was collected. Since valid quantitation requires recent recalibration of response factors on the specific instrument used for analysis, the numbers in this column should never be used to report concentrations of identified analytes. The error in these values can easiliy be a factor of 10 or higher. The purpose of the listed values is to give an approximate amount that can be used to guide standard preparation for quantitative calibration of the compound, if needed. The match value listed under the AMDIS column is the degree to which the extracted (deconvolved) spectrum of the peak at that RT matched the spectrum in the HCD AMDIS target library. The higher this number, the better the spectra agree. The
column R.T. Diff sec. lists the difference in seconds between the observed RT and that in the AMDIS target library. The lower this number, the better the RTs agree. The NIST column lists the reverse-match quality of the extracted spectrum compared with the NIST05 main library spectrum with the same CAS#. The entry Hit Num. is the number of the hit in the NIST search results that has the same CAS# as the identified compound. The higher the reverse-match value and the lower the hit number, the better the extracted spectrum matches with NIST05. The NIST column serves as a second opinion on the identity of the extracted spectrum. The analysis in Figure 2 is of course an easy one, but serves to demonstrate how the system works. All 14 spiked compounds were found by both the Agilent ChemStation and AMDIS. The certainty of identification is very high because: The target ion and three qualifier ions are present in appropriate ratios and at the appropriate time as determined by the Agilent ChemStation The deconvolved spectrum and the RT at which it appears closely matches the data in the AMDIS target library.
9
The extracted spectrum of the identified compound also matches the spectrum with the same CAS # in the NIST05 library. The compounds all have a significant response on the ECD, as expected from their halogen content. To challenge the system in a more realistic way, the effect of matrix and dilution of the analytes was studied. Additional samples were prepared that contained: the same 2-ppm mixture of analytes plus 100 ppm of pump gasoline; 100 ppb of analytes only; and 100 ppb of analytes plus 100 ppm of pump gasoline. Figure 4 shows the chromatograms from the 100 ppb of analytes with 100 ppm of gasoline. The complexity of the TIC chromatogram illustrates the severe matrix challenge presented by the thousand-fold excess of gasoline. In the ECD chromatogram, interference peaks are now apparent. However, with the exception of peaks 1, 2, 8, and 12, all of the analytes peaks are still visible above the matrix interferences.
Table 2 summarizes the results from the matrix and dilution experiments. In the sample that was 2 ppm of analytes with 100 ppm of gasoline, the Agilent ChemStation (column labeled Quant) found all but two of the compounds. Those two compounds had qualifier ions out of range due to interferences from the matrix. AMDIS successfully found all 14 compounds. Also, with the exception of compound 8, all of the analytes were clearly visible above the matrix responses on the ECD chromatogram. In the sample that contained 100 ppb of analytes but without gasoline, quant found 7 of the 14 analytes. Using full-scan data, the signal to noise ratio for most of the analytes at the 100-ppb level is very low. This results in difficulties with finding the qualifier ions in ratios that fall within the specified uncertainty range in the quant calibration table. AMDIS found 11 of the 14 compounds. Peak 3 was not found due to a severe overlap with the coeluting peak number 4. Peaks 9 and 13 were missed by AMDIS because the signal to noise ratio was too low.
TIC
1 2 5 6
8 7 11
12
ECD
9 10 13 14
3,4
FPD
2 4 6 8 10 12
5) 6) 7) 8) 9)
10) 11) 12) 13) 14)
Figure 4.
Ambient headspace analysis of volatile organics in water. Analytes at 100 ppb plus pump gasoline at 100 ppm.
10
Table 2.
Effect of Matrix and Concentration on DRS Results 2 ppm STD only 2 ppm STD with 100 ppm gasoline Quant AMDIS (ng) (match) 2.47 7.34 5.59 7.71 4.81 3.05 3.50 5.32 2.41 1.85 2.40 3.54 12 93 98 64 97 98 84 96 97 95 99 98 98 90 89 14 0.65 7 0.07 0.12 0.37 0.21 0.40 0.21 100 ppb STD only Quant (ng) AMDIS (match) 73 89 Overlap 90 88 53 72 66 S/N 89 48 79 S/N 75 10 0.36 7 0.31 0.19 0.14 0.22 0.30 0.23 100 ppb STD with 100 ppm gasoline Quant AMDIS (ng) (match) 65 85 Overlap 82 74 Overlap Overlap 46 66 88 53 75 59 52 11
RT (min) 1.491 1.536 1.793 1.863 2.317 2.658 2.735 2.938 3.250 4.003 5.151 5.283 8.208
Compound 1,2-Dichloroethane 1,1-Dichloropropylene 1,2-Dichloropropane Trichloroethylene cis-1,3-Dichloropropylene trans-1,3-Dichloropropylene 1,1,2-Trichloroethane 1,3-Dichloropropane 1,2-Dibromoethane 1,1,1,2-Tetrachloroethane 1,1,2,2-Tetrachloroethane 1,2,3-Trichloropropane 3-Chloro-1,2-dibromopropane Total Found
Peak Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Quant (ng) 2.27 7.60 4.92 7.58 4.39 3.30 2.82 3.39 2.60 5.15 2.38 1.89 1.62 16.46 14
AMDIS (match) 97 100 95 99 98 97 99 98 91 100 99 98 93 94 14
10.435 Hexachlorobutadiene
With 100 ppm of gasoline added to the 100-ppb sample, quant again found 7 of the 14 compounds and AMDIS again found 11 of the 14. Curiously, in both cases some of the compounds missed in the absence of matrix were now found. It is possible that the presence of matrix enhances the concentration of some of the analytes in the headspace. The compounds missed in quant were again the result of low signal to noise and/or interference. In AMDIS the three missed peaks were due to severe interferences from the gasoline. As indicated above, the ECD response from 10 of the 14 compounds was still visible above the peaks due to interferences. SIM/Scan The quant data in Table 2 was generated using full scan mode. Peak 13 was missed in quant due to low signal to noise ratio. SIM/Scan mode can be
used to collect SIM data simultaneously with the scan data. The 100 ppb plus 100-ppm gasoline sample was run in SIM/Scan mode with SIM groups for each of the 14 analytes. Figure 5 compares the target and qualifier extracted ion chromatograms in both modes with the ECD response for peak 13. The signal-to-noise (peak to peak) for the target ion increases from 34 in full scan mode to 433 in SIM mode. The peaks lost in quant due to low signal-to-noise were all recovered in SIM mode. This example demonstrates the power of SIM/Scan when looking for high-priority targets at low levels. If necessary, the ECD could also be used for quantitation, as it has a high signal to noise ratio and is free from interference.
11
Ion 157 Scan Ion 75 Scan Ion 155 Scan
s/n (pk-pk) = 34
Ion 39 Scan s/n (pk-pk) = 433
Ion 157 SIM Ion 75 SIM Ion 155 SIM
Ion 39 SIM
ECD
s/n (pk-pk) = 122
8.0
8.1
8.2
8.3
8.4
Figure 5.
Target and qualifier extracted ion chromatograms for peak 13 (3-Chloro-1,2-dibromopropane) in Figure 4. SIM, scan, and ECD data collected simultaneously.
AMDIS Figure 6 illustrates the ability of AMDIS to clean the interference ions from the spectrum of an analyte. The raw spectrum at the top of Figure 6 was taken at the apex of peak 13 in the 100 ppb plus 100-ppm gasoline sample. When searched against the NIST05 library using the NIST search program, the actual compound (3-Chloro-1,2-dibromopropane) was the 70th hit in the search results. Using manual subtraction of nearby spectra in the Agilent ChemStation data analysis program improved the quality of the spectrum so that it was now the second hit when searched in NIST. This is a tedious process, however, when dealing with a large number of analytes. The spectrum as deconvolved by AMDIS is shown in Figure 6 above the
NIST05 library spectrum. When this spectrum is searched, it is the first hit in the results. The automated deconvolution provided by AMDIS saves an enormous amount of time in the data review process. Fast Methods When a retention time locked database is constructed, the RTs are (or at least should be) collected under the highest resolution conditions expected for the application. If the database is collected under constant pressure mode, method translation can then be used to adjust the speed of the method to meet the needs of different situations.
12
100
155 44 75 55 81 97 117 132 188
50
Raw spectrum
211 227 246 261 280 300
0 157
100 75 38 0 30 60 90 120 58 95 115 132
50
Manual subtraction
188 150 157 180 211 210 230 246 240 261 270 282 300 300
100 75 50 39 49 0 49 50 100 30 50 39 75 70 90 110 130 61 85 93 93 105 136 119 129
Deconvolved spectrum
188 187 199
NIST 05 library
157 150 170 190
Figure 6.
Comparison of raw, manually subtracted, AMDIS deconvoluted, and NIST05 reference spectra for peak 13 (3-Chloro-1,2-dibromopropane) in Figure 4.
The 3X method uses RTs in its database that are simply the RTs from the 1X method divided by exactly 3. The 7X method likewise uses RTs that are 1/7 of those in the original database. The quality of RTs matching between the two new faster methods and the new divided databases is demonstrated in Figure 7. Three different mixtures containing 13 chlorinated hydrocarbons and 36 pesticides were run with the two methods. The RTs were compared to those in the two new databases. The graph at the top of Figure 7 plots the database RT on the x-axis versus the difference of the measured RT from the database on the y axis. If the RT matching were perfect, the plot would be a straight horizontal line at zero height on the y axis. The maximum deviation from the table values for the 3X method was 0.047 min. The plot
indicates that a peak recognition window of 0.1 min should be sufficient. The maximum deviation in the 7X plot at the bottom of Figure 8 is +0.032 min indicating that the same peak recognition window could be used here as well. In general the RTs in scaled methods agree very well with the predicted RTs. The conditions for the two higher-speed methods were chosen to increase speed while maintaining the same column capacity. The capacity is important for both the dynamic range of quantitative measurements and for minimizing analyte RT shifts in samples with high levels of matrix. In gas chromatography, the well-known triangle of speed, resolution, and capacity dictates that if the capacity is to be maintained and the speed is to be increased, then the resolution will decrease.
13
0.100
3X Measured vs. Table
0.050 Difference (min)
0.000
_0.050
_0.100 0.0 2.0 4.0 6.0 Table RT (min) 8.0 10.0 12.0
0.100
0.000 _0.050 _0.100 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Table RT (min)
Figure 7.
Difference between scaled HCD table and experimental retention times for 50 compound test set. Y axis is table value minus experimental, X axis is table RT. Top plot is 3X, bottom is 7X.
Figure 8 shows three sets of chromatograms using the HCD database at three different speeds. The sample consists of nine organophosphorus pesticides (identified in the caption to Figure 8) at 50 ppm and a matrix consisting of an equal volume mixture of gasoline, kerosene, and diesel fuel spiked at 50,000 ppm total mixture. The 1X and 3X data were collected on the three-way splitter instrument and the 7X was collected on the DFPD instrument. All nine compounds also contained sulfur as can be seen in the DFPD sulfur chromatogram at the bottom of Figure 8. Note that the sulfur tails somewhat compared to the phosphorus.
14
TIC: OP_50K_GKD1.D\DATA.MS
TIC
1 2 3 4 5 6 7 8 9
1X
P (FPD) ECD
12
16
20
24
28
TIC
3X
P (FPD) ECD
1 2 3 4 5 6 7 8 9
TIC: 50_OP_50K_GKD.D\DATA.MS
TIC
7X
P (FPD) S (FPD)
1 2 3 4
Peak identities
1) 2) 3) 4)
O,O,O-triethyl phosphorothioate Thionazin Sulfotepp Phorate
5) 6) 7) 8) 9)
Dimethoate Disulfoton Methyl parathion Parathion Famphur
Figure 8.
Comparison of 1X, 3X, and 7X chromatograms. 1X and 3X were run on GC/MS/ECD/FPD system, 7X on GC/MS/DFPD.
15
Figure 9 expands the RT region of the phosphorous chromatogram containing peaks 3, 4, and 5 from Figure 8. The decrease in resolution with increasing speed is clearly evident. If only the standard target and three qualifier ion approach is used, the loss in resolution causes a significant problems. With the 1X method, all nine of the analytes are identified and eight false positives are reported. With the 3X method, all analytes are again found but now with 25 false positives. With the significantly decreased resolution of the 7X method, only seven of the nine analytes are identified and 48 false positives are reported.
3
The situation is much different when using the approach described here. Even in the worst situation, the 7X method, AMDIS finds all nine analytes with high-quality matches and only three false positives. The DRS report for the 7X analysis is shown in Table 3. To simplify the table, the 48 false positives that only appear in the quant column are not shown. The analyte compounds are shown in bold. All show close RT and high-quality spectral matches to both the AMDIS target library and to the NIST05 library.
1X
16.50
17.75
3X
5.50
5.916
7X
2.365
2.544
Peak identities
3) Sulfotepp 4) Phorate 5) Dimethoate Figure 9. Comparison of FPD phosphorus chromatograms from 1X, 3X, and 7X runs in Figure 8.
16
Table 3.
DRS Report for 7x Analysis of 50 ppm Pesticides In 50,000 ppm Gasoline/Kerosine/Diesel Matrix Agilent ChemStation amount (ng) 13.92 AMDIS match 71 69 46 64 55 91 88 90 84 92 92 91 93 RT Diff (sec.) 9.5 0.7 7.6 0.6 2.3 0.5 0.5 0.6 0.7 0.6 0.6 0.7 0.8 NIST reverse match 74 71 74 80 85 85 83 85 85 88 82 85 85 Hit number 50 1 1 3 1 1 1 1 1 1 1 1 1
RT 0.973 1.380 1.520 1.520 2.113 2.138 2.138 2.275 2.417 2.427 2.485 2.619 2.748 2.901 3.360
Cas # 98862 126681 94597 52417502 132649 90437 2131411 297972 3689245 298022 60515 298044 298000 56382 52857
Compound name Acetophenone O,O,O-triethyl phosphorothioate Safrole Benzeneacetaldehyde, ,2,5-trimethylDibenzofuran o-Phenylphenol Naphthalene, 1,4,5-trimethylThionazin Sulfotepp Phorate Dimethoate Disulfoton Methyl parathion Parathion (ethyl) Famphur (48 quant-only hits not shown)
0.35
89.2 23.31 27.34 22.7 25.12
The peak at 0.973 minutes is a reasonable spectral match to acetophenone, but the large time difference and being the 50th hit in the NIST search results suggests that this is not the compound. The peak at 1.520 min is a poor spectral match with a large time difference. The absence of a NIST reverse search and hit entry means that the listed compound was not in the top 100 hits in the NIST search. The next compound listed at 1.520 min is the top entry from the NIST search. It is quite clear that safrole is not present. The peak at 2.113 min, dibenzofuran, was not one of the analytes added to the sample. However, it probably is present in the diesel fuel matrix. Its presence is supported by both reasonably good spectral matches and close time matching with a database. The last extraneous peak at 2.138 min is also questionable. The time match is somewhat poor and the NIST reverse search suggests the identification is not correct.
All nine analytes are detected with the FPD on both the phosphorus and sulfur chromatograms. All analytes except peak 1 are detected selectively on the ECD as well. These results suggest that while the loss of resolution in going to 7X is unacceptable when using only conventional screening approaches, with the method discussed here, it is a viable option. By using the DRS report combined with the selective detector data, the number of false positives and false negatives are significantly reduced. For those situations where speed is a critical factor, for example in response to homeland security incidents, the fastest method may be the one of choice. For many laboratories, the 3X method would be an attractive choice. It has higher resolution than the 7X and higher speed than the 1X and still allows the use of two GC detectors in parallel with the MSD. It also only requires a 240V oven, not the repositioning of the MSD to the back position.
17
Conclusions
The systems described here offer several advantages when screening samples for the presence of hazardous chemicals. The advantages derive from a combination of techniques that result in both faster and more accurate screening results. Retention time locked target database of 731 hazardous chemicals for screening with MS Microfluidic splitter - using selective detection simultaneous with MS data for added confirmation, finding non-target suspect compounds, and alternate quantitation SIM/Scan - Acquire SIM data on high priority targets simultaneously with scan data. Saves time by eliminating need to run samples in both modes. DRS - automated deconvolution dramatically increases accuracy of target identification, even in the most challenging matrices. The reduction of data interpretation from hours to minutes is especially useful for response to hazardous chemical incidents. Fast chromatography using shorter columns, faster ovens, and backflushing to greatly reduce run times. This combination of techniques offers a viable solution to the hazardous chemicals challenge.
3. Chin-Kai Meng, Improving Productivity with Synchronous SIM/Scan, Agilent Technologies, publication 5989-3108EN www.agilent.com/chem 4. Philip Wylie, Michael Szelewski, Chin-Kai Meng, Christopher Sandy, Comprehensive Pesticide Screening by GC/MSD using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1157EN www.agilent.com/chem 5. B. D. Quimby, L. M. Blumberg, M. S. Klee, and P. L. Wylie, Precise Time-Scaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking, Agilent Technologies, publication 5967-5820E www.agilent.com/chem 6. Michael J. Szelewski and Bruce Quimby, New Tools for Rapid Pesticide Analysis in High Matrix Samples, Agilent Technologies, publication 5989-1716EN www.agilent.com/chem 7. Michael J. Szelewski and Bruce D. Quimby, Ambient Headspace GC and GC-MSD Analysis of Nonpolar Volatiles in Water, Agilent Technologies, publication 5968-9455E www.agilent.com/chem

References
1. Retention time locking (RTL), Vince Giarrocco, Bruce Quimby, and Matthew Klee, Retention Time Locking: Concepts and Applications, Agilent Technologies, publication 5966-2469E www.agilent.com/chem 2. Chin Kai-Meng and Bruce Quimby, Identifying Pesticides with Full Scan, SIM, ECD, and FPD from a Single Injection, Agilent Technologies, publication 5989-3299EN www.agilent.com/chem
18
Appendix A

Volatiles: EPA 502/524, 60 compounds
1,1,1,2-Tetrachloroethane 1,1,1-Trichloroethane 1,1,2,2-Tetrachloroethane 1,1,2-Trichloroethane 1,1-Dichloroethane 1,1-Dichloroethylene 1,1-Dichloropropylene 1,2,3-trichlorobenzene 1,2,3-Trichloropropane 1,2,4-trichlorobenzene 1,2,4-trimethylbenzene 1,2-Dibromoethane 1,2-dichlorobenzene 1,2-Dichloroethane 1,2-Dichloropropane 1,3,5-trimethylbenzene 1,3-dichlorobenzene 1,3-Dichloropropane 1,4-dichlorobenzene 2,2-Dichloropropane 2-chlorotoluene 3-Chloro-1,2-dibromopropane 4-chlorotoluene Benzene Bromobenzene Bromochloromethane Bromodichloromethane Bromoform Bromomethane Carbon Tetrachloride Chlorobenzene Chlorodibromomethane Chloroethane Chloroform Chloromethane cis-1,2-Dichloroethylene cis-1,3-Dichloropropylene Dibromomethane Dichlorodifluoromethane Ethylbenzene Hexachlorobutadiene Isopropylbenzene Methylene Chloride m-xylene Naphthalene n-butylbenzene n-propylbenzene o-Xylene p-isopropyltoluene p-xylene Styrene tert-butylbenzene Tetrachloroethylene Toluene trans-1,2-Dichloroethylene trans-1,3-Dichloropropylene Trichloroethylene Trichlorofluoromethane Vinyl chloride 4,4'-DDD 4,4'-DDE 4,4'-DDT 4-aminobiphenyl 4-bromophenyl phenyl ether 4-chloro-3-methylphenol 4-chloroaniline 4-chlorophenyl phenyl ether 4-nitroaniline 4-nitrophenol 4-nitroquinoline-1-oxide 5-nitro-o-toluidine 7,12-dimethylbenz[a]anthracene a,a-dimethylphenethylamine Acenaphthene Acenaphthylene Acetone Acetophenone Aldrin Alpha-BHC (alpha-HCH) Aniline Anthracene Aramite (total) Benz[a]anthracene Benzene Benzo[a]pyrene Benzo[b]fluoranthene Benzo[ghi]perylene Benzo[k]fluoranthene Benzyl alcohol Beta-BHC (beta-HCH) Bis(2-chloroethoxy)methane Bis(2-chloroethyl) ether Bis(2-chloroisopropyl) ether Bis(2-ethylhexyl)phthalate Butyl benzyl phthalate Chlorobenzilate Chrysene Delta-BHC (delta-HCH) Diallate (total) Dibenz[a,h]anthracene Dibenzofuran Dieldrin Diethyl phthalate Dimethoate Dimethyl phthalate Di-n-butyl phthalate
Semivolatiles: EPA 8270C Appendix IX, 140 compounds

1,2,4,5-tetrachlorobenzene 1,2,4-trichlorobenzene 1,2-dichlorobenzene 1,3,5-trinitrobenzene 1,3-dichlorobenzene 1,4-dichlorobenzene 1,4-naphthoquinone 1-naphthylamine 2,3,4,6-tetrachlorophenol 2,4,5-trichlorophenol 2,4,6-trichlorophenol 2,4-dichlorophenol 2,4-dimethylphenol 2,4-dinitrophenol 2,4-dinitrotoluene 2,6-dichlorophenol 2,6-dinitrotoluene 2-acetylaminofluorene 2-chloronaphthalene 2-chlorophenol 2-methyl-4,6-dinitrophenol 2-methylnaphthalene 2-naphthylamine 2-nitroaniline 2-nitrophenol 2-picoline 3,3'-dichlorobenzidine 3,3'-dimethylbenzidine 3-methylcholanthrene 3-nitroaniline
19
Di-n-octyl phthalate Dinoseb Diphenylamine Disulfoton Endosulfan I Endosulfan II Endosulfan sulfate Endrin Endrin aldehyde Ethyl methanesulfonate Famphur Fluoranthene Fluorene Gamma-BHC (lindane) Heptachlor Heptachlor epoxide -isomer B Hexachlorobenzene Hexachlorobutadiene Hexachlorocyclopentadiene Hexachloroethane Hexachlorophene Hexachloropropene Indeno[1,2,3-cd]pyrene Isodrin Isophorone Isosafrole Kepone m-cresol (3-methylphenol) m-dinitrobenzene Methapyrilene Methoxychlor Methyl methanesulfonate Methyl parathion Naphthalene Nitrobenzene N-nitrosodiethylamine N-nitrosodimethylamine N-nitrosodi-n-butylamine N-nitrosodi-n-propylamine N-nitrosodiphenylamine N-nitrosomethylethylamine N-nitrosomorpholine (4-nitrosomorpholine) N-nitrosopiperidine (1-nitrosopiperidine) N-nitrosopyrrolidine (1-nitrosopyrrolidine) O,O,O-triethyl phosphorothioate o-cresol (2-methylphenol) o-toluidine p-(dimethylamino)azobenzene Parathion (ethyl) p-cresol (4-methylphenol) Pentachlorobenzene Pentachloroethane Pentachloronitrobenzene Pentachlorophenol Phenacetin 20
Phenanthrene Phenol Phorate p-phenylenediamine Pronamide Pyrene Pyridine Safrole Sulfotepp Thionazin
Chlorinated Dioxins and Furans: EPA 8282, 19 compounds

2,3,7,8-Tetrachlorodibenzo-p-dioxin 1,2,3,7,8-Pentachlorodibenzo-p-dioxin 1,2,3,4,7,8-Hexachlorodibenzo-p-dioxin 1,2,3,4,6,7,8-Heptachlorodibenzo-p-dioxin Octachlorodibenzo-p-dioxin 2,3,7,8-Tetrachlorodibenzofuran 1,2,3,7,8-Pentachlorodibenzofuran 1,2,3,4,7,8-Hexachlorodibenzofuran 1,2,3,4,6,7,8-Heptachlorodibenzofuran Octachlorodibenzofuran
Polychlorinatedbiphenyls: EPA 8082, 19 compounds

2-chlorobiphenyl 2,3-dichlorobiphenyl 2,2',5-trichlorobiphenyl 2,4',5-trichlorobiphenyl 2,2',5,5'-tetrachlorobiphenyl 2,2',3,5'-tetrachlorobiphenyl 2,3',4,4'-tetrachlorobiphenyl 2,2',4,5,5'-pentachlorobiphenyl 2,2',3,4,5'-pentachlorobiphenyl 2,3,3',4',6-pentachlorobiphenyl 2,2',3,5,5',6-hexachlorobiphenyl 2,2',4,4',5,5'-hexachlorobiphenyl 2,2',3,4,5,5'-hexachlorobiphenyl 2,2',3,4,4',5'-hexachlorobiphenyl 2,2',3,4',5,5',6-heptachlorobiphenyl 2,2',3,4,4',5',6-heptachlorobiphenyl 2,2',3,4,4',5,5'-heptachlorobiphenyl 2,2',3,3',4,4',5-heptachlorobiphenyl 2,2',3,3',4,4',5,5',6-nonachlorobiphenyl
Pesticides: Agilent RTL pesticide database (adapted), 567 compounds

1,2,4-Trichlorobenzene 1,2-Dibromo-3-chloropropane 17a-Ethynylestradiol 2-(1-naphthyl)acetamide 2-(2-Butoxyethoxy)ethyl thiocyanate 2-(Octylthio)ethanol
2,3,4,5-Tetrachlorophenol 2,3,4,6-Tetrachlorophenol 2,3,5,6-Tetrachlorophenol 2,3,5-Trichlorophenol 2,3,5-Trimethacarb 2,3,5-Trimethylphenyl methyl carbamate (Trimethacarb) 2,3,7,8-Tetrachlorodibenzofuran 2,3,7,8-Tetrachlorodibenzo-p-dioxin 2,4,5-T methyl ester 2,4,5-Trichlorophenol 2,4,6-Trichlorophenol 2,4-D methyl ester 2,4-D sec-butyl ester 2,4-DB methyl ester 2,4-Dichlorophenol 2,4-Dichlorophenyl benzenesulfonate 2,4-Dimethylaniline 2,6-Dichlorobenzonitrile 2,6-Dimethylaniline 2-[3-Chlorophenoxy]propionamide 2-Ethyl-1,3-hexanediol 2-Hydroxyestradiol 2-Methylphenol 2-Phenoxypropionic acid 3,4,5-Trimethacarb 3,4-Dichloroaniline 3,5-Dichloroaniline 3-Chloroaniline 3-Hydroxycarbofuran 4,4'-Dichlorobenzophenone 4,6-Dinitro-o-cresol (DNOC) 4-Chloroaniline 4-Methylphenol 5,7-Dihydroxy-4'-methoxyisoflavone 9,10-Anthraquinone Acephate Acetochlor Acifluorfen methyl ester Alachlor Aldrin Allidochlor Ametryn Amidithion Aminocarb Amitraz Ancymidol Anilazine Aniline Atraton Atrazine Azaconazole Azamethiphos Azinphos-ethyl Azinphos-methyl Aziprotryne Azobenzene Barban
Benalaxyl Benazolin-ethyl Bendiocarb Benfluralin Benfuresate Benodanil Bentazone Bentazone methyl derivative Benthiocarb Benzo(a)pyrene Benzophenone Benzoylprop ethyl b-Estradiol BHC alpha isomer BHC beta isomer BHC delta isomer Bifenox Bifenthrin Binapacryl Bioallethrin Bioallethrin S-cyclopentenyl isomer Bioresmethrin Biphenyl Bis(2-ethylhexyl)phthalate Bisphenol A Bitertanol I Bitertanol II Bromacil Bromobutide Bromocyclen Bromophos Bromophos-ethyl Bromopropylate Bromoxynil Bromoxynil octanoic acid ester Buprofezin Butachlor Butamifos Butoxycarboxim Butralin Butyl benzyl phthalate Butylate Butylated hydroxyanisole Captafol Captan Carbaryl Carbetamide Carbofuran Carbofuran-3-keto Carbophenothion Carbosulfan Carboxin Chinomethionat Chloramben methyl ester Chloranocryl Chlorbenside Chlorbromuron
Chlorbufam Chlordecone Chlordimeform Chlorfenethol Chlorfenprop-methyl Chlorfenson Chlorfenvinphos Chlorflurecol-methyl ester Chlormefos Chlornitrofen Chlorobenzilate Chloroneb Chloropropylate Chlorothalonil Chlorotoluron Chlorpropham Chlorpyrifos Chlorpyrifos Methyl Chlorthal-dimethyl Chlorthiamid Chlorthion Chlorthiophos Chlorthiophos sulfone Chlorthiophos sulfoxide Chlozolinate cis-Chlordane Clomazone Coumaphos Crimidine Crotoxyphos Crufomate Cyanazine Cyanofenphos Cyanophos Cycloate Cycluron Cyfluthrin I Cyfluthrin II Cyfluthrin III Cyfluthrin IV Cyhalothrin I (lambda) Cymoxanil Cypermethrin I Cypermethrin II Cypermethrin III Cypermethrin IV Cyprazine Cyprofuram Cyromazine d-(cis-trans)-Phenothrin-I d-(cis-trans)-Phenothrin-II Dazomet Decachlorobiphenyl Deltamethrin Demephion
Demeton-S Demeton-S-methylsulfon Desbromo-bromobutide Desmedipham Desmetryn Dialifos Di-allate I Di-allate II Diamyl phthalate Diazinon Dibrom (naled) Dicamba Dicamba methyl ester Dicapthon Dichlofenthion Dichlofluanid Dichlone Dichlormid Dichlorophen Dichlorprop Dichlorprop methyl ester Dichlorvos Diclobutrazol Diclofop methyl Dicloran Dicrotophos Dicyclohexyl phthalate Dicyclopentadiene Dieldrin Diethatyl ethyl Diethofencarb Diethyl dithiobis(thionoformate) (EXD) Diethyl phthalate Diethylene glycol Diethylstilbestrol Difenoconazol I Difenoconazol II Diflufenican Dimefox Dimethachlor Dimethametryn Dimethipin Dimethoate Dimethylphthalate Dimethylvinphos(z) Dimetilan Di-n-butylphthalate Diniconazole Dinitramine Dinobuton Dinocap I Dinocap II Dinocap III Dinocap IV
21
Dinoseb Dinoseb acetate Dinoseb methyl ether Dinoterb Dinoterb acetate Di-n-propyl phthalate Dioxacarb Dioxathion Dioxydemeton-S-methyl Diphacinone Diphenamid Diphenylamine Dipropetryn Disulfoton Ditalimfos Dithiopyr Diuron Dodemorph I Dodemorph II Drazoxolon Edifenphos Endosulfan (alpha isomer) Endosulfan (beta isomer) Endosulfan ether Endosulfan lactone Endosulfan sulfate Endrin Endrin aldehyde Endrin ketone EPN Epoxiconazole EPTC Erbon Esfenvalerate Esprocarb Etaconazole Ethalfluralin Ethiofencarb Ethiolate Ethion Ethofumesate Ethoprophos Ethoxyquin Ethylenethiourea Etridiazole Etrimfos Famphur Fenarimol Fenazaflor Fenbuconazole Fenchlorphos Fenfuram Fenitrothion Fenobucarb Fenoprop Fenoprop methyl ester Fenoxycarb 22
Fenpropathrin Fenson Fensulfothion Fenthion Fenthion sulfoxide Fenuron Fenvalerate I Fenvalerate II Fepropimorph Flamprop-isopropyl Flamprop-methyl Fluazifop-p-butyl Flubenzimine Fluchloralin Flucythrinate I Flucythrinate II Flumetralin Fluometuron Fluorodifen Fluotrimazole Flurenol-butyl ester Flurenol-methylester Fluridone Flurochloridone I Flurochloridone II Fluroxypyr-1-methylheptyl ester Flusilazole Flutolanil Flutriafol Fluvalinate-tau-I Fluvalinate-tau-II Folpet Fonofos Formothion Fuberidazole Furalaxyl Furathiocarb Furmecyclox Heptachlor Heptachlor epoxide Heptachlor exo-epoxide isomer B Heptenophos Hexabromobenzene Hexachlorobenzene Hexachlorophene Hexaconazole Hexazinone Hexestrol Imazalil Ioxynil Iprobenfos Iprodione Isazophos Isobenzan Isobornyl thiocyanoacetate Isocarbamide Isodrin
Isofenphos Isomethiozin Isoprocarb Isopropalin Isoprothiolane Isoproturon Isoxaben Isoxathion Jodfenphos Kinoprene Lenacil Leptophos Leptophos oxon Lindane Linuron Malathion Malathion-o-analog MCPA methyl ester MCPB methyl ester m-Cresol Mecarbam Mecoprop methyl ester Mefenacet Mefluidide Menazon Mephosfolan Mepronil Metalaxyl Metamitron Metasystox thiol Metazachlor Methacrifos Methamidophos Methfuroxam Methidathion Methiocarb Methiocarb sulfone Methiocarb sulfoxide Methomyl Methoprene I Methoprene II Methoprotryne Methoxychlor Methyl paraoxon Methyl parathion Methyl-1-naphthalene acetate Methyldymron Metobromuron Metolachlor Metolcarb Metribuzin Mevinphos Mirex Molinate Monalide Monocrotophos Monolinuron
Myclobutanil N,N-Diethyl-m-toluamide N-1-Naphthylacetamide Naphthalic anhydride Napropamide Nicotine Nitralin Nitrapyrin Nitrofen Nitrothal-isopropyl N-Methyl-N-1-naphthyl acetamide Norflurazon Nuarimol o,p'-DDD o,p'-DDE o,p'-DDT Octachlorostyrene o-Dichlorobenzene Omethoate o-Phenylphenol Oryzalin Oxabetrinil Oxadiazon Oxadixyl Oxamyl Oxycarboxin Oxychlordane Oxydemeton-methyl Oxyfluorfen p,p'-DDD p,p'-DDE p,p'-DDT Paclobutrazol Paraoxon Parathion p-Dichlorobenzene Pebulate Penconazole Pendimethalin Pentachloroaniline Pentachloroanisole Pentachlorobenzene Pentachloronitrobenzene Pentachlorophenol Pentanochlor Permethrin I Permethrin II Perthane Phenamiphos Phenkapton Phenoxyacetic acid Phenthoate Phorate Phosalone Phosfolan Phosmet Phosphamidon I
Phosphamidon II Phthalide Picloram methyl ester Pindone Piperalin Piperonyl butoxide Piperophos Pirimicarb Pirimiphos-ethyl Pirimiphos-methyl Plifenat p-Nitrotoluene Pretilachlor Probenazole Prochloraz Procymidone Profenofos Profluralin Promecarb Prometon Prometryn Propachlor Propamocarb Propanil Propargite Propazine Propetamphos Propham Propiconazole-I Propiconazole-II Propoxur Propyzamide Prothiofos Prothoate Pyracarbolid Pyrazon Pyrazophos Pyrazoxyfen Pyributicarb Pyridaben Pyridaphenthion Pyridate Pyridinitril Pyrifenox I Pyrifenox II Pyrimethanil Pyroquilon Quinalphos Quinoclamine Quizalofop-ethyl Resmethrin S,S,S-Tributylphosphorotrithioate Sebuthylazine Secbumeton Simazine Simetryn Sulfotep
Sulfur (S8) Sulprofos Swep Tamoxifen TCMTB Tebuconazole Tebutam Tecnazene Temephos Terbacil Terbucarb Terbufos Terbumeton Terbuthylazine Terbutryne Tetrachlorvinphos Tetradifon Tetraethylpyrophosphate (TEPP) Tetramethrin I Tetramethrin II Tetrapropyl thiodiphosphate Tetrasul Thenylchlor Thiabendazole Thiofanox Thiometon Thionazin Tiocarbazil I Tiocarbazil II Tolclofos-methyl Tolylfluanid trans-Chlordane Triadimefon Triadimenol Tri-allate Triamiphos Triazophos Tributyl phosphate Tributyl phosphorotrithioite Trichlorfon Trichloronate Triclopyr methyl ester Tricyclazole Tridiphane Trietazine Triflumizole Trifluralin Tryclopyrbutoxyethyl Tycor (SMY 1500) Uniconizole-P Vamidothion Vernolate Vinclozolin
23
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2006 Printed in the USA February 9, 2006 5989-4834EN
Identifying Pesticides with Full Scan, SIM, ECD, and FPD from a Single Injection Application
Food Safety, Environmental
Authors
Chin-Kai Meng and Bruce Quimby Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Abstract
In this application note, a gas chromatography/mass spectrometry (GC/MS) system capable of providing up to four signals from a single injection is described. When a three-way micro-fluidic splitter is added to the end of the column, two additional signals from GC detectors can be acquired together with the MS data from a single injection. This multi-signal configuration provides: full-scan data for library searching, selective ion monitoring (SIM) data for trace analysis, micro-electron capture detector and flame photometric detector data for excellent selectivity and sensitivity from complex matrices. A combination of element selective detectors, SIM/Scan, and deconvolution reporting software makes a very powerful pesticide analysis system. Examples for trace-level compound quantitation/confirmation or for screening are discussed.
Chromatograph (LC) depending on the nature of the compounds. For GC amenable compounds, the traditional detectors are NPD (Nitrogen Phosphorus Detector), ECD (micro-Electron Capture Detector), and FPD (Flame Photometric Detector) for their excellent sensitivity and selectivity. However, even with dual-column confirmation analysis, these GC detectors cannot be used to verify the identity of the compounds with high confidence. Full scan mass spectral data and library searching are typically used for final compound verification. However, full-scan analysis has a worse (higher) detection limit (DL) compared to selective detectors on a GC. To improve the DL, the technique selective ion monitoring (SIM) is often used. With SIM, the MS monitors only a few characteristic ions for each target compound within the retention time (RT) range that the target elutes from the column. By monitoring only a few specific ions, the signal-to-noise ratio (S/N) improves significantly. The ions monitored are time programmed in groups corresponding to the RTs of the targets. SIM analyses with closely eluting targets require precise alignment of chromatographic RTs with the time programming of SIM groups. The Retention Time Locking (RTL) technique can be applied to eliminate the need to adjust SIM group time-windows after column maintenance or replacement. In this application note, a GC/MS system capable of providing up to four signals from a single injection is described. The benefits of the multi-signal detection include: Confirmatory information Full-scan data for library search capability
Introduction
Many laboratories in the world are analyzing pesticide residue levels in both foods and the environment to protect human health. The process usually involves homogenizing the sample, extracting the pesticides, and analyzing the target compounds with a Gas Chromatograph (GC) or a Liquid
Maximum sensitivity SIM data enables trace analysis Excellent selectivity ECD and FPD detect trace-level hetero-compounds from complex matrices
Experimental
A recent technical note describes Synchronous SIM/Scan, which takes advantages of the Performance Electronics in the 5975 inert MSD to get both SIM and full-scan signals in a single run without sacrificing performance [1]. The SIM method can be easily developed automatically using the ChemStations AutoSIM tool [2]. By simply selecting a checkbox in the method, the SIM and fullscan data can be acquired together. The trade-off is giving up some cycles per second but gaining an additional signal (full-scan data or SIM data) for the whole analysis. With properly chosen acquisition parameters, for example, increasing the scan speed, the decrease of cycles per second is usually not significant and does not affect peak quantitation or the quality of results (for example, S/N).
EPC Injection MSD (SIM/Scan) Split vent Splitter ECD FPD (P)
Figure 2.
A close-up view of the micro-fluidic three-way splitter in the 6890 GC oven.
At the end of the column, effluent flow is split three ways according to the length and diameter of the capillary tubing (restrictor) used.
Figure 1.
A schematic of the multi-signal configuration. Note: the EPC flow adds to the column flow into the splitter.
The size of the micro-fluidic plate is 1.25 inches (3.2 cm) wide and 2.5 inches (6.4 cm) tall. The device was designed to eliminate the common problems of large thermal mass, excess dead volume, and leaky connection due to oven temperature cycling etc. The splitter's flow paths and connection points are laid out and etched onto a thin, stainless steel plate using photolithography and chem-milling technologies. The plate is diffusion bonded, mounted with column connectors, and surface deactivated, resulting in an integrated and compact micro-fluidic splitter. Metal ferrules are used at the connectors that are leak-free after temperature cycling and will not absorb solvents or sample matrix, improving sensitivity for trace analysis applications. Deactivated capillary tubing between the splitter and each detector was used as a flow restrictor. Aux EPC pressure and the restrictor dimensions were determined using a spreadsheet-like calculator program to achieve the proper split ratio among all detectors. The three-way splitter can easily turn into a two-way splitter when a connector is capped. Other advantages of a splitter include backflushing [3] and quick-swapping. The Aux EPC flow can be run-time programmed to a higher pressure, while at the same time the inlet pressure is lowered to near ambient. This causes the column flow to reverse direction, back-flushing the less volatile materials out of the split vent of the inlet. The Aux EPC on the splitter also allows column changing and inlet maintenance without cooling and venting the MSD. The splitters flow paths and connection points were designed in such a way
Besides the SIM/Scan data, the ChemStation software can simultaneously acquire up to two additional GC detector signals, for example, FPD (in phosphorus- or sulfur- mode) and NPD (nitrogenphosphorus detector) signals or both P- and S- signals from a dual-wavelength FPD (DFPD). See Figure 1. Figure 1 is a schematic for multi-signal detection. At the end of the column, a three-way micro-fluidic splitter was used to split the column effluent to different detectors [3]. For this study, an FPD and a ECD were installed. Notice on the figure that an Auxiliary Electronic Pneumatics Control (Aux EPC) gas channel was connected to the splitter to maintain the pressure at the end of the column so that the split ratios/flows are kept constant throughout a run. Figure 2 shows a close-up view of the microfluidic splitter installed in the GC oven.
2
that when the column fitting is removed, the helium gas from the Aux EPC purges the fitting, preventing air from entering the splitter/MSD. See Table 1 for hardware details and settings.
Table 1. GC Inlet
Gas Chromatograph, Mass Spectrometer, and Three-Way Splitter Operating Parameters Agilent Technologies 6890 EPC Split/Splitless Splitless, 1.0 L injected (7683 ALS) 280 C ~27 psi (chlorpyrifos methyl RT locked to 16.596 min) 50.0 mL/min 0.75 min 55.3 mL/min Off Helium Siltek Cyclosplitter, 4-mm id, Restek p/n 20706-214.1 C/min 25 3 8 Final (C) 70 150 200 280 0.5 min 325 C Agilent Technologies HP 5-ms, p/n 19091S-433 30.0 m 0.25 mm 0.25 m Constant pressure 2.5 mL/min Unspecified 3.8 psi (Aux EPC pressure to splitter) Sulfur mode Hydrogen flow: Oxidizer flow: Hold (min) 2.00 0.00 0.00 15
Mode Inlet temp Pressure Purge flow Purge time Total flow Gas saver Gas type Inlet liner Oven Oven ramp Initial Ramp 1 Ramp 2 Ramp 3 Total run time Equilibration time Oven max temp Column Length Diameter Film thickness Mode Nominal initial flow Outlet Outlet pressure
46.87 min (last standard elutes around 35 min)
Front detector (FPD) Phosphorus mode Hydrogen flow: Oxidizer flow: Temperature: Oxidizer gas type: Mode: Makeup flow: Makeup gas type: Lit offset: Data rate: 75.0 mL/min 100.0 mL/min 250 C Air Constant makeup flow 60.0 mL/min Nitrogen 2.00 5 Hz 50.0 mL/min 60.0 mL/min
Table 1.
Gas Chromatograph, Mass Spectrometer, and Three-Way Splitter Operating Parameters (Continued) 300 C Constant makeup flow 60.0 mL/min Nitrogen 5 Hz MSD Transfer line heater 280 C Helium 3.80 psi 0.00 min (this value will follow oven ramp) Agilent Technologies 5975 inert MSD Atune.U Scan 3.00 min Atune voltage 45 amu 555 amu 100 2 4 2.89 150 C 230 C Agilent 6890N Option 890, when installed on the GC during factory assembly 10:10:1 MSD:FPD:ECD 1.444 m 0.18-mm id Deactivated fused silica tubing 0.532 m 0.18-mm id Deactivated fused silica tubing 0.507 m 0.10-mm id Deactivated fused silica tubing 1.53 mL/min 1.53 mL/min 0.153 mL/min 1.38 mL/min
Back detector (ECD) Temperature: Mode: Makeup flow: Makeup gas type: Date rate: Thermal AUX 2 Use: Initial temp: Pressure AUX 5 Gas type: Initial pressure: Initial time: MSD Tune file Mode Solvent delay EM voltage Low mass High mass Threshold Sampling A/D Samples Scans/s Quad temp Source temp Three-way splitter Split ratio MSD restrictor FPD restrictor ECD restrictor Flow to MSD (at 280 C) Flow to FPD (at 280 C) Flow to ECD (at 280 C) Makeup flow (at 280 C)
Software Used in this Application Note GC/MSD ChemStation G1701DA Deconvolution Reporting Software (DRS) G1716AA NIST Library G1033A AMDIS (included for free with the NIST library CD)

Figure 3 shows four signals that were simultaneously acquired from a single injection of a pesticide mixture. Due to the high sensitivity of the ECD, the split ratios for the three detectors was set to MSD:FPD:ECD = 10:10:1. This split ratio distributes the sample of a 1-L splitless injection of a 1-ppm (1000 pg/L) sample to the different detectors as labeled in Figure 3.
Scan: ~ 480 pg
SIM: ~ 480 pg
ECD: ~ 48 pg
FPD (P): ~ 480 pg
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
Figure 3.
Signals acquired simultaneously from a 1-L splitless injection of 1-ppm standard. The split ratios were MSD:FPD:ECD = 10:10:1.
Figure 4 shows the signals when the pesticide standard was diluted 100-fold in a produce matrix. The total ion chromatogram (TIC) from full scan was not shown due to the lack of sensitivity. The FPD(P) and ECD were able to detect all the pesticides spiked in this extract. For trace-level target compound analysis, the SIM signal can be used for quantitation and the GC signals used for further confirmation.
1800 1700 1500 1300 1100 900 700 500 300 5.00 10.00 15.00 20.00 25.00 30.00 35.00
SIM
Produce extract 2688, spiked
700000 600000 500000 400000 300000 5.00 1240000 1200000 1160000 1120000 1080000 1040000 1000000 5.00
FPD(P)
10.00
15.00
20.00
25.00
30.00
35.00
ECD
10.00
15.00
20.00
25.00
30.00
35.00
Figure 4.
Data of a produce extract spiked at 10 ppb. FPD and ECD were able to detect the respective standards spiked into the extract.
Another application for this multi-signal system is for screening. In screening, no target list is available for the analysis; therefore, SIM acquisition or MS/MS is not possible. Figure 5 shows three signals (no SIM) from a produce extract.
800000 700000 600000 500000 400000 300000 200000 100000 5.00 50000 45000 40000 35000 30000 25000 20000 15000 10000 5.00 110000 100000 90000 80000 70000 60000 50000 40000 30000 5.00 10.00 15.00 20.00 25.00 30.00 35.00 10.00 15.00 20.00 25.00 30.00 35.00 10.00 15.00 20.00 25.00 30.00 35.00
Produce extract 13-10927
SCAN
40.00
45.00
FPD (P)
40.00 45.00
ECD
40.00
45.00
Figure 5.
Full-scan, FPD(P), and ECD data for extract 13-10927.
The Deconvolution Reporting Software (DRS) [3, 4] found several pesticides in the TIC as shown in Figure 6.
Figure 6.
Report for extract 13-10927 generated from DRS.
The possible pesticides in the sample were benzophenone, chlorpyrifos methyl, and thiabendazole. Propoxur and metamitron were not confirmed by both AMDIS and NIST; therefore, they were most likely false positives. Due to the complexity of the sample matrix and other interferences, it is sometimes difficult to get a high library match factor from peaks in the TIC, even after background subtraction. Therefore, element selective detectors would be very useful in providing the supporting information for compound confirmation. The multi-signal system was retention time locked, therefore, from the RT and the aligned peaks from the FPD(P) and the ECD responses, chlorpyrifos methyl (C7H7Cl3NO3PS) was confirmed.
It usually takes less than 3 minutes to turn off the FPD photomultiplier, swap the P-filter with the S-filter, and turn the photomultiplier back on. After the swap, adjust the detector gas flows to optimize the response in either P- or S- mode. A new injection of the same extract was made in FPD(S) mode. The FPD(S) result is shown with previously acquired signals in Figure 7. Two major peaks were seen on the FPD(S) chromatogram. From the peak RTs, they supported the presence of chlorpyrifos methyl and thiabendazole (C10H7N3S) respectively. Note that the full-scan TIC barely showed a peak for either compound, which made it impossible for traditional data analysis to identify both compounds. The FPD(S) mode is very selective, but it is not as sensitive as the FPD(P) mode. Although the ECD is very sensitive, it is not as selective as the FPD. A combination of GC detectors, SIM/Scan, and DRS makes a very powerful pesticide analysis system.
Conclusion
The Synchronous SIM/Scan provides users with library searchable full-scan spectra as well as trace level SIM data in a single analysis. When a threeway micro-fluidic splitter is added to the end of the column, two additional signals from element selective detectors can be acquired together with the MS data from a single injection. This configuration makes it very attractive for the analysis of trace-level pesticide residues in foods or environmental samples. This multi-signal configuration provides: full-scan data for library searching, SIM data for trace analysis, ECD and FPD data for excellent selectivity and sensitivity from complex matrices. In this application note, examples of ECD signal and FPD signal (P- or S- mode) were acquired together with the SIM/Scan data from a single injection for trace-level compound quantitation/confirmation, or for screening.
800000 700000 600000 500000 400000 300000 200000 100000 5.00 23000 21000 19000 17000 15000 13000 5.00 50000 45000 40000 35000 30000 25000 20000 15000 10000 5.00 10.00 15.00 20.00 25.00 30.00 10.00 15.00 20.00 25.00 30.00 10.00 15.00 20.00 25.00 30.00
Produce extract 13-10927
SCAN
35.00 40.00 45.00
Thiabendazole
FPD (S) (new injection)
Chlorpyrifos methyl
35.00
40.00
45.00
FPD (P)
35.00 40.00 45.00
Figure 7.
Full-scan, FPD(S), and FPD(P) data for extract 13-10927.
References
1. Chin-Kai Meng, Improving Productivity with Synchronous SIM/Scan, Agilent Technologies, publication 5989-3108, www.agilent.com/chem 2. Harry Prest and David W. Peterson, New Approaches to the Development of GC/MS Selected Ion Monitoring Acquisition and Quantitation Methods, Agilent Technologies, publication 5988-4188, www.agilent.com/chem 3. Michael J. Szelewski and Bruce Quimby, New Tools for Rapid Pesticide Analysis in High Matrix Samples Agilent Technologies, publication 5989-1716, www.agilent.com/chem 4. Philip L. Wylie, Michael J. Szelewski, Chin-Kai Meng, and Christopher P. Sandy, Comprehensive Pesticide Screening by GC/MSD Using Deconvolution Reporting Software Agilent Technologies, publication 5989-1157, www.agilent.com/chem

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2005 Printed in the USA July 6, 2005 5989-3299EN
A Blind Study of Pesticide Residues in Spiked and Unspiked Fruit Extracts Using Deconvolution Reporting Software Application
Food
Author
Christopher P. Sandy Agilent Technologies, Inc. Block A, CSC Eskdale Road Winnersh Triangle Estate Wokingham, Berk RG41 5DZ United Kingdom
Abstract
The Agilent Technologies mass selective detector (MSD) coupled with deconvolution reporting software (DRS) provides additional powerful data processing capabilities to the MSD ChemStation software. Reviewing full scan gas chromatography/mass spectrometry data for the confirmation of pesticide residues can be a labor-intensive and time-consuming process requiring great skill and concentration by an experienced analyst. The DRS is able to process a complex food extract total ion chromatogram in about 1 minute, whereas an experienced analyst may take more than 30 minutes to achieve the same quality result. Extensive data shown in this report supports the high confidence level that an analyst can have in results rapidly produced by the DRS.
ion ratios. It is sometimes very difficult to confirm target compounds from high matrix background because the matrix affects the ion ratios of the target compounds or complicates the spectrum with additional ions. To be certain of the results, background subtraction and manual integration are often practiced. It is, therefore, a timeconsuming process to confirm target compounds in a dirty matrix. It can take an experienced analyst 15 to 30 minutes to review/confirm one data file. Two powerful gas chromatography/mass spectrometry (GC/MS) techniques - Retention Time Locking (RTL) and deconvolution were combined to create a quantitation and screening tool that can identify 567 pesticides and endocrine disrupters from a single run in 12 minutes. The Agilent Technologies GC/MSD-DRS provides the additional functionality to the MSD ChemStation.
Experimental
DRS Overview A detailed overview of the DRS is given in an application note 5989-1157EN [1], available for download at www.agilent.com/chem. The operating principles of the DRS appear in Figure 1.
Introduction
Typical mass spectral pesticide residue analysis requires finding target ions and meeting qualifier
2.5E7 2E7 1.5E7 1E7 .5E7 5 10 15 20 25 30 35 40
Total ion chromatogram (TIC)
Targets are identified by comparison to locked retention times (RTs) and three qualifying ion ratios, quantified using target ion area versus internal standard (ISTD) calibration table Quant results
AMDIS 32 deconvolutes component spectra and searches target MS database, locked RT used as a qualifier
Deconvoluted target spectra confirmed by AMDIS searched against NIST02 MS database
Confirmed AMDIS hits
Confirmed NIST02 hits
Combined quantitative and qualitative HTML Summary report
Figure 1.
Schematic diagram summarizing the GC/MS DRS.
The quantitation capabilities of the MSD ChemStation are combined with the deconvolution power of the industry standard AMDIS program from NIST. AMDIS is able to separate spectra of interest from dirty matrix spectra present in samples analyzed for pesticides. A third level of confidence is obtained by sending the deconvoluted spectra for library searches of the NIST02 145,000 compound library. A comprehensive report is produced in about 1 minute.
Samples Six samples of fruit extracts, supplied in 90/10 iso-octane/toluene solvent were received for analysis by GC/MS. The samples were prepared by an accredited food pesticide laboratory based in Scandinavia. Three of the samples were spiked with a number of pesticides at varying concentration levels. Although the range of concentrations of the pesticides in each sample was given, neither the actual number of pesticides spiked into each control sample nor the identities were supplied. Details of the samples appear in Table 1. The other three samples were real, unspiked extracts.
Table 1. Sample number 1 2 3 4 5 6
Sample Details for Blind Study Matrix extracted Orange Lettuce Apple Grapes Orange Apple Number of pesticides 2040 2040 2040 24 24 24 Concn range (mg/Kg) 0.020.20 0.020.20 0.010.20 0.11.0 0.25.0 0.052.0 Comments Control sample - spiked Control sample - spiked Control sample - spiked Real sample Real sample Real sample
Instrumentation
The samples were analyzed by full-scan GC/MS using the analytical conditions given in Table 2. Data processing and reporting were performed using the default settings provided with the DRS.
Table 2.
RTL GC/MS Analysis Conditions for Fruit Extract Samples Agilent 6890N 30 m x 0.25 mm id x 0.25 m HP-5MS (p/n 19091S-433) Helium 1.9 mL/min at 70 C 18 psig, constant pressure mode Method RTLocked to methyl chlorpyrifos at 16.593 min PTV, septumless head 90 C (0.3 min) - 1720 C/min - 250 C 0.2 min 30 mL/min 0 psig 60 mL/min 1.0 min 50 L 15 L Empty multibaffle 70(2)-25-150(0)-3-200(0)-8-280(10)
Gas chromatograph Column Carrier gas Flow rate Head pressure Injector type Injector temperature (C), hold time (min), and ramp rate (C/min) Vent time Vent flow Vent pressure Purge flow Purge time Syringe volume Injection volume Liner Oven program: temperature (C), hold time (min), and ramp rate (C/min) MSD MS interface MS source MS quad Detection mode EM voltage
Agilent 5973 inert 280 C 230 C 150 C EI, Scan 40550 amu ATUNE value
Results
The results for the three spiked extracts appear in Table 3 - note that the details of which pesticides were added to the spiked samples were not supplied until after the results were shown to the customer. Those pesticides confirmed by the DRS, are shown lightly shaded. The analytes, shown darkly shaded, are not present in the Agilent RTL Pesticides database. Analyte entries left unshaded were not confirmed.
Table 3. MSD-DRS Results for Three Spiked Fruit Extract Samples Sample 2: Control-lettuce, spiked Added mg/kg 0.10 0.10 0.10 0.10 0.20 0.10 0.10 0.04 0.02 0.10 0.10 0.10 0.04 0.10 0.04 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.04 0.04 0.10 0.10 0.10 0.10 0.10 0.04 0.04 Pesticide Diphenylamine HCB Lindane (HCH-gamma) Diazinon Chlortalonil Vinclozolin Carbaryl Metalaxyl Pirimiphos-methyl Malathion Chlorpyrifos Cyprodinil Penconazole Captan Folpet** Procymidone Endosulfan-a pp-DDE Bupirimate Endosulfan-b Aclonifen Ethion Triazophos Endosulfan-sulfate Iprodione Bromopropylate Methoxychlor Phosalone Lambda-Cyhalothrin Permethrin Cypermethrin Fenvalerate Deltamethrin Added mg/kg 0.10 0.02 0.04 0.04 0.04 0.04 0.20 0.10 0.10 0.10 0.10 0.04 0.04 0.10 0.10 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.10 0.10 0.10 0.04 0.10 0.10 0.04 0.10 Sample 3: Control- apple, spiked Pesticide Mevinphos Trichlorfon Heptenophos Tecnazene HCH alpha HCH beta Dichloran Pyrimethanil Etrimphos Ethiofencarb Metribuzin Toclophos methyl Linuron Aldrin Diethofencarb Trichloronate Triadimenol Disulfoton sulfoxide Disulfoton sulfone Fluazinam Chlorbenzilate Oxadixyl Benalaxyl Dicofol Fenazaquin Pyrazophos Acrinathrin Bitertanol Cyfluthrin beta Alpha cypermethrin Added mg/kg 0.05 0.05 0.02 0.01 0.01 0.02 0.05 0.02 0.02 0.10 0.05 0.01 0.05 0.02 0.02 0.02 0.05 0.20 0.02 0.05 0.05 0.05 0.05 0.05 0.02 0.05 0.02 0.05 0.05 0.05
Sample 1: Control-orange, spiked Pesticide Methamidofos* Dichlorvos* Acephate* Omethoate Propachlor Chlorprofam Monocrotophos Dimethoate Quintozene Parathion-methyl Dichlofluanid Fenpropimorph Triadimefon Thiabendazole Tolylfluanid Mecarbam Methidation Vamidothion Imazalil Myclobutanil Kresoxim methyl Tebuconazole Phosmet Fenpropathrin Tetradifon Azinphos-methyl Fenarimol Azinpfos-ethyl Prochloraz Flucythrinate Esfenvalerate Azoxystrobin
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 * See Discussion item 1. ** See Discussion item 2.
The results for the three real extracts appear in Table 4. Those pesticides confirmed by the DRS are shown lightly shaded. The darkly-shaded analytes are not present in the Agilent RTL Pesticides database. Analyte entries left unshaded were not confirmed. Analytes with an associated concentration were confirmed as present by the customer using NPD/ECD. Lightly-shaded analytes without a concentration label were detected and confirmed by the DRS, but not by the customer.
Table 4. MSD-DRS Results for Three Real Fruit Extract Samples Sample 4: Grapes 0.68 mg/Kg Captan 0.21 mg/Kg Cyprodinil 0.27 mg/Kg Fludioxinil Diphenylamine Sample 5: Orange 2.5 mg/Kg Imazalil 0.25 mg/Kg Medidathion 3.0 mg/Kg Thiabendazole Sample 6: Apple 0.86 mg/Kg Diphenylamine 0.05 mg/Kg Chlorpyrifos 0.79 mg/Kg Thiabendazole Dimethoate Ethoxyquin Methyl parathion Endosulfan sulfate Propargite
2. Control - Lettuce spiked extract This control sample was spiked with 33 pesticides at levels ranging between 0.02 and 0.20 mg/kg. Twenty-nine pesticides were detected and confirmed by the DRS software, three were not reported since they are not present in the Agilent RTL Pesticide database and one was not detected. The one undetected analyte, (Folpet, marked with two asterisks in Table 3), was detected and confirmed if a higher sensitivity setting was used in the AMDIS deconvolution program. 3. Control - Apple spiked extract This control sample was spiked with 30 pesticides at levels ranging between 0.01 and 0.20 mg/kg. Twenty-two pesticides were detected and confirmed by the DRS software, six were not reported since they are not present in the Agilent RTL Pesticide database and two were not detected. Overall, of the 95 spiked analytes in the three control samples, 93% of the pesticides present in the Agilent RTL Pesticide database were detected and confirmed by full-scan library searching of the deconvoluted mass spectra. 4. Real Grape extract The customer had detected and confirmed three pesticide residues in the Grape extract sample Captan, Cyprodinil, and Fludioxinil. Of these three analytes, Captan was confirmed by the DRS and Cyprodinil and Fludioxinil are not entries in the Agilent RTL Pesticide database. However, DRS also confirmed an additional pesticide residue Diphenylamine, which was not reported by the customer. 5. Real Orange extract The customer had detected and confirmed three pesticide residues in the Orange extract sample Imazilil, Methidathion, and Thiabendazole. All three of these pesticides were confirmed by the DRS software and no other analytes were confirmed.
Discussion
1. Control - Orange spiked extract This control sample was spiked with 32 pesticides at levels ranging between 0.02 and 0.10 mg/kg. Twenty-six pesticides were detected and confirmed by the DRS software, two were not reported since they are not present in the Agilent RTL Pesticide database and four were not detected. The spiking was done to the raw matrix, not to a matrix extract. For the polar pesticides (methamidofos and acephate), the recovery was in the 20%30% range as confirmed by NPD/ECD. Therefore, that explains why these pesticides were not detected by DRS.
6. Real Apple extract The customer had detected and confirmed three pesticide residues in the Apple extract sample Diphenylamine, Chlorpyriphos, and Thiabendazole. All three of these pesticides were confirmed by the DRS. In addition, the DRS also confirmed the presence of five additional pesticide residues Dimethoate, Ethoxyquin, Methyl Parathion, Endosulfan Sulfate, and Progargite. These five pesticides had not been reported by the customer. The extensive data shown in this report, run under totally blind conditions, shows the high degree of confidence that an analyst can have in the results produced by the DRS in minutes.
Reference
1. Philip L. Wylie, Michael J. Szelewski, Chin-Kai Meng, and Christopher P. Sandy, Comprehensive Pesticide Screening by GC/MSD Using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1157EN, www.agilent.com/chem
Conclusions
The Agilent Technologies MSD-DRS provides additional powerful data processing capabilities to the MSD ChemStation software. Reviewing full scan GC/MS data for the confirmation of pesticide residues can be a labor-intensive and time consuming process requiring great skill and concentration by an experienced analyst. The DRS is able to process a complex food extract TIC in the order of 1 minute, whereas an experienced analyst may take more than 30 minutes to achieve the same quality result. The DRS software was proven to report the lowest number of false positives and false negatives in the shortest time period. In scan mode, the detection limit is not as low as in selected ion monitoring (SIM) mode; however, any prior knowledge of the target analytes (retention times or characteristic ions) is not required for the DRS.

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2004 Printed in the USA October 5, 2004 5989-1654EN
Comprehensive Pesticide Screening by GC/MSD using Deconvolution Reporting Software Application
Food Safety
Author
Philip L. Wylie, Michael J. Szelewski, and Chin-Kai Meng Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA Christopher P. Sandy Agilent Technologies, Inc. Block A, CSC Eskdale Road Winnersh Triangle Estate Wokingham, Berk RG41 5DZ United Kingdom
pesticides in food and environmental samples to track their distribution in the environment and to ensure a safe food supply. Current analytical methods target only a subset of the possible compounds. Whether for food or environmental samples, analyses are often complicated by the presence of co-extracted natural products. Food or tissue extracts can be exceedingly complex matrices that require several stages of sample cleanup prior to analysis [3]. Even then, it can be difficult to detect trace levels of contaminants in the presence of the remaining matrix. For efficiency, multiresidue methods (MRMs) must be used to analyze for most pesticides. Traditionally, these methods have relied upon gas chromatography (GC) with a constellation of element-selective detectors to locate pesticides in the midst of a variable matrix [4, 5, 6]. GC with mass spectral detection (GC/MS) has been widely used for confirmation of hits. Liquid chromatography (LC) has been used for those compounds that are not amenable to GC [2]. Today, more and more pesticide laboratories are relying upon LC with mass spectral detection (LC/MS) and GC/MS as their primary analytical tools [7, 8]. Still, most MRMs are target compound methods that look for a small subset of the possible pesticides. Any compound not on the target list is likely to be missed by these MRMs. Using the techniques of retention time locking (RTL) [9, 10, 11] and spectral deconvolution [12], a method has been developed to screen for 567 pesticides and suspected endocrine disrupters in a single GC/MS analysis. Spectral deconvolution
Introduction
According to The Pesticide Manual, more than 700 pesticides are currently approved for use around the world [1]. About 600 more were used in the past, but are either banned or no longer marketed. In spite of their discontinuance, some of these still persist in the environment where they may bioaccumulate in the flora and fauna. Many pesticides or their degradation products can be found at trace levels in food and beverages; in soil, water, and air; in aquatic and terrestrial flora and fauna; and in human blood, adipose tissue, and breast milk. The World Health Organization has classified pesticides into five groups based upon their acute toxicity to humans [2]. The categories range from Acutely Hazardous to those that are Unlikely to Present Acute Hazard in Normal Use. Certain pesticides are classified as persistent organic pollutants (POPs), carcinogens, teratogens, or endocrine disrupters. It is now common to analyze for
helps to identify pesticides even when they are buried under co-eluting matrix compounds. RTL helps to eliminate false positives and gives greater confidence in the results. Users can easily add compounds to the method if they wish.
Samples Vegetable extracts were obtained from Dr. Mark Lee and Stephen Siegel at The California Department of Food and Agriculture (CDFA; Sacramento, CA USA) and from Dr. J.G.J. Mol at TNO Nutrition and Food Research (Zeist, The Netherlands). Seventeen data files from the GC/MS analysis of surface water samples were also contributed by CDFA and were processed in this laboratory using the Deconvolution Reporting Software (DRS). GC/MS data files (locked to the Agilent Pesticide Method) for 17 crop extracts were supplied by NRM Laboratories, Berkshire, UK.
Experimental
Table 1 lists the instrumentation, software, and analytical parameters used by Agilent for pesticide analysis. Depending upon the desired injection volume, a PTV inlet or split/splitless inlet can be used.
Table 1.
Instrumentation and Conditions of Analysis Agilent 6890N Agilent 7683 Agilent PTV operated in the solvent vent mode Agilent 30 m 0.25 mm 0.25 m HP-5MS (p/n 19091S-433) Helium in the constant pressure mode Chlorpyrifos-methyl locked to 16.596 min (nominal column head pressure = 17.1 psi) 70 C (2 min), 25 C/min to 150 C (0 min), 3 C /min to 200 C (0 min), 8 C /min to 280 C (1015 min) Temp program: 40 C (0.25 min), 1600 C/min to 250 C (2 min); Vent time: 0.2 min; Vent flow: 200 mL/min; Vent pressure: 0.0 psi; Purge flow: 60.0 mL/min; Purge time: 2.00 min 15 L (using a 50-L syringe) Agilent 5973 inert 50550 amu 230, 150, and 280 C, respectively 4.00 min Autotune voltage
Gas chromatograph Automatic sampler Inlet Column Carrier gas RTL Oven temperature program PTV inlet parameters Injection volume Mass Selective Detector (MSD) Scan range Source, quad, transfer line temperatures Solvent delay Multiplier voltage Software GC/MSD ChemStation Deconvolution Reporting Software (DRS) Library searching software Deconvolution software MS Libraries
Agilent p/n G1701DA (Version D01.00 sp1) Agilent p/n G1716AA NIST MS Search (version 2.0) (included with NIST '02 mass spectral library, Agilent p/n G1033A) Automated Mass Spectral Deconvolution and Identification Software (AMDIS) (included with NIST '02 mass spectral library, Agilent p/n G1033A) NIST '02 mass spectral library (Agilent p/n G1033A); Agilent RTL Pesticide Library (p/n G1049A)

RTL and RTL Databases RTL is a technique developed by Agilent that allows users to match analyte retention times (RTs) on any Agilent 6890 GC, in any laboratory in the world, so long as the same nominal GC method and capillary column are used [13]. Using RTL, Agilent has developed several retention-timelocked databases for GC and GC/MS that include the locked retention time, compound name, CAS number, molecular formula, molecular weight, and mass spectrum (GC/MS databases only) for each entry [14]. The Agilent RTL Pesticide Library contains this information for almost all GC-amenable pesticides, as well as several endocrine disrupters - 567 compounds in all. For use with the DRS discussed below, this library was converted into the NIST format [15]. Separate Automated Mass Spectral Deconvolution and Identification Software (AMDIS) libraries for the RTs and compound information were created from the original RTL Pesticide Library. Users can easily augment these libraries with newer pesticides or other compounds of interest [15].
Basics of Deconvolution In GC/MS, deconvolution is a mathematical technique that separates overlapping mass spectra into cleaned spectra of the individual components. Figure 1 is a simplified illustration of this process. Here, the total ion chromatogram (TIC) and apex spectrum are shown. As is often the case, the peak is composed of multiple overlapping components and the apex spectrum is actually a composite of these constituents. A mass spectral library search would give a poor match, at best, and certainly would not identify all of the individual components that make up the composite spectrum. The deconvolution process finds ions whose individual abundances rise and fall together within the spectrum. In this case, it first corrects for the spectral skew that is inherent in quadrupole mass spectra and determines a more accurate apex RT of each chromatographic peak. As illustrated in Figure 1, deconvolution produces clean spectra for each overlapping component. These individual spectra can be library searched with a high expectation for a good match. The AMDIS that is incorporated into the Agilent DRS is supplied by the National Institute of Science and Technology (NIST) [12].
TIC and spectrum
Deconvoluted peaks and spectra
TIC Component 1 Component 2 Component 3 Matrix
Deconvolution
Interference
Target
Figure 1.
An illustration of mass spectral deconvolution process. 3
DRS Agilent's DRS results from the combination of three different GC/MS software packages: 1) the Agilent GC/MS ChemStation, 2) the NIST Mass Spectral Search Program with the NIST '02 MS Library, and 3) the AMDIS software, also from NIST. Included in the DRS, are mass spectral and locked RT libraries for 567 pesticides and suspected endocrine disrupters. Three separate, but complimentary, data analysis steps are combined into the DRS. First, the GC/MS ChemStation software performs a normal quantitative analysis for target pesticides using a target ion and up to three qualifiers. An amount is reported for all calibrated compounds that are detected. For other compounds in the database, an estimate of their concentration can be reported based upon an average pesticide response factor
(RF) that is supplied with the DRS software. The DRS then sends the data file to AMDIS, which deconvolutes the spectra and searches the Agilent RTL Pesticide Library (in AMDIS format) using the deconvoluted full spectra. A filter can be set in AMDIS, which requires the analyte's RT to fall within a user-specified time window. Because RTL is used to reproduce the RTL database RTs with high precision, this window can be quite small typically 20 seconds or less. Finally, the deconvoluted spectra for all of the targets found by AMDIS are searched against the 147,000-compound NIST mass spectral library for confirmation; for this step, there is no RT requirement. Once the appropriate method is loaded, the DRS report can be generated with a single mouse click as shown in Figure 2. The software can run automatically after each analysis or at a later time on a single file or a batch of files.
Figure 2.
ChemStation pull down menu showing options for running the DRS on single or multiple files.
Pesticides in an Herbal Mix Figure 3 shows a TIC from the extract of an herbal mix. Figure 4 shows the MSD Deconvolution Report for this sample, which is produced in html format so it can easily be emailed or copied into a spreadsheet. This sample was chosen because herbs are among the most difficult vegetable products to analyze. Their extracts contain a large number of natural products that interfere with pesticide analysis.
450000 400000 350000 Abundance 300000 250000 200000 150000 100000 50000 5.00 10.00 15.00
TIC: MOL_4A.D
20.00
25.00 Time
30.00
35.00
40.00
45.00
Figure 3.
TIC of an herbal mix.
Figure 4.
MSD Deconvolution Report generated for the herbal mix extract shown in Figure 3.
The DRS report in Figure 4 lists the RT, CAS number, and compound name for each hit. Phenanthrene-d10, listed at the bottom of the report, is the internal standard (ISTD) used by the ChemStation to estimate the quantity of each compound that it found. Since an average pesticide response factor was used for all 567 target compounds, the amounts listed in column 4 are only estimates. Experience has shown that most estimates reported using an average pesticide response factor fall within a factor of 10 of their actual values. True quantitation requires calibration with pesticide standards in the normal way, but this is not practical for all of the pesticides in the database. A laboratory would normally generate calibration curves for their target set of pesticides and use the average RF for the remaining compounds in the database. In this way, when a new compound is detected, the lab can immediately get a rough estimate of its concentration and decide if it should be added to the calibration list. Column 5 in the report shows the match factor obtained through AMDIS deconvolution and RTL Pesticide Library searching using the deconvoluted full spectra. In this case, several more targets were identified by AMDIS than were found by the ChemStation software (for example, Prometon and p,p-DDE), which is typical for complex samples. When locked RTs are available, it is a significant advantage to set a RT requirement in the AMDIS software. In this case, hits that did not fall within 10 seconds of the database RT were eliminated. Column 6 shows the RT difference (in seconds) between the compound's library RT and its actual value in the chromatogram. Figure 4 shows that the software identified two phthalates (suspected endocrine disrupters) in addition to the pesticides. Phthalates are ubiquitous in the environment and are extremely difficult to remove from the background. In this case, no attempt was made to determine if the phthalates were actually extracted from the sample or were introduced in the laboratory. The last two columns in the DRS report show the results from searching all of the AMDIS hits against the NIST 147,000-compound mass spectral library. When the NIST library search finds a compound in the top 100 matches (a user-settable value) that agrees with the AMDIS results, its match factor is listed in column seven. The hit number is shown in the last column, with 1 being the best match (highest match factor) in the NIST database. Occasionally, the NIST library search does not find the AMDIS hit among the top
6
100 spectral matches. In this case, the next line in the report shows the best library match for that spectrum. This is evident for fluvalinate-tau-I (Figure 4), which eluted at 34.779 min. The next line shows the best NIST library match for that spectrum - fluvalinate. In this case, no compound with the same CAS number as fluvalinate-tau-I is contained in the NIST mass spectral library. In fact, fluvalinate-tau-I is the D isomer, while fluvalinate is the DL isomer mixture. Blind Comparison Between DRS and Traditional Data Review Many comparisons have shown that the DRS is much better than conventional methods at identifying target compounds in complex samples, such as food and environmental extracts. Two such studies are described here. In the first case, 17 unspiked crop samples were analyzed by NRM Laboratories in Berkshire, UK using Agilent's RT-locked pesticide method. The data files, but not their list of pesticide hits, were sent to Agilent for analysis using the new DRS. Table 2 shows a comparison of the results from the two laboratories. Using manual data review, NRM identified 28 pesticides in the 17 samples, four of which were below their lowest calibration level. Using the same data files, the DRS identified 33 pesticides. Agilent's automated method did not identify azoxystrobin in the spring onion sample because it is not included in the RTL pesticide library. While it can be found in the NIST library, it has a molecular ion at 403 amu and method used at NRM only scanned to 400 amu. The DRS method confirmed all four pesticides that were below the NRM calibration range and found five more (terbacil, pyrimethanil, methiocarb, pyridaben, and propamocarb) that were not included in their method. The agreement between the manual and automated methods was excellent. However, the DRS looks for many more pesticides and was able to find several that were missed by the manual method. In addition, manual data review took a chemist about 7 hours for the 17 samples while the DRS finished the task in 50 minutes of unattended computer time.
Table 2.
A Comparison of the Pesticides Found in 17 Unspiked Crop Samples Using Conventional Data Review and Agilent's DRS. Pesticides that Were Found by Only One Method Are Underlined Agilent DRS results* Propyzamide Chlorthal-dimethyl p,p'-DDE Terbacil Pirimicarb Chlorthal-dimethyl Propyzamide Pyrimethanil Pirimicarb Metalaxyl Iprodione Not in DRS library Methiocarb Iprodione Procymidone Pyridaben Propamocarb Procymidone Buprofezin Endosulfan sulfate Iprodione Chlorthal-dimethyl Iprodione Dimethoate Metalaxyl Procymidone Terbuconazole Omethoate Procymidone lamda-Cyhalothrin Chloropropham Pirimicarb Pirimicarb NRM Manual Analysis** Propyzamide Chlorthal-dimethyl p,p'-DDE Not found*** Pirimicarb Chlorthal-dimethyl Propyzamide Not found*** Pirimicarb Metalaxyl Iprodione Azoxystrobin Not found*** Iprodione Procymidone Not found*** Not found*** Procymidone Buprofezin Endosulfan sulfate Iprodione Chlorthal-dimethyl Iprodione Dimethoate Metalaxyl Procymidone Terbuconazole Omethoate Procymidone lamda-Cyhalothrin Chloropropham Pirimicarb Pirimicarb
Sample Coriander
Rosemary
Spring Onion
Chives Cherry Tomato Courgette Aubergine
Flat Leaf Parsley Lambs Lettuce Cos Lettuce
Fine Endive Red Potato Fine Endive

* **

Pesticides found by re-analyzing NRM datafiles using Agilent's DRS software. Pesticides found by NRM using target compound analysis and manual verification. This compound is not included in the Agilent RTL Pesticide Library or the DRS software. Found by the Agilent ChemStation but not found by AMDIS or NIST library searching after deconvolution. After careful review of this hit, omethoate was judged not to be in the sample. Compound was detected but was below the calibration range.
*** This compound was not in the NRM target compound list.
Analysis of Surface Water Samples: In another study, the CDFA analyzed 17 surface water extracts for pesticides. TICs for two typical samples are shown in Figure 5. The CDFA used RTL and RTL database searching but without the benefit of spectral deconvolution. The same data files were then analyzed using the DRS for comparison. Table 3 shows the results from the CDFA manual analysis of the 17 samples compared to results using the DRS. The CDFA found 38 pesticide hits in the 17 samples, some of which were for the same pesticide in multiple samples. It took a skilled analyst about 8 hours to review the results, eliminate false positives, and verify all of the hits. The DRS found 37 of the compounds seen by the CDFA and identified one CDFA hit as a false positive. In addition, 34 more pesticides were
found for a total of 71 hits in the 17 samples. The process was fully automated and took about 20 minutes of unattended computer time to process all of the data files.
Table 3. A Comparison of Results from the Analysis of 17 Surface Water Samples by GC/MS. The CDFA Used RTL and RTL Database Searching, but No Deconvolution. Agilent's DRS Was Used to Analyze the Same Data Files CDFA Number of pesticide hits Number of false positives Time required for analysis 37 1 ~ 8 hours DRS Same 37 + 34 additional 0 20 minutes
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
Figure 5.
TICs of typical surface water extracts provided by the CDFA.
Conclusions
Agilent's new DRS solution for pesticide analysis offers laboratories a number of real benefits. Ease of use: This software solution is very simple to use and takes no more skill than is needed to operate the 6890N/5973 inert GC/MS system. There is no need for the user to learn about the intricacies of deconvolution or to master a new software package. Automation: The deconvolution report can be generated automatically after each run or a batch of samples can be processed all at once. Time savings: Data review is reduced from hours to minutes. Quality: It produces results with the fewest false positives and false negatives. Reproducibility: Results are not dependent upon the skill or experience of the operator. Accuracy: Comparisons such as those discussed in this application note show that the DRS finds pesticides with greater accuracy than manual methods of data analysis. It is particularly useful for relatively complex samples where co-eluting matrix components might obscure traces of target pesticides. Comprehensive: This method screens for almost all GC-amenable pesticides as well as several suspected endocrine disrupters in a single GC/MS run. With 567 compounds in the method, it is the most comprehensive pesticidescreening tool available. Users can add more compounds to the method as needed. Produces quantitative, semi-quantitative, and qualitative results: All calibrated compounds can be quantified. The concentrations of any other compounds can be estimated using an average pesticide response factor provided with the software. Use of the DRS is not limited to pesticide analysis. Other target compound mass spectral libraries can be converted into the AMDIS format and used with this software. For example, one could use existing libraries for forensic drugs, flavors and fragrances, organic pollutants, etc. Users can even generate their own libraries and use them with the DRS. While not required, it is a big advantage to have an RTL library with locked RTs for each entry, as this will give the fewest false positives.
Acknowledgements
The authors wish to thank Dr. Mark Lee and Stephen Siegel of the California Department of Food and Agriculture, Dr. J.G.J. Mol of TNO Research, The Netherlands, and the management and staff at NRM Laboratories, UK, for their contribution of samples and data.
References
1. C.D.S Tomlin, editor. The Pesticide Manual, 13th edition, British Crop Protection Council, Surry, UK (2003). 2. http://www.who.int/pcs/docs/ Classif_Pestic_2000-02.pdf 3. J. Cook, M.P. Beckett, B. Reliford, W. Hammock, and M. Engel (1999) J. AOAC Int., 82 (6), 14191435. 4. M.A. Luke, J.E. Froberg, and H.T. Masumoto, (1975) J. Assoc. Off. Anal. Chem. 58, 1020-1026. 5. M. Luke, J. Froberg, G. Doose, and H. Masumoto, (1981) J. Assoc. Off. Anal. Chem. 64, 1187-1195. 6. B. McMahon and N. Hardin (1994) Pesticide Analytical Manual, Vol. 1, 3rd Ed., U.S. Food and Drug Administration, Washington, DC. 7. J. Fillion, R. Hindle, M. Lacroux, and J. Selwyn (1995) J. AOAC Int. 78, 1252-1266. 8. J. Fillion, F. Sauv, and J. Selwyn (2000) J AOAC Int., 83, 698-712. 9. P. Wylie and B. Quimby, A Method Used to Screen for 567 Pesticides and Suspected Endocrine Disrupters, Agilent Technologies, publication 5967-5860E www.agilent.com/chem. 10.H. Prest, P. Wylie, K. Weiner, and D. Agnew, Efficient Screening for Pesticides and Endocrine Disrupters Using the 6890/5973 GC/MSD System, Agilent Technologies, publication 5968-4884E www.agilent.com/chem. 11.K. Weiner and H. Prest, Retention Time Locking: Creating Custom Retention Time Locked Screener Libraries, Agilent Technologies, publication 5968-8657E www.agilent.com/chem. 12.National Institute of Standards and Technology, AMDIS Literature and Downloads website: http://www.amdis.net/What_is_AMDIS/ AMDIS_Literature_and_Downloads/ amdis_literature_and_downloads.html.
13.V. Giarocco, B. Quimby and M. Klee, Retention Time Locking: Concepts and Applications, Agilent Technologies, publication 5966-2469E www.agilent.com/chem. 14.http://www.chem.agilent.com/cag/servsup/ usersoft/main.html#RTL. 15.M. Szelewski and C.K. Meng, Building and Editing RTL Screener Databases and Libraries, Agilent Technologies, publication 5989-0916EN www.agilent.com/chem.

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2004 Printed in the USA May 19, 2004 5989-1157EN
Analysis of Components, Contaminants, and Impurities in Fungicide Formulations by GC/MS and LC/MS Application
Agriculture, Specialty Chemical, Environmental, Ag Chem
Author
James D. McCurry Paul Zavitsanos Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Abstract
A commercially available fungicide formulation was analyzed by both gas chromatography/mass spectrometry (GC/MS) and electrospray ionization liquid chromatography/mass spectrometry (ESI-LC/MS). The GC/MS analysis provided a detailed look at the volatile components in the formulation, but did not yield any results for the active ingredient, triforine. The ESI-LC/MS provided information on the stereoisomers of triforine as well as the nonvolatile surfactants and contaminants in the formulation. This paper demonstrates the complementary nature of these two analytical techniques when trying to fully characterize a complex chemical formulation containing a broad range of components.
unexpected impurities and breakdown products that can affect product quality. However GC/MS can only provide meaningful information for compounds that are volatile, nonionic, thermally stable, and have relatively low molecular weight. Liquid chromatography is much better suited to analyzing compounds that are nonvolatile, ionic, polar, thermally labile, or have high molecular weight. This includes about 80% of all known organic compounds [1]. When coupled with a modern atmospheric pressure ionization (API) mass spectrometer, LC/MS offers a complementary tool to GC/MS in the chemical diagnostic laboratory. Commercial pest control formulations contain one or more active compounds along with a recipe of ingredients that can play an important role in the products efficacy. These inactive ingredients are often a combination of solvents and surfactants that allow for easy application and dispersal of the active ingredient onto the target substrate. For this work, an over-the-counter fungicide formulation was purchased at a local home products store. The active ingredient in this product is 6.5 % (wt) of N,N-[1,4-piperazinediylbis(2,2,2-trichloroethylidene)] bisformamide. This is also known as triforine (CAS registry number 26644-46-2), and the structure is shown in Figure 1. The inactive ingredients in this formulation are listed as cyclohexanone, N-methyl pyrrolidone, and Atlox 3406-F. The Atlox 3406-F is an agricultural dispersant that contains ionic and nonionic surfactants and mixed aromatic solvents.
Introduction
Gas chromatography/mass spectrometry (GC/MS) is an indispensable tool for solving complex problems in the chemical industry. This fast and powerful technique yields detailed information about the expected compounds in the mixture along with any
Cl Cl * Cl NH O
Table I.
GC/MS Analysis Conditions
Gas chromatograph conditions Column: Carrier gas: Flow rate: Inlet: 30 m 0.25 mm HP5-MS, 0.25 m (p/n 19091S-433) Helium at 13.00 psi 1.6 mL/min., constant flow mode Cool on-column at 50 C, oven track mode
N O HN Cl Cl N * Cl
*Optically active
Oven temperature program: 50 C for 3 min 10 C/min to 275 C 275 C for 4 min MS Transfer line: 280 C 1 L 1400 V 3 min 30 to 800 m/z 50 counts 2 1.95 scans/s Injection volume: Electron multiplier: Solvent delay: Scan range: Scan threshold: A/D Samples: Scan rate
Figure 1.
Chemical structure of triforine, the active ingredient in some commercial fungicides. The nominal molecular weight is 432, and the structure contains two optically active carbons.
Mass spectrometer conditions
A complete analysis of this formulation requires GC/MS to separate and identify the volatile components and LC/MS for the surfactants and polar components. Analysis of the active ingredient, triforine, presents a separate challenge. References for triforine analysis cite gas chromatography as the method of choice when analyzing environmental residues [2]. However, the melting point is reported to be 155 oC with decomposition, indicating that gas chromatography may only be possible with on-column injection.
Electrospray Ionization Liquid Chromatography/Mass Spectrometry (ESI-LC/MS) The same fungicide sample was run on the Agilent 1100 Series LC/MSD. This system included a vacuum degasser, a binary pump, an autoinjector, a thermostatted column compartment, and the LC/MSD SL quadrupole mass spectrometer. LC/MS instrument conditions for this analysis are shown in Table 2.
Experimental
Gas Chromatography/Mass Spectrometry (GC/MS) A 1% (v/v) solution of the triforine formulation was made in acetonitrile and the GC/MS analysis was performed with an Agilent 5973 GC/MS system. The components in this system were a 6890N gas chromatograph, a 7683 autoinjector, and a 5973 mass spectrometer. A cool-on-column inlet in the Agilent 6890 GC was used to avoid decomposition of the triforine. Instrument conditions for the GC/MS analysis are listed in Table 1.

Gas Chromatography/Mass Spectrometry (GC/MS) The complex nature of this fungicide formulation is revealed when one looks at the GC/MS data. Figure 2 shows the total ion chromatogram (TIC) of the fungicide sample. The volatile components
Table 2.
LC/MS Analysis Conditions
Liquid chromatograph conditions Column: Mobile phase A: Mobile phase B: Mobile phase gradient: Flow rate: Column temperature: Injection volume: Source: Drying gas flow: Nebulizer: Drying gas temperature: Vcap: Stepsize: Peak width: Time filter: Scan range Fragmentor 150 4.6 mm Zorbax XDB-C8, 5 m (p/n 993967-906) 0.1% Formic acid in water 0.1% Formic acid in acetonitrile 30% B at 0 min; 50% B at 7 min; 95% B at 10 min 1.0 mL/min 30 C 1 L Electrospray 12 L/min 40 psig 350 C 3500 V (positive) and 3000 V (negative) 0.1 amu 0.1 min On 120 to 1200 m/z Fixed at 60 V
in the formulation are easily identified from the mass spectral data. The major solvents, cyclohexanone and N-methyl-2-pyrrolidone, dominate the chromatogram while smaller amounts of C9 aromatics, C10 aromatics, and substituted napthalenes are easily separated and identified. There were no peaks in the TIC whose spectra matched the triforine reference spectra from the Wiley mass spectral library. An extracted ion profile using the triforine base peak of 203 m/z did not produce any chromatographic peak indicating the presence of triforine. From this data, it appears that the triforine did not elute from the column into the mass spectrometer. However, a spectral average of the large hump between 18 and 20 minutes shows an isotope pattern indicating one chlorine atom (Figure 3A). Since no chlorinecontaining species other than triforine are components in the formulation, the presence of chlorine and the broad peak shape indicates triforine decomposition in the gas chromatograph. The peak at approximately 20-minute retention time also has a mass spectrum containing an isotope pattern indicating the presence of two chlorine atoms in the structure (Figure 3B). This peak could be a decomposition product or a contaminant in the formulation.
5 2 4 1 2 3 4 5 6 7 1-hexanol Cyclohexanone C9-aromatics N-methyl-2-pyrrolidone C10-aromatics Naphthalenes Triforine decomposition
6 3
10
12
14
16
18
20
Figure 2.
GC/MS TIC showing the complex volatile components in the commercial fungicide formulation.
145 158 187 152
215
140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320 m/z
288
144
140 150
157
160 170 180
189
190 200 210
225
220 230 240 250
259
260 270 280 290 300 310 320 m/z
Figure 3.
(A) Average mass spectrum of broad hump between 18 and 20 minutes of TIC. Isotope patterns of the peaks at m/z 145, 158 and 187 indicate the presence of one chlorine atom. (B) Mass spectrum of the peak at 20 minutes shows the presence of two chlorine atoms in the structure.
Electrospray Ionization Liquid Chromatography/Mass Spectrometry (ESI-LC/MS) The positive ion ESI-LC/MS chromatogram is shown in Figure 4. Several major peaks are observed along with several minor components. The spectra of the three peaks eluting between 0 and 2 minutes are shown in Figure 5. Since electrospray is a soft ionization technique, these spectra do not exhibit the detailed fragmentation
needed to interpret structures for these three compounds. However, peak number 2 does have an isotopic pattern indicating the presence of two chlorine atoms in the structure. This compound could be a contaminant related to triforine production or a triforine decomposition product. Figure 6 shows the spectra of the three peaks between 10.5 and 13 minutes. These compounds are the various surfactants that make up the agricultural dispersant used in the formulation.
3 8
2 6
10
12
Figure 4.
TIC from positive ion ESI-LC/MS of fungicide formulation.
122
167
Peak 1
126 146 185 208
200 m/z
120
140
160
180
260 262
Peak 2
264
250 275
300
325
350
375
m/z
199
Peak 3
200
180 190 200
210
220
m/z
Figure 5.
Electrospray spectra from LC/MS peaks 1, 2, and 3. The spectra from peak 2 shows an isotope pattern indicating two chlorine atoms in this structure. This compound may be a contaminant in the formulation from the active ingredient triforine.
Peak 6
600
700
800
900
m/z
Peak 7
400
600
800
1000
m/z
Peak 8
400
500
600
700
800
m/z
Figure 6.
Electrospray mass spectra of LC/MS peaks 6, 7, and 8 from Figure 4. These compounds are the surfactants used in the formulation.
The spectra of LC/MS peaks 4 and 5 (Figure 7) are identical and correspond to the active ingredient, triforine. The protonated molecular ion is observed at m/z 433 along a sodium adduct at m/z 455. The multiplets for m/z 433 to 439 and m/z 455 to 461 exhibit an isotopic pattern consistent with six chlorine atoms. The ion at m/z 388 is due to a rearrangement and subsequent loss of a formamide group from the protonated molecular ion (m/z 433). This is also confirmed by the isotopic pattern indicating six chlorine atoms (m/z 388 to 396). The presence of two triforine peaks in Figure 4 can be explained by the stereochemistry of the structure. Triforine contains two optically active carbons that give rise to four stereoisomers. Figure 8 shows the four configurations that can be grouped into two pairs of mirror images that are diastereomers. The S,R and R,S configurations are mirror
images that are superimposable, resulting in a meso compound that exhibits no optical activity or differences in physical properties. Therefore, because the S,R and R,S configurations are identical, they will elute as one chromatographic peak. The second pair of mirror images are the R,R and S,S configurations. These are not superimposable and are, therefore, enatiomers that will exhibit different optical activity, but identical physical properties. Conventional reverse-phase liquid chromatography cannot separate these enantiomers, and they will co-elute as a single peak. However, these enantiomers are not mirror images of the meso compound and can be chromatographically separated from the meso compound. This is why there are two triforine peaks, one for the meso compound and one for the enatiomers. Without pure standards of the stereoisomers, it is not possible to determine which configurations can be attributed to the observed chromatographic peaks.
(M+H)+ -43
Peak 4
7.554 min
(M+H)+ (M+Na)+
390 392
Peak 5
7.757 min
388
394
433 435 437 439
400
420
440
457 459 461

460 m/z
Figure 7.
Electrospray mass spectra of peaks 4 and 5 from Figure 4. Both spectra show a protonated molecular ion at m/z 433 representing the active ingredient triforine. There is also a sodium adduct (m/z 455) of triforine observed for both peaks. A rearrangement and loss of a formamide group from the protonated molecular ions give rise to the multiplet at m/z 388 to 396. Mirror Mirror
O HN H
(S)
396
O NH HN H Cl3C
(R)
O NH
CCl3
Cl3C
(R)
(S)
CCl3
N H
(R)
N CCl3 Cl3C
(S)
N H H
(R)
N CCl3 Cl3C
(S)
NH O (S,R)
HN O (R,S) O
NH
HN O
(R,R)
(S,S)
Meso structure no optical activity Figure 8.
Enantiomers optically active
The four triforine stereoisomers arising from the two chiral carbons in the structure. These two pairs of mirror images account for the two triforine peaks observed in the chromatogram (Figure 4). 7
The fungicide formulation was also run by ESI-LC/MS in the negative ion mode. The results of this analysis are shown in Figure 9. The negative ion mass spectra for these two peaks are shown in Figure 10. For both triforine peaks, the most stable negative ion species is the chloride adduct (m/z 467). However, the spectra for the first peak contains a
7.554
deprotonated molecular ion (m/z 431) and a formate adduct (m/z 477) that is not observed in the spectra of the later eluting peak. This selective adduct formation is likely related to the stereochemstry of the triforine, but again, without pure standards, the correct configurations cannot be assigned to the chromatographic peaks.
7.757
10
12
Figure 9.
TIC from negative ion ESI-LC/MS of fungicide formulation.

_ (M+Cl)
469 471
RT = 7.554 min
(M+HCO2)
479 _
(M+TFA)
_ 545 547
467
473
477
483
(M_H)
(M+NH4CO2)
494 494 494 494
431 433 435 437
RT = 7.757 min
425
450
475
500
525
550
551 m/z
Figure 10. Negative ion electrospray mass spectra of the two triforine peaks. The spectra from peak at 7.554 minutes shows a deprotonated molecular ion (m/z 431) and a formate adduct (m/z 477) that is not seen in the later eluting peak (7.757 minutes).
549
481
Conclusions
This paper demonstrates the complimentary nature of GC/MS and LC/MS when trying to characterize a formulation that is composed of many different chemical species. The volatile compounds in the formulation can be easily separated and identified by GC/MS. In this case, polar solvents such as cyclohexanone and N-methyl-2-pyrrolidone were the major components while 1-hexanone, C9 aromatics, C10 aromatics, and substituted naphthalenes were present as minor components or contaminants. However, GC/MS did not yield any information on the active fungicidal ingredient, triforine, a hexachlorinated compound. This was most likely due to thermal decomposition during GC/MS analysis. Evidence for this was seen in a broad chromatographic hump containing chlorine-containing constituents. The nonvolatile components in this fungicide were quickly analyzed by ESI-LC/MS. This analysis yielded information on several polar contaminants, some containing chlorine, which may be by-products of triforine production or triforine breakdown products. Also observed were several surfactants that are used in agricultural products as dispersants. The LC/MS analysis did yield significant information on the triforine active ingredient, showing a distribution of stereoisomers in the formulation.
References
1. Willoughby, R., Sheehan, E, and Mitrovich, S. A., Global View of LC/MS: How to Solve your Most Challenging Problems, p. 80, Global View Publishing, 1998. 2. The Pesticide Manual 9th Edition, Worthing, C.R. and Hance R.J. eds., p. 853, The British Crop Protection Council, 1991.

Identification and Quantitation of Pesticides in the Parts-per-Trillion Range Using Retention Time Locking and GC/MS Application
Environmental, Food
Author
C. Kai Meng Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Abstract
The typical pesticide quantitation limit for a mass spectrometer in the Scan mode is in the sub-ppm range. By using a selected ion monitoring method, a lab can lower the target compound quantitation limit to the low parts-per-billion (pg/L) range using a retention time locked gas chromatography/mass spectrometry method. By adding large volume injection capability to the method, target compounds at parts-per-trillion can be quantified. A specially developed 567-compound retention time locking pesticide mass spectral library can automatically screen an acquired samples data file for all 567 compounds in seconds. The library can also be applied for rapid screening of samples acquired in selected ion monitoring method. Using the compound library information, a selected ion monitoring method for 80 target compounds was created in less than 2 hours without running any analyses.
detection limits and are easy to operate, the data do not provide sufficient information to confirm a compounds presence with confidence. Due to the universal nature of mass spectrometric detection, a mass spectrometer (MS) provides additional information and increased confidence in the assignment of compound identity. With recent advances in GC/MS hardware and software and the decrease in cost of ownership, more and more laboratories are routinely analyzing pesticide residue samples with MS detection. To match the GC/ESD detection limits and/or to eliminate sample concentration steps, a user must lower the MS detection limit by 2 to 3 orders of magnitude. This application note, discusses the following approaches. Run the MS in single ion monitoring (SIM) mode Make large volume injections (LVIs) Use higher electron multiplier voltage (EMV) For compound identification, a specially developed 567-compound retention time locking (RTL) [1] pesticide library could perform the entire 567-compound screening in seconds using Scan data. A subset of the library could be screened in seconds from SIM data.
Introduction
Most pesticides are typically analyzed on a gas chromatograph (GC) with element-selective detectors (ESDs). Although these ESDs provide low ppb
Experimental
A pesticide standard mixture was used to compare the lowest detection limits of splitless injection and LVI under Scan and SIM modes.
System Configuration for Screening and Quantitation: 6890 GC with a programmable temperature vaporizer (PTV) [2,3] inlet 5973 Mass Selective Detector (MSD) 7683 Automatic Liquid Sampler (ALS) tray and autoinjector HP-5MS capillary column (30 m 0.25 mm 0.25 m), P/N 19091S-433 G1701BA version B.00.00 MSD ChemStation software or higher G1049A MSD RTL Pesticide Database/Library
Table 3. MS Method Parameters Solvent delay 3 min Tune file Atune.u Transfer line 280 C MS Quad 150 C MS source 230 C Threshold 150 Sample # 2 Scan range 35 to 500 amu (in Scan mode) Forty (40) SIM groups (in SIM mode)
Table 4. Pesticide Screening Parameters for the SIM Method Extraction window Qualifier mode Qualifier % Zero qualifiers Subtraction mode Screen database 0.100 minute Absolute 30 Included Average start/stop Rtlpest.SCD
Table 1. GC Method Parameters Oven Inlet 70(2)/25/150(0)/3/200(0)/8/280(10) = 41.87 min PTV
Inlet pressure 17.30 psi (locked to methyl chlorpyrifos at 16.593 min), constant pressure mode

RTL [1] was used to: 1. Expedite data comparison in overlay format
Table 2. Injection Parameters Injection mode Injection volume (syringe) Injection speed Inlet temp Solvent vent 25 L (50-L syringe, P/N 5183-0318) Inject @ 100 L/min Draw @ 300 L/min Dispense @ 4500 L/min 40(0.35)/600/320 (3)/50/200 (Hold until end) Vent time = 0.29 min Vent flow = 150 mL/min Vent pressure = 0.00 psi 60 mL/min @ 2 min Deactivated, Multi Baffled (P/N 5183-2037) Liquid CO2 60 mL/min @ 2 min Deactivated, Multi Baffled (P/N 5183-2037) None Splitless 1 L (10-L syringe, P/N 9301-0713) Fast
2. Achieve lower target compound detection limit 3. Allow rapid pesticide screening using the RTL pesticide database/library 4. Help to differentiate isomers by their retention time (RT) differences 5. Eliminate the tedious SIM method RT updating process after column maintenance
280 C
6. Simplify the editing of the SIM ion groups A mixture from the California Department of Food and Agriculture (CDFA) of 80 pesticides at 5000 pg/L each was used as the stock solution for this study. The mixture contained carbamate, organochlorine, organophosphorus, and organonitrogen pesticides. Figure 1 is an offset overlay of three total ion chromatograms (TIC) with 50, 100, and 500 pg of each of the pesticides injected. These TICs were obtained in the Scan mode from 1-L spiltless injections. For many of these pesticides the quantitation limit in the Scan mode is about 500 pg on column.
Vent
Purge Liner
Inlet cooling
240000
50 pg on column
220000
180000
100 pg on column
140000
100000
60000
500 pg on column
20000
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
Figure 1.
Total ion chromatograms from 1-L splitless injections of 80 pesticides with 50, 100, and 500 pg of each compound injected.
SIM Mode To lower the detection limit, a SIM method was created. Instead of the traditional way of making a SIM method, a user can use the information in the RTL Pesticide Database to build a SIM method
without running an analysis. Here are the steps for editing SIM ion group parameters: 1. List the MSD RTL Pesticide Database from the ChemStation (Figure 2 is a partial listing) and paste the complete listing into a spreadsheet.
Figure 2.
A partial listing of the pesticide screener database. The listing includes the compound number, compound name, target ion, expected retention time, and three qualifier ions. 3
2. In the spreadsheet, delete the rows of the compounds not needed in the method. 3. Separate target compounds into groups (see the added Group # column on Figure 3) using these criteria: One to three compounds in each group, and The RTs of the adjacent compounds in adjacent groups are at least 0.2 minute apart. For example, compounds 42 and 51 are more than 0.2 minute apart, so they are in different groups. Compounds 51 and 55 are less than 0.2 minute apart, so they are in the same group. 4. Use the average RT of the adjacent compounds in adjacent groups as the SIM group RT (see the added Group RT column on Figure 3). For example, the average retention time of compound 42 (7.91 min, in group 2) and compound 51 (8.78 min, in group 3) is 8.35 minute which is used as the starting retention time of group 3. When all the group numbers and respective starting retention times are determined, make a hardcopy of the spreadsheet for easy entry into the MS SIM/Scan Parameters in the next step. 5. Enter the target ion and qualifier ion(s) (Q1, Q2, and/or Q3) of all compounds into the respective ChemStation SIM group (Figure 4). Notice that all the information for building the SIM groups came from Figure 3.
# 24 35 42 51 55 76 82 92 98 102 103 104 111 113 117 120 122 124 129 Compound Name 2,6-Dichlorobenzonitrile Mevinphos Propham o-Phenylphenol Pentachlorobenzene Propoxur Diphenylamine Chlorpropham Ethalfluralin Bendiocarb Trifluralin Benfluralin Phorate BHC alpha isomer Hexachlorobenzene Dicloran Demeton-S Dimethoate Simazine MSD_RT 6.75 7.60 7.91 8.78 8.95 10.35 10.52 11.05 11.28 11.54 11.64 11.73 11.96 12.09 12.38 12.56 12.63 12.68 12.91 T 171 127 93 170 250 110 169 127 276 151 306 292 75 181 284 206 88 87 201 Q1 173 192 179 169 252 152 168 213 316 126 264 264 121 219 286 176 60 93 186 Group # Group RT 1 3.00 2 3 4 5 6 7.75 8.35 9.60 10.76 11.41
The number of qualifier ions used in a SIM method depends on the number of analytes of interest. For a method monitoring 20 to 30 compounds, all three qualifier ions should be used in the SIM method. As the list of target compounds grows, fewer qualifier ions should be used in the method to maintain a reasonable and comparable ion dwell time and sampling rate. In general, 10 scans (cycles) per peak are recommended for quantitation purposes. For example, if an analyte peak is 6 seconds wide, about 1.7 cycles per second should be maintained for that SIM ion group. Once the number of cycles per second is determined, the dwell time of the ions can be varied to meet that. As the dwell time is entered for each ion, the ChemStation automatically shows the number of cycles per second. In Figure 4, Group 6 has 3.03 cycles per second.
Figure 4.
A screen capture of the MSD ChemStation showing the MS and SIM parameters. The SIM parameters (group ID, group retention time, and ions) were all derived from Figure 3.
Figure 3.
A spreadsheet of target compounds separated into different SIM groups with RTs of the adjacent compounds in adjacent groups at least 0.2 minute apart. The starting retention time of each group was determined by calculating the average RT of the adjacent compounds in adjacent groups.
Figure 5 shows two chromatograms obtained from 1-L splitless injections at 50 pg/L using both Scan and SIM modes. The Scan mode has significantly higher baseline noise than the SIM mode. Some of the compounds, especially the late eluters, were not detected in the Scan mode. When the Scan method was changed to a SIM method at this concentration, the signal-to-noise ratio (S/N) increased by a factor of 100. It is worth pointing out that a SIM method does not record background ions from the sample matrix, therefore minimizing the baseline noise and improving the S/N.
21000 19000 17000 15000 13000 11000 9000 7000 5000 3000 1000 5.00 10.00 15.00 20.00 25.00 30.00 Injection volume: 1 L Concentration: 50 pg/L PTV mode: Splitless
Scan
SIM
35.00 40.00
Figure 5.
Chromatograms of 1-L splitless injections at 50 pg/L from Scan and SIM modes.
In a SIM method, the retention times of the ion groups normally need updating after column maintenance. By using RTL, a user can not only eliminate the tedious RT updating process [4] but also decrease the detection limit. With reproducible known RTs of target compounds, the start and end time of each ion group can be determined optimally. By narrowing the time windows of an ion group to monitor only one or two compounds at a time, the MS can monitor fewer ions in each window, allowing more sampling time for the target ions. Ideally, a SIM method will have the maximum number of ion groups and the minimum number of ions in each group. In this way, each ion group can get more scans per unit time resulting in better peak shape and more accurate quantitation.
LVIs To decrease the detection limit further, a user can put more sample on column using the LVI technique. The typical solvent-vent approach is to inject the sample slowly into a PTV inlet at a temperature just below the solvent boiling point and let solvent evaporate before ramping up the inlet temperature to move the compounds onto the capillary column. Figure 6 compares a 1-L splitless injection with a 25-L solvent-vent injection. Both injections resulted in 50 pg per compound on column. Note that the solvent-vent ion chromatogram is plotted upside down for ease of comparison with the splitless ion chromatogram. It is obvious from the figure that the two techniques provide very similar results. This demonstrates that the solvent-vent technique is a viable approach for sample introduction.
7000 6000 5000 4000 3000 2000 1000 0 -1000 -2000 -3000 -4000 -5000 6.00 8.00
SIM
Injection volume: 1 L Concentration: 50 pg/L PTV mode: Splitless
Injection volume: 25 L Concentration: 2 pg/L PTV mode: Solvent vent
10.00
12.00
14.00
16.00
18.00
20.00
22.00
24.00
26.00
28.00
30.00
32.00
Figure 6.
SIM results of 50 pg on column using either a 1-L splitless or a 25-L solvent-vent injection.
Higher EMV It is known that the signal increases with higher EMV on the MS. In Figure 7, the upper signal, after 10-fold magnification, is a 25-L LVI of 0.5 pg/L at tune voltage. The bottom signal is the same injection with the electron multiplier set to tune +400 V. Adding 400 V to the EMV increases
the signal by 10X, which makes the integration more accurate. However, the baseline noise also increases by 10X, so the S/N stays the same. Although increasing the EMV does help to bring small peaks over the detection threshold, it shortens the life of the multiplier. In general, the EMV should be kept at the tune voltage.
40000 35000 30000 25000 20000 15000 10000 5000 0 -5000 -10000 -15000 -20000 -25000 -30000 -35000 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00
Tune voltage
(Magnified 10X)
Injection volume: 25 L Concentration: 0.5 pg/L
Tune voltage + 400 V
Figure 7. SIM results of 12.5 pg on column using either EMV at Tune voltage or Tune +400 V.
LVIs in Combination with SIM Methods Combining LVI and SIM, Figures 8 and 9 show quantifiable peaks of three compounds at as low as 5 pg on column. In Figure 8, ion chromatograms of endosulfan sulfate and p,p-DDT at 0.2 and 500 pg/L are shown. The top chromatogram was from a 25-L solvent-vent SIM method and the bottom chromatogram was from a 1-L splitless Scan method. By using LVI and SIM, it is interesting to see that similar S/N ratios were achieved
even with a 2500-fold decrease (from 500 to 0.2 pg/L) in sample concentration. By increasing the injection volume to 100 L, samples at concentration as low as 0.05 pg/L can also be quantified as shown in Figure 9. The top portion shows the chlorthal-dimethyl extracted ion chromatograms (EIC) of mass 299 and 301 from a 100-L full Scan run. The bottom portion shows the same ions from a 100-L SIM run. The SIM method shows better peak shape and lower baseline noise.
280 250 220 190 160
Injection volume: Concentration:
25 L 0.2 pg/L (5 pg on column)
SIM
30000 24000 18000 12000 6000 26.30
Injection volume: Concentration:
1 L 500 pg/L (500 pg on column)
p,p-DDT
Scan
Endosulfan sulfate
26.40
26.50
26.60
26.70
26.80
26.90
27.00
27.10
27.20
27.30
Figure 8.
Ion chromatograms of endosulfan sulfate and p,p-DDT at 0.2 and 500 pg/L. The top chromatogram was from a 25-L solvent-vent SIM method and the bottom chromatogram was from a 1-L splitless Scan method.
1400 1200 1000 800 600 400 200 0 1400 1200 1000 800 600 400 200 0 18.6 18.8 19.0 19.2 19.4 19.6 19.8 20.0 20.2 20.4
Ion 301
Scan
Concentration: 0.05 pg/L
Ion 299
1000 800 600 400 200 0 1000 800 600 400 200 0 18.6 19.0 19.2
SIM
Ion 301
Concentration: 0.05 pg/L
Ion 299 *SIM group start time
*
19.4 19.6
*
19.8 20.0 20.2
Figure 9.
Ion chromatograms of 100-L chlorthal-dimethyl injected at 0.05 pg/L. The top portion was from a full Scan run and the bottom portion was from a SIM run.
Target Compound Screening Combing RTL and the G1049A MSD RTL Pesticide Database/Library, a user can screen for 567 pesticides and suspected endocrine disrupters from any Scan run [5]. A user can screen a subset of the library with improved sensitivity using a SIM method. The MSD ChemStation can generate a
567-compound screening report automatically in less than 30 seconds. Figure 10 is a report of the 0.5 pg/L sample (25 L injected in SIM mode) that lists the probable hits (marked with an x) and possible hits (marked with a ?). All target compounds at this 12.5 pg on column level were found by the software.
Figure 10. Typical report from the GC/MS pesticide screener showing probable "hits" (marked with an x) and possible hits (marked with a ?). Other information includes the library retention time followed by the RT difference in this chromatogram, the target ion, its abundance, out of range qualifier(s), and a cross correlation value with the library spectrum.
Conclusions
Using the information (compound names, retention times, and ion masses) in the RTL pesticide database, a SIM method of 80 target compounds can be created in less than 2 hours without running any analyses. The examples show that both LVI and SIM are effective techniques to decrease the quantitation limit of target compounds from sub-ppm to ppt. Any lab can decrease the quantitation limit by a factor of 100 without any hardware modification. Lowering the quantitation limit from 500 pg down to 5 pg on column can be done using a SIM method and RTL. By adding LVI to the system, target compounds in femtogram/L can be quantified.
Acknowledgement
The author would like to acknowledge Alex Chung and Mark Lee at CDFA for providing the pesticide mixture used in this study.
References
1. Vince Giarrocco, Bruce Quimby, and Matthew S. Klee, Retention Time Locking: Concepts and Applications, Agilent Technologies Application note, 5966-2469E, printed March 2000, www.agilent.com/chem 2. Bill Wilson, Philip L. Wylie, and Matthew S. Klee, Large Volume Injection for Gas Chromatography Using a PTV Inlet, Agilent Technologies Application note, 5965-7770E, printed February 2000, www.agilent.com/chem
3. Philip L. Wylie, Trace Level Pesticide Analysis by GC/MS Using Large-Volume Injection, Agilent Technologies Application note, 5966-1214E, printed April 2000, www.agilent.com/chem 4. Prest, H. and P. Cormia, Retention Time Locking: Advantages in GC/MS SIM Analysis, Agilent Technologies Application Note, 5968-3797E, printed December 1999, www.agilent.com/chem 5. Harry Prest, Philip L. Wylie, Ken Weiner, and Doug Agnew, Efficient Screening for Pesticides and Endocrine Disrupters Using the 6890/5973 GC/MSD System, Agilent Technologies Application note, 5968-4884E, printed December 1999, www.agilent.com/chem

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2001 Printed in the USA November 14, 2001 5988-4392EN
Fast Screening of Pesticides and Endocrine Disrupters Using the Agilent 6890/5973N GC/MSD System, Part II
Application
Gas Chromatography May 2000
for 567 of the most common pesticides and endocrine disrupters of concern worldwide. A GC/MSD method was developed based on the standard 42-min method1 to screen for all 567 of the most common analytes. A specific combination of column stationary phase, carrier gas flow rate, and oven temperature programming is required to lock all the compounds to an expected retention timetable2. Compound identification based only on spectral searching alone is difficult when analyzing extracts containing significant sample matrix content because of overlapping peaks and noisy baselines. The new screening tool, integrated within Agilent's ChemStation for MSD, searches for all 567 compounds. It first checks and integrates four characteristic ions within the expected time window and then prints a report showing "hits" and "possible hits" (ratios of characteristic ions that do not match the expected values in the library within specified limits). In Part I of the MSD fast screening application brief 3 , a 10 m 0.1 mm 0.1 m Agilent HP-5 column was used to increase analysis speed up to fourfold. In this application brief, a 15 m 0.25 mm 0.25 m Agilent HP-5MS column was used. The faster methods were scaled exactly as predicted by using a combination of Agilent's method translation (MTL) and RTL software. Because scaling was exact, these faster methods can be used with precisely-scaled pesticide libraries, making the screening process even more powerful and adaptable to individual needs.
Authors C. Kai Meng Agilent Technologies, Inc. 2850 Centerville Road Wilmington, Delaware 19808-1610 USA Michael J. Szelewski Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Abstract Agilent Technologies' new, fast GC/MSD method can significantly speed up the screening of pesticides. Agilent's GC Method Translation software (available free from the Agilent Technologies Web site, http://www.chem.agilent.com/cag/ servsup/usersoft/main.html#mxlator) was used in developing the new method based on the standard 42-min method. A 15 m 0.25 mm 0.25 mm Agilent HP-5MS column was used to increase analysis speed up to fourfold. The time savings were implemented in increments (down to 10.5 minutes) to verify the predictability of scaling and the effect of scaling on the signal-to-noise ratio. Key Words
RTL, pesticide, environmental, screening, fast GC, method translation, 5973, 6890, MTL
Introduction
Analysts want faster analyses to improve laboratory productivity. Often, when speeding up GC methods, an analyst will trade resolution for increased analysis speed. This loss of resolution can complicate peak identification, even with a mass selective detector (MSD). Agilent Technologies has developed new techniques to solve the peak identification problem based on Agilent's retention time locking (RTL) and a new mass spectral library that contains the locked retention times and characteristic ions
Experimental
The GC method translation software tool was used to find operating conditions for the faster methods. Figure 1 is a screen capture of MTL software data entry showing the original conditions and the new chromatographic conditions for a fourfold speed gain. The column flow rate, which is helpful to avoid exceeding MSD pumping capacity 4, also is found in the table. In this study, a turbo pump was used, which could handle the 3.8 mL/min carrier flow. The program also determined the required column head pressure and corresponding oven ramp. The Agilent 6890 GC fast oven option (220/240V in the U.S.) was required for the faster oven ramp used in this study. General chromatographic conditions are listed in table 1. The standard used was a mixture of 26 pesticides at 10 ppm. A 15 m 0.25 mm 0.25 m Agilent HP-5MS column (part number 19091S-431) was used. The head pressure determined by the method translation software (18 psi) was used as the starting point for retention time locking. The column head pressure required to lock retention times of the compounds to the library (the original retention time divided by 4) was determined using the automated RTL process integrated within the Agilent ChemStation for MSD.
Figure 1. Screen capture showing the method translation (MTL) software data entry used in a 4X speed gain translation.
This process (first translate the method then lock the retention times) was repeated for the 2.5X time reductions. Figure 2 shows the results of the shortened analysis times. The three chromatograms look extremely similar, except that the time axis is scaled proportionally. Because MTL followed by RTL scales methods very precisely, scaled screening libraries for corresponding time reductions can be obtained by dividing the retention times in the library by the speed gain (which does not have to be an integer). Using the same injection method (1-L splitless), the peak heights of the faster runs were twice those from the original
Table 1 Chromatographic Conditions
Speed GC Column Injection mode Column head pressure Column flow (mL/min) Inlet control mode Carrier gas Injector Temp. Oven Temp. Ramp 1 Ramp 2 Ramp 3 Oven equilibration Injection volume Liner MS Conditions (Turbo pump) Solvent delay Tune file Low mass High mass Threshold Sampling Scans/sec Quad Temp. Source Temp. Transfer line Temp. Acquisition mode
Onefold 110 V 30 m 0.25 mm 0.25 mm HP-5MS (P/N 19091S-433) Splitless 18.0 psi 1.9 Constant pressure Helium 250 C 70 (2 min) 25 C/min 150 (0 min) 3 C/min 200 (0 min) 8 C/min 280 (10 min) 2 min 1 mL 5183-4647
Two and a half fold 220/240 V 15 m 0.25 mm 0.25 mm HP-5MS (P/N 19091S-431) Splitless 5.74 psi 1.49 Constant pressure Helium 250 C 70 (0.8 min) 62.5 150 (0 min) 7.5 200 (0 min) 20 280 (4 min) 2 min 1 mL 5183-4647
Fourfold
18.0 psi 3.8
70 (0.5 min) 100 150 (0 min) 12 200 (0 min) 32 280 (2.5 min)
3 min Atune.u 35 amu 500 amu 150 2 3.15 150 C 230 C 280 C Scan (EI)
1.44 min Atune.u 35 amu 450 amu 250 2 3.50 150 C 230 C 280 C Scan (EI)
0.9 min
1 6.54
analysis. A faster oven ramp and the shorter column made the peaks narrower and higher, so an improvement in the signal-to-noise ratio is realized with the faster methods.
Abundance 7000000 6000000 5000000 4000000
TIC: RTLDEMO.D
1X
Conclusion
The highly accurate and reproducible pressure and temperature control of the Agilent 6890 GC allows precise scaling of the standard 42-min GC/MSD pesticide method. Run time was shortened to 10.5 minutes using a fast oven ramp rate and a 15-meter, 250-micron column. The combination of MTL and RTL facilitated scaling and yielded exact scaling. RTL libraries can be scaled accurately to correspond to the faster analyses.
3000000 2000000 1000000 0 Time 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00
Agilent HP-5MS, 30 m 0.25 mm 0.25 m

Abundance 18000000 16000000
TIC: 15MMIX1A.D
References 1. B. D. Quimby, L.M. Blumberg, M. S. Klee, and P. L. Wylie, "Precise Time-Scaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking," Application Note 228-401, Agilent publication number 5967-5820E, May 1998. 2. H. Prest, P. L. Wylie, K. Weiner, and D. Agnew, "Efficient Screening for Pesticides and Endocrine Disrupters Using the HP 6890/ 5973 GC/MSD System," Agilent publication number 5968-4884E, April 1999.
3. C. K. Meng and M. Szelewski, "Fast Screening of Pesticide and Endocrine Disrupters Using the Agilent 6890/5973N GC/MSD System", Agilent publication number 5968-9220, January 2000. 4. H. Prest, "GC Column Selection and Pumping Considerations for Electron and Chemical Ionization MSD operation," Agilent publication number 5968-7958E, November 1999.
14000000 12000000 10000000 8000000 6000000 4000000 2000000 Time 0 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00
2.5X
Abundance 18000000 TIC: 4X-MIX1A.D 16000000 14000000 12000000 10000000 8000000 6000000 4000000 2000000 0 Time 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00
4X
Agilent HP-5MS, 15 m 0.25 mm 0.25 m

Figure 2. The TICs of the 2.5X and 4X speedups. The standard analysis (1X) was 42 minutes long.
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Copyright 2000 Agilent Technologies, Inc. Printed in the USA 5/00 5980-1057E
Trace Level Pesticide Analysis by GC/MS Using Large-Volume Injection
Application
Gas Chromatography September 1997
Author
Philip L. Wylie Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
were injected into a single PTV liner. This application note includes recommendations for doing LVI using the PTV/6890/5973 GC/MSD system.
Introduction
More than 700 pesticides are registered for use in the world1 , and many more continue to persist in the environment, even though they are no longer being applied. For the protection of human health and the environment, pesticide residues are routinely monitored in food, water, soil, and tissue samples. "Acceptable" residue limits have been set for various foods and environmental samples by agencies such as the United States Environmental Protection Agency (U.S. EPA), the Codex Alimentarius Commission2 , and many other governmental organizations around the world. A great many methods have been developed to screen for pesticides in food3-7 and the environment8-10 to ensure that risks associated with pesticide use are minimized.
Abstract
Large-volume injection (LVI) using the Agilent programmable temperature vaporizing (PTV) inlet can improve gas chromatography system detection limits by one to two orders of magnitude over standard methods that call for 1- or 2- L injections. An Agilent 6890 Series gas chromatograph (GC), configured with a PTV inlet, a 6890 Series automatic liquid sampler (ALS), and an Agilent 5973 mass selective detector (MSD), was used for the analysis of pesticides in standards and several food extracts. By making 100- L injections, several pesticides could be identified by scanning gas chromatography/mass spectrometry (GC/MS) at the 100 ppt (100 ng/L) level. The PTV inlet tolerated dirty food extracts very well; more than 1,500 L of such samples
Recently, concern has increased that certain pesticides and other synthetic chemicals may be acting as pseudo hormones which disrupt the normal function of the endocrine system in wildlife and humans. Birth defects, behavioral changes, breast cancer, lowered sperm counts, and reduced intelligence are among the many disorders that have been blamed on these "endocrine disrupting" compounds, though much research must be done to verify these assertions. In 1996, Colborn, Domanoski, and Myers11 brought these issues into the public spotlight with the publication of their book Our Stolen Future. Recently, the United States Congress passed legislation calling for increased testing of suspected endocrine disrupters and monitoring their levels in food12 and water13 supplies. Because the endocrine system can be exquisitely sensitive to extremely low hormone concentrations, there is a need to measure concentrations of suspected endocrine disrupters (many of which are pesticides) at very low levels. Initiatives such as the Pesticide Data Program, developed by the United States Department of Agriculture14 , seek to
determine the lowest measurable pesticide levels in various foods to develop a total exposure model. Clearly, there is pressure to push pesticide detection limits to even lower levels than are routinely achieved today. Most residue measurements are made by gas chromatography using a variety of element-selective or mass spectral detectors (GC/MS). Therefore, to achieve lower detection limits, it is necessary to improve the detection limits of these GC methods. In GC, there are primarily four ways to improve method detection limits: 1) increase the concentration of analytes in a sample, usually by reducing the volume of an extract; 2) increase the sensitivity of the detector; 3) increase the selectivity of the detector to reduce chemical background "noise" or 4) increase the volume of sample injected. Because GC/ MS can be highly selective and extremely sensitive, it is often the method of choice for pesticide analysis and/or confirmation. However, for the reasons discussed above, there are occasions when even greater sensitivity is required. This application note describes a method for increasing GC/MS system detection limits by making large-volume injections (LVI) using Agilent's new programmable temperature vaporizing (PTV) inlet. Because this LVI technique is detector-independent, it is applicable to other GC configurations that may be used for pesticide residue analysis.
I, p,p'-DDE, propargite, iprodione, methoxychlor, and fenvalerate (mix of isomers I and II) to individual 20mL vials and diluting with 10.0 mL of acetone. Permethrin was obtained as a mixture of permethrin I and permethrin II comprising 32 percent and 27 percent of the sample, respectively, so 16.95 mg of this mixture was diluted with 10 mL of acetone giving a solution in which the combined permethrins represented 1 mg/mL. A stock mixture was prepared by adding 4 mL of the permethrin and fenvalerate solutions and 1 mL of each of the other stock solutions to a 100-mL volumetric flask and diluting to volume with acetone. The resultant solution contained 40 ng/ L each of the combined permethrin and fen-
valerate isomers and 10 ng/ L each of the other 12. This sample was diluted further with acetone to prepare standards that were analyzed by LVI. All these pesticides were obtained in neat form from Chem Service (West Chester, PA USA).
Extracts
Fruit and vegetable extracts were obtained from the Florida Department of Agriculture and Consumer Services (Tallahassee, FL USA). Commodities were extracted using a version of the Luke procedure15-17 that gave a final sample representing 1.75 g of the commodity per mL of extract.
Table 1.
Instrumentation and Conditions Used for Pesticide Samples

6890 Series GC 6890 Series ALS 5973 Series MSD PTV with CO2 cooling HP Vectra XU 6/200 G1701AA Version A.03.00 running Microsoft Windows 95 30 m x 0.25 mm x 0.25 m Agilent HP-5MS
GC/MS System Gas chromatograph Automatic liquid sampler Mass spectral detector Programmable temperature vaporizing inlet Computer for data acquisition and analysis Software Column Instrumental Conditions GC Parameters Carrier gas Inlet liner Syringe size Injection volume Injection delay Inlet temperature program Vent flow Purge flow to split vent Column head pressure Oven temperature program MSD Parameters Acquisition mode Temperatures
Experimental
Pesticide Standard Solution
Stock solutions of 14 pesticides were prepared at 1 mg/mL by adding 10 mg each of trifluralin, hexachlorobenzene, pentachloronitrobenzene, dichloran, chlorothalonil, chlorpyrifosmethyl, chlorpyrifos, endosulfan
Helium Prototype deactivated borosilicate with fritted glass on interior walls (part no. 5183-2041) 50 L 100 L (Inject 10 L 10 times) 12 sec 40 C (4.2 min), 200 C/min to 320 C (2 min) 400 mL/min Vent pressure 0.0 psi for 4.00 min 50.0 mL/min at 6.50 min 0 psi (4 min) then 17.3 psi (constant pressure) 50 C (6.13 min), 30 C/min to 150 C (2 min), 3 C/min to 205 C (0 min), 10 C/min to 250 C (20 min) Scan (35-550 amu) Transfer line = 280 C, MS quad = 150 C, MS source = 230 C
Instrumentation
Table 1 lists the instrumentation and chromatographic conditions used for LVI and GC/MS analysis of pesticide samples.
manual injections are usually impractical and good precision may be hard to achieve. The 6890 Series ALS is designed to make one or more injections of up to 25 L into the PTV inlet. After the desired number of injections has been made, the inlet is heated and the chromatography begins. Though the system controls allow up to 99 injections, a reasonable upper limit is about 10, making 250 L the typical injection volume limit for this system. For even larger injections, the controlled speed injector19 should be used. For all of the analyses described below, 100 L were injected by making 10 sequential injections of 10 L each.
How the PTV Works in the Solvent Vent Mode

Figure 1 shows a diagram of the PTV inlet. For large-volume injections, three steps are required. These are: 1) injection and solvent elimination; 2) splitless sample transfer to the GC column; and 3) chromatographic separation and, if desired, a simultaneous inlet bake-out step. The steps are described more completely below.
Brief PTV Tutorial

Before focusing on the PTV/GC/ MS analysis of pesticides, it is important to understand how the PTV inlet operates in the solvent vent mode for large-volume injections.
The PTV Inlet

The PTV inlet has the same basic functions as the split/ splitless inlet except that it is temperature programmable from -60 C (using CO2 cooling) or -160 C (using liquid N2 cooling) to 450 C at rates up to 720 C/min. However, the PTV's design has been optimized for its main uses-LVI and cold split/splitless injection. Although hot split and splitless injections may be made with or without a pressure pulse, care must be taken not to exceed the small internal volume of the PTV inlet. In practice, it is best to choose the Agilent split/splitless inlet for hot injections and the PTV inlet for LVI and cold split/ splitless techniques. Most GC pesticide methods call for injecting 1-2 L; splitless injection is used because it is compatible with dirty extracts of food, soil, or water. Pulsed splitless injection allows one to make injections of up to 5 L using standard equipment18. Enormous gains in system sensitivity can be realized by using the PTV inlet in the "solvent vent" mode, which is compatible with injections of 5-1,000 L. These large injections may be made manually or automatically using either a standard 6890 Series ALS in the multiple injection mode or by using a controlled speed injector available from Gerstel19. Because the injection process may take several minutes,
Injection and Solvent Elimination (Step 1)

During injection, the column head pressure is set to 0 psi to eliminate or, in the case of GC/MS, reduce the flow through the column. When mass spectral detection is used, there is still
Septumless Sampling Head Carrier Gas Line Coolant Liner Seal Heating Coil
Split/Splitless Solenoid Proportional Valve
Glass Inlet Liner
Capillary Column
Figure 1. The PTV inlet shown with the septumless head. The inlet is also available with a septum head that may be equipped with a standard septum or a Merlin Microseal. (Figure reproduced with permission of Gerstel GMBH.)
some flow because the column outlet is under vacuum. At the same time, a steady stream of carrier gas passes through the inlet and out through the split vent. This flow is typically between 100 and 500 mL/min. The sample is injected into the cool liner where it remains as a liquid, dispersed over the liner walls or any packing material that may be in the liner. The steady flow of carrier gas through the liner causes the solvent (and any volatile fraction of the sample) to evaporate and be swept with the carrier gas out through the split vent. This is analogous to "blowing down" a sample with a stream of inert gas, except that this takes place inside the PTV inlet. When most of the solvent has evaporated, the next injection is made and the evaporation process repeats, accumulating more sample in the inlet. To recover an analyte completely, its boiling point should be at least 100 C greater than that of the solvent; most pesticides fall into this category.
Normal
The timing of these multiple injections can be important. If the sample is introduced too rapidly, the liner may become flooded and liquid will be forced out through the split vent. Chromatographically, this shows up as reduced area counts for all analytes (see figure 2A). If there is too much time between injections, all of the solvent may evaporate and more of the volatile analyte fraction may be lost too. This results in poor recovery of volatiles but 100 percent recovery of the less volatile compounds (see figure 2B). Set-points such as inlet temperature, vent flow, and injection delay times can affect recovery of volatiles. Note that for 100 percent recovery, an analyte should have a boiling point at least 100 C greater than the solvent. One can adjust the delay between injections by entering the desired value in the ChemStation software. Some experimentation is usually necessary when setting this delay for a new method. It will be dependent upon such factors as the solvent type, injection volume, vent flow, and inlet temperature.
a splitless injection, except that instead of flash vaporization, the sample is transferred as the inlet temperature is programmed up. For the most gentle treatment of labile analytes, slow ramp rates may be used. This allows analytes to be flushed into the column at the minimum temperature needed for volatilization. When sample decomposition is not a problem, the inlet may be heated as fast as 720 C/min.
Chromatographic Separation (Step 3)

During sample transfer, the oven temperature is usually held between 30 C below and 20 C above the solvent's atmospheric boiling point, depending on whether the solvent effect is needed to focus the more volatile fraction of the analytes. Again, some experimentation is necessary to optimize peak shapes. After the sample has been transferred in step 2, the oven temperature is programmed up and chromatography begins. After the inlet has reached its maximum temperature and sufficient time has elapsed to transfer the sample to the column, a purge flow of 30-50 mL/min is restored to the split vent. If desired, one can set a very large split flow for a few minutes and bake out the inlet at a higher temperature to remove nonvolatile impurities. To conserve carrier gas, gas saver should be turned on at the end of this bake-out step.
Splitless Sample Transfer to the GC Column (Step 2)

Inlet Flooding
Once the desired number of injections has been made, the column head pressure is restored and the vent flow is tur ned off. At this point, the inlet temperature is programmed up to a value that is sufficient to transfer all of the desired analytes to the GC column. This step is similar to
A Sample is injected too rapidly
Normal
Volatiles Lost
B Solvent evaporates completely between injections
Figure 2. Chromatograms A and B illustrate the result of poor timing of multiple injections.
Entering PTV Inlet Parameters into the Agilent ChemStation

When preparing the PTV portion of a GC method, one should first decide on the sample size and how many injections are required. In this work, ten 10- L injections were made for a total of 100 L. When entering parameters into the ChemStation screen, the Injector icon is first selected (figure 3) under the "GC edit parameters" menu. Next, the Configure button is pressed to enter the syringe size and enable multiple injections. From the main injector screen, the injection volume (10 L) and number of injections are entered10 . For this work, a 12-second delay was chosen between injections to allow for solvent evaporation. The estimated total injection time is listed on the Inlets screen (figure 4). This is helpful when setting the inlet and oven parameters. First, the vent flow rate (400 mL/min for these analyses) is chosen, which sets the vent pressure to 0 psi until the injection sequence is done and solvent from the last injection has largely evaporated (4.00 min in figure 4). This is done by entering these values in the following fields: Vent Flow 400 mL/min Vent pressure 0.0 psi until 4.00 min Next, the purge flow and elapsed time are set by entering values in the following field: Purge Flow to Split Vent 50.0 mL/min @ 6.50 min Note that as an aid in setting up the method, the "estimated total injection time" is shown just above the previous data entry fields.
In this example, the normal column head pressure was restored and the vent flow was turned off at 4.00 min. This prepares the inlet for the splitless transfer of the sample to the column. The vent flow remained off until it was set to 50 mL/min at 6.5 min. Thus, there is a 2.5-min period for inlet temperature
programming and splitless sample transfer to the column. In this example, the inlet was held at 40 o C for 4.2 min, enough time to make 10 injections, turn off the purge flow, and restore the column head pressure; the PTV was then programmed to 320 o C at 200 o C/min (figure 4).
Figure 3. The injector screen from Agilent GC and GC/MS ChemStation software showing the setpoints available for multiple injections. To configure the sampler for multiple injections, set the syringe size, and choose slow injection, click on the Configure button.
Figure 4. The inlets screen from Agilent GC and GC/MS ChemStation software showing the setpoints available for operation of the PTV inlet in the solvent vent mode.
Although not done for these analyses, the inlet could be baked out by setting the "purge flow to split vent" to a large value (perhaps 500 mL/min) at the end of the splitless time (6.50 min) and at the same time, program the inlet to a higher temperature. After the bake-out period, the inlet temperature is programmed downward and gas saver is turned on. Normally, the GC oven is held at its starting temperature until the splitless injection is complete (6.50 min in this case) at which time oven temperature programming is begun. For this work, the oven temperature program was begun at 6.13 min so that the pesticide retention times would match a retention time data base that was in use. Figure 5 diagrams the PTV and GC oven setpoints used for this work.
PTV Inlet Liner Considerations

The correct liner choice is critical to the success of any pesticide analysis by PTV injection. The liner must be thoroughly deactivated or many labile pesticides may decompose or adsorb in the inlet. In general, any liner containing glass wool will be unsatisfactory for the analysis of labile pesticides, whether or not the glass wool is deactivated. At this time, two PTV liners are suggested for pesticide analysis: Part no. 5183-2037 is a deactivated, open multibaffled liner with no internal packing that may be used for single or multiple injections of 5 L or less. This liner gives very good recovery for pesticides, even extremely difficult ones such as acephate and methamidophos.
Part no. 5183-2041 is a deactivated liner with an internal coating of sintered glass to give it more surface area and is, therefore, suitable for single or multiple 25- L injections. This liner gives better than 70 percent recovery for most pesticides, although tests have shown that acephate and methamidophos cannot be analyzed using this liner, and that recoveries of guthion are often less than 50 percent. A prototype version of this liner was used for all of the work described in this application note.
Multiple injections
12 sec injection delay
PTV purge flow 400 mL/min 4.00 min 0 psi 4.00 min 40 C 4.20 min 50 C 6.13 min
Column head pressure PTV temperature Oven temperature
Figure 5. Illustration of the GC and sampler setpoints used for 100- L injections of pesticide samples. Note that normally, the GC oven hold period would have been at least 6.5 min for this method. A value of 6.13 min pesticide retention times to a data base.
}
n 0 mL/min 280 C 200 C/min 50 mL/min 6.50 min 30 C/min
6

When compared to a typical 2-L splitless injection, 100- L PTV injections can often result in a 50-fold improvement in system detection limits. Selective detectors such as the MSD can help the analyst to realize the full measure of this sensitivity improvement by excluding background that may be introduced from solvent impurities, vial cap extract, and indigenous compounds coextracted with the analytes. In this application, it was possible to see most of the pesticides in the 14-component mixture at 100 ppt in the scan mode (400 ppt for the isomer mixes of permethrin and fenvalerate). Figure 6 shows extracted ion chromatograms for trifluralin and hexachlorobenzene (HCB) at 100 ppt. Library searching gave a match quality of 93 for the HCB peak. Fenvalerate isomers I and II were found in the solution in a ratio of about 78:22. Figure 7 shows extracted ion chromatograms for fenvalerate I at a concentration of 311 ppt.
Trifluralin (100 ppt) m/z 306
m/z 264
A Extracted ion current chromatograms of trifluralin
Hexachlorobenzene (100 ppt) Extracted ions 284, 286, and 282 Match quality = 93
B Extracted ion current chromatogram of HCB with its mass spectrum and library match
Figure 6. Scanning GC/MS results for a pesticide standard containing Trifluralin and Hexachlorobenzene at 100 ppt. (Ten 10- L injections were made using the PTV inlet.)
Fenvalerate I (311 ppt) m/z 167
m/z 125
m/z 225
Figure 7. Extracted ion current chromatograms of Fenvalerate I at a concentration of 311 ppt in a pesticide standard. (Ten 10- L injections were made using the PTV inlet.)
Analysis of a bell pepper extract revealed several pesticide residues. As seen in figure 8, chlorpyrifos and the endosulfans were easily detected. The Florida Department of Agriculture determined the concentration of chlorpyrifos, alpha-endosulfan, betaendosulfan, and endosulfansulfate to be 0.210, 0.011, 0.018, and 0.013 ppm, respectively. It is important to note that these compounds could be detected with very high selectivity by extracting high mass ions that are characteristic of these pesticides but not of the matrix. Using LVI, there is ample signal from these less abundant ions for good quantitation. With normal injection volumes, selectivity may have to be compromised and the most abundant ions extracted in a pesticide spectrum to gain sensitivity. Phosmet, captan, and propoxur were all easily detected in a pear sample. The total ion current chromatogram (TIC) is shown in figure 9 along with spectrum obtained for captan juxtaposed with the library spectrum. Figure 10 shows the propoxur peak along with 2,4,6-tribromoanisole and 2,4,6-tribromophenol, two other compounds that were surprising to find in a pear sample. Though the origin of these brominated compounds is not known, a recent paper by Hoffmann and Sponholz 20 suggests that tribromophenol is used to treat storage palettes for the prevention of fire and mold growth, and that the anisole is formed from the phenol microbiologically. Perhaps these pears were shipped in containers that had been similarly treated.
Figure 8. GC/MS Analysis of a bell pepper extract. (Ten 10- L injections were made using the PTV inlet.) Using LVI, there was sufficient signal to use high mass ions with smaller abundances to achieve greater selectivity.
Figure 9. TIC of a pear extract resulting from a 100- L Injection (10 x 10 L). Captan was easily detected, and its spectrum gave a library match quality of 96.
Figure 10. TIC of a pear extract resulting from a 100- L Injection (10 x 10 L). Propoxur and two brominated phenolics were easily identified.
A single sintered glass coated liner of the type described above (part no. 5183-2041) was used for about ten 50- and ten 100- L injections (ca. 1,500 L total) of vegetable extracts before it was replaced. All of the extracts were rather dirty, and an inlet bake-out step was not used. Although the liner looked somewhat discolored for about 2 cm where injections were made, it still performed well at the time it was replaced.
Conclusion
Using the PTV inlet in the solvent vent mode, it is relatively simple to increase system detection limits by one or two orders of magnitude. When combined with the Agilent 6890 Series automatic liquid sampler,
multiple injections of up to 25 L each into the inlet can be made, allowing the solvent to vent while pesticides and other less volatile analytes accumulate. After the desired sample volume has been introduced (typically 5-250 L), the solvent vent is closed and the sample is transferred to the column in a temperature-programmed splitless injection. By making 100- L injections into a PTV-equipped Agilent 6890 Series GC coupled to the Agilent 5973 MSD, it was possible to see several pesticides at the 100 ng/L level (100 ppt) in the scan mode. With such low detection limits, less abundant ions can be used to identify and quantitate pesticides at low ppb levels, thereby gaining in selectivity as well.
When performing LVI, there are several parameters to adjust and some method development time is usually required. However, the method described herein worked well and can be duplicated for the PTV/GC/MS analysis of pesticides in food.
Acknowledgment
The author wishes to thank Ms. Joanne Cook of the Florida Department of Agriculture and Consumer Services for supplying the food extracts used in these experiments and Dr. Bill Wilson (Agilent Technologies) for supplying liner deactivation test results.
References
1. Tomlin, Clive, ed (1994), The Pesticide Manual, Tenth Edition, British Crop Protection Council, Surry, UK. 2. Miller, R. W., This is Codex Alimentarius, Secretariat of the Joint FAO/WHO Food Standards Programme, Food and Agriculture Organization of the United Nations, Rome. 3. McMahon, B. M. and Hardin, N. F. eds. (1994), Pesticide Analytical Manual, Vol I, Third Edition, U.S. Food and Drug Administration, Washington, DC. 4. Lee, S. M., Papathakis, M. L., Feng, H.-M. C., Hunter, G. G., and Carr, J. E. (1991), Fresenius' A. Anal Chem 339, 376-383. 5 Fillion, J., Hindle, R., Lacroix, M., and Selwyn, J. (1995), J AOAC Int 78, 1252-1266.
7. Luke, M. A., Froberg, J. E., Doose, G. M., Masumoto, H. T. (1981), J Assoc Off Anal Chem 64, 1187-1195. 8. Stan, H. J., ed. (1995), Analysis of Pesticides in Ground and Surface Water II, SpringerVerlag, Berlin, Germany. 9. Wagner, R. E., Kotas, W., and Yogis, G. A., eds. (1994), Guide to Environmental Analytical Methods, 2nd Edition Genium, Schenectady, NY. 10. U.S. Environmental Protection Agency, Test Methods for Evaluating Solid Waste, SW-846, Draft Method 8085: Pesticides by GC/AED. 11. Colborn, T., Dumanoski, D., and Myers, J. P. (1996), Our Stolen Future, Penguin, New York, NY. 12. Food Quality Protection Act of 1996, Public Law 104-170, Congressional Record pp. H8127-H8141. 13. Safe Drinking Water Act Amendments of 1996, Public Law 104-182, Congressional Record pp. H9680-H9700.
14. Pesticide Data Program Annual Summary Calendar Year 1994, U.S. Department of Agriculture, Agricultural Marketing Service, Washington, DC. 15. Luke, M. A., Froberg, J. E., Doose, G. M., and Masumoto, H. T. (1981), J Assoc Off Anal Chem 64, 1187-1195. 16. Luke, M. A., and Doose, G. M. (1983), Bull Environ Contamin Toxicol 30, 110-116. 17. Sawyer, L. D. (1985), J Assoc Off Anal Chem 68, 64-71. 18. Wylie, P. L., Phillips, R. J., Klein, K. J., Thompson, M. Q., and Hermann, B. W. (1991), J High Resol Chromatog 14, 649-655. 19. The controlled speed injector is available from Gerstel US, 1510 Caton Center Dr., Baltimore, MD 21227 USA. 20. Hoffmann, A. and Sponholz, W. (March 1997), American Laboratory, 22-23.
6. Working Group on Development and Improvement of Residueanalytical Methods (1996), Analytical Methods for Pesticide Residues in Food-stuffs, General Inspectorate for Health Protection, Ministry of Public Health, Welfare & Sport, The Netherlands.
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Microsoft is a U.S. registered trademark and Windows is a U.S. trademark of Microsoft Corporation. HP is a registered trademark of Hewlett-Packard Company. Copyright 2000 Agilent Technologies, Inc. Printed in the USA 4/2000 5966-1214E
Fast Screening of Pesticide and Endocrine Disrupters Using the Agilent 6890/5973N GC/MSD System, Part I
Application
Gas Chromatography January 2000 retention times and characteristic ions for 567 of the most common pesticides and endocrine disrupters of concern worldwide. A GC/MSD method was developed based on the standard 42-min method1 to screen for all 567 of the most common analytes. A specific combination of column stationary phase, carrier gas flow rate, and oven temperature programming is required to lock all the compounds to an expected retention timetable2. Compound identification based only on spectral searching alone is difficult when analyzing extracts containing significant sample matrix content because of overlapping peaks and noisy baselines. The new screening tool, integrated within Agilents ChemStation for MSD, searches for all 567 compounds by first checking and integrating four characteristic ions within the expected time window, and second by printing out a report showing hits and possible hits (ratios of characteristic ions that do not match the expected values in the library within specified limits). In one application, the analysis time of the standard pesticide method was reduced by one half, two-thirds, and three-fourths. The faster methods were scaled exactly as predicted by using a combination of Agilents method translation (MTL) and RTL software. Because scaling was exact, these faster methods can be used with precisely-scaled pesticide libraries, making the screening process even more powerful and adaptable to individual needs.
Authors C. Kai Meng Agilent Technologies 2850 Centerville Road Wilmington, Delaware 19808-1610 USA Michael Szelewski Agilent Technologies 2850 Centerville Road Wilmington, Delaware 19808-1610 USA
Abstract Agilent Technologies new, fast GC/MSD method can significantly speed up the screening of pesticides. Agilents GC method translation software (available free from the Agilent Technologies Web site, http://www. chem. agilent.com/cag/ servsup/usersoft/main.html#mxlator) was used in developing the new method based on the standard 42-min method. A 10 m x 0.1 mm x 0.1 m HP-5 column was used to increase analysis speed up to fourfold. The time savings were implemented in increments (down to 10.5 minutes) to verify the predictability of scaling and the effect of scaling on the signal-tonoise ratio. Key Words RTL, pesticide, environmental, screening, fast GC, method translation, 5973, 6890, MTL Introduction Analysts want faster analyses to improve laboratory productivity. Often, when speeding up GC methods, an analyst will trade resolution for increased analysis speed. This loss of resolution can complicate peak identification, even with a mass selective detector (MSD). Agilent Technologies has developed new techniques to solve the peak identification problem based on Agilents retention time locking (RTL) software and a new mass spectral library that contains the locked
Experimental The GC method translation software tool was used to find operating conditions for the faster methods. Figure 1 is a screen capture of MTL software data entry showing the original conditions and the new chromatographic conditions for a twofold speed gain. The column flow rate, which is helpful to avoid exceeding MSD pumping capacity3, is also found in the table. A 16:1 split ratio was suggested in the table as a proportional scaling from the original column to the smaller i.d. column with corresponding lower capacity. The program also determined the required column head pressure and corresponding oven ramp. The Agilent 6890 GC fast oven option (220/240V in the U.S.) was required for the faster oven ramp used in this study.
Figure 1. Screen capture showing the method translation (MTL) software data entry used in a twofold speed gain translation.
General chromatographic conditions are listed in table 1. The standard used was a mixture of 26 pesticides at 10 ppm. A 10 m x 0.1 mm x 0.1 m HP-5 column (part number 19091J141) was used. The head pressure determined by the method translation software (30.72 psi) was used as the starting point for retention time locking. The column head pressure required to lock retention times of the compounds to the library (the original retention time divided by 2) was determined using the automated RTL process integrated within the Agilent ChemStation for MSD. This process (first translate the method then lock the retention times) was repeated for the threefold and fourfold time reductions.
Table 1. Chromatographic Conditions Speed GC Column Injection mode Column head pressure Column flow (mL/min) Inlet control mode Carrier gas Injector temperature Oven temperature Ramp 1 Ramp 2 Ramp 3 Oven equilibration Injection volume Liner MS Conditions Solvent delay Tune file Low mass High mass Threshold Sampling Scans/sec Quad temperature Source temperature Transfer line temperature Acquisition mode Onefold (1X) 110 V 30 m x 0.25 mm x 0.25 m HP5-MS (P/N 19091S-433) Splitless 18.0 psi 1.5 Constant pressure Helium 250 C 70 (2 min) 25 C/min 150 (0 min) 3 C/min 200 (0 min) 8 C/min 280 (10 min) 2 min 1 L 5183-4647 Twofold (2X) Threefold (3X) 220/240 V 10 m x 0.1 mm x 0.1 m HP-5 (P/N 19091J-141) 16:1 split 36.55 psi 63.17 psi 0.4 0.8 Constant pressure Helium 250 C 70 (1 min) 70 (0.67 min) 50 75 150 (0 min) 150 (0 min) 6 9 200 (0 min) 200 (0 min) 16 24 280 (5 min) 280 (3.33 min) 2 min 1 L 5183-4647 Fourfold (4X)
90.0 psi 1.5
70 (0.5 min) 100 150 (0 min) 12 200 (0 min) 32 280 (2.5 min)
3 min Atune.u 35 amu 500 amu 150 2 3.15 150 C 230 C 280 C Scan (EI)
1.8 min Atune.u 35 amu 450 amu 250 2 3.50 150 C 230 C 280 C Scan (EI)
1.2 min
0.9 min
1 6.54
1 6.54
Figure 2 shows the results of the shortened analysis times. The three chromatograms look extremely similar, except that the time axis is scaled proportionally. Because MTL followed by RTL scales methods very precisely, scaled screening libraries for corresponding time reductions can be obtained by dividing the retention times in the library by the speed gain (which does not have to be an integer). The peak heights from all the methods are very similar. Although the sample was split 16:1 for the smaller column, the small column i.d. and faster oven ramp combination made the peaks narrower and higher, so there was minimal loss in the signal to noise ratio. Conclusion The highly accurate and reproducible pressure and temperature control of the Agilent 6890 GC allows precise scaling of the standard 42-min GC/MSD pesticide method. Run time was shortened to 10.5 minutes using a fast oven ramp rate and a 10-meter 100-micron column. The combination of MTL and RTL facilitated scaling and yielded exact scaling. RTL libraries can accurately be scaled to correspond to the faster analyses. References 1. B. D. Quimby, L.M. Blumberg, M. S. Klee, and P. L. Wylie, Precise TimeScaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking, Application Note 228-401, Agilent publication number 5967-5820E, May 1998. 2. H. Prest, P. L. Wylie, K. Weiner, and D. Agnew, Efficient Screening for Pesticides and Endocrine Disrupters Using the HP 6890/ 5973 GC/MSD System, Agilent publication number 5968-4884E, April 1999. 3. H. Prest, GC Column Selection and Pumping Considerations for Electron and Chemical Ionization MSD operation, Agilent publication Number 5968-7958E, November 1999.
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Copyright 2000 Agilent Technologies Printed in USA 2/00 5968-9220E
Abundance 2,400,000 2,200,000 2,000,000 1,800,000 1,600,000 1,400,000 1,200,000 1,000,000 800,000 600,000 400,000 200,000 Time 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00
2X Speed
18 min
Abundance 6,000,000 5,500,000 5,000,000 4,500,000 4,000,000 3,500,000 3,000,000 2,500,000 2,000,000 1,500,000 1,000,000 500,000 Time 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00
3X Speed
Abundance 6,000,000 5,500,000 5,000,000 4,500,000 4,000,000 3,500,000 3,000,000 2,500,000 2,000,000 1,500,000 1,000,000 500,000 Time 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
12 min
4X Speed
9.00
10.00
9 min
Figure 2. Three TICs of the 2X, 3X, and 4X speedups. The standard analysis (1X) was 42 minutes long. The two vertical lines on the figure are used as references to show the similarity of the TICs.
LC/MS and LC/MS/MS

Application Notes
Pesticide Analysis using an Agilent 1290 Infinity LC System with an Agilent 6140 Single Quadrupole LC/MS System
Application Note
Environmental
Author
Rob Sample Littleover Derby, DE23 3XH, UK Edgar Naegele Agilent Technologies Waldbronn, Germany
Abstract
This Application Note describes the use of the Agilent 1290 Infinity LC system in combination with the Agilent 6140 Single Quadrupole LC/MS system for cost-effective LC/MS analysis of pesticides in clean to moderately dirty water samples such as river water or drinking water. A separation method for the analysis of water samples with relatively clean matrices such as those that might be prepared by a simple concentration and cleanup step using a solid phase extraction (SPE) cartridge was developed. This demonstrates how the run time for such a method can be shortened.
Introduction
In pesticide analysis literature, there are many examples of liquid chromatography combined with triple quadrupole mass spectrometry. While these techniques are extremely sensitive for detecting pesticides in dirty matrices such as soil or crop extracts the single quadrupole MS detector is a more cost-effective method for the analysis of clean to moderately dirty samples such as drinking water or river water. Addition of a single quadrupole MS detector to a UV diode array detector-based system is a common first step into mass spectrometric detection. This Application Note develops a separation method for the analysis of water samples with relatively clean matrices, prepared by a simple concentration and cleanup step using a solid phase extraction (SPE) cartridge. It also demonstrates how the run time for this method can be shortened. Simple techniques for converting methods to faster methods are illustrated.
Experimental
All analyses were performed using the Agilent 1290 Infinity LC system including binary pump, autosampler, thermostatted column compartment and diode array detector in series with the Agilent 6140 Single Quadrupole LC/MS system equipped with an atmospheric pressure electrospray ionization (APESI) source. Agilent ZORBAX Rapid Resolution High Definition (RRHD), 1.8 m columns were used with the Agilent 1290 Infinity LC system. The column dimensions were chosen to handle operating pressures up to 1200 bar. For the initial method development, an Agilent ZORBAX StableBond C18 RRHD 100 mm 2.1 mm, 1.8 m column was chosen.
Component Atrazinedesethyl
m/z [M+H]+ 188
Structure
NH N Cl N N NH2 N N CH3 Cl N H CH3 H H2C N N H3C N N Cl N H CH3 N N O N O Cl Cl C N C H CH3 CH2 CH2
Component Methabenzthiazuron (also known as methibenzuron)
m/z [M+H]+ 222
Structure
O S N O N C N CH3 H
CH3
Atrazine
216
Cl N
Metobromuron
259
CH3 N O CH3
N H Cl
Br
H N CH
Chlorotoluron
213
CH3 N
CH3 O C
Metolachlor
284
Cl CH2 C H3C
CH3
CH3
N CH CH2 O CH2 CH3
CH3
Cyanazine
241
H3C
Metoxuron
229
CH3 N
O N N
Cl O CH3
CH3 H O
Diuron
233
CH3 N
O C
Monolinuron
215
CH3 N O CH3 Cl
N H H
Cl
Samples
A standard mixture containing 10 ng/L each of 17 pesticides in acetonitrile solution was obtained from Dr. Ehrenstorfer GmbH (Pesticide Mix 44, part no. 18000044 - Dr. Ehrenstorfer GmbH Bgm.-Schlosser-Str. 6 A, 86199 Augsburg, Germany). Aliquots of this solution were diluted to 10% with water to prepare standards containing 1 ng/L (1 ppm) of each component for development of the separation method. The mixture contained herbicides from the triazine, phenylurea and chloroacetanilide classes. The components are listed in Table 1. These components are all expected to form positive ions under the acidic conditions of the mobile phase.
Hexazinone 253
CH3 CH3 H3C N
N N H N N
N CH2 CH3
Sebuthylazine
230
CH CH2 CH3 CH3
Isoproturon
207
CH3 N
O C N CH CH3 CH3 Cl N H O Cl CH2 H3C C N CH2 N CH3 N Cl
Simazine
202
H Cl N H N N N CH2 CH3 H Cl N H N N N CH2 CH3 CH3 CH3 CH3 N C N CH2 CH3
CH3 H
Linuron
249
CH3 N O CH3
O C
Terbuthylazine
230
Metazachlor
278
Table 1 Components in pesticide test mixture.

Standard method development
An Agilent ZORBAX StableBond C18 RRHD 100 mm 2.1 mm, 1.8 m column was chosen for the separation method development. Conditions were adjusted to obtain a target analysis of the 17 components in about 10 minutes. Formic acid (0.1% v/v) was added to the acetonitrile/water mobile phase
to promote positive ion formation for mass spectrometric detection using the atmospheric pressure electrospray source. Fifteen of the seventeen components were separated in 10 minutes with the remaining two (diuron and isoproturon), coeluting at about 5.8 minutes. An advantage of using the mass spectro2
metric detector is that two overlapping peaks may be separated by their mass/charge (m/z) ratios into different chromatograms. The maximum pressure observed during the gradient separation was 546 bar, which is well within the operating range of 1200 bar. Table 2 shows the peak assignments and Table 3 summarizes the LC/MS method conditions used in this Application Note.
The Agilent 6140 Single Quadrupole LC/MS system can output four chromatographic signals simultaneously (MSD1 to MSD4). MSD1 was set to scan mode so that the mass spectrum of each peak in the resulting total ion current (TIC) chromatogram (Figure 1), could be used to verify that molecular ions ([M+H]+) were produced for each component in order to identify and quantify the peaks. High quality spectra were obtained for all components and two examples are shown in Figures 4a and 4b. For quantification of components, the use of TICs from selected ion monitoring (SIM) mode yields higher selectivity and much higher sensitivity than in scan mode because the mass detector spends more time in each measurement cycle collecting ions at a specified mass/charge (m/z) ratio. A signal in SIM mode can be used to monitor a number of discrete masses for greater flexibility. The coeluting peaks can be separated by appropriate use of SIM signals or by extracted ion chromatograms (EIC) for individual ion m/z values. In this case, m/z 207 and 233 signify isoproturon and diuron (Figure 2). The expanded section of the TIC chromatogram overlaid with the extracted ion chromatograms in Figure 3 allows closer examination of the coeluting compounds. The selected mass spectra of diuron and isoproturon are shown in Figures 4a and 4b. The spectrum of diuron (Figure 4a) shows a good resolution of the isotope pattern of this doubly chlorinated compound and allows clear differentiation from the isotopic pattern of the non-chlorinated coeluting compound isoproturon (Figure 4b).
105 25 20 15 10 5 0 2 4 6 1.789 3.541 3.321 3.110 3.877 4.989 5.132 5.333 5.890 6.777 5.615 6.201 7.077
7.772
9.313
7.936
min
Figure 1 Chromatogram of the pesticide mixture.
104 20 15 10 5 0 105 0 10 6 2 4 EIC 233 Diuron
5.847
6 5.893
min
EIC 207 Isoproturon 2 0 2 4 6 8 min
Figure 2 Extracted ion chromatograms EIC 233 and EIC 207 for the pesticide mixture.
106 2 1.5 1 0.5 0 5.5 5.6 5.7 5.8 5.9 6 6.1 6.2 min 5.890 5.615 5.893
Figure 3 Detail view of coeluting diuron and isoproturon peaks showing the scan TIC and extracted ion chromatograms overlaid.
5.847
6.201
100 80 60 40 20 0 150
m/z = 233 Diuron
233.0 235.0
Peak Number Component 1 2 3 4 5 6 7 8 9 10* 11* 12 13 14 15 16 17 Atrazine-desethyl Metoxuron Hexazinone Simazine Cyanazine Chlorotoluron Atrazine Monolinuron Diuron Isoproturon Metobromuron Metazachlor Sebuthylazine Terbuthylazine Linuron Metolachlor
Retention (mins) 1.791 3.111 3.323 3.542 3.878 5.133 5.336 5.616 5.847 5.893 6.203 6.777 7.080 7.773 7.936 9.313
m/z [M+H]+ 188 229 253 202 241 222 213 216 215 233 207 259 278 230 230 249 284
Max: 94376
A
207.1 216.1 236.0 255.0
Methabenzthiazuron 4.990
175
200
225
250
275
m/z
Figure 4a Mass Spectrum of Diuron.

100 80 60 m/z = 207 Isoproturon
207.1
Max: 1.25235e+006
20 0 160 180 200
208.1
40
220
240
260
280
m/z
Figure 4b Mass Spectrum of Isoproturon. Agilent 1290 Infinity LC/UV Conditions Column 1 Column 2 Column Temperature Mobile Phase A Mobile Phase B Chromatogram shown in: Column Length Flow Rate Mobile Phase Gradient: 20% B 40% B 70% B Injection Volume Injection Needle Wash Diode-array Detector Signal A Diode-array Detector Signal B Spectrograph Slit Width Ion source Ion polarity Capillary Voltage Drying gas flow Drying gas temperature Nebulizer pressure MSD Signal 1 MSD Signal 2 MSD Signal 3 Fragmentor voltage Agilent ZORBAX RRHD SB-C18 2.1 mm 100 mm, 1.8 m Agilent ZORBAX RRHD SB-C18 2.1 mm 50 mm, 1.8 m 40 C Water + 0.1% formic acid Acetonitrile + 0.1% formic acid Figure 1 100 mm 0.5 mL/min 0.00 min 6.50 min 10.00 min 10 L Figure 5 100 mm 1.0 mL/min 0.00 min 3.25 min 5.00 min 10 L Figure 7 50 mm 0.5 mL/min 0.00 min 3.25 min 5.00 min 5 L Figure 8 50 mm 1.0 mL/min 0.00 min 1.63 min 2.50 min 5 L
* Diuron and isoproturon coelute Table 2 Peak assignments.
Development of a fast method using a 100 mm column

Agilent ZORBAX C18 RRHD 1.8 m stationary phase material offers high efficiency separations and supports faster flow rates due to the flatter Van Deemter profile associated with the small particle size. The 10-min method was converted to a 5-min method by doubling the flow rate to 1.0 mL/min and adjusting the gradient time (halving it) to keep the same gradient steepness in terms of the number of column volumes passing through the column during the gradient (Figure 5).
In Flush Port, 10 s, acetonitrile/water (50/50) 242 nm, bandwidth 4 nm. Reference Off. 226 nm, bandwidth 4 nm. Reference Off 4 nm Atmospheric pressure electrospray (API-ESI) Positive 4000 V 12 L/min 350 C 30 psi @ 0.5 mL/min; 50 psi @ 1.0 mL/min Scan m/z 150 350 SIM m/z 188, 202, 207, 216, 222, 229, 249, 278 (5 ms dwell each) SIM m/z 213, 215, 230, 233, 241, 253, 259, 284 (5 ms dwell each) 90 V, all signals
Agilent 6140 Single Quadrupole MS Detector Conditions
Table 3 Agilent 1290 Infinity LC/UV conditions.
2.510 2.561 2.649 2.759 2.938 3.051
1.706 1.777 1.930
0.922
1.578
3.354
3.905
4.598
The relationship between gradient steepness and flow is shown in the following equation
3.504
106 2.5 2 1.5 1 0.5 106 5 4 3 2 1 0 106 6 4

1.578 1.707 2.560
3.839
0.5
0.923
1.5
1.778
2.5
2.648 2.511
3
2.938
3.5
4.5
min
where: %B is the range of the stronger mobile phase component across the gradient VM is the delay volume of the column, or the volume of mobile phase in the column F is the flow rate tG is the time range (duration) of the gradient. This is linked to the concept of gradient retention factor, k* in the following equation, and illustrates how a constant gradient steepness keeps the relative spacing of the peaks constant:
1.578
3.356
2.757
3.906
0.5
1.5
2.5
3.5
3.506 3.840
4.5
min
1.931
2 0 0 0.5 1
2.759 2.910 3.052
4.599
1.5
2.5
3.5
4.5
min
Figure 5 Pesticide mixture separated on 100 mm column in a 5 min gradient at 1 mL/min flow rate.
S is a compound-specific factor for each solute. As expected, the overall appearance of the chromatogram is the same but the retention times are half of the original. Closer examination reveals some slight selectivity changes. The diuron and isoproturon coelution peak at 2.9 minutes shows a distinct shoulder indicating slightly more separation between these two compounds. Peaks 6 and 7 now slightly overlap (compare to Figure 1). Because the gradient slope remained constant, it is concluded that frictional heating is generated from the flow rate and pressure. The separation was run at temperatures ranging from 30 to 45 C to test the effect of temperature variation. As expected, the general
trend was to reduce retention as temperature increased. It can also be seen that the selectivity between various peak pairs is temperature sensitive. The following changes in selectivity were noted as the temperature increased: methabenzthiazuron and chlortoluron (peaks 6 and 7) decreased; chlortoluron and atrazine (peaks 7 and 8) increased; diuron and isoproturon (peaks 10 and 11) increased; and terbuthylazine and linuron (peaks 15 and 16) decreased. The selectivity change observed in changing the flow rate and pressure may be indicated by a frictional heating effect. This can be estimated by empirical comparison at 1 to 2 C. In this case the effect could be offset by lowering the thermostat temperature by 1-2 C as reported elsewhere1. Frictional heating can cause band broadening if a radial temperature gradient is created in the column. This is influenced by the type of thermostatting employed in the system. The design of
the 1290 Infinity LC column compartment takes care to avoid this. Therefore no extra dispersion was indicated by the observed peak widths. Increasing the temperature, which reduces the viscosity of the mobile phase causes an observed difference in pressure readings (for example, 1110 bar at 30 C and 920 bar at 45 C). This effect is often used to create more headroom for pressure and flow rate increases (Figure 6).
Development of a fast method using a 50 mm column

Agilent ZORBAX C18 RRHD 1.8 m columns offer high efficiency separations. Examination of the 100 mm column resolution data suggests that the separation is also adequate on a 50 mm column with the same packing material. Transferring a method from a 100 mm column to a 50 mm column follows the same approach as discussed above. Because the stationary phase is identical, the separation should be the same as long as the gradient is the same. The flow rate remains the same and because the column is half the length the peaks will elute in half the time. To maintain gradient slope, the time steps (tG) in the timetable are divided by 2 because the column volume (VM) is divided by 2. The 10-min method on the 100 mm column is reduced to a 5-min method on the 50 mm column. The resulting chromatogram is shown in Figure 7. It is similar to the chromatogram shown in Figure 1 for the 100 mm column. It can be expected that the efficiency of the column measured in the number of theoretical plates (N) under isocratic conditions will be reduced by a factor of 2 for the 50 mm column compared to the 100 mm column. The effect on resolution is a factor of about 1.5 and it can be seen from the experimental results in Table 4 that peak width, resolution and peak capacity number are reduced by this factor. The most critical resolution for peaks separated in the scan chromatogram is between peaks 14 and 15 with a resolution of 1.2. However, the two peaks are totally separated by use of the SIM signals. The maximum separation pressure reached in the separation using the 50 mm column was 290 bar compared with 550 bar for the 100 mm column. Although 290 bar is attainable on a conventional (400 bar) HPLC system such as the Agilent 1200 HPLC system, the increased system delay volume intro-
106 2 1 0 106 2 1 0 106 2 1 0 106 2 1 0 45 C 0.5 1 1.5 2 2.5 3 3.5 4 4.5 min 40 C 0.5 1 1.5 2 2.5 3 3.5 4 4.5 min 35 C 0.5 1 1.5 2 2.5 3 3.5 4 4.5 min 30 C 0.5 1 1.5 2 2.5 3 3.5 4 4.5 min
Figure 6 Effect of temperature on the separation: Chromatograms at 30C (pressure, P = 1110 bar), 35C (P = 1045 bar), 40C (P = 970 bar), and 45C (P = 920 bar).
2.462 2.538 2.658 2.796 2.906
1.534 1.627 1.768
3.350
3.518
3.865
0.902
1.926
1.5 1 0.5 0 106 0.5
3.086
3.939
4.624
106
1.5
1.769
2.5
2.659
3.5
4.5
min
0.902
3.350
2 1 0 0 106 0.5
1.535
2.462
2.910
3.940
1.5
2.5
3.5
4.5
min
1.628
2.538
3.518
3 2 1 0 0 0.5 1
3.865
1.925
2.799 2.892
1.535
3.086
4.625
1.5
2.5
3.5
4.5
min
Figure 7 Pesticide mixture separated on 50 mm column in a 5 min gradient at 0.5 mL/min flow rate.
duces a longer isocratic delay at the start of the gradient. In addition, increased dispersion in the larger flow cell reduces resolution. As with the 100 mm column, the same technique of increasing the flow rate and reducing the gradient times to maintain gradient slope was employed. Figure 8 shows the chromatogram at 1 mL/min and 2.5 minutes gradient for the 50 mm column with an observed maximum pressure of 550 bar. While the scan TIC chromatogram shows some loss of resolution, the SIM chromatograms still give good separation for all peaks. The maximum pressure (1200 bar) of the Agilent 1290 Infinity LC allows the flow rate to be increased by a factor of two. The electrospray source has a maximum flow rate of 1 mL/min, however, so additionial increases require the use of a flow splitter to maintain good MS detection.
Peak 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
100 mm column, 10min gradient, 0.5 ml/min Ret Resol- Peak Time Width ution Capacity 1.789 3.110 3.321 3.541 3.877 4.989 5.132 5.333 5.615 5.890 6.201 6.777 7.077 7.772 7.936 9.313 0.043 0.046 0.046 0.053 0.050 0.053 0.051 0.054 0.050 0.064 0.049 0.065 0.062 0.055 0.051 0.047 18.0 2.7 2.6 3.8 12.7 1.6 2.2 3.0 2.7 3.0 5.7 2.8 7.0 1.8 16.3 173 163 162 142 150 141 147 138 152 117 154 116 121 136 146 159
50 mm column, 5 min gradient, 0.5 ml/min Ret Resol- Peak Time Width ution Capacity 0.902 1.534 1.627 1.768 1.926 2.462 2.538 2.658 2.796 2.906 3.086 3.350 3.518 3.865 3.939 4.624 0.030 0.033 0.033 0.037 0.035 0.036 0.035 0.038 0.040 0.044 0.037 0.043 0.041 0.037 0.034 0.033 12.0 1.7 2.4 2.6 9.0 1.3 1.9 2.1 1.6 2.6 3.9 2.3 5.2 1.2 11.8 123 114 112 99 107 102 107 99 92 85 101 86 91 101 109 113
Table 4 Comparison of peak parameters on 100 mm and 50 mm columns.
1.341
1.935
1.249 1.277
0.472
4 0 106 3
0.473 0.786
1.534
1.390
1.678
12
0.785 0.845 0.901 0.971
1.465
2.315
105
Ar Are 1.953 ea: 1 .77 a: 4 886 272 e+ 41 0
1.766
06
0.25
0.5
0.75
0.901
1.25
1.340
1.5
1.467
1.75
2.25
min
1.250
2 1 0 106 4 3 2 1 0 0 0.25 0 0.25
1.678
1.387
1.962
0.5
0.75
0.846
1.25
1.5
1.75
1.766 1.936
2.25
2.315
min
0.971
1.278 1.390 1.450
0.785
1.534
0.5
0.75
1.25
1.5
1.75
2.25
min
Figure 8 Pesticide mixture separated on 50 mm column in 2.5 minutes gradient at 1 ml/min flow rate.
Conclusion
This Application Note shows a method for the separation of a mixture of common herbicides from the triazine, phenylurea and chloroacetanilide classes in less than 10 minutes using a high efficiency sub-2-m column with the Agilent 1290 Infinity LC system and Agilent 6140 Single Quadrupole LC/MS system. Several aspects of UHPLC method development, including frictional heating effects, are highlighted. Using normal method transfer techniques it is shown that this separation using MS detection can be accelerated to complete in less than 2.5 minutes.
Reference
1. Pat Sandra, "Increasing productivity in the analysis of impurities in meto clopramide hydrochloride formula tions using the Agilent 1290 Infinity LC system", Agilent Application Note 5990-3981EN, 2009.
www.agilent.com/chem/1290
Agilent Technologies, Inc., 2010 May 1, 2010 Publication Number 5990-5794EN
Pesticide Dynamic MRM Compound Database for Screening and Identification Using the Agilent Triple Quadrupole LC/MS Systems
Technical Note
Introduction
Over the last century more than 1000 pesticides have been in common use for crop protection. Beyond approved and recommended usage, there always exists the possibility that any of these chemicals can be found in the environment and make their way into the food supply. To protect both the environment and human health, detection and identification in survey type monitoring is very important. Liquid chromatography/tandem mass spectrometry with a triple quadrupole LC/MS meets this need by providing the most sensitive and highly selective detection in complex samples. The system must be run in multiple reaction monitoring (MRM) mode, in order to obtain the maximum sensitivity and selectivity from this technology. Although the triple quadrupole LC/MS is the most sensitive method for multiresidue analysis, the technique can only detect the pesticides that have been included in the methodology. Each pesticide must contain its predetermined precursor ion and an indicative product ion. This single precursor/product ion pair or transition is required for screening in the MRM mode. For confirmation of a compound at least two transitions must be included in the method so that their presence and correct ratio in a sample can be determined along with the correct chromatographic retention time. Because every compound is different there are specific instrument conditions that will provide an appropriate response for each transition. On the Agilent systems this includes both fragmentor voltage, optimizing transmission of the precursor ion into the mass spectrometer, and the collision energy, optimizing the maximum intensity for a specific product ion. Excellent results are obtained using all other mass spectrometer settings provided by the system's Autotune program. A powerful tool called MassHunter Optimizer has been added to the Agilent 6400 Series triple quadrupole LC/MS systems. This software allows automated optimization of compound specific parameters, and it is within this tool that the Pesticide Database operates. Any compound that the user optimizes can be saved to a Project or to a Database. Agilent has created a pesticide database containing the operating parameters for over 700 pesticides. The Pesticide Database is read-only and can be saved to any name for customization by re-optimization of compounds in
the database or addition/deletion of those present. Presently, modifications can be done only in MassHunter Optimizer or by saving method conditions in MassHunter Acquisition, not in the database browser. The power of the database is due to the fact that the conditions included will provide good results without the user needing to optimize each and every compound. In addition, methods are included that now contain retention times for Dynamic MRM [1,2] acquisition. For analysts needing customized methods with hundreds of pesticides per method, the database will allow fast startup and provide good results without the need to optimize the compounds in the database. This does not negate the need to run standards and validate results with good QC/QA procedures. Compounds that the user needs to analyze but are not in the database will need to be optimized. This is readily facilitated by MassHunter Optimizer, and the compounds can then be added to the user's customized database.
Description
The Pesticide Dynamic MRM Database requires MassHunter Acquisition and MassHunter Optimizer 3.01 or later. The link to the database is from the MassHunter Acquisition software or MassHunter Optimizer. It is here that the user can import selected compounds to rapidly develop a customized method, which meets the analytical needs of a specific analysis. This can include compounds in the supplied Pesticide MRM Database and those added by the user, importing to a custom designed chromatographic method for a specific matrix, or a host of other needs for customization. Figure 1 is a screen capture of the MS QQQ Acquisition setup tab for controlling any of the Agilent 6400 Series triple quadrupole LC/MS systems. Right-clicking on the white or grey area of the "Scan Segments" section of the screen displays the pull-down menu where "Import from optimizer" is accessed. When this is selected, the database browser is opened with the default database. The Agilent pesticide database or a customized version of it can be made the default (please note again that the supplied database is read-only). The user can then select the compounds and product ions to import into the acquisition method of his or her choice. The user can also save the compounds in the method to the database with the retention times used in Dynamic MRM.
Figure 1.
MS QQQ Acquisition tab of MassHunter showing the Import from optimizer function.
Once in the Database Browser the Pesticide Database can be opened. Figure 2 shows the Database Browser with the Pesticide Database loaded. The pull-down menu at the top left of the screen in Figure 2 or the Database Save icon allows the user to save the database to a new name. This is a necessary step for changes to be made because the Pesticide Dynamic MRM Database is read-only. Compounds can be deleted from the database in the Database Browser, but they cannot be added. That must be done from MassHunter Optimizer (see below). The copied database can now be customized and set to the default. The database contains the compound name, its formula, the nominal monoisotopic mass (nominal mass plus one decimal place where the second is not significant) of the compound, and the method(s) that were used for analyses. The parameters for analysis include the precursor that gave the optimal signal and its associated fragmentor voltage, at least two product ions (if the compound did produce two significant product ions), and the optimized collision energy for each product ion. In addition, the abundance and response factor of each ion is shown so that the user can distinguish between the quantitation ion and the qualifier by their response. In addition, acquisition methods and retention times with retention time windows are given for Dynamic MRM. The user may select a method and import compounds and their associated retention times with that method. The LC portion of this method must be used or the retention times become invalid. If not all the compounds desired are found in a specific method, the user must find the retention times for those compounds using the desired method and then import the other compounds from the database with the same method. The method must also use the same LC configuration used in the database method. Therefore, if a user has a 1200 SL pump, a method using the Agilent Infinity 1290 LC should not be used without expecting to modify the retention times. If only some of the compounds in a method or compounds from various methods are desired, they can be added to the "Import List" as shown in Figure 3. This shopping cart of compounds allows mixing sources of the compounds found by the various search filters provided (Figure 2).
Figure 2.
Database Browser view of Pesticide Database.
A good way to search for compounds is to make a list in Excel and copy that list to the search text window (Figure 3). However, the search filter desired MUST be checked. For example, if the user has a list of compound names, then the compound name box must be checked. Likewise if the list is CAS numbers then the "CAS" box must be checked. Selected compounds can be added to the Import list and removed from that list if necessary. However the database itself cannot be edited from the database browser, even in a user-created database.
Figure 3.
Import list where compounds selected from different search filters can be added to a Dynamic MRM method.
Finally, the other access point for the database is from MassHunter Optimizer program. It is from this access that the user can add compounds to their customized database. Figure 4 shows the initial screen for the MassHunter Optimizer. The icons across the top allow import and export to and from Excel, naming and saving projects and compounds, starting an optimization or breakdown profile, and access to the databases. The circled icon in Figure 4 allows the import from an acquisition method. Figure 5 with "show results" from an import demonstrates that not only the operating parameters, but the retention times and the retention time windows are imported from the method. By selecting the icon circled in Figure 5 the user can then export this to the default database (their custom database). Again, the Pesticide Dynamic MRM database is read-only but once accessed it can be saved as a customized database and then set as the default. As the default, compounds and projects can be saved from MassHunter Optimizer to the database. Compounds can be deleted directly from the Database Browser but can only be added or changed using MassHunter Optimizer. It will be useful for the user to save compounds not in the provided database and at times to re-optimize certain compounds for specific user conditions.
Figure 4.
First tab of MassHunter Optimizer where parameter range and method settings are made. Note icon "Import from Database" provides access to Database Browser.
Figure 5.
Import from an Acquisition method with "Show Results" displayed. The circled icon allows export of all compound information to the default database.
Summary
The Dynamic MRM Pesticide Database provides over 700 compounds with specific parameters for all Agilent 6400 Series triple quadrupole LC/MS systems. It is designed to meet the needs of laboratories analyzing hundreds of pesticides in one analysis. It allows re-optimization of compounds through the MassHunter Optimizer program and incorporation of the compounds into data acquisition methods for multi-residue analysis where Dynamic MRM is most useful. Its benefits to the analyst are: Fast method development with compound-specific parameters for hundreds of compounds Storage and retrieval of compounds added to the supplied database Customization to meet specific needs of laboratories and their analyses The database and its functionality will continue to evolve to provide greater search and retrieval capabilities and faster development for Dynamic MRM methods.

The Dynamic MRM Pesticide Database is included in only the Pesticide Application kit (Agilent part number, G1733AA). Details of how to use the database are given in the Quick Start Guide that is included in the application kit. For more information on our products and services, visit our Web site at www.agilent.com/chem.
Agilent Technologies, Inc., 2010 Printed in the USA April 28, 2010 5990-4255EN
Agilent 6540 Accurate-Mass Q-TOF: superior mass accuracy for high-confidence characterization of unknowns
Application Note
Author David Weil Agilent Technologies Schaumburg, IL USA
Introduction Quadrupole time-of-flight (Q-TOF) instruments are invaluable for characterization of unknown compounds. They offer a combination of unbeatable separation speed and resolution as well as superior qualitative MS and MS/MS capabilities. However, even with their ability to give high resolution accurate mass data, it is common to get multiple candidate molecular formula for the [M+H]+ ion, or several elemental compositions for a given ion in the MS/MS spectra. High quality Q-TOF data helps to eliminate unlikely and incorrect formula. The Agilent 6540 Accurate-Mass Q-TOF LC/MS System provides outstanding mass accuracy and isotopic fidelity, which reduces the number of formula for unknown samples. The systems high resolution can also resolve charge states or pinpoint small changes in structure.
Excellent MS mass accuracy limits number of possible formula To test the capability of the instrument, a laboratory submitted a sample described as a compound with a nominal molecular weight 218. The test was to determine the correct molecular formula. The sample was analyzed using the Agilent 6540 Accurate-Mass Q-TOF.
Figure 1 shows the resulting MS spectrum, and analysis using Agilent Molecular Formula Generation (MFG) software,1 part of Agilent MassHunter Qualitative Analysis software. The theoretical value for [M+H]+ ion is m/z 219.0764. Because the mass error was only 0.33 ppm and the isotope abundance match was excellent, MFG was able to select the correct formula, i.e. C11H10N2O3, from among the many potential candidates.
[M+H]+ 0.33 ppm mass accuracy in MS mode
Figure 1. Superb MS mass accuracy leads to the correct formula.
Outstanding MS/MS mass accuracy enables greater confidence in structure elucidation Figure 2 shows the analysis of the same unknown in MS/MS mode. At the top is the MS/MS spectrum of the [M+H]+ ion, m/z 219.0763. The inset table shows the MFG results. At m/z values greater than 100, MS/MS mass accuracy is better than 2 ppm. Even below 100, the mass accuracy is excellent, with values in the millidalton range. Mass accuracy maintained even at the high data rates needed for UHPLC The 6540 Accurate-Mass Q-TOF produces very accurate mass measurements for all MS acquisition rates, important when you want to use it with fast chromatography of complex mixtures of unknown compounds. With high resolution separations from ultra high performance liquid chromatography (UHPLC), peak widths can be as narrow as 0.5 second. Very narrow peaks reduce ion suppression by resolving components that might interfere during the ionization process. However, narrow peaks mean that mass spectrometers used for UHPLC must acquire spectra rapidly without loss of mass accuracy. Unlike some competing accurate-mass instruments, the Agilent Q-TOF systems meet this challenge by providing the speed needed to realize the benefits of todays fastest UHPLC separations.
MS/MS mode Mass accuracy above m/z 100 better than 2 ppm
[M+H]+
Figure 2. This outstanding MS/MS mass accuracy provides more confident ion assignments, and a clearer picture of the substructures in an unknown compound.
For this test, the same unknown was run at various spectral acquisition rates in MS-only mode. Table 1 shows the differences in ppm between calculated and measured masses. Regardless of
spectral acquisition rate, the mass error was extremely small, ranging from 0.02 to 0.46 ppm, making the Agilent 6540 Accurate-Mass Q-TOF an ideal detector for fast chromatography.
Scan rate (Hz) 1 2 4 6 8 10
m/z 219.07642 219.07641 219.07647 219.07639 219.07632 219.07633
Mass error (ppm) -0.02 0.05 -0.24 0.11 0.46 0.41
Table 1. Excellent mass accuracy across spectral acquisition rates makes the Agilent 6540 Accurate-Mass Q-TOF a perfect match for UHPLC analysis of unknown compounds.
Conclusion The Agilent 6540 Accurate-Mass Q-TOF delivers clear and unambiguous information for unknown samples. Whether analyzing PTMs, profiling biomarkers, identifying metabolites, screening for pesticides, or characterizing intact proteins, Agilent Q-TOF solutions deliver the data quality demanded of the most critical science. The theoretical value for [M+H]+ ion is m/z 219.0764. In this example, mass accuracy was better than 0.5 ppm in MS mode, and better than 2 ppm above m/z 100 in MS/MS mode. This outstanding performance, together with the unique Agilent Molecular Formula Generation software, produced the correct molecular formula for an unknown compound and its substructures. Mass accuracy was independent of spectral acquisition speeds, which means the instrument can be used for reliable identification of unknowns from UHPLC analyses. The Accurate-Mass Q-TOF platform provides sensitivity, accurate mass, dynamic range, and resolutionall fully compatible with ultra fast UHPLC separations.
Reference
1.E. Darland, D. McIntyre, D. Weil, F. Kuhlmann, and X. Li, Superior Molecular Formula Generation from Accurate-Mass Data, Agilent publication number 5989-7409EN, 2008.
www.agilent.com/chem/qtof
This item is intended for Research Use Only. Not for use in diagnostic procedures. Information, descriptions and specifications in this publication are subject to change without notice. Agilent Technologies shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance or use of this material. Agilent Technologies, Inc. 2010 Published in the U.S.A., March 25, 2010 5990-5405EN
Determination of pesticides in baby food by UHPLC/MS/MS using the Agilent 1290 Infinity LC system and the Agilent 6460 triple quadrupole LC/MS
Application Note
Food
Authors
Gerd Vanhoenacker, Frank David, Pat Sandra Research Institute for Chromatography Kennedypark 26 B-8500 Kortrijk Belgium
x10 2 0.8 0.6 0.4 0.2 0 0 1 2 3 4 5 6 7 Counts vs. Acquisition Time (min)
Abstract
The qualitative and quantitative analysis of pesticides at trace levels in baby food matrices using UHPLC and triple quadrupole MS is demonstrated. Sample preparation is performed using an Agilent SampliQ QuEChERS kit for extraction and dispersive SPE. The extracts are analyzed by LC/MS/MS on an Agilent 1290 Infinity LC system coupled to an Agilent 6460 triple quadrupole LC/MS using Dynamic MRM. The method and extraction performance were evaluated in terms of repeatability, linearity and sensitivity. Moreover the influence of the additional dispersive SPE cleanup was investigated. Detection limits were between 500 ng/kg and 10 ng/kg (ppt), which is much lower than the maximum residue level (MRL) of 10 g/kg (ppb) imposed by the European Union.
Introduction
Due to diversity of pesticides used in food protection and the globalization of the food industry, the monitoring of programs that cover a large number of pesticides is important. The application of UHPLC systems combined with the new generation triple quadrupole mass spectrometers facilitate the analysis of pesticides in challenging matrices such as food samples. As a result of the high sensitivity and the high scan rate capabilities of the Agilent 6460A triple quadrupole mass spectrometer, the simultaneous qualitative and quantitative multiresidue analysis of a large set of pesticides at trace levels can be performed. The high sensitivity is essential for the analysis of these compounds in derived products, where the concentrations will be a fraction of the concentration in the raw material. In this respect, baby food is a challenging matrix. This application notes describes the quantitative analysis of 40 pesticides in baby food at levels below the maximum residue level (MRL) (10 g/kg fruit or vegetable) specified in EC Regulation 396/2005 which was implemented in September 2008.1 A QuEChERS extraction and dispersive SPE method was applied to isolate the pesticides from the baby food matrix. An Agilent 1290 Infinity LC was used to perform the separation on a Rapid Resolution High Definition (RRHD) ZORBAX Eclipse Plus column. The total analysis time was 10 min (including 1.5 min re-equilibration) and detection limits ranged from 10 to 500 ng/kg using Dynamic MRM and two transitions (quantifier and qualifier) per compound. Three different baby food compositions were analyzed. Extraction performance criteria such as repeatability, recovery (accuracy) and sensitivity were investigated.
Experimental
Instrumentation
An Agilent 1290 Infinity LC system and an Agilent 6460A triple quadrupole LC/MS with Agilent jet stream technology were used. The 1290 Infinity LC system was configured as follows:
Part number G4220A G4226A G1316C G4212A Description Agilent 1290 Infinity Binary Pump with integrated vacuum degasser Agilent 1290 Infinity Autosampler Agilent 1290 Infinity Thermostatted Column Compartment Agilent 1290 Infinity Diode Array Detector
Method parameters: Column Mobile phase Flow rate Gradient Agilent ZORBAX Eclipse Plus RRHD C18, 150 mm L 2.1 mm id, 1.8 m dp A = 0.05% (w/v) ammonium formate + 0.01% (v/v) formic acid in water B = Methanol 0.5 mL/min Min 0 to 5 5 to 6.5 6.5 to 8.5 8.5 to 10 45 C 2 L, with needle wash (flushport, 5 s, water/methanol 1/1) MS/MS Electrospray, positive ionization 250 C 10 L/min 30 psig 11 L/min 4500 V 500 V See Table 1 50 200 ms %B 10 to 65 65 to 95 95 10
Temperature Injection Detection Ionization Jet Stream parameters Drying gas temperature Drying gas flow Nebulizer pressure Sheath gas flow Capillary voltage Nozzle voltage Acquisition Dynamic MRM Delta EMV Cycle time
Sheath gas temperature 340 C
Compound Cyromazine Cyromazine Flonicamid Flonicamid Thiamethoxam Thiamethoxam Monocrotofos Monocrotofos Dicrotofos Dicrotofos Ethiofencarb-sulfone Ethiofencarb-sulfone Imidacloprid Imidacloprid Clothianidin Clothianidin Ethiofencarb-sulfoxide Ethiofencarb-sulfoxide Methiocarb-sulfoxide Methiocarb-sulfoxide Thiofanox-sulfone Thiofanox-sulfone Trichlorfon Trichlorfon Vamidothion Vamidothion Acetamiprid Acetamiprid Carbofuran-3-OH Carbofuran-3-OH Fenthion-oxon-sulfoxide Fenthion-oxon-sulfoxide Carbendazim Carbendazim Fenthion-oxon-sulfone Fenthion-oxon-sulfone Cymoxanil Cymoxanil Oxycarboxin Oxycarboxin Chlothiamid Chlothiamid Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q
Precursor ion (m/z) 167.1 167.1 230.1 230.1 292.2 292.2 224.1 224.1 238.1 238.1 258.1 258.1 256.1 256.1 250.0 250.0 242.1 242.1 242.0 242.0 251.1 251.1 256.9 256.9 288.1 288.1 223.1 223.1 238.1 238.1 279.0 279.0 192.1 192.1 295.0 295.0 199.2 199.2 268.1 268.1 205.9 205.9
Product ion (m/z) 85.1 125.1 203.1 174.1 211.0 181.0 127.0 193.0 112.1 127.0 107.1 201.1 175.1 209.0 169.1 132.1 107.1 185.1 185.1 170.0 57.1 76.1 109.0 221.0 146.1 118.1 126.0 56.0 163.1 181.1 104.1 121.1 160.1 132.1 217.1 104.1 128.0 111.0 175.0 146.9 189.0 172.0
Fragmentor (V) 100 100 80 80 85 85 85 85 90 90 80 80 90 90 90 90 80 80 80 90 100 100 100 100 80 80 100 100 85 85 125 125 100 100 125 125 65 100 100 100 85 85
Collision energy (V) 25 25 15 15 4 16 10 5 5 15 10 10 20 15 7 15 15 15 10 15 15 15 15 15 10 20 15 15 5 5 30 30 15 25 25 25 5 20 10 25 20 20
Retention time (min) 1.20 1.20 2.85 2.85 2.92 2.92 3.11 3.11 3.41 3.41 3.47 3.47 3.55 3.55 3.58 3.58 3.60 3.60 3.79 3.79 3.80 3.80 3.92 3.92 3.94 3.94 3.94 3.94 3.96 3.96 4.03 4.03 4.11 4.11 4.18 4.18 4.24 4.24 4.27 4.27 4.29 4.29
Retention time window (min) 1.5 1.5 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
Table 1 Dynamic MRM data acquisition parameters for the compounds under investigation. Q = quantifier, q = qualifier.
Compound Thiacloprid Thiacloprid Florasulam Florasulam Tricyclazole Tricyclazole Butocarboxim Butocarboxim Thiabendazole Thiabendazole Aldicarb Aldicarb DMSA DMSA Propoxur Propoxur Carbaryl Carbaryl Monolinuron Monolinuron Fluazifop Fluazifop Spiroxamine Spiroxamine Pyrimethanil Pyrimethanil Fenhexamid Fenhexamid Fenbuconazole Fenbuconazole Iprodion Iprodion Kresoxim-methyl Kresoxim-methyl Penconazole Penconazole TPP TPP Pyraclostrobin Pyraclostrobin Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q Q q
Precursor ion (m/z) 253.1 253.1 360.0 360.0 190.1 190.1 213.1 213.1 202.0 202.0 208.0 208.0 201.0 201.0 210.1 210.1 202.1 202.1 215.2 215.2 328.1 328.1 298.4 298.4 200.1 200.1 302.1 302.1 337.2 337.2 330.0 330.0 314.2 314.2 284.1 284.1 327.1 327.1 388.2 388.2
Product ion (m/z) 126.0 186.0 129.1 191.9 163.2 136.2 75.1 156.1 175.0 131.0 116.0 89.1 92.1 137.1 111.1 93.0 145.1 127.0 126.0 148.1 282.1 254.1 144.2 100.2 107.1 82.0 97.0 142.1 125.0 194.1 244.9 287.9 116.0 222.0 69.9 158.8 77.0 151.9 194.1 296.2
Fragmentor (V) 100 100 100 100 100 100 110 110 120 120 70 70 85 85 50 50 50 50 100 100 120 120 100 100 100 100 120 100 120 120 110 110 70 70 85 85 180 180 100 100
Collision energy (V) 20 10 20 10 20 25 15 5 25 35 0 5 15 10 10 20 2 20 20 20 20 20 10 10 25 30 10 5 15 15 10 5 10 10 15 30 40 40 10 10
Retention time (min) 4.34 4.34 4.51 4.51 4.62 4.62 4.66 4.66 4.69 4.69 4.73 4.73 4.76 4.76 5.36 5.36 5.62 5.62 5.75 5.75 5.99 5.99 6.54 6.54 6.61 6.61 6.88 6.88 6.94 6.94 6.98 6.98 7.08 7.08 7.11 7.11 7.14 7.14 7.18 7.18
Retention time window (min) 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
Table 1 Dynamic MRM data acquisition parameters for the compounds under investigation. Q = quantifier, q = qualifier. (continued)
Solutions and Samples

Stock solutions of the pesticides were prepared in acetonitrile. These solutions were diluted to the appropriate concentration (range 0.05 ppb to 1 ppm) in 1% v/v acetic acid in acetonitrile. An internal standard solution of triphenylphosphate (TPP, 20 g/mL) was prepared in the same solvent.
The sample preparation was performed using Agilent SampliQ QuEChERS AOAC kits for extraction and dispersive SPE cleanup. The procedure is described below. Extraction 1. Weigh 15 g of sample into a 50-mL centrifuge tube. 2. Add 100 L TPP solution. 3. Add spiking solution, if necessary. 4. Vortex for 30 s. 5. Add 15 mL of 1% v/v acetic acid in acetonitrile and the SampliQ AOAC extraction salt (p/n 5982-5755). 6. Cap tubes and shake vigorously by hand for 1 min. 7. Centrifuge at 4,000 rpm for 5 min. 8. Filter 1 mL of sample through a syringe filter (0.2 m pore size, regenerated cellulose, p/n 5061-3366) and analyze directly (no SPE) or (additional clean-up). Dispersive SPE 1. Transfer 8 mL from the centrifuged extract into a 15-mL SampliQ AOAC dispersive SPE tube for fatty samples (p/n 5982-5158). 2. Vortex for 30 s. 3. Centrifuge at 13,000 rpm for 2 min. 4. Filter 1 mL through a syringe filter (0.2 m pore size, regenerated cellulose, p/n 5061-3366) and analyze.
Sample Preparation
Three baby food products were obtained from a local supermarket. According to the labels, the samples were composed of the following ingredients: Sample 1: carrots (40%), potatoes (18%), tomatoes (18%), beans (13%), beef (10%) Sample 2: water (37%), potatoes (30%), spinach (17%), chicken (10%) Sample 3: carrots (54%), potato (23%), water (16%), rice (7%)

State-of-the-art LC/MS/MS equipment enables fast multiresidue analysis of pesticides at low levels in complex matrices. The Agilent 1290 Infinity LC provides the necessary power to perform analysis of the 40 selected pesticides within the 10-min total analysis time (run time and equilibration time). A 15 cm column was preferred above a 10 or 5 cm column because of the higher resolving power. This is useful to minimize ion suppression or response enhancement due to matrix effects. Methanol was chosen as an organic modifier because of the significantly improved sensitivity compared to acetonitrile for this analysis. During the analysis, a total of 82 transitions (2 per solute + 2 for IS) had to be performed. The dynamic MRM function allows MRM transition lists to be built based on a retention time window specified for each analyte. Consequently, the pesticides are only monitored during that elution window in the analytical run. This approach leads to equivalent or better results in terms of sensitivity and quantification (data points) compared to the traditional time segment based methods 2. With the Dynamic MRM enabled, the maximum number of concurrent MRMs was 32. Using an MRM cycle time of 200 ms, the minimal and maximal transition dwell times were 2.75 and 96.50 ms
(values given by MassHunter acquisition software), respectively. The resulting number of data points across the peaks was above 20 for all compounds which is largely sufficient for quantitation purposes. The performance of the LC/MS/MS method was tested by the analysis of standard solutions. The chromatogram (overlaid MRMs of quantification ions) for a 10 ppb solution is shown in Figure 1. Figures of merit are summarized in Table 2. The injection precision was tested at two concentration levels (1 and 10 ppb). The standard solutions were each injected five consecutive
times. The linearity of the method was evaluated between 0.05 and 20 ppb at eight levels (0.05, 0.10, 0.20, 0.50, 1,2,10, and 20 ppb). Each solution was injected once. The lowest level is below the detection limit for some compounds. For these analytes, the calibration curve was started at the limit of detection. The sensitivity was excellent and all compounds could be analyzed at the sub-ppb level. An example of the ion traces (quantification ion transition and qualifier ion transition) and the corresponding calibration curves for fluazifop
x10 2 0.8 0.6 0.4 0.2 0 0 1 2 3 4 5 6 7 Counts vs. Acquisition Time (min)
Figure 1 MRM of a 10 ppb standard solution (only quantifier transitions are shown).
Compound Acetamiprid Aldicarb Butocarboxim Carbaryl Carbendazim Carbofuran-3-OH Chlothiamid Clothianidin Cymoxanil Cyromazine Dicrotofos DMSA Ethiofencarb-sulfone Ethiofencarb-sulfoxide Fenbuconazole Fenhexamid Fenthion-oxon-sulfone Fenthion-oxon-sulfoxide Flonicamid Florasulam Fluazifop Imidacloprid Iprodione Kresoxim-methyl Methiocarb-sulfoxide Monocrotofos Monolinuron Oxycarboxin Penconazole Propoxur Pyraclostrobin Pyrimethanil Spiroxamine Thiabendazole Thiacloprid Thiamethoxam Thiofanox-sulfone Trichlorfon Tricyclazole Vamidothion
* ** *** 1
Repeatability of injection (% RSD) 1 ppb 10 ppb 2.20 4.82 19.93 1.70 2.93 14.68 20.64 7.69 7.30 2.02 3.69 5.13 2.69 6.24 11.29 5.19 13.96 13.13 10.87 9.51 5.77 3.31 24.53 4.46 4.02 1.71 0.67 5.92 2.02 0.70 1.23 5.55 0.91 2.99 1.57 1.38 5.13 6.34 1.66 4.56 1.62 2.03 2.36 1.73 1.28 2.50 7.28 2.14 3.88 1.08 1.01 2.36 2.25 2.02 1.24 4.93 7.10 2.90 2.55 3.25 3.28 1.15 4.28 1.16 2.85 1.45 0.27 1.93 1.60 0.94 0.93 0.60 0.87 0.96 1.10 1.89 2.14 4.31 0.85 1.16
Linearity (R) 0.9999 0.9999 0.9996** 0.9996 0.9997 0.9996* 0.9979* 0.9999* 0.9998* 0.9993 0.9994 0.9996 0.9998 0.9991* 0.9994** 0.9988** 0.9988 0.9986 0.9980 0.9999 0.9998** 0.9998 0.9984*** 0.9999 0.9991* 0.9996 0.9999 0.9991 0.9997 0.9998 0.9996 0.9997 0.9997 0.9999 0.9995 0.9998 0.9998 0.9988 0.9999 0.9997
Detection limit (ppb) Q q 0.02 0.01 0.20 0.01 0.01 0.10 0.20 0.10 0.10 <0.501 0.01 0.05 0.05 0.10 0.20 0.20 0.05 0.05 0.05 0.05 0.20 0.05 0.50 0.01 0.10 0.01 0.05 0.05 0.01 0.01 0.01 0.02 <0.01 0.02 0.02 0.01 0.05 0.05 0.02 0.01 0.02 0.02 0.50 0.01 0.05 0.10 1.00 0.20 0.50 0.50 0.02 0.20 0.20 0.10 1.00 1.00 0.05 0.10 0.20 0.20 0.50 0.05 5.00 0.05 0.20 0.02 0.05 0.05 0.02 0.01 0.02 0.05 <0.01 0.02 0.05 0.05 0.10 0.20 0.02 0.01
Detection limit is 0.10 ppb, calibration starts at 0.10 ppb. Detection limit is 0.20 ppb, calibration starts at 0.20 ppb. Detection limit is 0.50 ppb, calibration starts at 0.50 ppb. High due to interference of a system peak.
Table 2 Method performance results.
(a compound with relatively low sensitivity) and for propoxur (a compound with relatively good sensitivity) are shown in Figures 2 and 3, respectively. Most of the compounds have detection limits below 0.05 ppb. The sensitivity for spiroxamine is below the lowest level injected (0.01 ppb) which is significantly better compared to the other pesticides. No accurate detection limit could be determined for cyromazine due to a system peak that interfered at low levels.
10 1 5.1 5 Counts 4.9 4.8 4.7 4.6 4.5 5.7
*5.975
10 1 5
*5.972
Quantifier 328.1 > 282.1

Counts
4.9 4.8 4.7 4.6
Qualifier 328.1 > 254.1
5.8 5.9 6 6.1 6.2 Acquisition time (min)
6.3
5.7
6.3
10 3 4 Relative responses 3.5 3 2.5 2 1.5 1 0.5 0 _1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Calibration R 2 = 0.9998
Concentration (ppb)
Figure 2 Ion traces for two transitions at the LOD (0.5 ppb standard solution) and calibration curve for fluazifop.
Relative responses
The QuEChERS sample preparation procedure was applied to three baby food samples. Extracts were analyzed with and without additional dispersive SPE cleanup. There were no target compounds detected above the LOD in nonspiked samples. The resulting chromatogram, shown as an overlay of quantification transitions for a sample spiked at 1-ppb level with all 40 pesticides, is depicted in Figure 4. The signals for the quantifier and qualifier transitions for fluazifop and propoxur in the spiked sample at 1-ppb level are shown in Figure 5. From these traces it is clear that excellent selectivity and sensitivity are obtained. The relative response of the quantification transition and qualifier transition are clearly within the limits for positive identification.
10 1 6.6 6.4 6.2 6 5.8 5.6 5.4 5.2 5 4.8 4.6 4.4 5 5.1
5.358
10 1 5.8
*5.370

Counts
5.6 5.4 5.2 5 4.8
Qualifier 210.1 > 93.1
Counts
5.2 5.3 5.4 5.5 5.6 Acquisition time (min)
5.7
5.1
5.7
10 -1 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 _0.2
Calibration R 2 = 0.9998
_1
10 11 12 13 14 15 16 17 18 19 20 21
Concentration (ppb)
Figure 3 Ion traces for two transitions at the LOD (0.01 ppb standard solution) and calibration curve for propoxur.
10 3 2
1,5
0,5
3 4 Counts vs. acquisition time (min)
Figure 4 MRM of an extract of sample 2 spiked with 1 ppb (only quantifier transitions are shown). No dispersive SPE performed on the sample. The transition for the internal standard is not shown.
The performance criteria of the sample preparation and analysis method are summarized in Table 3. The extraction repeatability is calculated on sample 1, spiked at the 10-ppb level and repeated (extraction + analysis) five times. Most RSDs are below 10%, with the exception of iprodione and fluazifop, where higher values are obtained after SPE. The average recovery (response spike sample / response calibration sample) for the three different samples was between 70% and 110% at 1 and 10 ppb spike level in most cases. No significant differences were observed between the different matrices. The recovery is satisfactory even at the 1-ppb level and in most cases there is no significant difference between extracts that have been subjected to the additional SPE procedure and those that have not. For cyromazine, better values are obtained after SPE. For fluazifop, on the other hand, very low recoveries (and high RSD) are obtained when additional dispersive SPE is used. In this case, analysis without additional SPE is recommended.
10 1
Fluazifop
*5.975
10 1
Counts
Counts
6 Ratio = 45.9
5.6 10 2 6 5 4 Counts 3 2 1 0 5
5.7
5.8 5.9 6 6.1 6.2 Acquisition time (min) 5.361
6.3
5.6
5.7
6.3
Propoxur
10 2 6 5
Counts
4 3 2 1 0 Ratio = 48.1
5.1
5.7
5.1
5.7
Figure 5 Ion traces for 2 transitions for fluazifop and propoxur in an extract of sample 2 spiked with 1 ppb. No dispersive SPE performed on the sample. The uncertainty was set at 20% (dotted lines).
10
Repeatability of extraction (% RSD)1 Compound Acetamiprid Aldicarb Butocarboxim Carbaryl Carbendazim Carbofuran-3-OH Chlothiamid Clothianidin Cymoxanil Cyromazine Dicrotofos DMSA Ethiofencarb-sulfone Ethiofencarb-sulfoxide Fenbuconazole Fenhexamid Fenthion-oxon-sulfone Fenthion-oxon-sulfoxide Flonicamid Florasulam Fluazifop Imidacloprid Iprodion Kresoxim-methyl Methiocarb-sulfoxide Monocrotofos Monolinuron Oxycarboxin Penconazole Propoxur Pyraclostrobin Pyrimethanil Spiroxamine Thiabendazole Thiacloprid Thiamethoxam Thiofanox-sulfone Trichlorfon Tricyclazole Vamidothion
1 2 3 4
Average recovery (%)2 1 ppb SPE 107.0 91.9 95.4 92.3 88.6 85.8 87.5 87.0 101.5 108.7 102.9 96.3 92.1 90.8 107.5 74.0 91.4 122.8 94.4 72.2 14.4 115.2 87.0 74.0 94.5 90.2 90.0 89.8 73.4 94.7 86.1 85.5 79.6 92.7 96.7 104.8 99.0 86.9 91.3 92.4 No SPE 83.4 82.8 79.6 75.4 79.4 114.4 105.1 117.7 72.1 87.1 83.0 109.9 85.6 88.9 80.6 100.8 78.4 96.5 86.1 103.3 117.7 114.9 90.8 80.6 98.5 81.6 80.9 107.4 78.2 83.0 89.3 84.4 91.7 74.5 90.7 108.5 93.3 86.5 72.0 79.2 SPE 99.3 92.4 91.8 93.2 90.5 100.5 91.8 103.7 99.0 73.0 93.5 103.0 91.8 92.1 90.8 91.5 89.8 100.8 94.3 73.8 18.8 111.5 89.8 71.8 93.1 90.9 92.9 101.0 76.9 95.6 84.8 86.8 78.4 91.1 94.3 112.4 91.5 99.7 90.9 90.0 10 ppb No SPE 92.2 81.8 86.9 80.1 78.1 98.2 70.1 103.2 98.1 57.8 83.5 108.1 84.2 83.2 99.8 83.5 89.8 90.0 91.8 115.9 92.0 117.6 91.0 80.7 87.1 83.1 84.3 105.6 83.6 84.9 90.9 78.3 85.8 78.7 86.4 108.7 84.1 92.4 75.6 81.8
Lowest level detected in extract (ppb)3
No IS SPE 1.50 1.74 4.41 1.84 1.87 6.13 4.59 3.01 5.27 1.70 2.73 2.13 1.52 2.33 3.77 5.22 6.08 2.63 2.77 5.98 20.71 2.98 14.30 4.16 3.06 1.54 1.66 2.04 4.25 1.61 3.62 1.99 3.91 1.29 2.51 2.09 2.70 6.94 0.90 1.75
No IS No SPE 1.55 1.69 4.17 1.30 1.23 4.62 9.54 1.70 5.11 0.54 1.98 1.73 2.96 0.74 6.28 7.29 4.49 0.90 3.02 3.76 1.45 2.35 4.37 3.82 1.63 0.72 0.71 1.74 2.77 0.25 4.59 2.22 1.50 1.52 1.74 1.12 1.48 1.93 1.58 2.63
0.10 0.10 1.00 0.10 0.10 1.00 1.00 1.00 1.00 <10.004 0.10 0.10 1.00 0.10 1.00 1.00 1.00 0.10 0.10 0.10 1.00 0.10 1.00 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10 1.00 0.10 0.10
Sample 1, spiked with 10 ppb, extracted 5 times. 1 injection per extract. Average of samples 1 to 3, spiked at 1 ppb and at 10 ppb and extracted once. 1 injection per extract. Samples were spiked at 0.1, 1, and 10 ppb level. Lowest detected level is reported. High due to interference of a system peak.
Table 3 Extraction performance.
11
Conclusion
The multiresidue LC/MS/MS method enabled the analysis of 40 pesticides at low levels in baby food. Sample preparation was performed using an Agilent SampliQ QuEChERS AOAC kit. The total analysis time using the Agilent 1290 Infinity LC system and the Agilent 6460A triple quadrupole LC/MS was 10 min. All compounds could be detected at 0.5 g/kg or lower in the samples, which is 20 times lower than the MRL for these compounds in baby food according to EU regulation. The extraction repeatability and recovery were good. No difference on extraction and analytical performance due to differences in sample matrix were observed. The optional dispersive SPE cleanup procedure can be applied but for some solutes larger standard deviation and lower recoveries were observed after SPE.
References
1. European Parliament and Council Regulation No. 396/2005 on Maximum Residue Levels of Pesticides In or On Food and Feed of Plant and Animal Origin and Amending Council Directive 91/414/EEC, 23 February 2005. 2. P. Stone, T. Glauner, K. Kuhlmann, T. Schlabach, K. Miller, "New Dynamic MRM Mode Improves Data Quality and Triple Quad Quantitation in Complex Analyses," Agilent Technologies, Technical Overview, 5990-3595EN, June 2009.
www.agilent.com/chem/1290
Agilent Technologies, Inc., 2009 December 1, 2009 Publication Number 5990-5028EN
Multi-Residue Pesticide Analysis with Dynamic Multiple Reaction Monitoring and Triple Quadrupole LC/MS/MS Fast and Effective Method Development Using an Application Kit and a Pesticides Compound Parameter Database
Application Note
Authors
Jerry Zweigenbaum, Michael Flanagan, Peter Stone, Thomas Glauner, Limian Zhao Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
The analysis of pesticide residues in food and environmental samples is challenging due to the low concentrations and large number of analytes that need to be monitored and quantified. In addition, method development for Liquid Chromatography/Mass Spectrometry/Mass Spectrometry (LC/MS/MS) with a triple quadrupole instrument is laborious and time consuming because of the compound dependent parameters that need to be optimized. This application note describes how pesticide residue LC/MS/MS methods can be set up quickly and efficiently using the Agilent Pesticides Application Kit. This Application Kit contains a pesticide test mix, a 600compound pesticide MRM database, a quick start guide and several dynamic Multiple Reaction Monitoring (MRM) methods, which can easily be incorporated into a specific method for pesticide residue analysis. The Pesticides Dynamic MRM database contains compounds commonly monitored around the world and provides fast, customized method development of the analysts' list of pesticides. Results from a 100 and 300-compound mixture are demonstrated with an Agilent 1200 SL Series Rapid Resolution LC and the Agilent 6460 Series Triple Quadrupole LC/MS System with Agilent Jet Stream Technology. The 300-compound mixture was also analyzed using an Agilent 1290 Infinity Ultra High Pressure Liquid Chromatograph (UHPLC) and a 6460 LC/MS. With the higher pressure capabilities of the Agilent 1290 Infinity UHPLC, rapid separations with higher peak capacity and less peak overlap than the Agilent 1200 Series RRLC were produced. Using a spinach matrix spiked with 16 pesticides, the performance of a complete method with the SampliQ extraction and dispersive SPE kits and the Agilent LC/MS/MS triple quadrupole on a typical food matrix was
Introduction
The analysis of target pesticide residues has traditionally been performed using Gas Chromatography/Mass Spectrometry (GC/MS) or Liquid Chromatography/Mass Spectrometry (LC/MS) methods. Because of the number of pesticides used and the sensitivity needed for monitoring hundreds of pesticides in a single analysis, both techniques are a requirement. GC/MS is needed for the less polar, more volatile pesticides and LC/MS for pesticides that are more polar or thermally labile and there is much overlap between them. However, many of the pesticides developed over the last 20 years are most amenable to LC/MS. The method of choice for trace analysis in complex matrices uses a triple quadrupole (QQQ) mass spectrometer incorporating multiple reaction monitoring (MRM). During an MRM analysis the QQQ monitors the product ions produced by collisions of precursor ions in the central quadrupole (the collision cell) of the mass spectrometer, as seen in Figure 1. An MRM analysis can generate a very sensitive and specific analysis of target
affect the sensitivity and specificity of the analysis. A new technique, Dynamic Multiple Reaction Monitoring (MRM) alleviates these limitations and also allows easier method development and future modifications of the method, such as the addition of new pesticides to be analyzed. Using Dynamic MRM, analyte ions are only monitored while they are eluting from the LC. This significantly improves the MS duty cycle time for very complex samples when compared with the time segment method and improves the sensitivity and specificity of an analysis.[1] One of the challenges in developing an MRM method, whether it is a time segment or Dynamic MRM, is creating the time sequence of MS/MS events and mass spectrometer conditions necessary to maximize sensitivity and specificity. It is essential to generate a list of two or more MRM transitions and compound specific parameters, fragmentor voltage and collision energy for each compound being analyzed. The availability of a database containing over 600 pesticides with the MS/MS instrumental information that can be used with all Agilent triple quadrupoles eliminates the need to create this information via tedious manual procedures. The database allows easy import of selected compounds into the user's analytical method. A portion of this database is shown in Figure 2. In addition to creating custom methods, the readonly database allows the user to copy their customized database to meet his or her specific needs. A technical note describes this database in detail. [2] The Agilent Pesticides Application Kit also includes a pesticide test mixture that is used to demonstrate the performance of the system and pretested methods, allowing faster method development. Neither the kit nor the test mixture diminishes the need for each laboratory to define suitable QC/QA procedures and perform validation. Each laboratory must have QC tests fit-forpurpose and run analytical standards to validate analytical results. This application note will demonstrate the use of the Agilent Pesticide Application Kit with a 600-compound parameter database and Dynamic MRM for the analysis of complex pesticide mixtures. The liquid chromatographic separations are performed using an Agilent 1200 SL Series RRLC or an Agilent 1290 Infinity UHPLC with an Agilent 6460 QQQ incorporating Jet Stream technology.[3] The methods described in the note are straightforward to generate using the Agilent MassHunter data analysis software and the Pesticide Dynamic MRM Database. Some limits of detection (LOD) of 100 fg or less were achieved using these methods with the Agilent 6460 Series QQQ LC/MS system. These methods are also compatible with all Agilent 6400 series LC/MS systems.
Figure 1.
A schematic diagram of MRM mode on a triple quadrupole instrument. The precursor ion is selected in Q1, fragmentation occurs in Q2, and the product is selected by Q3. Since two stages of mass selectivity are used, there is very little interference from background matrix resulting in excellent sensitivity.
compounds. Over time regulating agencies have continually increased the number of pesticides and residues that must be monitored. It is now common that hundreds of residues need to be analyzed in a single LC/MS analysis. To address this challenge the MRM transitions that need to be monitored are switched using programmed time segments. This is called time segmented MRM. It is accomplished by programming the QQQ to monitor specific product ions in time segments during the LC/MS analysis. However, the method requires well defined elution time boundaries and must avoid time segment switches when compounds elute from the LC. If a time segmented MRM analysis is generated for a sample that contains hundreds of residues, the time segmented MRM analysis becomes subject to cycle and dwell time limitations that
Figure 2.
Compound Parameter Database with over 600 pesticides entries.
Experimental
Reagents and Chemicals
Agilent Pesticide Test Mix, p/n 5190-0469 acid and base diluted separately as instructed to 10 ppb in 10% acetonitrile/90% water An Agilent SampliQ QuEChERS AOAC Extraction kit, p/n 5982-5755. Agilent SampliQ QuEChERS AOAC Dispersive SPE kits for Highly Pigmented Fruits and Vegetables, p/n 5982-5321 (2 mL) and p/n 5982-5356 (15 mL) Multiple pesticide standards were obtained from Sigma, Chemservice, and Dr. Erhenstofer *Appendix I: LC/MS/MS Conditions for Test mix Positive and Negative Ion Samples
Appendix II: LC/MS/MS Conditions for a 100 Pesticide Methods *Appendix III: LC/MS/MS Conditions for 300-Pesticide Methods using the Agilent 1200 Series SL Appendix IV: LC/MS/MS Conditions for the 300Pesticide Methods using the Agilent 1290 Infinity LC Appendix V: LC/MS/MS Conditions for Pesticides in Spinach using QuEChERS Extraction. *Appendix VI: LC/MS/MS Conditions for the 165Pesticide Methods using the Agilent 1200 Series SL *Appendix VII: LC/MS/MS Conditions for the 224Pesticide Methods using the Agilent 1200 Series SL Appendix VIII: LC/MS/MS Conditions for the 224Pesticide Methods using Agilent 1290 Infinity LC
Instrument Settings
*Each of these methods are included with the Application Kit
Spinach Sample Preparation

Weigh 15 g (0.1 g) of homogenized spinach sample. Spike standards or IS solution if necessary. Vortex 30 s. Add 15 mL of 1% acetic acid in acetonitrile. Add 1 bag of extraction kit (p/n 5982-5982-5755) buffered QuEChERS extraction tubes, AOAC Method 2007.01 to 6 g MgSO4 and 1.5 g NaAc. Cap and hand-shake vigorously for 1 min. Centrifuge at 4000 rpm for 5 min. Transfer 1 mL or 8 mL upper layer to the dispersive SPE kit (p/n 5982-5321 or p/n 5982-5356) for highly pigmented fruits and vegetables. Vortex 1 min. Centrifuge 2-mL tubes at 13000 rpm for 2 min, or 15 mL tubes at 4000 rpm for 5 min. Transfer 200 L of the upper layer to the autosampler vial. Add 800 L of water or appropriate standard spiking solution. Vortex 1 min, to prepare for LC/MS/MS analysis.
10 5 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1

Positive and Negative Ion Test Mix
In addition to the 600-compound database, the Agilent Application Kit for pesticide residue analysis also includes a positive and negative ion test mix, with their analysis methods shown in Appendix I. The methods contain compound names, MRM transitions, fragmentor voltages, collision energies, and retention times for the Dynamic MRM. The test mix and the supplied method allow the analyst to demonstrate that the system is operating properly for pesticide analysis immediately after installation. The LC/MS/MS extracted ion chromatograms (EIC) from the test mix analyzed in the positive and negative ion mode using Dynamic MRM is shown in Figures 3 and 4. The Application Kit Quick Start Guide [4] shows the analyst how to run the test mixes and create a Dynamic MRM method. To create new methods, standards are analyzed at higher concentrations with a one segment MRM method. The data is processed using the Agilent MassHunter Quantitative Data Analysis software to generate a custom report that now includes analyte retention times. A Dynamic MRM method is generated by importing the results from the custom report and specifying a delta retention time window. This process will be automated in the near future. Table 1 shows a partial listing of
3.128
8.038
5.992 7.019 10.776 4.072 5.064 10.679 3.959 5.918 7.472 7.437 8.557 9.133
8 8.5 9
9.615
10 10.5 11
0.5
1.5
2.5
3.5
4.5 5 5.5 6 6.5 7 Counts vs. acquisition time (min)
7.5
9.5
Figure 3.
Positive ion test mix extracted ion chromatogram (see Appendix 1 for list of compounds matching retention times given in chromatogram).
10 3 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 Counts vs. acquisition time (min) 7.5 8 8.5 9 9.5 10 6.572 8.047 8.805 9.650
10.877
10.503
10.5
11
Figure 4.
Negative Ion Test Mix extracted ion chromatogram (see Appendix 1 for list of compounds matching retention times given in chromatogram).
the acquisition parameters from a Dynamic MRM method. Note in this example the retention time window (Delta RT) is 2 min which is large for narrow peaks. A window this wide can be used to run standards where retention times have shifted and need to be updated in the users customized method.
Table 1. Dynamic MRM Screen Capture of Acquisition Parameters
Fast and effective screening of a 100-compound pesticide mix using Dynamic MRM
A 100-compound mix of pesticides was used to demonstrate the effectiveness of the Dynamic MRM. Appendix II contains the LC/MS/MS conditions and a partial listing of the Dynamic MRM method used to analyze a 100-pesticide mixture at the 100 pg/compound level. Note that the column used was 50 mm in length so faster analysis and less efficiency is obtained. The LC/MS/MS extracted ion chromatogram shown in Figure 5 illustrates the performance of the system. The complete LC analysis took less than 15 minutes. Figure 6 shows a 1-min time window where 11 compounds (22 MRM's) are eluting. Figure 7 shows the 1-min delta retention time window for each Dynamic MRM transition. Note the
many peak overlaps in the chromatograms. This necessitates the use of dynamic transitions instead of time segmented transitions in order to achieve the needed cycle time so that each peak can have enough data points to adequately describe the peak for quantitation. Furthermore time segmented MRM has an inherent "dead time" data loss when monitoring analyte peaks eluting near or between time segment boundaries. Time segmented MRM methods may require duplicate monitoring of specific analytes which elute over adjoining time segments. In addition, Dynamic MRM maximizes the dwell times for overlapping peaks enhancing the signal-to-noise while maintaining constant cycle time. Note that the cycle time selected should ideally provide about 20 data points across a peak with a minimum of 64 data points in the retention time window (Delta Ret Window).
Figure 5.
Extracted Ion Chromatograms of 100 compound pesticide mixture (100 pg level).
Compound name Cinosulfuron Cinosulfuron (Q) Chlorotoluron Chlorotoluron (Q) Atrazine Atrazine (Q) Carbaryl Carbaryl (Q) Carboxin Carboxin (Q) Chlorsulfuron Chlorsulfuron (Q) Ethiofencarb Ethiofencarb (Q) Dodemorph Dodemorph (Q) Diuron (Q) Cyprodinil Cyprodinil (Q) Difenoxurone Difenoxurone (Q) Figure 6.
Precursor ion 414.1 414.1 213.1 213.1 216.1 216.1 202.1 202.1 236.1 236.1 358.0 358.0 226.1 226.1 283.3 282.3 233.0 226.1 226.1 287.1 287.1
Product ion 183 157 72 140 174 132 145 117 143 87 167 141 107 164 116 98 160 108 93 123 72
Retention time 5.579 5.579 5.642 5.642 5.682 5.682 5.736 5.734 5.836 5.836 5.896 5.896 5.937 5.936 6.073 6.074 6.101 6.245 6.246 6.509 6.509
Left: Table of 11 compounds monitored during a 1 minute time window. Right: Dynamic MRM of compounds being monitored.
Figure 7.
Dynamic MRM windows for each MRM transition.
Typical results achieved with the method are shown in Figure 8. It illustrates the results from one of the compounds, atrazine, in the 100-compound mixture. Note the 20 data points that were collected during the elution of atrazine. This provides a sufficient number of data points to assure quantitative accuracy and shows the effectiveness of Dynamic MRM.
Average signal height: Average signal area: RSD: Estimated LOQ:
15,650 50,966 3.2% 100 fg or less
67 data points above FWHM 3 sec FWHM, 6 sec at 10% valley 20 data points baseline-to-baseline
Figure 8.
Typical analytical results shown with 10 pg of atrazine visualizing the effectiveness of Dynamic MRM.
The calibration data from four compounds in the mixture are illustrated in Figure 9. R2's = 0.0998 are achieved for each pesticide. With constant cycle time maintained, the quantitative results with Dynamic MRM are excellent.
Desisopropyl atrazine
Desisopropylatrazine - 7 Levels, 7 Levels Used, 14 Points, 14 Points Used, 0 QCs y = 331.9557 * + 728.9399 R^2 = 0.99913754
Cymoxanil
Cymoxanil - 7 Levels, 7 Levels Used, 14 Points, 14 Points Used, 0 QCs y = 169.5944 * - 1245.2250 R^2 = 0.99873310
10 5 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 -0.2 _100
10 2 9 8 7 6 5 4 3 2 1 0
+ MRM (174.1 -> 132.0) PS-G12_DMRM-CCT_LM110308P.d
1.286
R2 = 0.99883
10 5 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1
10 2 9 8 7 6 5 4 3 2 1 0
1.286
R2 = 0.99873
Counts
Responses
Responses
Counts
10 pg
10 pg
0.9
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Acquisition time (min)
Cl N H3C N H N N NH2
0.9
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Acquisition time (min)
O NC H3 OC N N H
O NH CH3
100
200
300
400
500
600
700
800
900
1000
1100
_100
100
200
300
400
500
600
700
800
900
1000
1100
Concentration (ng/mL)
Butylate
Butylate - 7 Levels, 7 Levels Used, 14 Points, 14 Points Used, 0 QCs y = 370.5459 * - 2068.1506 R^2 = 0.99830331
Bromuconazol (diasteromers)
Bromuconazole - 7 Levels, 7 Levels Used, 14 Points, 14 Points Used, 0 QCs y = 313.5297 * - 1582.6245 R^2 = 0.99858585
10 5 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 _0.2
10 2 4.8 4.4 4 3.6 3.2 2.8 2.4 2 1.6 1.2 0.8 0.4 0
3.545
R2
= 0.99830
Counts
10 pg
3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 Acquisition time (min)
CH3 H3 C H3C CH CH CH3 CH2 N CH2 C S CH2 CH3 O
10 5 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 _0.2
10 2 8 7 6 5 4 3 2 1 0 7
7.663 8.121
R2 = 0.99859
Counts
10 pg
Responses
Responses
7.2
7.4 7.6 7.8 8 Acquisition time (min)
8.2
8.4
N N N Br
O Cl
0 100 200 300 400 500 600 700 800 900 Concentration (ng/mL)
Cl
1000 1100
_100
100
200
300
400
500
600
700
800
900
1000
1100
_100
Figure 9.
Dynamic MRM Calibration Plots, 10 pg1 ng (7 levels).
Sharper peaks are produced with a 300-pesticide mix using the new Agilent 1290 Infinity LC
Appendix III contains a partial listing of the Dynamic MRM method used to analyze a 300-pesticide mixture at the 100 pg/compound level. The LC/MS/MS extracted ion chromatogram is shown in Figure 10. The analysis took less than 20 minutes using the Agilent 1200 Series SL RRLC and an Eclipse Plus C18 2.1 mm 100 mm, 1.8 m column at a flow rate of 0.5 mL/min. The same mixture was separated using an Agilent 1290 Infinity UHPLC with an Eclipse-Plus C18,
2.1 mm 150 mm, 1.8 m column. Figure 11, an extracted ion chromatogram and Figure 12, an expanded portion of the chromatogram, demonstrate that this complex mixture has been analyzed in about 15 minutes which is approximately 25% faster than with the Agilent 1200 Series SL RRLC. The Agilent 1290 Infinity UHPLC also produced a separation with higher peak capacity and less peak overlap than the Agilent 1200 Series SL RRLC. Typical peak heights using atrazine as an example with the Agilent 1290 Infinity UHPLC are 1.8 s. This is because the longer column provides higher efficiency and the Agilent 1290 Infinity LC can operate at the pressure these conditions incurred (~900 bar).
10 6 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.6 0.9 1.4 1 1.9 2.2 2 2.7 3 3.2 3.8 4.1 4 4.7 5 6 6.8 7 7.2 17.0 8 9 10 11 12 13 14 15 16 17 18.2 18.7 18 19 19.7 3.6 5.4 7.1 6.2 4.5 5.0 5.9 6.5 7.4 8.0 8.4 8.9 9.4 5.2 9.1 11.7
10.9 1 12 23 34
15.5
16.6
12.4 7.9
10.2
13.0 12.1 14.0 14.7 15.1 11.4 13.4 19.0 14.3 17.6 16.1
5.7
9.3
10.5
9.9
Figure 10. EIC of 300 compound pesticide mixture using an Agilent 1200 Series SL RRLC.
10
10 6 1 0.9 0.8 0.7
13.9 11.9 12.1 14.8
10.7 8.5
9.5 0.6 0.5 0.4 0.3 0.2 0.1 0.7 1.0 1.4 1 1.8 2.4 3 4 4.3 2.8 2.5 3.2 3.8 4.0 4.8 5 6 7 5.2 5.4 6.3 6.7 15.3 8.2 7.5 16.1 8 9 10 11 12 13 14 15 16 17 17.5 14.4 5.7 6.0 7.0 7.3 9.1 10.1 10.5 12.7 8.7 9.8 11.5 13.0
Figure 11. EIC of 300-compound pesticide mixture using the Agilent 1290 Infinity UHPLC.
10 5 1.6
1.4
1.2
0.8
0.6
0.4
0.2
8.2
8.4
8.6
8.8
9.2
9.4
9.6
9.8
10
10.2
10.4
10.6
10.8
11
11.2
11.4
11.6
11.8
12
Figure 12. Expanded EIC of 300-compound pesticide mixture using an Agilent 1290 Infinity UHPLC illustrating the high peak capacity of the Agilent 1290 Infinity.
11
Faster analysis with a 224-pesticide mix using the new Agilent 1290 Infinity LC
Another advantage of the Agilent 1290 Infinity LC with the Agilent 6460 Series QQQ LC/MS is the ability to increase flow and decrease analysis time. Using the 1200 Series SL the analysis of 225 pesticides is performed in 15 min and shown in Figure 13. The method for this analysis is given in Appendix IV. With the Agilent 1290 Infinity LC the flow can be doubled and the gradient completed in half the time. This provides the
same separation in less than 7 min as shown in Figure 14. The method for this analysis is given in Appendix V. Analyzing hundreds of pesticides in one run, it is best to obtain the highest peak capacity as shown in the 300-pesticide example. However, if speed of analysis is absolutely necessary, it is shown that the higher pressure capability of the Agilent 1290 Infinity LC and the higher pressure capability of the HD columns provide the performance needed.
10 5 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
16.450
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Figure 13. EIC of 224 pesticides using the Agilent 1200 Series SL LC and the Agilent 6460 QQQ LC/MS.
12
10 4 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2
0.2 0.4 0.6 0.8
1.2 1.4 1.6 1.8
2.2 2.4 2.6 2.8
3.2 3.4 3.6 3.8
4.2 4.4 4.6 4.8
5.2 5.4 5.6 5.8
6.2 6.4 6.6 6.8
7.2 7.4
Figure 14. EIC of 224-pesticide mix analyzed with Agilent 1290 Infinity LC and the Agilent 6460 QQQ LC/MS.
13
Pesticides Application Kit in a food matrixSpinach using SampliQ Extraction and Dispersive SPE Kits
To demonstrate the use of the Agilent application kit for the analysis of a typical food product with Agilent's easy to use SampliQ extraction and dispersive SPE kits, a spinach matrix was spiked with 10 ppb of the 16 pesticides listed in Table 2. Triphenylphosphate (TPP) is the internal standard.
Table 2.
List of 16 Pesticides and Instrument Parameters Spiked into Spinach Matrix at 10 ppb MRM channel (m/z) Quantifier Qualifier 184.0 > 94.9 142.0 > 94.0 218.1 > 105.0 192.1 > 160.0 256.1 > 209.1 202.1 > 175.0 210.1 > 111.0 343.1 > 151.0 202.0 > 145.0 243.1 > 130.9 297.1 > 158.9 284.1 > 158.9 226.1 > 93.0 333.0 > 123.0 314.0 > 222.1 347.0 > 136.9 327.1 > 77.0 184.0 > 110.0 142.0 > 124.9 218.1 > 78.0 192.1 > 105.0 256.1 > 175.0 202.1 > 131.0 210.1 > 92.9 343.1 > 117.9 202.0 > 115.0 243.1 > 172.9 297.1 > 200.9 284.1 > 172.9 226.1 > 108.0 333.0 > 223.9 314.0 > 235.0 347.0 > 238.0 327.1 > 151.9 Fragmentor (V) 60 60 115 95 60 110 50 105 50 80 80 80 120 85 70 60 70 Collision energy (V) Quantifier Qualifier 3 8 20 18 12 27 12 17 3 15 22 32 35 28 10 25 45 15 8 50 40 18 38 15 65 40 15 15 32 35 5 10 3 45 Retention Time (min) 2.55 2.54 2.97 5.07 5.53 5.65 6.89 7.08 7.30 8.50 8.52 8.95 9.23 9.40 9.44 9.73 9.49
Analyte Acephate Methamidophos Pymetrozine Carbendazim Imidacloprid Thiabendazole Propoxur Thiophanate methyl Carbaryl Ethoprophos Imazalil Penconazole Cyprodinil Dichlorfluanid Kresoxim methyl Tolyfluanid TPP (IS)
14
Figure 15 shows the EIC of the spinach sample spiked at the 10-ppb pesticide level. All the pesticides are easily detected at this level with a total analysis time less than ten minutes.
10 2 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1
+ESI MRM Frag=120.0V CID@35.0 (226.1 -> 93.0) 1mL std 1-1.d 12 23 7 34 4
10
11 13
1 2
12 5 8 6 6.5 7 7.5 8 8.5 9
15 14 9.5 16
0.5
1.5
2.5
3.5
4.5
5.5
10
10.5
11
11.5
12
12.5
10 ppb of pesticides spiked into the spinach matrix. Peak identification: 1. 6. 11. 16. Acephate Thiabendazole Imazalil Tolyfluanid 2. Methamidophos 7. Propoxur 12. Penconazole 3. Pymetrozine 8. Thiophanate methyl 13. Cyprodinil 4. Carbendazim 9. Carbaryl 14. Dichlorfluanid 5. Imidacloprid 10. Ethoprophos 15. Kresoxim methyl,
Internal standard peak (TPP) is not shown to get clear peak profile of other small peaks.
Figure 15. EIC of 10 ppb pesticides into spinach matrix..
15
An example of the linearity achieved for the spiked spinach matrix is shown in Figure 16. The calibration range was 5 250 ng/g and seven levels were used to generate the curve, 5, 10, 25, 100, and 250 ng/g. The curve was generated by plotting the ratio of the analyte peak area, carbaryl, to the internal standard (IS) peak area with the ratio of the analytes concentration to IS concentration. The R2 = 0.998.
1.15 1.1 1.05 1 0.95 0.9 0.85 0.8 0.75 0.7 Relative responses 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 _ 0.05
Carbaryl - 7 Levels, 7 Levels Used, 14 Points, 14 Points Used, 18 QCs y = 0.4243 * - 0.0013 R^2 = 0.99787782
_0.1
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.1
2.2
2.3
2.4
2.5
2.6
2.7
Relative concentration
Figure 16. Carbaryl calibration curve.
16
Conclusions
The Agilent Pesticide Application Kit for LC/QQQ provides the user with fast method development for hundreds of pesticides with multiple transitions and the ability to develop those methods customized to his or her specific analytical needs. This application note demonstrates the use of the Agilent Application Kit for Pesticides using several Agilent technologies for screening large numbers of compounds. The following technologies are used: 600 compound pesticide MRM database and the Agilent MassHunter Data Acquisition and Analysis software. The combination gives users the ability to generate acquisition and analysis methods quickly. The methods can be easily customized and rapidly modified to meet the needs of future analyses. Dynamic MRM which maximizes the detection capability of the QQQ when hundreds of residues are being analyzed. Agilent 1200 Series SL RRLC interfaced to the Agilent 6400 series triple quadrupoles for fast and high resolution LC/MS/MS analysis. Use of the Agilent 6460 QQQ with Agilent's Jet Stream Electrospray Ion Source ensures lowest levels of detection of the pesticides. However, any of the Agilent 6400 series LC/QQQ will provide excellent results. Easy to use SampliQ QuEChERS sample preparation kits included in the Application Kit provide a fast and reproducible method to extract pesticide residues from complex food matrixes in a few simple steps. Ready to use methods with retention times for Dynamic MRM using the Agilent 1200 Series SL LC system. See all * Appendix methods.[4]
References
1. Application Note 5990-3595EN, New Dynamic MRM Mode Improves Data Quality and Triple Quad Quantification in Complex Samples. Technical Note 5990-4255EN Pesticide Dynamic Multireaction monitoring Database. Technical Note 5990-3494EN Agilent Jet Stream Thermal GradienFocusing Technology. Agilent Publication 5990-4262EN Pesticide analysis with DRMRM database quick start guide.
2. 3. 4.

Use of these technologies allows methods to be quickly developed and enables screening of complex matrices containing hundreds of potential residues at femtomole concentrations. This kit is compatible with all Agilent 1200 Series LC and 6400 series QQQ MS systems and will enable the user to quickly get started running multi-residue pesticides. For the most demanding analyses, the Agilent 1290 Infinity LC with the 6460 QQQ should be considered. Additional methods for this system should be available in the near future.
17
Appendix I
LC/MS/MS Conditions for Test mix Positive and Negative Ion Samples
Agilent 1200 Series SL LC Parameters Column: Agilent ZORBAX Eclipse Plus C18, 2.1 mm 100 mm 1.8 m Agilent p/n 959764-902 35 5 Ambient 5 s with methanol A = 5 mM acetic acid in water B = 100% acetonitrile 0.3 mL/min 5% B at t = 0 to 95% B at t = 12 min 12 min Time 3 min Jet Stream Conditions Gas temperature: Gas flow: Nebulizer: Sheath gas temperature: Sheath gas flow: Capillary + ion: Nozzle voltage: Capillary ion: Nozzle voltage: 250 C 7 L/min 40 psi 325 C 1 L/min 3500 V 0V 2500 V 1500 V
Column temperature: Injection volume: Autosampler temperature: Needle wash: Mobile phase: Flow Rate: Gradient: Stop Time: Post:
MS/MS Scans for positive ions

Compound Name Aminocarb Imazapyr Thiabendazole Dimethoate Imazalil Metoxuron Carbofuran Atrazine Metosulam Metazachlor Molinate Malathion Pyraclostrobin Diazinon ISTD? q q q q q q q q q q q q q q Prec Ion 209 262 202 230 297 229.1 222 216 418 278.1 188.1 331 388 305 MS1 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Prod Ion 137 217 131 171 159 72.1 123 132 175 134.1 55.1 99 163 153 MS2 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Frag (V) 120 160 120 80 160 93 120 120 144 75 78 80 120 160 CE (V) 20 15 30 10 20 14 15 20 26 18 22 10 20 20 Ret Time 3.128 3.959 4.072 5.064 5.918 5.992 7.019 7.437 7.472 8.038 9.113 9.615 10.679 10.776 Ret Window 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Polarity Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive
MS/MS Scans for negative ions

Compound Name Bentazon 2,4,5-T Silvex Acifluorfen Dinoseb Hexaflumuron ISTD? q q q q q q Prec Ion 239.1 252.9 266.9 360 239.1 459 MS1 Res Unit Unit Unit Unit Unit Unit Prod Ion 132 194.8 194.9 315.9 207 438.9 MS2 Res Unit Unit Unit Unit Unit Unit Frag (V) 80 76 90 78 154 102 CE (V) 32 9 5 5 21 5 Ret Time 6.572 8.047 8.805 9.650 10.503 10.877 Ret Window 1 1 1 1 1 1 Polarity Negative Negative Negative Negative Negative Negative
18
Appendix II
LC/MS/MS Conditions for 100-Pesticide Methods
Agilent 1200 Series LC Parameters Column: Agilent ZORBAX Eclipse Plus-C18, 2.1 mm 50 mm, 1.8 m Agilent p/n 959741-902 35 C 1.0 L 6 C Flushport (MeOH:H2O 75:25), 5 s A = 0.1% formic acid in water B = 0.1% formic acid in 95:5 acetonitrile:water 0.6 mL/min Time 0 10 15 20 5 %B 10 70B 90B 10B Jet Stream Conditions Drying gas temperature: Drying gas flow (nitrogen): Nebulizer gas pressure (nitrogen): Capillary voltage: Sheath gas temperature: Sheath gas flow: Nozzle voltage: Agilent 6460A QQQ settings MS1 and MS2 resolution: Time Filtering: Dynamic MRM transitions: Constant cycle time: Delta EMV: Unit Peak width = 0.03 min 200 373 ms 400 V 325 C 6 L/min 35 psig 4000 V 400 C 12 L/min Off
Column temperature: Injection volume: Autosampler temperature: Needle wash: Mobile phase:
Flow rate: Gradient
Stop time Post time
Note that example transitions, fragmentor voltages, and collision energies for this method are shown in Figure 7.
19
Appendix III
LC/MS/MS Conditions for 300-Pesticide Methods using the Agilent 1200 Series SL
Agilent 1200 Series LC Parameters Column: Agilent ZORBAX Eclipse Plus-C18, 2.1 mm 100 mm, 1.8 m Agilent p/n 959764-902 35 C 1.0 L 6 C Flushport (MeOH:H2O 75:25), 5 s A = H2O w/5 mM ammonium formate + 0.01% formic acid B = 5 mM ammonium formate + 0.01% formic acid in 95:5 acetonitrile:water 0.5 mL/min Pressure 600 600 600 600 10%B Solv ratio B 6 95 95 6 Jet Stream Conditions Drying gas temperature: Drying gas flow (nitrogen): Nebulizer gas pressure (nitrogen): Capillary voltage: Sheath gas temperature: Sheath gas flow: Nozzle voltage: 325 C 6 L/min 35 psig 4000 V 400 C 12 L/min Off
Flow rate: Gradient pump time table Time Flow 0.5 No change 18 No change 20 No change 20.01 No change Stop time 20 min Post time 5 min
Ten representative MS/MS Transitions from 300-Compound Methods

Compound Name Promecarb Promecarb Flurtamone Flurtamone Isoxaflutole Isoxaflutole Dimethenamide Dimethenamide Diethofencarb Diethofencarb ISTD? q q q q q q q q q q Prec Ion 208.1 208.1 334.1 334.1 377.1 360.1 276.1 276.1 268.2 268.2 MS1 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Prod Ion 151 109 303 247 360.1 251 244 168 226 152 MS2 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Frag (V) 80 80 120 120 100 120 120 120 80 80 CE (V) 5 10 20 30 5 10 10 15 5 20 Ret Time 11.635 11.635 11.644 11.644 11.669 11.669 11.683 11.683 11.706 11.706
20
Appendix IV
LC/MS/MS Conditions for 300-Pesticide Methods using the Agilent 1290 Infinity LC
Agilent 1290 LC Parameters Column: Agilent ZORBAX Eclipse Plus-C18, 2.1 mm 150 mm, 1.8 m RRHD 1200 Series bar columns Agilent p/n 959759-902 60 C 35 L (stacked injection, 5 L sample + 30 L H2O 6 C Flushport (MeOH:H2O 75:25 + 0.01% formic acid), 10 s A = H2O w/5 mM ammonium formate + 0.01% formic acid B = MeOH w/5 mM ammonium formate + 0.01% formic acid 0.5 mL/min 6% B (T = 0) to 98% B (T = 15 min), hold 3 min MS Parameters Sheath gas flow: Sheath gas heater: Charging Electrode: Capillary voltage: Nebulizer pressure: Drying gas temperature: Drying gas flow: 11 L/min 375 C 300 V (pos ion mode) 4 kV (pos ion mode) 35 psig 325 C 8 L/min
LC flow rate: LC gradient:
Ten representative MS/MS Scan Segments from 300-Compound Methods

Compound Name Chloridazon Chloridazon Aminocarb Aminocarb Fluroxypyr Fluroxypyr Acetamiprid Acetamiprid Vamidothion Vamidothion ISTD? q q q q q q q q q q Prec Ion 222 222 209.1 209.1 255 255 223.1 223.1 288 288 MS1 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Prod Ion 104 92 152.1 137 209 181 126 56 146 118 MS2 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Frag (V) 120 120 120 120 80 80 80 80 80 80 CE (V) 25 30 10 20 10 15 15 15 10 20 Ret Time 5.841 5.841 5.841 5.841 5.845 5.845 5.858 5.858 5.996 5.996
21
Appendix V
LC/MS/MS Conditions for Pesticides in Spinach using QuECHERS Extraction
Agilent 1200 Series HPLC conditions Column: Agilent ZORBAX Eclipse Plus Phenylhexyl, 150 mm 3 mm, 3.5 m Agilent p/n 959963-312 30 C 10 L A = 5 mM ammonium acetate, pH 5.0 in 20:80 MeOH/H2O B = 5 mM ammonium acetate, pH 5.0 in ACN Time (min) (min) 0 0.5 8.0 10.0 10.1 12.0 13.0 min 4 min 17 min 1:1:1:1 ACN/MeOH/IPA/H2O w/0.2% FA Flow rate %B (mL/min) 20 0.3 20 0.3 100 0.3 100 0.3 20 0.5 100 0.5
Column temperature: Injection volume: Mobile phase:
Needle wash: Gradient:
Stop time: Post run: Total cycle time: Positive mode Gas temperature: Gas flow: Nebulizer: Capillary:
Agilent 6410 MS conditions 350 C 10 L/min 40 psi 4000 V
22
Appendix VI
LC/MS/MS Conditions for 165-Pesticide Methods using the Agilent 1200 Series Infinity SL
Agilent 1200 Series Infinity SL LC Parameters Column: Agilent ZORBAX Eclipse Plus C18, 2.1 mm 100 mm 1.8 m Agilent p/n 959764-902 35 C 5.0 L 6 C Flushport (MeOH:H2O 75:25) 5 s A = H2O w/5mM ammonium formate + 0.01% formic acid B = 5 mM ammonium formate + 0.01% formic acid in methanol Jet Stream Conditions Spray Chamber Conditions Gas temperature: Dry gas : Nebulizer: Sheath gas temperature: Sheath gas flow: Positive cap voltage: Nozzle voltage: 200 C 6 L/min 35 psi 250 C 12 L/min 4000 V 300 V
Gradient
Pump Time Table Time (min) Solv ratio B (%) 0.00 10 1.00 10 18.00 100 20.00 100 20.10 10 25.00 10
Ten Representative MS/MS Transitions from 167-Compound Methods

Compound Name Ethiofencarb-sulfon Ethiofencarb-sulfon Clothianidin Clothianidin Imidacloprid Imidacloprid Ethiofencarb-sulfoxid Ethiofencarb-sulfoxid Monalide Monalide ISTD? q q q q q q q q q q Prec Ion 275 275 250 250 256.1 256.1 242 242 257.1 257.1 MS1 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Prod Ion 201 107 169 132 209 175.1 185 107 200.1 137.1 MS2 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Frag (V) 80 80 90 90 80 80 80 80 105 105 CE (V) 0 10 5 15 15 20 15 5 4 8 Ret Time 6.89 6.89 7.064 7.064 7.071 7.071 7.153 7.153 7.165 7.165 Ret Window 1 1 1 1 1 1 1 1 1 1
23
Appendix VII
LC/MS/MS Conditions for 224-Pesticide Methods using the Agilent 1200 Series SL
Agilent 1200 Series LC Parameters Column: Agilent ZORBAX Eclipse Plus-C18, 2.1 mm 100 mm, 1.8 m Agilent p/n 959764-902 55 C 5.0 L 6 C Flushport (MeOH:H2O 75:25), 5 s A = H2O w/5 mM ammonium formate + 0.01% formic acid B = 5 mM ammonium formate + 0.01% formic acid in 95:5 acetonitrile:water 0.3 mL/min Pressure 600 600 600 Solv ratio B 6 95 95 Jet Stream Conditions Drying gas temperature: Drying gas flow (nitrogen): Nebulizer gas pressure (nitrogen): Capillary voltage: Sheath gas temperature: Sheath gas flow: Nozzle voltage: 225 C 10 L/min 25 psig 4500 V 350 C 11 L/min 500 V
Flow rate: Gradient pump time table Time Flow 0.5 No change 14 No change 17 No change Stop time 17 min Post time 3 min
Ten representative MS/MS Transitions from 224-Compound Methods

Compound Name Buprofezin Buprofezin Sulprofos Sulprofos Eprinomectin B1a Eprinomectin B1a Chlorfluazuron Chlorfluazuron Fenpyroximat Fenpyroximat ISTD? q q q q q q q q q q Prec Ion 306.2 306.2 323 323 914.6 914.6 540 540 422.2 422.2 MS1 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Prod Ion 201.1 57.2 247.1 219 468.3 330.3 383 158 366.2 135 MS2 Res Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit Frag (V) 115 115 130 130 150 150 115 115 130 130 CE (V) 4 16 5 12 5 10 16 16 15 40 Ret Time 14.321 14.321 14.327 14.327 14.372 14.372 14.402 14.402 14.428 14.428 Ret Window 1 1 1 1 1 1 1 1 1 1
24
Appendix VIII
LC/MS/MS Conditions for 224-Pesticide Methods using the Agilent 1290 Infinity LC
Agilent 1200 Series LC Parameters Column: Agilent ZORBAX Eclipse Plus-C18, 2.1 mm 100 mm, 1.8 m Agilent p/n 959764-902 55 C 5.0 L 6 C Flushport (MeOH:H2O 75:25), 5 s A = H2O w/5mM ammonium formate + 0.01% formic acid B = 5 mM ammonium formate + 0.01% formic acid in 95:5 acetonitrile:water 0.6 mL/min Pressure 600 600 600 Solv ratio B 6 95 95 Jet Stream Conditions Drying gas temperature: Drying gas flow (nitrogen): Nebulizer gas pressure (nitrogen): Capillary voltage: Sheath gas temperature: Sheath gas flow: Nozzle voltage: 225 C 10 L/min 25 psig 4500 V 350 C 11 L/min 500 V
Flow rate: Gradient pump time table Time Flow 0.5 No change 7 No change 10 No change Stop time 10 min Post time 3 min
25
Agilent Technologies, Inc., 2009 Printed in the USA October 13, 2009 5990-4253EN
Determination of Acidic Herbicides using an Agilent 6460 Triple Quadrupole LC/MS Equipped with Agilent Jet Stream Technology and Direct Aqueous Injection, for Potable and Environmental Samples
Application Note
Environmental
Authors
Neil Cullum Anglian Water Services Laboratory Huntingdon Cambridgeshire United Kingdom Peter JW Stone Agilent Technologies Inc Stevens Creek Boulevard Santa Clara, CA, 95051 USA
Abstract
The analysis of a suite of acidic herbicides by direct aqueous injection in several water matrices is presented, therefore negating the need for solid phase extraction. Standard deviations in the range of 2.5% to 6.6% for analyte precision and recoveries of between 91.5% and 105.1% were observed over 11 batches of samples. Limits of detection were less than 10 ng/L (10 ppt) for all the compounds in the suite thus meeting current industry standards.
Introduction
Acidic herbicides cover a broad range of compounds which are widely used in crop protection and general weed control. Therefore, reliable and robust methods are required for their detection in potable and environmental waters. Currently, typical sample preparation and analytical techniques include solid phase extraction (SPE) followed by derivatisation for GC/MS or simply SPE for LC/MS analyses. Both methods are time consuming and the derivatising agents used in GC/MS techniques can be particularly hazardous. Acidic herbicides are predominately ionized in negative ionization mode via LC/MS. They give a relatively low response when compared to other herbicides, such as the phenyl urea and triazine compounds that typically undergo positive ionization. Therefore, acidic herbicides traditionally require some form of sample enrichment before analysis. With the introduction of the new Agilent 6460 Triple Quadrupole LC/MS with Jet Stream Technology an increase in sensitivity of up to 5 to 10 times can be realized when compared to the Agilent 6410 Triple Quadrupole LC/MS. This increase in sensitivity allows the analysis of this class of compounds to be performed by direct aqueous injection, eliminating costly consumables such as SPE cartridges and other time-consuming sample preparation steps. [1] The aim of this application note is to demonstrate a reliable and robust analytical method for the analysis of acidic herbicides in potable and environmental water samples, with an industry performance criteria of 12.5% analyte precision, analyte recoveries in the range of 90 to 110% and limits of detection 10 ng/L (10 ppt). The method presented here describes the analysis of a mixture of acidic herbicides (Figure 1) in different water matrices by direct aqueous injection. An overview of the full validation data is summarized.
Cl O Cl Cl O OH O OH O H3C O CH3 N H N S O O CH3 CH3
Experimental
This analysis was performed using an Agilent 6460 Triple Quadrupole Mass Spectrometer coupled to an Agilent 1200 Series LC system. The LC system consisted of a binary pump (G1312B), vacuum degasser (G1379B), automatic liquid sampler (G1367C), thermostatted column compartment (G1316B) and MassHunter data system.
Sample Preparation
Minimal sample preparation was required, which was simple acidification of all standards and samples. The samples were acidified using a concentration of 0.1% formic acid as the modifier.
Instrumentation
LC Conditions Column: Mobile phase Agilent ZORBAX SB C18, 2.1 mm 100 mm, 1.8 m, thermostatted at 60 C A: 0.2% acetic acid in HPLC water B: acetonitrile Time (min) Initial 0.5 8.0 20.0 20.1 100 L 26.0 min Electrospray interface (Positive ionization): Gas temperature: 275 C Drying gas: Nitrogen 8 L/min Nebulizer pressure: 45 psig 3500 V Vcap voltage: Sheath gas temp: 300 C Sheath gas flow: Nitrogen 11 L/min Nozzle voltage: 500 V Electrospray interface (Negative ionization): Gas temperature: 250 C Drying gas: Nitrogen 8 L/min Nebulizer pressure: 45 psig 2000 V Vcap voltage: Sheath gas temp: 300 C Sheath gas flow: Nitrogen 11 L/min Nozzle voltage: 500 1000 V Fragmentor voltage: See Table 1 See Table 1 MRM parameters: A (%) 95 95 68 35 95 B (%) 5 5 32 65 5 Flow rate mL/min 0.3 0.3 0.3 0.3 0.3
Gradient program:
Injection volume: Total run time: Spray chamber conditions:
Triple Quadrupole MS Conditions
2, 4-DB
Mecoprop
Bentazone
Figure 1.
Molecular structures of selected acid herbicides.
MRM Parameters
Table 1. Time seg 1 2 3 4 5 6 MRM Transitions for Herbicide Suite Time (min) 0.0 2.8 4.9 7.0 9.0 11.1 Delta EMV (V) 400 500 500 300 0 500 Compound Aminopyralid +ve ion Clopyralid +ve ion Picloram +ve ion Imazapyr +ve ion Quinmerac +ve ion Bromacil +ve ion Fluroxypyr +ve ion Benazolin +ve ion Bentazone +ve ion Bromoxynil -ve ion 2,4-D -ve ion MCPA -ve ion Ioxynil -ve ion Triclopyr -ve ion 2,4,5-T -ve ion Dichloroprop -ve ion Mecoprop -ve ion 2,4-DB -ve ion MCPB -ve ion Propyzamide -ve ion Precursor ion ion (m/z) 206.8 192.0 241.0 262.2 222.0 261.1 255.0 244.0 241.1 275.9 219.1 199.2 369.9 254.1 253.1 233.1 213.2 247.1 227.2 254.2 Product ions (m/z) 161.0 146.2 195.0 234.3 204.0 205.1 209.1 170.0 199.2 79.0 161.0 141.1 127.0 196.1 195.1 161.0 141.1 161.0 141.1 228.2 Fragmentor voltage (V) 75 65 75 95 75 85 90 90 80 130 80 90 135 65 75 75 80 65 65 125 Collision energy (V) 15 19 18 14 6 11 9 20 6 40 7 7 50 6 7 5 9 7 0 9 Dwell time (msec) 500 500 500 500 500 100 300 100 100 100 300 100 125 125 125 125 125 300 300 500
13.8
400
15.3
400
17.3
400
10
18.6
400

The TIC MRM chromatogram for a 0.5 g/L (500 ppt) standard consisting of this acidic herbicide suite is shown in Figure 2 which also illustrates the positioning of the time segmentation.
Figure 2. Total ion MRM chromatogram of 0.5 g/L standard.
Five levels of calibration standards were used to prepare the calibration curves over the concentration range of 0.0, 0.05, 0.10, 0.30 and 0.50 g/L (ppb). Selected and representative calibration curves are shown in Figures 3a through 3c.
(a)
(b)
(c)
Figure 3.
Calibration curves of (a) Bentazone, (b) MCPA and (c) MCPB.
Validation of the method was carried out on 11 batches of samples. Borehole groundwater, potable water (which was from a surface water source) and river water were spiked at two levels 0.01 g/L (10 ppt) and 0.10 g/L (100 ppt). Deionized water was spiked at three levels with Aqc material at 0.01 g/L, 0.10 g/L and 0.40 g/L. Each batch of samples was analyzed in duplicate and in a random order. The limit of detection (LOD) for each herbicide was calculated from the within-batch standard deviation (5 sw) of the deionized water spiked at 0.01 g/L. Recoveries for the groundwater, potable water and river water were calculated using each respective 0.1 g/L (100 ppt) matrix spike. The overall experimental and validation results are shown in Table 2.
Table 2. Compound Clopyralid Picloram Benazolin Fluroxypyr Bromacil Bentazone Bromoxynil 2,4-D MCPA Triclopyr Ioxynil Dichloroprop Mecoprop 2,4,5-T 2,4-DB MCPB Propyzamide Overall Suite Validation data %Recovery, %RSD and the Limit of Detection (LOD) Borehole %Rec 96.6, 4.6 102.1, 3.5 97.0, 5.1 100.0, 6.3 99.0, 3.6 99.2, 4.4 99.0, 4.4 96.9, 3.9 97.7, 4.1 97.8, 5.0 99.7, 4.8 98.3, 4.9 97.4, 3.4 96.2, 4.5 100.0, 4.0 98.5, 4.1 99.2, 4.2 98.5, 4.4 Tap Water %Rec 91.5, 3.4 105.1, 2.5 95.2, 3.4 96.9, 3.5 94.9, 3.6 97.1, 3.7 97.1, 5.8 96.9, 3.9 94.4, 4.4 94.9, 5.4 101.3, 4.7 96.6, 3.1 97.6, 3.5 95.1, 4.6 98.1, 3.5 97.1, 3.4 99.0, 2.5 97.0, 3.8 River Water %Rec 91.9, 6.6 99.9, 4.5 96.2, 5.3 97.3, 4.0 92.3, 4.7 96.6, 3.3 94.8, 4.4 96.4, 2.9 95.8, 3.7 97.0, 4.0 98.6, 3.7 96.7, 3.0 96.2, 2.5 96.3, 3.4 97.2, 3.5 97.4, 3.1 99.8, 3.0 96.5, 3.9 LOD (ug/L) 0.004 0.003 0.005 0.003 0.003 0.004 0.005 0.005 0.004 0.006 0.004 0.002 0.003 0.004 0.004 0.003 0.002 0.004
(a)
(b)
(c)
Aminopyralid, imazapyr and quinmerac are included in the suite but for screening analysis only. Selected and representative examples of MRM chromatograms derived from real sample matrices are shown in Figure 4.
Figure 4.
MRM chromatogram (EIC) of (a) MCPA in river water; (b) mecoprop in river water; (c) dichloroprop in river water.
Conclusions
The data show that this method is capable of a sensitive and quantitative analysis for the acidic herbicides in a single analytical suite by a direct aqueous injection of 100-L sample volumes onto the analytical column. This can be achieved without a need for SPE sample preparation; only sample acidification was undertaken as a preparation stage. All of the current industry method performance criteria are met, which include 12.5% analyte precision, recoveries in the range of 90 to 110% and limits of detection 10 ng/L (10 ppt). We demonstrate in this application note that direct aqueous injection of 100-L samples onto the analytical column achieves the required analytical method performance levels, through sensitivity increases and selectivity with the Agilent 6460 Triple Quadrupole LC/MS instrumentation equipped with Agilent Jet Stream Technology. The net benefits of such an approach to this methodology are direct and significant cost reduction in consumable items (solid phase cartridges), which are no longer required and significant labor reductions with minimal sample preparation necessary.

For more information on Agilent products and services visit our website www.agilent.com/chem For more details concerning this application note, please contact Neil Cullum at Anglian Water Services Laboratory, Huntingdon, Cambridgeshire, UK. Reference herein to any specific commercial products or nonprofit organization, process, or service by trade name, trademark, manufacturer, or other-wise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government and shall not be used for advertising or product endorsement purposes.
References
1. "Achieving the Desired Prescribed Sensitivities of Selected Herbicides by Direct On-Column Aqueous Injection of Potable and Environment Samples using the Agilent 6410BA LC/QQQ," Agilent Technologies publication 5990-3762EN.
An Application Kit for Multi-Residue Screening of Pesticides using LC/TOF or Q-TOF with a Pesticide Personal Compound Database
Application Note
Authors
Jerry Zweigenbaum and Peter Stone Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
An application kit for pesticide screening has been developed for the Agilent time-offlight (TOF) and quadrupole time-of-flight (Q-TOF) mass spectrometers using a database with almost 1600 entries. It can be quickly and easily used for both food and environmental samples where the ability to detect and identify a large number of pesticides is necessary. The system allows the user to create custom databases containing retention times of compounds of interest for targeted analysis. Screening with this database thus provides both targeted and non-targeted pesticide detection. A test mix for both positive ion and negative ion modes is provided to demonstrate the functionality of the kit. An example of a general method for pesticide screening is given along with an example of a spinach extraction using the Agilent SampliQ extraction and dispersive SPE kits for complete food analysis.
Introduction
Because over 1000 pesticides have been in use over the last century and new pesticides are being developed, there is a great need to perform both targeted and non-targeted screening in food and the environment. The Agilent time-of-flight (TOF) mass spectrometers provide both high mass resolution and mass accuracy that allow comparison of the measured mass to the exact mass of an ionized compound. In addition, the tandem hybrid quadrupole time-of-flight (Q-TOF) mass spectrometer provides the capability of both screening and confirming compounds in one instrument.[1] Both liquid chromatography (LC) combined with TOF MS and Q-TOF MS provide a robust and sensitive means to perform this type of screening at levels required by the international community. Because TOF is a pulsing instrument the resulting data is always full spectra, which allows the screening of compounds that are sought (targeted) and those that may not be expected (non-targeted).[2] In contrast LC/MS/MS with a triple quadrupole in its most sensitive mode, multi-reaction monitoring (MRM), provides targeted screening and confirmation only.[3] Recently Agilent has introduced the Pesticide Personal Compound Database (PCD) consisting of 1600 compounds and pesticides. With PCD the analyst can use the pesticide database as is for non-targeted screening or create custom databases from the read-only supplied database. The custom database can be edited by changing entries, adding, and deleting entries. In addition, a powerful feature of updating retention times allows the users custom database to be modified with retention times from the users chromatographic conditions.[4] The analyst can create as many custom databases with LC-dependent retention times as needed. This allows easy targeted (compounds verified with standards run with specific conditions) and non-target analysis (compounds in the database that have not been verified). The ability to detect and identify compounds not being sought in food and environmental samples can be very important. However, this ability must not be confused with affirmation that compounds not detected are not present. This can only be done by validation studies showing that the specific LC/MS method employed on specific matrices can detect the compounds reported as not present at the levels of concern. As an example, the pesticide database contains compounds not amenable to LC/MS such as hexachlorobenzene. These are included for the added information of the user. In addition, confirmation of positives would always require standards run with chromatographic conditions that would provide indicative retention times and additional structural information that can be obtained from fragments generated by MS/MS. Even with these analytical considerations, screening for a large list
2
of pesticides as enabled by the LC/TOF or Q-TOF with the Agilent Pesticide PCD can be very valuable in detecting and identifying compounds that should not be present.
Experimental
Pesticide standards were from a variety of sources: Sigma, Ultra-Scientific, ChemService, and Dr. Ehrenstorfer. For trace analysis the highest purity mobile phases are recommended. B&J LC/MS grade acetonitrile and methanol are used here. Buffers should be prepared from the highest quality chemicals such as GFS doubly distilled acetic acid, formic acid and ammonium hydroxide. If solid ammonium acetate and ammonium formate is used it should be prepared in a concentrated solution and then any particulates removed with 0.2-m filters. Agilent Pesticide Test Mix, p/n 5190-0469 acid and base diluted separately as instructed to 10 ppb in 10% acetonitrile/90% water. An Agilent SampliQ QuEChERS AOAC Extraction kit, p/n 5982-5755. Agilent SampliQ QuEChERS AOAC Dispersive SPE kits for Highly Pigment Fruits and Vegetables, p/n 5982-5321 (2 mL) and p/n 5982-5356 (15 mL).
LC/MS methods are given in the Appendices:

Appendix I, LC/MS/MS Conditions for Test Mix Positive and Negative Ion Samples. Appendix II. Agilent 1200 Series SL LC Parameters. Appendix III, Agilent 1290 Infinity LC Parameters.
Spinach sample preparation

Weigh 15 g (0.1 g) of homogenized spinach sample. Spike standards or IS solution if necessary. Vortex 30 s. Add 15 mL of 1% acetic acid in acetonitrile. Add 1 bag of extraction kit (p/n 5982-5982-5755) buffered QuEChERS extraction tubes, AOAC Method 2007.01 with 6 g MgSO4, 1.5 g NaAcetate. Cap and hand shake vigorously for 1 min. Centrifuge at 4000 rpm for 5 min. Transfer 1 mL or 8 mL of the upper layer to the dispersive SPE kit (p/n 5982-5321 or p/n 5982-5356) for highly pigmented fruits and vegetables. Vortex 1 min. Centrifuge 2-mL tubes at 13000 rpm for 2 min, or 15-mL tubes at 4000 rpm for 5 min.
Transfer 200 L of the upper layer to the autosampler vial. Add 800 L of water or appropriate standard spiking solution. Vortex 1 min, and prepare for LC/MS/MS analysis.

Fast and easy startup with Agilent test mix
To facilitate fast startup for pesticide screening, a positive and negative ion compound test mix is included with the Agilent Application Kit. This type of screening depends on obtaining accurate mass results and the TOF or Q-TOF should be operated with appropriate reference ions so that the best results will be obtained. Each of these test mixtures are prepared with a final injection concentration at 10 ppb, the accepted limit for pesticides worldwide. The extracted ion chromatogram (EIC) for each of the pesticides in the positive ion mix is shown in Figure 1. A method is provided with the kit that will allow the user to repeat this analysis. This method is an acquisition only method. Similar results demonstrate that
the system is working properly. There are also two methods provided for work list automation data analysis that will generate the summary report of a database search of the Pesticide PCD. One method is the MFE_pesticide and this uses the find compounds by molecular feature extraction (MFE) algorithm in MassHunter Qualitative Analysis, a powerful data mining tool. This unique data mining program searches the data for all ions that can be associated with a real chromatographic peak and that may represent a "feature" of a molecule. This excludes reference ions and constant background ions and "spikes" that do not represent real compounds in the data file. MFE will create a compound list of all peaks in the data file that it has determined to represent real molecules. This algorithm is fast and generates good results with appropriate settings. The resulting report is shown in Table 1 for the positive test mix. This mixture contains only the compounds highlighted in the report. Please note that the database search screen does not confirm the presence of compounds and that compounds listed in the results does not indicate that they are conclusively present. Compounds listed could be from the blank, carry over or other sources.
10 5 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11
1
11.492
6.740 3.456 4.601 8.821 7.803 5.861 7 6.578 8.127 . 11.480
4.531 53
9.919 10.428
11.5
Figure 1.
Extracted ion chromatogram of the positive ion test mix .
Table 1. Data File
Find Compounds by Molecular Feature Extractor with Pesticide Database Search Report for Positive Ion Test Mix TestMix_pos_1.d Sample CAS6530_1 Test_Mix_Pos.m Success Sample Name Position User Name Acquired Time DA Method 6/1/2009 3:28:51 PM MFE_Pesticide.m Test_Mix_pos_1 P1-F2
Sample Type Instrument Name Acq Method IRM Calibration Status Comment
Compound Table
Compound Label Cpd 19: Cpd 40: Cpd 41: Cpd 52: Cpd 62: Cpd 65: Cpd 66: Cpd 68: Cpd 85: Cpd 88: Cpd 89: Cpd 90: Cpd 91: Cpd 92: Cpd 93: Cpd 97: Cpd 99: Cpd 107: Cpd 111: Cpd 113: Cpd 121: Cpd 123: Cpd 125: Cpd 127: Aminocarb Imazapyr Thiabendazole Ethiofencarb sulfoxide Dimethoate Imazalil Imazalil Metoxuron Carbofuran Atrazine DEET Tibenzate Metosulam Fluoroglycofen Tibenzate Tibenzate Metazachlor Molinate Malathion Phenylacrylicacid Tri-n-butyl phosphate Tri-n-butyl phosphate Pyraclostrobin Diazinon RT 3.472 4.543 4.612 5.176 5.866 6.549 6.579 6.746 7.805 8.138 8.2 8.323 8.33 8.33 8.433 8.527 8.837 9.927 10.448 10.558 11.177 11.272 11.477 11.497 Mass 208.1213 261.1113 201.036 241.0777 228.9998 296.0488 296.0485 228.0666 221.1054 215.094 191.1309 228.0607 417.0069 419.0033 228.0608 228.0609 277.0983 187.1027 330.036 148.0522 266.1645 266.1646 387.0989 304.1012 Name Aminocarb Imazapyr Thiabendazole Ethiofencarb sulfoxide Dimethoate Imazalil Imazalil Metoxuron Carbofuran Atrazine DEET Tibenzate Metosulam Fluoroglycofen Tibenzate Tibenzate Metazachlor Molinate Malathion Phenylacrylicacid Tri-n-butyl phosphate Tri-n-butyl phosphate Pyraclostrobin Diazinon DB Formula C11H16N2O2 C13H15N3O3 C10H7N3S C11H15NO3S C5H12NO3PS2 C14H14Cl2N2O C14H14Cl2N2O C10H13ClN2O2 C12H15NO3 C8H14ClN5 C12H17NO C14H12OS C14H13Cl2N5O4S C16H9ClF3NO7 C14H12OS C14H12OS C14H16ClN3O C9H17NOS C10H19O6PS2 C9H8O2 C12H27O4P C12H27O4P C19H18ClN3O4 C12H21N2O3PS DB Diff (ppm) 0.44 0.03 0.2 1.91 0.75 1.58 0.65 0.09 1.05 0.92 0.53 1 0.98 3.28 0.39 0.12 0.53 2.02 0.2 1.59 0.58 0.32 0.9 0.56
Database Search Results

Compound Aminocarb Compound Aminocarb Hits 1 Best TRUE Formula C11H16N2O2 Mass 208.1213 Tgt Mass 208.1212 Diff (ppm) 0.44 RT 3.472

Compound Imazapyr Compound Imazapyr Hits 1 Best TRUE Formula C13H15N3O3 Mass 261.1113 Tgt Mass 261.1113 Diff (ppm) 0.03 RT 4.543

Compound Thiabendazole Compound Thiabendazole Hits 1 Best TRUE Formula C10H7N3S Mass 201.036 Tgt Mass 201.0361 Diff (ppm) 0.2 RT 4.612

Compound Ethiofencarb sulfoxide Compound Thiabendazole Hits 1 Best TRUE Formula C10H7N3S Mass 201.036 Tgt Mass 201.0361 Diff (ppm) 0.2 RT 4.612

Compound Ethiofencarb sulfoxide Compound Ethiofencarb sulfoxide Methiocarb sulfoxide Hits 2 Best TRUE FALSE Formula C11H15NO3S C11H15NO3S Mass 241.0777 241.0777 Tgt Mass 241.0773 241.0773 Diff (ppm) 1.91 1.91 RT 5.176 5.176

Compound Dimethoate Compound Dimethoate Hits 1 Best TRUE Formula C5H12N03PS2 Mass 228.9996 Tgt Mass 228.9996 Diff (ppm) 0.75 RT 5.866

Compound Imazalil Compound Imazalil Hits 1 Best TRUE Formula C14H14Cl2N2O Mass 296.0488 Tgt Mass 296.0483 Diff (ppm) 1.58 RT 6.549

Compound Imazalil Compound Imazalil Hits 1 Best TRUE Formula C14H14Cl2N2O Mass 296.0485 Tgt Mass 296.0483 Diff (ppm) 0.65 RT 6.579

Compound Metoxuron Compound Metoxuron Hits 1 Best TRUE Formula C10H13ClN2O2 Mass 228.0666 Tgt Mass 228.0666 Diff (ppm) 0.09 RT 6.746

Compound Carbofuran Compound Carbofuran Hits 1 Best TRUE Formula C12H15NO3 Mass 221.1054 Tgt Mass 221.1052 Diff (ppm) 1.05 RT 7.805

Compound Atrazine Compound Atrazine Hits 1 Best TRUE Formula C8H14ClN5 Mass 215.094 Tgt Mass 215.0938 Diff (ppm) 0.92 RT 8.138

Compound DEET Compound DEET Hits 1 Best TRUE Formula C12H17NO Mass 191.1309 Tgt Mass 191.131 Diff (ppm) 0.53 RT 8.2

Compound Tibenzate Compound Tibenzate Hits 1 Best TRUE Formula C14H12OS Mass 228.0607 Tgt Mass 228.0609 Diff (ppm) 1 RT 8.323

Compound Metosulam Compound Metosulam Hits 1 Best TRUE Formula C14H13Cl2N5O4S Mass 417.0069 Tgt Mass 417.0065 Diff (ppm) 0.98 RT 8.33

Compound Fluoroglycofen Compound Fluoroglycofen Hits 1 Best TRUE Formula C16H9ClF3NO7 Mass 419.0033 Tgt Mass 419.002 Diff (ppm) 3.28 RT 8.33

Compound Tibenzate Compound Tibenzate Hits 1 Best TRUE Formula C14H12OS Mass 228.0608 Tgt Mass 228.0609 Diff (ppm) 0.39 RT 8.433

Compound Tibenzate Compound Tibenzate Hits 1 Best TRUE Formula C14H12OS Mass 228.0609 Tgt Mass 228.0609 Diff (ppm) 0.12 RT 8.527

Compound Metazachlor Compound Metazachlor Hits 1 Best TRUE Formula C14H16ClN3O Mass 277.0983 Tgt Mass 277.0982 Diff (ppm) -0.53 RT 8.837

Compound Molinate Compound Molinate Hits 1 Best TRUE Formula C9H17NOS Mass 187.1027 Tgt Mass 187.1031 Diff (ppm) 2.02 RT 9.927

Compound Malathion Compound Malathion Hits 1 Best TRUE Formula C10H19O6PS2 Mass 330.036 Tgt Mass 330.0361 Diff (ppm) 0.2 RT 10.448

Compound Phenylacrylicacid Compound Phenylacrylicacid Hits 1 Best TRUE Formula C9H8O2 Mass 148.0522 Tgt Mass 148.0524 Diff (ppm) 1.59 RT 10.558

Compound Tri-n-butyl phosphate Compound Tri-n-butyl phosphate Tri-iso-butyl phosphate Hits 2 Best TRUE Formula C12H27O4P C12H27O4P Mass 266.1645 266.1645 Tgt Mass 266.1647 266.1647 Diff (ppm) 0.58 0.58 RT 11.177 11.177

Compound Tri-n-butyl phosphate Compound Tri-n-butyl phosphate Tri-iso-butyl phosphate Hits 2 Best TRUE Formula C12H27O4P C12H27O4P Mass 266.1646 266.1646 Tgt Mass 266.1647 266.1647 Diff (ppm) 0.32 0.32 RT 11.272 11.272

Compound Pyraclostrobin Compound Pyraclostrobin Hits 1 Best TRUE Formula C19H18ClN3O4 Mass 387.0989 Tgt Mass 387.0986 Diff (ppm) 0.9 RT 11.477

Compound Diazinon Compound Diazinon Hits 1 Best TRUE Formula C12H21N2O3PS Mass 304.1012 Tgt Mass 304.1011 Diff (ppm) 0.56 RT 11.497
The second method is Find_formula_pesticide. This method uses the find by formula algorithm of MassHunter Qualitative Analysis. This algorithm searches the data for the ions specified for each molecule in the database. For the supplied database this would entail generating extracted ion chromato-grams for each entry times each adduct (1600 for H+, 1600 for Na+, etc.). This is thorough but slower. However, if these searches are done automatically in a worklist, the processing time is reasonable. The analyst must determine what is the best fit-for-purpose procedure. Note that automatic database searching can be done during the worklist acqui-
sition or after. Using the worklist run parameter of MassHunter acquisition, acquisition and data analysis can be selected, or data analysis only after the data has been collected. The data analysis methods can be added to the worklist by adding the column "Override DAMethod" to the MassHunter worklist and inserting the method to be used. (The qualitative analysis methods can be saved to the name of the acquisition method eliminating the need to add the "Override" column. (However, keeping acquisition methods and data analysis methods separate provides more flexibility.) All methods can be customized to meet the needs of a particular analysis.
Figure 2 shows the chromatogram of the negative ion compound test mix. Table 2 shows the results automatically generated for the negative ion mixture using the MFE_pesticide method. The report is generated using "Find and Identify" selection of compound automation and the 1600-compound pesticide database, pesticides.mtl, is searched. The worklist automation uses the "Compound automation and report" selection. To obtain the report shown, the "CompoundReportwithIdentificationHits.xltx" template of the "Common Reporting Options" in the General Navigation bar of MassHunter Qualitative analysis must be selected. This is important because as shown in the compound list of Table 2, the wrong isomer, dinoprop, is listed. This is the first isomer found in the database. The selected report template then lists
the results of each database hit and the three isomers in the database are shown under this heading in the report. If the data were analyzed with the "Find by Formula" algorithm, the report would include all the isomers in the database in the main body of the report. If a retention time that matched the compound were in a custom database, only that isomer would be reported (targeted analysis). (Note that for the find by formula method to work within a worklist the Worklist Actions of the method should separately list "Compound Automation without report" and then "Generate Compound Report.") The compound actually present is dinoseb and if this were a non-targeted analysis the analyst would need to confirm which one was present.
10 6 4.6 1 4.4 4.2 4 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7
11.267
1
7.496
11.526 8.934 9.629 10.713
7.5
8.5
9.5
10
10.5
11
11.5
Figure 2.
Extracted ion chromatogram of the negative ion test mix.
Table 2. Data File
Find compounds by Molecular Feature Extractor with Pesticide Database Search Report for Negative Ion Test Mix Test_mix_neg_01.d Sample CAS6530_1 Test_mix_neg.m Success Sample Name Position User Name Acquired Time DA Method 6/1/2009 1:33:54 PM MFE_Pesticide.m Test Mix Neg 1 P1-F1
Compound Table
Compound Label Cpd 12: Cpd 15: Cpd 24: Cpd 26: Cpd 32: Cpd 37: Cpd 39: Cpd 41: Cpd 42: Cpd 51: Cpd 52: Cpd 56: Bentazone Dibutyl succinate 2,4-D Methyl ester 2,4,5-T Silvex Citronellal hydrate Citronellal hydrate Acifluorfen Citronellal hydrate Alantolactone Dinoprop (see below) Hexaflumuron RT 7.491 7.904 8.768 8.934 9.623 10.219 10.37 10.716 10.736 11.249 11.267 11.53 Mass 240.0573 230.1517 233.9847 253.9306 267.9465 172.1465 172.1464 360.9967 172.1466 232.1462 240.075 459.982 Name Bentazone Dibutyl succinate 2,4-D Methyl ester 2,4,5-T Silvex Citronellal hydrate Citronellal hydrate Acifluorfen Citronellal hydrate Alantolactone Dinoprop Hexaflumuron DB Formula C10H12N2O3S C12H22O4 C9H8Cl2O3 C8H5Cl3O3 C9H7Cl3O3 C10H20O2 C10H20O2 C14H7ClF3NO5 C10H20O2 C15H20O2 C10H12N2O5 C16H8Cl2F6N2O3 DB Diff (ppm) 1.69 0.5 1.7 0.72 1.6 1.25 0.47 0.55 1.37 0.35 1.72 0.76

Compound Bentazone Compound Bentazone Hits 1 Best TRUE Formula C10H12N2O3S Mass 240.0573 Tgt Mass 240.0569 Diff (ppm) 1.69 RT 7.491

Compound Dibutyl succinate Compound Dibutyl succinate Hits 1 Best TRUE Formula C12H22O4 Mass 230.1517 Tgt Mass 230.1518 Diff (ppm) 0.5 RT 7.904
10

Compound 2,4,5-T Compound 2,4,5-T Tricamba Hits 2 Best TRUE Formula C8H5Cl3O3 C8H5Cl3O3 Mass 253.9306 253.9306 Tgt Mass 253.9304 253.9304 Diff (ppm) 0.72 0.72 RT 8.934 8.934

Compound Silvex Compound Silvex Hits 1 Best TRUE Formula C9H7Cl3O3 Mass 267.9465 Tgt Mass 267.9461 Diff (ppm) 1.6 RT 9.623

Compound Citronellal hydrate Compound Citronellal hydrate Hits 1 Best TRUE Formula C10H20O2 Mass 172.1465 Tgt Mass 172.1463 Diff (ppm) 1.25 RT 10.219


Compound Acifluorfen Compound Acifluorfen Hits 1 Best TRUE Formula C14H7ClF3NO5 Mass 360.9967 Tgt Mass 360.9965 Diff (ppm) 0.55 RT 10.716


Compound Alantolactone Compound Alantolactone Hits 1 Best TRUE Formula C15H20O2 Mass 232.1462 Tgt Mass 232.1463 Diff (ppm) 0.35 RT 11.249
11
Database Search Results (Note that the following are isomers with the same formula even though the compound present is listed as "FALSE.")
Compound Dinoprop Compound Dinoprop Dinoseb Dinoterb Hits 3 Best TRUE FALSE Formula C10H12N2O5 C10H12N2O5 C10H12N2O5 Mass 240.075 240.075 240.075 Tgt Mass 240.0746 240.0746 240.0746 Diff (ppm) 1.72 1.72 1.72 RT 11.267 11.267 11.267

Compound Hexaflumuron Compound Hexaflumuron Hits 1 Best TRUE Formula C16H8Cl2F6N2O3 Mass 459.982 Tgt Mass 459.9816 Diff (ppm) 0.76 RT 11.53
Customized databases with user added retention times

One of the powerful benefits of the supplied database is that it can be saved to a user customized database. To create a read-write customizable database, the user selects the "File" menu item and the "New Database." The software then allows selection of an existing database and then the naming of a new database. A description can also be given. When
"Create" is selected the database with the new name contains all the entries of the selected database. In this way multiple custom databases can be created. The technical note on the Pesticide PCD [4] shows how the user can run standards with unique chromatographic conditions and easily update retention times in their custom database. This is shown in Figures 3 and 4 for the positive and negative ion test mix respectively.
Figure 3.
Pesticide Personal Compound Database (now with Library PCDL, not shown) customized with retention times from positive ion test mix.
12
Figure 4.
Pesticide Personal Compound Database (PCDL) customized with retention times for negative ion test mix.
If the analysis were for targeted compounds where retention times are known, dinoseb would be chromatographically separated from the other isomers. It is a simple exercise to take the results of the test mix, create a custom database from the provided pesticide database and update retention times. This would now create a targeted analysis. Either data analysis method can be modified for a targeted and non-targeted analysis by selecting "mass and retention time (optional)" for the search criteria. Targeted only analysis would be performed
if "mass and retention time (required)" was checked. A report for a targeted and non-targeted analysis of the negative test mix with the method Find_by Formula and a custom database with the retention time for dinoseb would only list that compound. In this result, only dinoseb is reported because it is the compound in the custom database that matches the retention time. Even with retention times, identified compounds in the database must be confirmed. Both screening and confirmation can be done with the LC/Q-TOF.
13
Table 3. Data File
Find compounds by Formula with Pesticide Database Search Report for Negative ion Test Mix Test_mix_neg_01.d Sample CAS6530_1 Test_mix_neg.m Success Sample Name Position User Name Acquired Time DA Method 6/1/2009 1:33:54 PM find_by_formula_pesticids.m Test Mix Neg 1 P1-F1
Compound Table
Compound Label Cpd 1: Cpd 2: Cpd 3: Cpd 6: Cpd 7: Cpd 4: Cpd 5: Cpd 9: Cpd 8: Dichloromethoxybenzene Bentazone Dibutyl succinate Dichloroprop Disugran Dichlorophenol 2,42,4-D Methyl ester Tricamba 2,4,5-T RT 5.583 7.492 7.904 8.764 8.764 8.764 8.764 8.941 8.941 9.613 9.624 10.708 10.708 10.708 10.732 11.251 11.262 11.262 11.262 11.262 11.262 11.533 Mass 175.9796 240.0573 230.1517 233.9845 233.9845 161.9633 233.9845 253.9306 253.9306 195.9245 267.9468 360.9966 317.0062 317.0061 172.1466 232.1462 241.078 241.078 240.0752 240.0752 240.0752 459.9823 Abund 9712 108523 7790 33463 33463 6051 33463 15646 15646 5877 18804 18261 10928 9536 132453 10414 23056 23056 249379 249379 249379 19824 Name Dichloromethoxybenzene Bentazone Dibutyl succinate Dichloroprop Disugran Dichlorophenol 2,42,4-D Methyl ester Tricamba 2,4,5-T Trichlorophenol, 2,4,6Silvex Acifluorfen Nitrofluorfen Azinphos-methyl Citronellal hydrate Alantolactone Ethiofencarb sulfoxide Methiocarb sulfoxide Dinoprop Dinoseb Dinoterb Hexaflumuron Formula C7H6Cl2O C10H12N2O3S C12H22O4 C9H8Cl2O3 C9H8Cl2O3 C6H4Cl2O C9H8Cl2O3 C8H5Cl3O3 C8H5Cl3O3 C6H3Cl3O C9H7Cl3O3 C14H7ClF3NO5 C13H7ClF3NO3 C10H12N3O3PS2 C10H20O2 C15H20O2 C11H15NO3S C11H15NO3S C10H12N2O5 C10H12N2O5 C10H12N2O5 C16H8Cl2F6N2O3 Tgt Mass 175.9796 240.0569 230.1518 233.985 233.985 161.9639 233.985 253.9304 253.9304 195.9249 267.9461 360.9965 317.0067 317.0058 172.1463 232.1463 241.0773 241.0773 240.0746 240.0746 240.0746 459.9816 DB Diff (ppm) 0.11 1.69 0.5 2.39 2.39 3.59 2.39 0.75 0.75 2.12 2.64 0.43 1.41 1 1.39 0.54 2.86 2.86 2.22 2.22 2.22 1.44
Cpd 10: Trichlorophenol, 2,4,6Cpd 11: Silvex Cpd 14: Acifluorfen Cpd 13: Nitrofluorfen Cpd 12: Azinphos-methyl Cpd 15: Citronellal hydrate Cpd 16: Alantolactone Cpd 20: Ethiofencarb sulfoxide Cpd 21: Methiocarb sulfoxide Cpd 19: Dinoprop Cpd 17: Dinoseb Cpd 18: Dinoterb Cpd 22: Hexaflumuron
14
The power of Q-TOF for screening and confirmation

As an example of the power of this technique, a strawberry extract was spiked and analyzed using an Agilent 1200 Series SL LC with an Agilent 6520 Q-TOF. The extracted ion chromatogram of the over 100 pesticides spiked into this sample is shown in Figure 5. A pesticide screen with a Q-TOF is the same as with a TOF. However, LC/Q-TOF MS offers the highly selective MS/MS with accurate mass measurement that provides a workflow for both screening and confirmation. [1]
with the Agilent Pesticide Screen Application Kit for TOF and Q-TOF. The highest quality results are obtained with good chromatographic and mass spectral resolution. The ability to detect and identify thousands of compounds lies in both these parameters and accurate mass measurement. However, for any given real food sample only a few pesticides will be found. This may not be the case for environmental samples but the possibility of no more than 10 to 20 per site would be realistic. Given this reality, the need to be able to validate that hundreds of compounds can be detected in a fast analysis would provide this capability. Figure 6 shows a 3-minute run of over 100 pesticides using the new Agilent 1290 Infinity LC connected to the new Agilent 6540 Q-TOF. Given the chromatographic resolution achieved and the mass spectral resolution obtained, this analysis is reasonable for screening pesticides in food and environmental samples. The quality of the mass spectral data is shown in Figure 7 and this was collected at rate of 10 spectra per second.
Screening hundreds of target and non-target pesticides using the Agilent 1200 Series SL with 6230 TOF
A standard of over 200 pesticides is run in a similar fashion and the EIC generated from the pesticides detected in a find compounds by molecular feature extractor with database search is shown in Figure 4. This method employs the Agilent 1200 Series SL and the Agilent 6230 TOF with Jet Stream Technology. This is the preferred configuration as it provides additional sensitivity to meet the demanding needs of multiresidue analysis. The method for this analysis is also provided
10 6 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 Counts vs. acquisition time (min) 11 12 13 14 15 16 Cpd 323: Tribufos: +ESI ECC Scan Frag=150.0V SeqR_16-r003.d
15.212 Cpd 323: Tribufos
Figure 5. Extracted compound chromatogram (from compounds found by MFE) of 200 pesticides using the Agilent 1200 Series SL LC with the Agilent 6230 TOF.
15
10 5 2 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Counts vs. acquisition time (min) 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 0.649
Figure 6. Extracted compound chromatogram of 100 pesticides in 3 min using the new Agilent 1290 Infinity LC with the new Agilent 6540 Q-TOF.
10 4 2.3 2.2 2.1 2 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
294.1366
m/z 294.1366 295.1396 296.1337
Abund 21636 3591 6601
Abund % 100 16.6 30.51
Charge 1 1 1
Sat
Width 0.0095 0.0101 0.01
Rea 31044 29122 29608
Cl CH3 N OH N N CH3 N OH N N
Cl
H3C H3C
H3C H3C
296.1337
295.1396
294
294.25
294.5
294.75
295
295.25
295.5
295.75
296
296.25
296.5
296.75
297
297.25
297.5
297.75
Counts vs. mass-to-charge (m/z)
Figure 7.
Example mass spectrum from data on 3 min run with Agilent 1290 Infinity LC and Agilent 6540 Q-TOF. Note the mass resolution at 10 spectra per second.
16
Extraction to analysis with SampliQ extraction and SPE Kits

Finally, as an example of a complete analysis a spinach sample was spiked with pesticides at the 10-ppb level and extracted using the SampliQ QuEChERS Kit p/n 5982-5755. Then the Agilent SampliQ QuEChERS AOAC Dispersive SPE kit for highly pigmented fruits and vegetables, p/n 5982-5356 (15 mL), was used for clean-up. In addition, a reagent blank was prepared and run using an Agilent 1200 Series SL/6530 LC/Q-TOF and the standard screened with find by molecular feature extractor and the Pesticide database (not customized). The resulting mass list from the reagent blank was placed in the MFE settings by exporting the mass list to a .csv file,
Table 4. Data File Sample Type Instrument Name Acq Method IRM Calibration Status Comment Neat Pesticide Standard for Spinach Extract 2ppb neat std .d Sample CAS6530_1 MFE_Compound_report.m Success
selecting "exclude these masses" under "Filter Mass", and using the exported .csv file as the database. In this way all the ions in the reagent blank will be removed from standards and samples processed with this method. The spiking solution (neat standard) was analyzed using the same acquisition method and the Worklist Automation. The results are given in Table 4 and represent the pesticides in the standard. It should be noted that if background removal is performed, the mass list should be searched by the database to make sure that compounds of concern will not be excluded. The .csv file is editable in Excel and masses can be removed from the exclusion list if necessary (for example, if pesticides are found in the blank).
Sample Name Position User Name Acquired Time DA Method
2ppb neat in 20:80 ACN/H2O P1-F1 Jaz 7/10/2009 12:43:56 PM MFE_Pesticide_report.m
Compound Table
Compound Label Cpd 12: Cpd 15: Cpd 18: Cpd 24: Cpd 25: Cpd 29: Cpd 35: Cpd 44: Cpd 51: Cpd 54: Cpd 55: Cpd 86: Cpd 87: Cpd 89: Cpd 91: Cpd 92: Cpd 97: Cpd 99: Methamidophos Acephate Acephate Pymetrozine Pymetrozine Carbendazim Thiabendazole Imidacloprid Imazalil Dicyclanil Thiophanate-methyl Propoxur Pyrocatechol Norethynodrel Carbaryl Naphthol, 1Ethoprop Penconazole RT 2.053 2.467 2.632 3.242 3.361 4.259 4.633 5.564 6.587 7.172 7.426 7.621 7.621 7.631 7.994 7.995 9.908 10.219 10.482 10.924 Mass 141.0012 183.0115 183.0119 217.0967 217.0965 191.0695 201.0359 255.0527 296.0489 190.0968 342.0463 209.1053 110.0368 298.192 201.079 144.0575 242.0569 283.0649 225.1268 313.132 Name Methamidophos Acephate Acephate Pymetrozine Pymetrozine Carbendazim Thiabendazole Imidacloprid Imazalil Dicyclanil Thiophanate-methyl Propoxur Pyrocatechol Norethynodrel Carbaryl Naphthol, 1Ethoprop Penconazole Cyprodinil Kresoxim methyl DB Formula C2H8NO2PS C4H10NO3PS C4H10NO3PS C10H11N5O C10H11N5O C9H9N3O2 C10H7N3S C9H10ClN5O2 C14H14Cl2N2O C8H10N6 C12H14N4O4S2 C11H15NO3 C6H6O2 C20H26O2 C12H11NO2 C10H8O C8H19O2PS2 C13H15Cl2N3 C14H15N3 C18H19NO4 DB Diff (ppm) 0.96 2.24 0.02 1.54 0.5 0.06 0.81 1.43 1.92 0.77 1.97 0.35 0.59 4.2 0.07 0.01 2 2.26 0.68 1.93 1 1 1 1 1 1 1 1 1 7 1 1 1 1 1 Hits (DB) 1 1 1 1
Cpd 101: Cyprodinil Cpd 105: Kresoxim methyl
17
Figure 8 shows the extracted compound chromatogram from the Molecular Feature Extractor of the spinach extract. Even with the clean-up procedure and background ions removed, this is a complex sample. Table 5 shows the database search result for the spinach extract and all compounds detected in the standards were detected in the extract. If this were an analysis done for targeted and non-targeted analysis, all non-target positives (those without matching retention times) should be examined in MassHunter Qualitative Analysis before further analysis.
Figure 8.
Extracted compound chromatogram of spinach sample with over 1200 compound features found .
18
Table 5. Data File
Results of Spinach Screen using Molecular Feature Extractor Spinach AOAC 10ppb.d Sample CAS6530_1 MFE_Compound_report.m Some Ions Missed Sample Name Position User Name Acquired Time DA Method Spinach AOAC 10 ppb (2 ppb in sample) P1-A4 Jaz 7/10/2009 12:40:59 PM MFE_Pesticide_report.m
Compound Table
Compound Label Cpd 36: Cpd 42: Cpd 44: Cpd 74: Cpd 94: Cpd 97: Cpd 103: Cpd 119: Cpd 121: Cpd 136: Cpd 140: Cpd 161: Cpd 198: Cpd 207: Cpd 221: Cpd 284: Cpd 303: Cpd 310: Cpd 316: Cpd 323: Cpd 338: Cpd 368: Cpd 386: Cpd 455: Cpd 461: Cpd 492: Cpd 584: Methamidophos Carbofuran-3-OH-7-phenol Metolcarb Acephate Decarbofuran Quinacetol sulfate 3,5-Xylyl methylcarbamate Carbofuran, - 3 hydroxy Pymetrozine Propoxur 3,5-Xylyl methylcarbamate 8-Hydroxyquinoline Metalaxyl Phenyl isocyanate Aspidinol Phenoxyacetic acid Dimethyl phthalate Trinexapac Carbendazim Geraniol Dimethyl phthalate Propoxur Aldicarb Phenoxyacetic acid Thiabendazole Butopyronoxyl Tiocarbazil RT 2.042 2.226 2.233 2.614 3.127 3.14 3.197 3.323 3.36 3.445 3.471 3.59 3.805 3.817 3.946 4.155 4.188 4.21 4.254 4.282 4.316 4.372 4.443 4.617 4.628 4.723 4.904 Mass 141.001 180.0786 165.0789 183.0116 207.0896 187.0631 179.0945 237.1003 217.0963 209.1053 179.0953 145.0526 279.1478 119.0373 224.1049 152.0472 194.0579 224.0685 191.0695 154.1357 194.0579 209.1054 190.0776 152.0474 201.0362 226.1207 279.167 Name Methamidophos Carbofuran-3-OH-7-phenol Metolcarb Acephate Decarbofuran Quinacetol sulfate 3,5-Xylyl methylcarbamate Carbofuran, - 3 hydroxy Pymetrozine Propoxur 3,5-Xylyl methylcarbamate 8-Hydroxyquinoline Metalaxyl Phenyl isocyanate Aspidinol Phenoxyacetic acid Dimethyl phthalate Trinexapac Carbendazim Geraniol Dimethyl phthalate Propoxur Aldicarb Phenoxyacetic acid Thiabendazole Butopyronoxyl Tiocarbazil DB Formula C2H8NO2PS C10H12O3 C9H11NO2 C4H10NO3PS C11H13NO3 C11H9NO2 C10H13NO2 C12H15NO4 C10H11N5O C11H15NO3 C10H13NO2 C9H7NO C15H21NO4 C7H5NO C12H16O4 C8H8O3 C10H10O4 C11H12O5 C9H9N3O2 C10H18O C10H10O4 C11H15NO3 C7H14N2O2S C8H8O3 C10H7N3S C12H18O4 C16H25NOS DB Diff (ppm) 2.37 0.12 0.43 1.72 0.45 1.39 0.78 0.88 0.44 0.42 3.52 0.82 2.52 1.32 0.24 1.13 0.28 0.08 0.21 0.55 0.1 1.02 0.09 0.33 0.66 0.65 4.82
19
Compound Label Cpd 587: Cpd 641: Cpd 642: Cpd 644: Cpd 720: Cpd 721: Cpd 740: Cpd 743: Cpd 804: Cpd 816: Cpd 830: Cpd 858: Cpd 976: Kresoxim methyl Pyrethrin I Allethrin Spiromesifen Phosfon Santonin Dimethyl phthalate Metaldehyde Phosfon Allethrin Buthiobate Imidacloprid Kresoxim methyl
RT 4.905 5.022 5.022 5.022 5.19 5.197 5.255 5.265 5.439 5.454 5.498 5.57 6.15 6.366 6.463 6.521 6.595 6.838 6.93 7.152 7.22 7.419 7.596 7.614 7.615 7.7 7.778 7.996 7.996 8.128 8.278 8.364 8.621 8.752 8.797 8.81 8.908 9.184 9.22 9.279 9.883
Mass 313.132 328.2041 302.1884 370.2148 396.1312 246.1259 194.0576 176.1044 396.1301 302.1882 372.1696 255.0527 313.1321 313.132 232.1463 246.1258 296.049 239.1522 226.1206 274.1937 190.0968 342.0457 228.1141 209.1052 110.0369 274.1933 228.1147 144.0574 201.0791 717.4462 731.4619 297.2675 745.4775 294.1838 274.1935 246.1261 274.1935 274.1936 363.8501 363.8502 289.1417
Name Kresoxim methyl Pyrethrin I Allethrin Spiromesifen Phosfon Santonin Dimethyl phthalate Metaldehyde Phosfon Allethrin Buthiobate Imidacloprid Kresoxim methyl Kresoxim methyl Alantolactone Santonin Imazalil Salbuterol Butopyronoxyl Cinmethylin Dicyclanil Thiophanate-methyl Bisphenol A Propoxur Pyrocatechol Cinmethylin Bisphenol A Naphthol, 1Carbaryl Spinosyn B Spinosyn A Spiroxamine Spinosyn D Embelin Cinmethylin Santonin Cinmethylin Cinmethylin Bromophos Bromophos Imazethapyr
DB Formula C18H19NO4 C21H28O3 C19H26O3 C23H30O4 C19H32Cl3P C15H18O3 C10H10O4 C8H16O4 C19H32Cl3P C19H26O3 C21H28N2S2 C9H10ClN5O2 C18H19NO4 C18H19NO4 C15H20O2 C15H18O3 C14H14Cl2N2O C13H21NO3 C12H18O4 C18H26O2 C8H10N6 C12H14N4O4S2 C15H16O2 C11H15NO3 C6H6O2 C18H26O2 C15H16O2 C10H8O C12H11NO2 C40H63NO10 C41H65NO10 C18H35NO2 C42H67NO10 C17H26O4 C18H26O2 C15H18O3 C18H26O2 C18H26O2 C8H8BrCl2O3PS C8H8BrCl2O3PS C15H19N3O3
DB Diff (ppm) 1.89 0.8 0.79 1.01 1.33 1.43 1.68 2.75 1.47 0.08 0.67 1.47 2.14 1.78 0.23 0.98 2.29 0.34 0.52 1.56 0.54 0.24 4.06 0.1 1.11 0.16 1.28 0.69 0.39 1.38 1.43 2.27 1.29 2.25 0.9 2.01 0.82 1.21 2.55 2.57 3.08
Cpd 1047: Kresoxim methyl Cpd 1063: Alantolactone Cpd 1075: Santonin Cpd 1089: Imazalil Cpd 1124: Salbuterol Cpd 1144: Butopyronoxyl Cpd 1212: Cinmethylin Cpd 1242: Dicyclanil Cpd 1274: Thiophanate-methyl Cpd 1331: Bisphenol A Cpd 1335: Propoxur Cpd 1337: Pyrocatechol Cpd 1348: Cinmethylin Cpd 1357: Bisphenol A Cpd 1394: Naphthol, 1Cpd 1395: Carbaryl Cpd 1410: Spinosyn B Cpd 1429: Spinosyn A Cpd 1438: Spiroxamine Cpd 1477: Spinosyn D Cpd 1494: Embelin Cpd 1501: Cinmethylin Cpd 1506: Santonin Cpd 1526: Cinmethylin Cpd 1574: Cinmethylin Cpd 1578: Bromophos Cpd 1604: Bromophos Cpd 1715: Imazethapyr
20
Compound Label Cpd 1728: Ethoprop Cpd 1746: Imazethapyr Cpd 1782: Penconazole Cpd 1797: Cinmethylin Cpd 1829: Cyprodinil Cpd 1833: Chenodeoxycholic acid Cpd 1838: Embelin Cpd 1885: Chenodeoxycholic acid Cpd 1908: Cinmethylin Cpd 1927: Kresoxim methyl Cpd 1933: Cinmethylin Cpd 1951: Cinmethylin Cpd 1978: Carbofuranphenol Cpd 2017: Spiroxamine Cpd 2044: Spiroxamine Cpd 2054: Spiroxamine Cpd 2062: Spiroxamine Cpd 2096: Etacelasil Cpd 2205: Spiroxamine Cpd 2206: Ivermectin B1b Cpd 2253: Ivermectin B1b Cpd 2314: Ivermectin B1b Cpd 2474: Ivermectin B1b
RT 9.911 10.027 10.224 10.325 10.49 10.526 10.541 10.749 10.852 10.933 10.967 11.089 11.235 11.536 11.731 11.846 11.955 12.24 12.797 12.806 13.128 13.372 14.122
Mass 242.0566 289.1419 283.0649 274.1931 225.1271 392.291 294.1837 392.2909 274.1936 313.1321 274.1934 274.1939 164.0841 297.2666 297.2663 297.2673 297.2672 316.1098 297.267 860.488 860.4891 860.4896 860.4906
Name Ethoprop Imazethapyr Penconazole Cinmethylin Cyprodinil Chenodeoxycholic acid Embelin Chenodeoxycholic acid Cinmethylin Kresoxim methyl Cinmethylin Cinmethylin Carbofuranphenol Spiroxamine Spiroxamine Spiroxamine Spiroxamine Etacelasil Spiroxamine Ivermectin B1b Ivermectin B1b Ivermectin B1b Ivermectin B1b
DB Formula C8H19O2PS2 C15H19N3O3 C13H15Cl2N3 C18H26O2 C14H15N3 C24H40O4 C17H26O4 C24H40O4 C18H26O2 C18H19NO4 C18H26O2 C18H26O2 C10H12O2 C18H35NO2 C18H35NO2 C18H35NO2 C18H35NO2 C11H25ClO6Si C18H35NO2 C47H72O14 C47H72O14 C47H72O14 C47H72O14
DB Diff (ppm) 0.93 2.59 1.99 0.68 2.07 4.36 2.12 4.45 1.24 2.18 0.53 2.13 2.38 0.5 1.49 1.65 1.26 3.42 0.58 4.9 3.61 3.04 1.85
21
Conclusions
The Agilent TOF and Q-TOF Pesticide Application Kit has been developed to provide comprehensive screening of pesticides for both targeted and non-targeted compounds. The database includes almost 1600 compounds and gives the user great flexibility in its use. The kit offers: Fast and easy startup of complex analyses A comprehensive pesticide database of almost 1600 compounds including: Chemical structures, formulas and exact masses Direct Chemical Internet links to PUBCHEM and Chemspider IUPAC Names The ability to create spectral libraries Completely customizable additions/deletions and retention time additions for chromatographic conditions developed by the user
References
1. Agilent Technologies publication 5990-3935EN, Q-TOF LC/MS Screening and Confirming of Non-Targeted Pesticides in a Strawberry Extract. Agilent Technologies publication 5989-5496EN, Automated Screening of 600 Pesticides in Food by LC/TOF MS Using a Molecular-Feature Database Search. Agilent Technologies publication 5990-4253EN, MultiResidue Pesticide Analysis with Dynamic Multiple Reaction Monitoring and Triple Quadrupole LC/MS/MS. Agilent Technologies publication 5990-3976EN, Pesticide Personal Compound Database for Screening and Identification.
2.
3.
4.

Results can be searched directly from the PCDL software Results can be data-mined with powerful searching tools such as, the Molecular Feature Extractor and Find by Formula Searches of the database can be partially or completely automated using MassHunter Qualitative Analysis and the MassHunter Acquisition Worklist
22
Appendix I
LC/MS/MS Conditions for Test mix Positive and Negative Ion Samples
Agilent 1200 Series SL LC Parameters Column: Agilent ZORBAX Eclipse Plus C18, 2.1 mm 100 mm, 1.8 m Agilent p/n 959764-902 35 C 5 Ambient 5 s with methanol A = 5 mM acetic acid in water B = 100% acetonitrile 0.3 mL/min 5% B at t = 0 to 95% B at t = 12 min 12 min 3 min
Column temperature: Injection volume: Autosampler temperature: Needle wash: Mobile phase: Flow rate: Gradient: Stop time: Post time: Agilent 6530 Q-TOF Parameters Jet Stream Conditions Gas temperature: Gas flow: Nebulizer: Sheath gas temperature: Sheath gas flow: Capillary + ion: Nozzle voltage: Capillary ion: Nozzle voltage: Acquisition Mode: Min Range Max Range Scan Rate Reference Masses:
250 C 7 L/min 40 psi 325 C 11 L/min 3500 V 0V 2500 V 1500 V MS1 100 m/z 1100 m/z 1.4 per s Positive ion 121.050873 (M+H+ for purine) 922.009798 (M+H+ for HP-921) Negative ion 119.0362 (MH for purine) 980.016375 (M+C2H3O2 for HP-921 acetate adduct)
Reference Masses:
23
Appendix II
Agilent 1200 Series SL LC Parameters
Agilent 1200 Series LC Parameters Column: Agilent ZORBAX Eclipse Plus C18, 2.1 mm 100 mm, 1.8 m Agilent p/n 959764-902 55 C 5.0 L 6 C Flushport (MeOH:H2O 75:25), 5 s A = H2O w/5 mM ammonium formate + 0.01% formic acid B = 5 mM ammonium formate + 0.01% formic acid in 95:5 acetonitrile:water 0.3 mL/min Pressure 600 600 600 Solv ratio B 6 95 95
Flow rate: Gradient pump time table Time Flow 0.5 No change 14 No change 17 No change Stop time 17 min Post time 3 min Agilent 6230 TOF Parameters Jet Stream Conditions Drying gas temperature: Drying gas flow (nitrogen): Nebulizer gas pressure (nitrogen): Capillary voltage: Sheath gas temperature: Sheath gas flow: Nozzle voltage: Acquisition Mode MS1 Min Range Max Range Scan Rate Reference Masses:
225 C 9 L/min 25 psig 4500 V 350 C 11 L/min 500 V 25 m/z 3200 m/z 3 Positive ion 121.050873 (M+H+ for purine) 922.009798 (M+H+ for HP-921)
24
Appendix III
Agilent 1290 Infinity LC Parameters
Column: Agilent ZORBAX Eclipse Plus C18 HD, 2.1 mm 100 mm, 1.8 m Agilent p/n 60 C 5.0 L 6 C Flushport (MeOH:H2O 75:25) 5 s A = H2O w/5 mM ammonium formate + 0.01% formic acid B = 5 mM ammonium formate + 0.01% formic acid in 95:5 acetonitrile:water 1.0 mL/min Pressure 600 600 600 Solv ratio B 6 95 95
LC flow rate: Gradient pump time table Time Flow 0.15 No change 2.1 No change 3 No change Stop time 3 min Post time 1 min 6540 Q-TOF Parameters Jet stream conditions Drying gas temperature: Drying gas flow (nitrogen): Nebulizer gas pressure (nitrogen): Capillary voltage: Sheath gas temperature: Sheath gas flow: Nozzle voltage: Acquisition Mode: Min Range Max Range Scan Rate Reference Masses:
325 C 8 L/min 60 psig 4000 V 350 C 12 L/min 500 MS1 100 m/z 1000 m/z 10 per s Positive ion 121.050873 (M+H+ for purine) 922.009798 (M+H+ for HP-921)
25
Agilent Technologies, Inc., 2009 Printed in the USA August 5, 2009 5990-4251EN
Pesticide Personal Compound Database for Screening and Identification
Technical Note
Introduction
The Pesticide Personal Compound Database has been developed to provide rapid screening for a large number of compounds. The content of the database is described along with its usage. In addition to providing the necessary information for screening and identification of pesticides using time-of-flight mass spectrometry (TOF), the database allows for rapid and semi-automated customization of "target" compounds through updating retention times. In addition to screening, important information is provided to the pesticide analyst. This includes structure, a link to the NIH PUBCHEM database, and a host of useful internet links. Over the last century more than 1000 pesticides have been in common use for crop protection. Beyond approved and recommended usage, there always exists the possibility that any of these chemicals can be found in the environment and make their way into the food supply. Detection and identification in survey type monitoring is very important to protect both the environment and human health. Liquid chromatography/time-of-flight mass spectrometry (LC/TOFMS) provides sensitive and highly selective detection with full spectrum accurate mass measurement which helps meet this need. In addition, Agilent has developed an Accurate Mass Retention Time (AMRT) database providing over 1600 pesticides and related compounds to complement this capability. The database contains common names, molecular formulas, structures, and CAS registry numbers. It also provides the ability to easily and semi-automatically update retention times with the chromatographic RTs that have been obtained by the user's conditions. This technical note describes how to use the database to search processed LC/TOFMS data directly from Agilent's Personal Compound Database (PCD) software. It also describes how to manually and automatically search the database from Agilent's MassHunter qualitative data processing software.
Description
Personal Compound Database
The Personal Compound Database software user interface is shown in Figure 1. Four tabs are available to the user: Single Search, Batch Search, Batch Summary, and Edit Compounds. The first tab is displayed in Figure 1 and demonstrates the functions and data available in the pesticide database. Functions on this tab include searching by mass (protonated, neutral, and proton abstraction), retention time (optional or required), formula, and CAS registry number. Any or all of these parameters can be included in a single search while allowing cations and/or anions. The data includes the common name of the pesticide or chemical, its molecular formula, exact mass, structure, and CAS registry number. The registry number is linked to the on-line National Institutes of Health
database PUBCHEM and if clicked will launch the users internet browser and search PUBCHEM for that CAS number. Please note that not all compounds are listed in PUBCHEM, or if there, may not contain the CAS registry number. In addition, not all compounds in the Pesticide database have been registered in Chemical Abstracts. For example, Figure 1 shows Atrazine D-5 with the CAS number missing. Two other important points need to be made about the Pesticide database. One is that it includes non-pesticides that may be commonly found in samples screened for pesticides such as bisphenol A. The other is that it also includes pesticides, such as hexachlorobenzene, that will not respond to an LC/MS analysis. This makes it a very inclusive database of information. The retention time column is not populated in the database as this would be highly dependent on the user's LC conditions and configuration. However, this provides a very powerful feature that will be described in detail.
Figure 1.
Single Search tab of the Personal Compound Database for Pesticides showing the results for the "Atrazine" name search.
Batch Search
The second tab, Batch Search, is shown in Figure 2. This tab allows the user to import a mass list with retention times from data collected on the user's instrumentation. Generating mass lists and searching the database was created for use with the Agilent LC/MS TOF or LC/MS QTOF systems collecting accurate mass data. Nominal mass lists collected with other instruments will provide too many hits as there are many nominal mass isobaric compounds in the database. In the example shown, isomers are listed. A mass list is typically generated from standards and samples that have been processed in the "Find Compounds - Molecular Feature Extractor" function of the Agilent MassHunter Qualitative Analysis software. The mass list can either be imported as a .csv or .txt file. The mass list can also be pasted directly from a copied list in the MassHunter Qualitative Analysis software as shown in Figure 2. Once a mass list, such as one from pes-
ticide standards, is imported compounds found with similar masses will appear in red. This is indicated by arrows in Figure 2, a black-and-white screen shot of the Personal Compound Database software. These conflicting hits must be resolved manually by the user. In this example, three masses within the 10 ppm mass tolerance selected (mass 151.0631, 151.0641, and 151.0634) are shown having different retention times. It should be noted that the masses imported from MassHunter Molecular Feature Extractor are the calculated masses of the molecule, not the ion so all adduct possibilities are taken into account. These could be isomers or different compounds. The Agilent TOF and QTOF instruments typically provide mass accuracy better than 3 ppm, so a lower tolerance such as 3 ppm would avoid some conflicts. Since these are standards, the analyst should know the compounds injected and be able to select the correct compound and retention time.
Figure 2.
Batch Search tab of the Personal Compound Database for Pesticides showing the results for a mass list generated from a mixture of pesticide standards. Hits in red (indicated by the arrows) show multi-results of mass and retention times and represent conflicts.
Batch Summary
The third tab, Batch Summary, is shown in Figure 3. The capabilities provided in this tab are very powerful. Once all conflicts are resolved with the Batch Search, the user can update retention times of important targeted compounds under the chromatographic conditions selected by the user. This information is then added to the users custom database. The original Agilent Pesticide Database is Read Only and cannot be edited. However, if the user creates a new database based on the Agilent Pesticide database, then this custom file can be edited, retention times updated and compounds added (or deleted). The process can be repeated creating a custom database for each chromatographic set of conditions that the user requires for each different analyses of matrix compound combinations. This provides the analyst "targeted" analyses tailored to their specific needs.
Edit Compounds
The last tab, Edit Compounds, allows the user to change the information in their custom database, individually edit names,
structures, retention times, and other parameters. This tab is shown in Figure 4. Compounds can be added or deleted and the database must be updated from this screen. Once a custom database is created, it can be searched by the methods described above or searched directly in the MassHunter Qualitative Analysis software. This will be described in more detail. In summary, the Personal Compound Database for Pesticides provides an informative, user friendly interface that allows customization to meet the needs of the user. Links directly to the internet include direct PUBCHEM searching of each compound in the database with CAS numbers. This gives direct links to TOXNET and the literature enabling the user to efficiently access information about pesticides of interest or concern. The Links pull-down menu of the Personal Compound Database software provides access to useful websites, PUBCHEM, Chem Industry, Compendium of Pesticides, EPA - Pesticides, NIPC, NIH NLM, TOXNET, PAN Database, and WHO - Pesticides.
Figure 3.
Batch Summary tab of the Personal Compound Database for Pesticides showing the update of retention times from the results obtained in the Batch Search tab.
Figure 4.
Edit Compounds tab of the Personal Compound Database for Pesticides that allows for editing entries, adding entries, deleting entries and updating a customized database.
Searching the PCD Pesticides

The power of the information rich database lies in the ability to manually, semi-automatically or automatically search LC/Single MS TOF or QTOF accurate mass data for the compounds contained therein. Having made this statement, the user is reminded that of the 1600 compounds provided in the database, there are pesticides that do not ionize by the ionization methods commonly employed by LC/MS. For example, compounds such as hexachlorobenzene will not be detected in a datafile collected by electrospray ionization LC/MS TOF. However, there are still more than a thousand compounds that would be detected at different detection limits in different matrices. Although obvious, it must also be stated that compounds that ionize well only in positive ionization mode will be detected in that mode and those ionized well only in negative ion mode will only be detected in that mode. Certainly, pesticides that have been analyzed under specific conditions where retention times and detection limits have been recorded, under good laboratory practices and quality control procedures will most likely be detected in samples that contain those compounds. Even with these caveats, there is still great analytical power in the ability to screen samples for such a large list of pesticides to determine if a compound not being sought (non-targeted) is there. The ability to collect data with mass accuracy routinely better than 3 ppm and search an exact mass database makes this
5
possible. The technology to detect and identify compounds that may be in real samples at significant levels but are not being sought, is a needed and worthwhile endeavor. Searching the Personal Compound Pesticide Database can be done with data from real samples in the following ways. Single m/z values from LC/MS TOF or QTOF data can be input directly into the Single Search screen of the Personal Compound Database software. Retention time can be included in the search. It should be made optional so that all hits with and without retention times that match in m/z value will be returned. Using the Personal Compound Database software with the provided or customized pesticide database, a mass list created in MassHunter Qualitative software from sample data can be searched using the Batch Search tab by either importing or pasting the mass list. This is the same procedure used for standards to update the retention times. However, the most efficient search of the database is from MassHunter Qualitative software. Once a search method has been developed, this method can be incorporated into the MassHunter Acquisition worklist so that the data analysis (DA) will be done automatically, and the search done after each data acquisition run. The ability to create customized databases with the chromatographic conditions needed for specific matrices allows the user to conduct targeted screening. Target screening is a specific search for pesticides where standards have been analyzed under defined conditions and
retention times entered into the custom database. In addition, searching the entire database with optional retention times provides the ability to detect non-targeted compounds, or those where no standards have been run but are in the database and have been ionized and detected.
Example of MassHunter Qualitive Analysis Search Methods

As described above, the pesticide analyst may be concerned with three types of compound identification: targeted, nontargeted, and unknowns. For targeted compound determination a custom database may be created starting with the
supplied pesticide database of 1600 compounds. A standard or series of standards would then be run on the LC/MS TOF or QTOF. An example chromatogram of pesticide standards is shown in Figure 5. Using the "Find Compounds" and the "Molecular Feature Extractor" of MassHunter a masslist of compounds found in the standard analysis is created and copied to the "Batch Search" tab of the Personal Compound Database. If there are any conflicts as shown in Figure 2, they must be resolved by checking the correct isomer. In this example a conflict of two isomers, acetachlor and arachlor is shown in Figure 6 and are resolved. After resolving conflicts the retention times are updated on the Batch Summary tab as demonstrated in Figure 7 thus saving the updates to the custom database.
x10 6 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 1
7.656
5.469 4.581
6.463 2.593 3.295 2.394 2.858 5.137 6.118 5.575 6.834 7.046 4.647 6.118 6.291
Figure 5.
Total ion chromatogram of a pesticide standard mixture run on an Agilent LC/MS TOF system and processed with MassHunter Qualitative Analysis software.
Figure 6.
Batch search results of the data file represented by the chromatogram in Figure 5.
Figure 7.
Batch summary results with retention times updated.
Once the analyst has customized his or her database with specific retention times for the chromatographic conditions used for targeted pesticides or has decided to simply search the database in a non-targeted analysis for all possible compounds that may be ionized, that search can be done automatically in MassHunter Qualitative Analysis. The search can be done automatically from a worklist or directly from MassHunter Qualitative Analysis software. In either case the first step is to create a DataAnalysis method (DA). Figure 8 shows the total ion chromatogram of a bell pepper extract run under the same analytical conditions as the standard. The chromatogram shows a very complex matrix. Using the "Molecular Feature Extractor" in MassHunter over 4000 compounds were found in this data file. A search of the
customized pesticide database using mass and optional retention times (Figure 9) quickly provides a determination of all "targeted" compounds in the sample or those being sought with retention times. It also provides a determination of "nontargeted" compounds detected that are in the database. The results for this example are shown in Figure 10. Note that the compounds with retention times are scored > 90. Compounds with no retention times are scored 50 or less. Scores are based on mass accuracy of all isotopes, abundance of the isotopes as compared to theoretical values, and the spacing of the isotopes as compared to theoretical spacing. The scoring can be used by the analyst to access whether the hit is a real positive.
x10 7 1 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8
Figure 8.
Total ion chromatogram of a bell pepper extract run on the same Agilent LC/MS TOF system.
Figure 9.
The database search criteria in MassHunter Qualitative Analysis using the customized pesticide database with > 1600 compounds and retention times added from standards analyses. Note these settings will provide for "targeted" (compounds sought with retention times) and "non-targeted" (compounds in database for which no standards have been run) determination.
Figure 10. Search results from MassHunter Qualitatve Analysis of bell pepper extract showing positive targeted analysis hits (those with retention times) and non-targeted analysis hits.
Summary
The Agilent Personal Compound Database for Pesticides has been created for analysts to screen for a large list of targeted pesticides and non-targeted compounds easily and accurately. The database provides: An information rich resource Easy and powerful customization Seamless integration with Agilents MassHunter Qualitative analysis for Automated determinations

Agilent Technologies, Inc., 2009 Printed in the USA May 14, 2009 5990-3976EN
Q-TOF LC/MS Screening and Confirming of Non-Targeted Pesticides in a Strawberry Extract

Application Note
Food Safety
Authors
Chin-Kai Meng and Jerry Zweigenbaum Agilent Technologies Inc. 2850 Centerville Road Wilmington, Delaware 19808 USA Peter Frst and Eva Blanke Chemisches Landes- und Staatliches Veterinruntersuchungsamt Joseph-Knig-Strae 40, D-48147 Mnster, Germany
Abstract
A process for rapid screening and confirming of non-target pesticides is presented using a fruit-extract example. A strawberry extract was rapidly screened for pesticide content using the Molecular Feature Extraction (MFE) algorithm of the Agilent MassHunter software, a 600-pesticide database, and accurate mass spectra using liquid-chromatography, electrospray, quadrupole-time-of-flight mass spectrometry (Q-TOF LC/MS). Accurate masses of the ions detected and identified by MFE were compared to the exact masses of compounds in the pesticide database. Positive compounds were then confirmed using the MS/MS (Q-TOF LC/MS) mode.
Introduction
As the food industry has globalized in recent years, government regulations to enforce food safety standards have increased. The United States has more than 1000 registered pesticides. Of these, approximately 400 have EPA-established and FDA-enforced tolerance levels in foods. Over 1100 pesticides are monitored in food and animal feed in Europe as a result of the European Union's strict food pesticide regulations. Over 500 pesticides have established maximum residue levels (MRLs) in Japan [1]. Consumer concern along with these types of regulations has led to an increased demand for data on pesticide levels in foods. This includes both target pesticides, which the analyst investigates, as well as, nontarget pesticides, which the analyst might not anticipate. LC/MS/MS technology using LC in tandem with a triple quadrupole mass spectrometer (LC/QQQ) has become the approach of choice for target analyses of LC-amenable analytes in food. This highly selective and sensitive methodology has proven remarkably rugged and rapid for analyzing targeted pesticides at trace levels in complex matrices. Now, with the advent of Q-TOF LC/MS as a complementary form of LC/MS/MS, rapid screening for non-target pesticides at trace levels in foods becomes feasible. LC/QQQ uses the Multiple Reaction Monitoring (MRM) process to essentially remove chemical background from the sample matrix. However, LC single and triple quadrupole instruments cannot readily be used for nontarget identifications for two reasons. Full scan mode is required to acquire the full spectra needed to identify unknowns, but in scanning, the instrument sensitivity is significantly reduced. In this mode, LC single and triple quadrupole instruments do not have the selectivity gained by the MRM approach that focuses on ion transitions. Without good selectivity, characteristic ions of an analyte are sometimes overwhelmed by chemical noise. Plus, common LC-MS spectral libraries are unavailable due to difficulties in standardizing and reproducing fragmentation energies among instruments from different vendors. In contrast, the Q-TOF LC/MS application acquires accurate-mass spectra providing superior selectivity while maintaining high sensitivity. LC/Q-TOF, first introduced commercially in the mid-nineties, is similar to LC/QQQ except the third quadrupole is replaced by an analyzer known for delivering accurate mass, the TOF. This configuration gives excellent selectivity from high mass resolution and high mass accuracy with full spectra [2, 3, 4]. The Agilent 6510 Q-TOF LC/MS system used for this work has demonstrated mass accuracy values of < 2 ppm for small molecules. With this extraordinary accuracy comes the ability to distinguish between compounds of very similar molecular
2
mass. Since the exact or theoretical mass is uniquely determined by the molecular formula, obtaining accurate experimental masses allows an analyst to quickly narrow a search to a limited number of possible molecular formulas. In addition, using the accurate mass and relative abundance of isotopes further limits the possible molecular formulas. Unlike fragmentation patterns, molecular formulas by exact mass are reliably catalogued in databases and can be easily and quickly searched using computerized algorithms. For confirmation of identity, compounds can be further analyzed using the collision induced dissociation (CID) in Q-TOF MS/MS mode. Comparison of the structure of the proposed compound with the fragments obtained can confirm the identity. Accurate mass data and isotopic distributions for the precursor and product ions can be compared to spectral data of reference compounds, if available, obtained under identical conditions for final confirmation. Several unique features allow the Agilent 6510 Q-TOF LC/MS system to achieve the 2 ppm level of mass accuracy. One feature is automatic internal referencing. The Agilent 6510 adjusts for variation in conditions by using internal reference standards. In this technique, two compounds of known mass are introduced continuously in the Q-TOF ion source. The software automatically calibrates the mass axis of every spectrum. A second feature includes a shift from a time-to-digital converter (TDC) to analog-to-digital converter (ADC). Older TDC technology only registers an ion's arrival if its signal is above a certain intensity and gives the same response whether the signal is the result of one or many ions. As a result, the TDC approach cannot maintain mass accuracy over a wide dynamic range. ADC technology creates a continuous digital representation of the detector's signal and can provide up to three decades of dynamic range, thereby allowing components to be found in the presence of higher-abundance components [5]. Another benefit of the Agilent 6510 Q-TOF LC/MS system is temperature stability. The flight tube in this model is constructed from a metal alloy with an extremely low coefficient of thermal expansion. This effectively insulates the system from temperature fluctuations such that 1 to 2 ppm errors in mass accuracy can be readily attained [5]. While excellent MS and MS/MS mass-accuracy enables confident, simultaneous screening for thousands of compounds using confirmatory MS/MS data [6], the quantity and complexity of data produced is significant. Powerful software tools are needed to efficiently process the data. Agilent's Molecular Feature Extraction (MFE) algorithm reduces data
processing time by automatically finding all sample components down to the lowest-level abundance and extracting all relevant spectral and chromatographic information [6, 7]. In addition, the proprietary, Molecular Formula Generator (MFG) software provides high-confidence identification of unknowns, using multiple dimensions of information to generate and score lists of possible molecular formulas. The MFG reduces the number of plausible formulas by using the accurate-mass of adduct ions and their isotopes [6].
Screening and Confirming Workflow

This application note describes a "screen and confirm" workflow for the identification of nontargeted pesticides in a strawberry extract. The process starts with an injection of an aliquot of the sample extract into the Q-TOF LC/MS. The analytes from the chromatographic column are ionized and passed through the first quadrupole and into the TOF without CID. The resulting data file is cleaned of extraneous background noise and unrelated ions by the Molecular Feature Extraction (MFE) tool. The MFE then creates a listing of all possible components as represented by the full TOF mass spectral data. An exact mass database is then searched for hits to identify the pesticides in the data file.
To confirm the identity, a "Targeted MS/MS" method is developed for the pesticides that were found in the MFE and database search. In the method, the retention time and precursor ion for each target are entered. A second injection using the targeted MS/MS method is made. In the second analysis, the experimental accurate-mass MS/MS information is used to calculate formulas with the molecular formula generation (MFG) tool. This confirmation step compares not only exact mass and retention times but also fragment patterns and isotope ratios. A diagram depicting this "screen and confirm" workflow is shown in Figure 1.
Experimental
LC/MS
Agilent 1200 Series Rapid Resolution LC system with: Agilent 1200 Series binary pump SL and degasser Agilent 1200 Series high performance autosampler SL (ALS SL) Agilent 1200 Series thermostatted column compartment (TCC) Agilent 1200 Series diode-array detector SL (DAD SL) - for method development and troubleshooting
1. Screen
QuEChERS extraction aliquot LC Q-TOF Full Single Mass Spectra Data File Molecular Feature Extractor (MFE) List of components Targeted MS/MS Data File Match vs. database of pesticides
2. Confirm, if necessary
Background noise and unrelated ions
aliquot
LC
Q-TOF Molecular Formula Generator (MFG)
Pesticide confirmed by formula
Identification of pesticide
Analyst reviews ion fragments to confirm identity

Confirmation
Figure 1.
Screen and Confirm - LC/Q-TOF analysis and software workflow.
Agilent 6510 Accurate-Mass Q-TOF LC/MS Column: 2.1 mm 100 mm, 1.8 m ZORBAX Eclipse Plus C18, RRHT, 600 bar (p/n, 959764-902)
LC (1200) and MS (6510 QTOF) Method Parameters

The HPLC and MS were operated under the following conditions: Flow Rate 0.3 mL/min Injection Volume 10 L Solvent A 0.1% formic acid in water Solvent B 100% acetonitrile Gradient Time (min) Solvent B 0 10% 20 95% 25 95% Ion Source ESI Drying Gas Temperature 325 C Drying Gas Flow 10 L/min Nebulizer 50 psi 4000 V VCap Fragmentor 175 V Reference Masses 121.050873 and 922.009798 Acquisition Mode MS1 Min Range 100 Max Range 1000 Scan Rate 1 Acquisition Mode Targeted MS/MS MS Min Range 100 MS/MS Min Range 100 MS Max Range 1000 MS/MS Max Range 1000 MS Scan Rate 1.4 MS/MS Scan Rate 0.7 Max Time Between MS 10 Varied Collision Energy with Mass Slope 5 Offset 5
QuEChERS Sample Preparation Method

QuEChERS is the acronym for the sample preparation method, which stands for quick, easy, cheap, effective, rugged, and safe. It is a method that is widely receiving acceptance for rapid extraction of pesticides in food (8, 9).
Extraction:
Chop samples into small pieces and freeze in a bag overnight before grinding. Dry ice should be added during grinding. Weigh 10 g of an homogenized sample into a 50-mL Teflon centrifuge tube. Add 10 mL of acetonitrile (and ISTD solution, if used). Add 4 g of anhydrous magnesium sulphate, 1 g of sodium chloride, 1 g of trisodium citrate dehydrate, and 0.5 g of disodium hydrogen citrate sesquihydrate to the tube. Adjust the pH to 5-5.5 using 5 N NaOH. Shake the sample vigorously for 1 min by using a vortex mixer at maximum speed or by hand shaking. Centrifuge for 5 minutes at 3000 rpm.

A sample of strawberries was extracted with acetonitrile using the QuEChERS protocol and analyzed as described in the screening workflow section. The raw TOF (MS1 Mode) Full Spectrum TIC for the sample is shown in Figure 2. The TOF data was processed using the Molecular Feature Extraction (MFE) algorithm of the MassHunter Workstation software to find likely compounds. This algorithm is designed to find all ions in the data file that represent real compounds [7]. Figure 3 shows the molecular feature method editor menu and specific search criteria for the strawberry extract.
Cleanup:
Transfer 6 mL of supernatant into a 12-mL polypropolyene centrifuge tube which contains 150 mg of primary-secondary amine (PSA) adsorbent and 900 mg of MgSO4. Shake for 30 s. Centrifuge for 5 min at 3000 rpm. Adjust the pH of the cleaned extract to 5.0 for analysis, if necessary.
Figure 2.
The raw TOF (MS1 Mode) Full Spectrum TIC for an acetonitrile extract of a sample of strawberries.
Extraction Ion species Charge state Compound filters Mass filters Mass defect Results
Peak filter: Use peaks with height 1000 Positive Ions: H, Na, K, NH4 Peak spacing tolerance: 0.0025 m/z, plus 7.0 ppm Limit assigned charge states to a maximum of 1 Relative height 0.2 % None Filtering not used Delete previous compounds Highlight all compounds
Figure 3 .
MFE Method Editor Menu and settings for pesticide screening of the strawberry extract.
The MFE produced 822 potential compounds. Figure 4 shows the TIC, the hyperlinked extracted compound chromatogram (ECC), and the mass spectrum for one of these compounds. Since ECCs are created when one of the MFE Find Compounds algorithms is run, the ECC consists of all the related ions and no chemical noise. The accurate mass of each of these compounds was subsequently searched against a working exact mass database of 600 pesticides1. The criteria used in this search are shown in Figure 5. Twenty-six of the 822 compounds had mass matches (3 ppm tolerance) with pesticides in the database. Three plausible exact-mass match compounds, cyprodinil, azoxystrobin,
and boscalid were then selected for further confirmation using MS/MS (Q-TOF) analysis with the same instrument. The ECC and mass spectrum for each of the three are provided in Figure 6 along with the database search results showing a difference of less than 1 ppm in experimental and database masses. The precursor ion (M + H)+ masses chosen for the MS/MS analysis of the strawberry extract were exact masses from the database: 226.13395, 404.12410, and 343.03995 for cyprodinil, azoxystrobin, and boscalid, respectively. Criteria for finding compounds in the resulting MS/MS chromatogram are listed in Figure 7. Using the accurate MS/MS masses for the fragment ions, formulas were generated for each compound found in this step.
Figure 4.
The ECC and mass spectrum are shown for one of 822 compounds found using the MFE software along with the TIC.
Search criteria Database Peak limits Positive Ions Charge state range No neutral losses Search results Figure 5.
Match Mass only with 3.00 ppm tolerance Your Exact Mass compound database 10 H, Na, K, NH4 1-2 Limit to the best 5 hits
Exact mass database search menu and settings.
A 1600-compound Mass Hunter Personal Pesticide Database (G6854AA) is available for Exact Mass Searching.
Figure 6.
The hyperlinked ECC and mass spectrum for three positive pesticides, cyprodinil, azoxystrobin, and boscalid, found in the strawberry extract, are shown along with the exact mass database search results (a portion of the results is shown).
Integrator MS/MS Integrator Processing Maximum chromatogram peak width 0.25 min Cpd TIC Peak Filters Filter on peak area Limit (by height) to the largest 10 peaks Peak Spectrum Spectra to include average scans > 10% of peak height Exclude TOF spectra anywhere if above 40.0% of saturation MS/MS peak spectrum background: None Results Delete previous compounds Highlight all compounds Extract MS/MS chromatogram Extract MS/MS spectrum
Figure 7. Software settings for Find Compounds by Targeted MS/MS.
The confirmation process for azoxystrobin will be further discussed as an example. In Figure 8, the best-fit (with mass accuracy of 0.26 ppm and isotopes) formula generated from the Targeted MS/MS analysis for one of the compounds was C22H17N3O5, the formula for azoxystrobin. The two associated fragment masses for this peak had less than 1 ppm difference in mass (0.31 and 0.2 ppm) when compared to the database masses for fragments expected from the C22H17N3O5 parent formula. In addition, the three isotope masses for the molecular ion all differed by less than 1 ppm. The table outlined in
Figure 8 shows that the experimental isotope abundances of the three isotopes match well with the calculated (theoretical) abundances. The boxes in Figure 9 surrounding the isotopes represent the theoretical isotope abundances. Final confirmation of the structure is obtained by comparing the experimental fragment masses to likely theoretical fragment ion masses. Analysis of the structural formula, as seen in Figure 10, shows two likely fragments, with masses of 344.10351 and 372.09843 that match closely in mass to the
Figure 8.
Results from the formula analysis of MS and MS/MS data of a compound (azoxystrobin) in the strawberry extract.
Figure 9.
The experimental isotope abundances of the three isotopes match well with the theoretical abundances (outlined in boxes).
two fragment ions found in the MS/MS data: 344.10286 and 372.09781 amu. The difference between experimental and calculated is 0.31 and 0.20 ppm, respectively. The accuracy of this comparison, along with the MS data identification and MS/MS formula generation results all strongly suggest the presence of azoxystrobin in the analyzed strawberry extract.
Conclusion
This application note demonstrates that the high degree of mass accuracy (and selectivity) now available in combination with powerful database searching tools can be used to successfully identify nontargeted pesticides in food. The measured mass is used to generate a few potential molecular formulas and thereby frees the analyst from labor-intensive manual comparison of fragmentation patterns. These formulas, when combined with one or two fragment ions of accurate mass, can be used to quickly and confidently identify nontargeted pesticides in food samples.
344.10351 O
H 3C
O O
CH 3
O N N
372.09843
Figure 10. Structural analysis for Azoxystrobin fragments.
References
1. P. Wylie, J. Zweigenbaum, M. Churley, C.-K. Meng, and C. Zhe, "Comprehensive Screening, Confirmation, and Quantification of Organic Pesticides in Foods by GC-MS and LC-MS," (2008) Current Trends in Mass Spectrometry, November. I. V. Chernushevich, A. V. Loboda, and B. A.Thomson, "An Introduction to Quadrupole-Time-of-Flight Mass Spectrometry," (2001) J. Mass Spectrom. 36: 849-865. H. R. Morris, T. Paxton, A. Dell, J. Langhorne, M. Berg, R. S. Bordoli, J. Hoyes, and R. H. Bateman, "High Sensitivity Collisionally-activated Decomposition Tandem Mass Spectrometry on a Novel Quadrupole/Orthogonal-acceleration Time-of-flight Mass Spectrometer," (1996) Rapid Commun. Mass Spectrom. 10: 889. A.A. Shevchenko, I.V. Chernushevich, W. Ens, K.G.Standing, B, Thomson, M.Wilm, and M. Mann, "Rapid de novo peptide sequencing by a combination of nanoelectrospray, isotopic labeling and a quadrupole/time-of-flight mass spectrometer," (1997) Rapid Commun. Mass Spectrom. 11: 1015. K. Imatani, "Advances in Accurate-Mass TOF and Q-TOF LC-MS Systems," (2008) American Laboratory, 40:12-16.
3.
4.
5.
6. "Achieving Unmatched Speed and Confidence for Complex Sample Analyses: Agilent True High-Definition TOF Technology," Agilent Technologies, publication 5990-3168EN. 7 E. M. Thurman, I. Ferrer, and J. A. Zweigenbaum, "Automated Screening of 600 Pesticides in Food by LC/TOF MS Using a Molecular-Feature Database Search," Agilent Technologies, publication 5989-5496EN. M. Anastassiades, S. J. Lehotay, D. Stajnbaher, and F. J. Schenck, "Fast and Easy Multiresidue Method Employing Acetonitrile Extraction/Partitioning and "Dispersive Solid-Phase Extraction" for the Determination of Pesticide Residues in Produce," (2003) Journal of AOAC International, 86:412-431. S. J. Lehotay, K. Matovsk, and A. R. Lightfield, "Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables," (2005) Journal of AOAC International, 88:615-629.
8.
9.

Agilent Technologies, Inc., 2009 Printed in the USA May 5, 2009 5990-3935EN
Achieving the Desired Prescribed Sensitivities of Selected Herbicides by Direct On-Column Aqueous Injection of Potable and Environment Samples Using the Agilent 6410BA LC/QQQ
Application Note
Environmental
Authors
Neil Cullum Anglian Water Services Laboratory Huntingdon Cambridgeshire United Kingdom Paul Stephens Agilent Technologies UK Ltd. Winnersh Berkshire United Kingdom Dr Peter J. W. Stone Agilent Technologies Inc. Stevens Creek Boulevard Santa Clara, CA, 95051 USA
Abstract
Here we describe the analysis of 20 selected herbicides by direct on-column aqueous injection of environmental water samples of several matrices with little pretreatment. We demonstrate that this approach fulfills sufficient sensitivity requirements outlined by the UK Drinking Water Inspectorate. Precision data obtained were typically in the range of 2.2 to 7.0% and analyte recoveries were between 90.2 to 104.7%. Limits of detection were less than 10 ng/L (10 ppt) for all of the compounds in this suite.
Introduction
Several sample preparation/analysis approaches are available for the determination of herbicides in water samples, typically LC/MS after solid phase extraction (SPE) and even LC/MS employing on-line analyte enrichment [1]. Solid phase extraction is time consuming and adds an expense to the method with consumable materials and additional man-hours. On-line enrichment [ibid], on the other hand, requires the purchase of additional hardware, such as switching valves and an additional LC pump. With the introduction of affordable, reliable and sensitive LC/QQQ instrumentation such as the Agilent 6410BA triple quadrupole LC/MS system, it is now possible to achieve prescribed analysis requirements by injecting
CH3 O O Cl CH3 H3C NH N N N NH CH3 HN O N NH CH3
aqueous samples directly onto the analytical column using conventional injection volumes of up to 100 L. The aim of this application note is to demonstrate a reliable and robust analytical method for the analysis of 20 selected herbicides in potable and environmental water samples, with a performance criteria of < 12.5% analyte precision, analyte recoveries in the range of 90 to 110%, and limits of detection < 10 ng/L (10 ppt). The method presented here describes the analysis of a mixture of 20 acidic, neutral, and basic herbicides (Figure 1) in different water matrices by direct aqueous injection. An overview of the full validation data is summarized.
H3C HN N O Cl NH2 Cl
CH3 O Cl O N OH Cl
CH3
Atrazine
Carbetamide
H3C CH3 O HO NH CH3 Cl Cl
Chloridazon
H3C H3C N CH3 O NH N
Chlorotoluron
H3C HN CH3 O
Clopyralid
CH3 H3C HN O O
Cl N C H3C CH3 NH N N N
HN
Cl H3C CH3 Cl
Cyanazine
Diuron
H3C HN N CH3 O Cl CH3 Cl O OH
Imazapyr
Isoproturon
Linuron
NH2 Cl N Cl H3C CH3 NH N
CH3 N CH3 NH CH3 H3C CH3 NH N
Cl N N NH CH3 CH3
N O N
NH2
Metamitron
CH C H3C O NH CH3
Monuron
Picloram
Prometryn
Propazine
Cl N N N NH CH3 H3C H3C CH3 NH N
Cl N N NH CH3 H3C H3C CH3 NH N
CH3 N N NH CH3 H3C H3C N
Cl N N NH CH3
Cl
Cl
H3C
NH
Propyzamide
Simazine
Terbuthylazine
Terbutryn
Trietazine
Figure 1.
Suite of neutral, acidic, and basic herbicides.
Experimental
This analysis was performed using an Agilent LC/QQQ 6410BA mass spectrometer upgraded with a hotbox kit coupled to an Agilent 1200 Series LC system. The LC system consisted of a binary pump (G1312B), vacuum degasser (G1379B), automatic liquid sampler (G1367C), thermostatted column compartment (G1316B), and MassHunter data system. The hotbox upgrade kit (G2573A) comprised an additional MS turbo-pump with controller and replacement entrance and exit lenses for the collision cell.
Instrumentation
LC Conditions Column: Mobile phase Gradient program: Agilent ZORBAX SB-C18, 2.1 100 mm 1.8 m thermostatted at 70 C A: 0.1% formic acid in HPLC water B: methanol Time (min) Initial 0.5 1.0 20.0 20.1 100 L 26.0 min A (%) 95 95 80 20 95 B (%) 5 5 20 80 5 Flow rate (mL/min) 0.3 0.3 0.3 0.3 0.3
Sample Preparation
Minimal sample preparation was required, which was simple acidification of all standards and samples. These were acidified to a concentration of 0.1% formic acid, which was used as the pH modifier.
Injection volume: Total run time: QQQ MS Conditions Source Conditions Electrospray AP-ESI: Drying gas temperature and flow: Nebulizer gas pressure: Capillary voltage: Fragmentor voltage: MRM parameters:
Positive ionization polarity 300 C, 10 L/min 40 psi 3000 V See Table 1 See Table 1
MRM Parameters
Table 1. Time seg 2 3 4 MRM Transitions for Herbicide Suite Time (min) 0.2 6.4 7.6 Delta EMV (V) 600 600 400 Compound Clopyralid Clopyralid Picloram Picloram Metamitron Metamitron Imazapyr Imazapyr Chloridazon Chloridazon 5 10.0 400 Carbetamide Carbetamide Monuron Monuron Cyanazine Cyanazine Simazine Simazine 6 7 14.5 15.6 400 400 Chlorotoluron Chlorotoluron Diuron Diuron Atrazine Atrazine Isoproturon Isoproturon 8 17.0 400 Prometryn Prometryn Terbutryn Terbutryn Linuron Linuron Propazine Propazine Terbuthylazine Terbuthylazine Propyzamide Propyzamide 9 19.6 400 Trietazine Trietazine *(q) = Qualifier ion Precursor ion (m/z) 192.0 192.0 241.0 241.0 203.1 203.1 262.2 262.2 222.1 222.1 237.1 237.1 199.1 199.1 241.2 241.2 202.1 202.1 213.1 213.1 233.1 233.1 216.2 216.2 207.2 207.2 242.2 242.2 242.2 242.2 249.1 249.1 230.2 230.2 230.2 230.2 256.1 256.1 230.2 230.2 Product ions (m/z) 146.2 174.2 (q)* 223.1 195.0 (q)* 175.1 104.1 (q)* 234.3 217.2 (q)* 104.2 92.1 (q)* 192.3 72.2 (q)* 72.2 126.1 (q)* 214.2 104.1 (q)* 132.2 104.1 (q)* 72.2 140.2 (q)* 72.2 160.3 (q)* 174.2 104.1 (q)* 72.2 165.3 (q)* 200.3 158.2 (q)* 186.2 91.2 (q)* 182.1 160.3 (q)* 188.2 146.1 (q)* 174.2 104.1 (q)* 190.1 173.0 (q)* 202.2 99.2 (q)* Fragmentor voltage (V) 75 75 95 95 115 115 130 130 135 135 80 80 105 105 125 125 125 125 110 110 110 110 120 120 110 110 135 135 120 120 105 105 125 125 110 110 95 95 130 130 Collision energy (V) 19 6 9 18 14 22 14 17 22 27 2 22 16 25 12 31 16 27 21 24 22 26 15 32 22 10 17 24 17 30 18 12 15 24 15 30 12 22 18 24 Dwell time (msec) 400 100 400 100 90 90 90 90 90 90 70 70 70 70 70 70 70 70 250 250 90 90 90 90 90 90 30 30 30 30 100 100 30 30 30 30 30 30 250 250

The total ion chromatogram (TIC) for a 0.5 g/L (500 ppb) standard consisting of this 20 herbicide suite is shown in Figure 2, which also illustrates the positioning of the time segmentation.
Five levels of calibration standards were used to prepare the calibration curves over the concentration range of 0.0, 0.05, 0.10, 0.30, and 0.50 g/L. Selected and representative calibration curves are shown in Figures 3a through 3c.
Figure 2.
MRM overlay showing 20 herbicides from 0.5 g/L standard.
Monuron R2 = 0.99999
Figure 3a. Monuron calibration curve.
Simazine R2 = 0.99998
Figure 3b. Simazine calibration curve.
Propyzamide R2 = 0.99996
Figure 3c. Propyzamide calibration curve.
Validation of the method was carried out on 11 batches of samples. Borehole groundwater, potable water (which was from a surface water source), and river water were spiked at two levels (0.01 and 0.10 g/L). Deionized water was spiked at three levels with analytical quality control material at 0.01, 0.10, and 0.40 g/L. Each batch of samples was analyzed in duplicate and in a random order. The limit of detection (LOD) for each herbicide was calculated from the withinbatch standard deviation (5 x sw) of the deionized water spiked at 0.01 g/L. Recovery for the groundwater, potable water, and river water was calculated from the 0.1 g/L spike. Experimental results are shown in Table 2.
Table 2.
Validation Data: %Recovery, %RSD, and Limit of Detection (LOD) Borehole groundwater %Rec 100.8 5.7 99.7 4.0 100.5 4.3 101.7 3.4 99.7 3.5 98.0 5.2 99.8 3.3 99.4 4.5 100.1 2.9 99.5 3.2 98.3 3.7 99.4 2.2 99.1 3.8 99.7 2.9 99.0 2.9 99.3 5.8 99.6 3.2 99.8 3.8 101.4 4.4 99.9 2.8 99.7 3.8 Potable water River water %Rec %Rec 104.7 5.7 94.2 4.0 96.2 3.9 97.9 3.2 93.0 4.5 90.2 4.7 92.5 3.8 91.0 4.5 98.9 2.9 99.9 3.8 100.2 5.0 99.4 2.9 99.7 3.7 100.1 3.0 99.1 3.4 100.2 3.3 99.9 3.3 99.0 3.0 99.4 3.3 100.0 2.7 97.8 3.7 101.4 7.0 94.3 5.5 97.1 3.4 97.3 3.9 92.9 4.5 93.8 3.9 90.8 3.5 92.7 3.2 98.2 3.1 99.7 3.7 98.9 5.3 100.5 3.5 99.0 3.9 100.5 3.5 99.7 3.3 102.4 6.4 99.4 2.9 100.6 2.9 99.8 3.8 101.0 2.5 98.0 4.0 LOD (g/L) 0.007 0.005 0.003 0.005 0.004 0.009 0.005 0.004 0.004 0.003 0.006 0.002 0.003 0.003 0.002 0.003 0.002 0.003 0.004 0.002 0.004
Compound Clopyralid Picloram Metamitron Imazapyr Chloridazon Carbetamide Monuron Cyanazine Simazine Chlorotoluron Diuron Atrazine Isoproturon Prometryn Terbutryn Linuron Propazine Terbuthylazine Propyzamide Trietazine Overall suite
Selected and representative examples of MRM chromatograms derived from real sample matrices are shown in Figure 4.
Carbetamide 0.1204 g/L
Figure 4a. MRM chromatogram of carbetamide (river water matrix).
Simazine 0.0263 g/L
Figure 4b. MRM of simazine (river water matrix).
Trietazine 0.0369 g/L
Figure 4c. MRM of trietazine (river water matrix).
Conclusions
The data show that the method herein presented is capable of sensitive and quantitative analysis for the 20 herbicides in a single analytical suite by a direct aqueous injection of 100 L sample volumes onto the analytical column. Only sample acidification was undertaken as a preparation stage. All the method performance criteria are met, which are < 12.5% analyte precision, recoveries in the range of 90 to 110%, and limits of detection < 10 ng/L (10 ppt). We demonstrate in this application note that direct aqueous injection of 100 L samples onto the analytical column achieves the required method performance levels and is possible due to the sensitivity and selectivity of the Agilent LC/QQQ 6410BA instrumentation. The net benefit of such an approach to this methodology is a direct cost reduction in the form of consumable items (solid phase cartridges), which are no longer required, together with significant man-hour cost reductions since only minimal sample preparation is undertaken (acidification).
Reference
1. "Determination of Phenyl Urea and Triazine Herbicides in Potable and Groundwaters by LC/MS Using AP-ESI Selective Ion Monitoring and Direct Large Volume Aqueous Injection," Agilent Technologies publication 5989-0813EN.

For more details concerning this application note, please contact Neil Cullum at Anglian Water Services Laboratory, Huntingdon, Cambridgeshire, UK. Reference herein to any specific commercial products or nonprofit organization, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government and shall not be used for advertising or product endorsement purposes.

Agilent Technologies, Inc., 2009 Printed in the USA March 31, 2009 5990-3762EN
Multiresidue Analysis of 301 Pesticides in Food Samples by LC/Triple Quadrupole Mass Spectrometry
Application Note
Pesticides
Authors
E. Michael Thurman and Imma Ferrer Department of Civil, Environmental and Architectural Engineering Center for Environmental Mass Spectrometry University of Colorado ECOT 441, 428 UCB Boulder, CO 80309 USA Jerry A. Zweigenbaum Agilent Technologies Inc. 2850 Centerville Road Wilmington, Delaware 19806 USA
Abstract
An analytical methodology for screening and confirming the presence of 301 pesticides in vegetable samples was developed using the Agilent G6410A Triple Quadrupole Mass Spectrometer (QQQ). We found that, of the 301 compounds, 90% could be identified using this procedure with a limit of detection (LOD) in vegetable matrices of 0.01 mg/kg (ppm) or below, which is the level for baby food and banned substances and is the maximum residue level (MRL) used by the European Union. These levels were reached in a single analysis using positive ion electrospray with 99 transitions per segment and a quantifying and confirming ion for each compound. The analytical performance of the method was evaluated for different types of vegetables (tomato and green pepper), showing little or no matrix effects. Linearity of response over 2 orders of magnitude was demonstrated (r2 > 0.99).
Introduction
Currently more than 900 pesticides are used worldwide, both legally and illegally, on food products and in the treatment of soil and crops. Most of these pesticides have maximum residue levels (MRLs) for both food and water to protect the consumer. The MRL concentrations have to be monitored as part of the quality control of food, especially fruits and vegetables; thus, multiresidue methods with hundreds of pesticides are needed for quality control. However, the ability to monitor hundreds of pesticides in a single analysis is a challenging problem both for chromatography and mass spectrometry. In this application we evaluate the Agilent 6410 QQQ to not only screen but also to confirm 301 pesticides in a single analysis using a combination of the new 1.8-micron LC columns (for maximum peak capacity) and eight time segments with 100 transitions per segment in order to have both a quantifying ion and also a qualifier ion, which satisfies the European Union (EU) specifications for unequivocal identification by mass spectrometry. This study extends the tested capabilities of the instrument for limits of detection and speed of response, as well as for chromatography [1]. This study is one of the first of its kind to fully examine the Agilent QQQ for the analysis of pesticides in food for hundreds of pesticides in a single analysis. This topic was chosen because of the relevance of these compounds and their significant use on food commodities. The sensitivity of the QQQ easily meets the levels required by the regulations on pesticides in food for 90% of the compounds studied.
LC/MS/MS Instrumentation
Agilent SB-C-18, 4.6 mm x 150 mm, 1.8 m (p/n 829975-902) Column temperature: 25 C Mobile phase: 10% ACN and 90% H2O with 0.1% HCOOH Flow-rate: 0.6 mL/min Gradient: Time 0 = 10% ACN linear to Time 28 = 98% ACN Time 30 = 100% ACN Time 31 = 100% ACN Injection volume: 10 L MS Conditions Mode: Nebulizer: Drying gas flow: V capillary: Drying gas temp: Fragmentor voltage: Collision energy: MRM: Dwell time: Positive ESI using the Agilent G6410AA Triple Quadrupole Mass Spectrometer 40 psig 9 L/min 4000 V 350 C 70120 V 530 V 2 transitions for every compound as shown in Table 1 10 msec LC Conditions Column:

Optimization of LC/MS/MS Conditions The initial study consisted of two parts. First was to optimize the fragmentor voltage for each of the 301 compounds in order to produce the greatest signal for the precursor ion. Typically the protonated molecule was used for the precursor ion. Each compound was analyzed separately using an automated procedure to check the fragmentor at each voltage. The data were then selected for optimal fragmentor signal and each compound was injected in a programmed run at a concentration of 10 g/mL to determine collision energies for both the quantifying and qualifying ions. Various collision energies (5, 10, 15, 20, 25, and 30 V) were applied to the compounds under study. The energies were then optimized for each of the ions and the voltages that gave the best sensitivity were selected. The MRM transitions used for this study are shown in Tables 1A and 1B along with the list of the 301 compounds that were studied.
Experimental
Sample Preparation Pesticide analytical standards were purchased from Chem Services, Inc. (Philadelphia, USA) and Sigma Aldrich (Louisville, USA). Individual pesticide stock solutions (approximately 1,000 g/mL) were prepared in pure acetonitrile or methanol, depending on the solubility of each individual compound, and stored at 18 C. From these mother solutions, working standard solutions were prepared by dilution with acetonitrile and water. Vegetable samples were obtained from the local markets. Blank vegetable extracts were used to prepare the matrixmatched standards for validation purposes. In this way, two types of vegetables (green peppers and tomatoes) were extracted using the QuEChERS method [2]. The vegetable extracts were spiked with the mix of standards at different concentrations (ranging from 0.1 to 100 ng/mL or ppb) and subsequently analyzed by LC/MS/MS.
2
Table 1A. Analytical Conditions, MRMs, Limits of Detection, and r2 for Compounds Tested
Compound 3,4,5-Trimethacarb 3-Hydroxycarbofuran Acephate Acetamiprid Aclonifen Aldicarb Aldicarb sulfone Aldicarb sulfoxide Ametryn Aminocarb Atrazine Azamethiphos Azinphos-methyl Azoxystrobin Benalaxyl Bendiocarb Bensulfuron-methyl Benzoximate Bifenox Bitertanol Bromacil
Precursor ion 194 238 184 223 265 116 223 207 228 209 216 325 318 404 326 224 411 364 342 338 261
Product ions 137 122 163 220 143 125 126 56 248 193 89 70 148 76 89 132 186 96 137 152 174 132 183 139 132 160 372 344 148 294 109 167 149 182 199 105 189 310 99 269 205 188
Pepper LOD (ppb) 0.9 5.0 9.0 0.2 20.0 1.5 6.0 4.0 4.3 4.5 0.8 0.9 1.0 0.5 1.0 1.0 0.5 1.0 25.0 0.6 2.0
Tomato LOD (ppb) 0.5 5.0 5.0 0.5 10.0 5.0 1.0 1.0 4.3 2.0 0.6 0.9 1.0 0.5 0.5 1.0 0.4 0.5 40.0 2.0 1.0
r2 0.999 0.999 1.000 1.000 0.979 1.000 1.000 0.999 0.999 1.000 1.000 0.993 0.975 0.973 0.992 0.996 0.998 0.999 *** 0.995 0.999
Continued 3
Compound Bromuconazole 1 Bromuconazole 2 Bupirimate Buprofezin Butocarboxim Butocarboxim-sulfoxide Buturon Butylate Carbaryl Carbendazim Carbetamide Carbofuran Carboxin Carfentrazone-ethyl Chlorbromuron Chlorfenvinphos Chlorfluazuron Chloridazon Chloroxuron Chlorpropham Chlorpyrifos methyl
Precursor ion 376 376 317 306 213 207 237 218 202 192 237 222 236 412 293 359 540 222 291 214 322
Product ions 159 70 159 70 166 272 201 116 156 75 75 132 84 126 57 156 145 117 160 132 118 192 165 123 143 87 346 366 204 182 155 127 383 158 104 92 72 218 172 154 125 290
Pepper LOD (ppb) 1.0 4.0 1.0 0.7 5.0 5.0 1.0 5.0 5.0 0.5 0.5 1.0 0.5 8.0 7.0 2.0 1.0 0.9 0.5 1.0 7.0
Tomato LOD (ppb) 1.0 2.0 5.0 0.8 *** 5.0 0.5 45.0 5.0 1.0 0.5 1.0 0.5 8.0 10.0 1.0 1.0 1.0 2.0 1.0 35.0
r2 1.000 0.983 0.999 0.999 0.940 0.997 0.999 0.990 0.999 0.997 0.996 0.999 0.993 0.998 0.996 0.990 0.989 1.000 0.983 0.998 0.978
Continued 4
Compound Chlorsulfuron Cinosulfuron Clethodim Clodinafop-propargyl Clofentezine Clomazone Cloquintocet-mexyl Coumaphos Cyanazine Cycloate Cymoxanil Cyproconazole Cyprodinil Cyromazine Daminozide Deethylatrazine Deethylterbuthylazine Deisopropylatrazine Demeton-S-methyl-sulfone Desmedipham Diazinon
Precursor ion 358 414 360 350 303 240 336 363 241 216 199 292 226 167 161 188 202 174 263 301 305
Product ions 141 167 183 157 164 268 266 238 138 102 125 89 238 192 227 307 214 174 83 154 128 111 70 125 93 108 85 125 143 61 146 104 146 110 96 132 169 125 136 182 169 153
Pepper LOD (ppb) 1.0 3.0 5.0 5.0 6.0 0.9 0.1 1.0 2.0 1.0 6.0 1.0 5.0 10.0 50.0 1.0 1.0 4.0 0.5 1.0 0.5
Tomato LOD (ppb) 1.0 1.0 10.0 1.0 10.0 5.0 0.5 5.0 1.0 10.0 5.0 2.0 7.0 10.0 40.0 1.0 0.5 5.0 0.5 10.0 0.2
r2 0.999 0.998 0.940 0.993 0.998 0.999 0.990 1.000 0.999 0.999 1.000 0.993 1.000 0.997 0.999 0.999 0.999 0.998 0.990 0.999 1.000
Continued 5
Compound Dichlofenthion Dichlofluanid Dichlorvos Diclobutrazol Diethofencarb Difenoconazole-1 Difenoconazole-2 Difenoxuron Diflubenzuron Diflufenican Dimefuron Dimethachlor Dimethenamid Dimethoate Dimethomorph 1 Dimethomorph 2 Diniconazole Diphenylamine Disulfoton Diuron Dodemorph EPN
Precursor ion 315 333 221 328 268 406 406 287 311 395 339 256 276 230 388 388 326 170 275 233 282 324
Product ions 287 259 123 224 109 145 70 159 152 226 251 337 251 337 72 123 158 141 266 246 72 167 224 148 244 168 199 171 301 165 301 165 70 159 93 65 89 61 72 160 116 98 157 296
Pepper LOD (ppb) 10.0 10.0 5.0 2.0 0.5 0.3 0.3 0.6 6.0 5.0 0.5 0.5 0.6 0.7 0.6 0.6 1.0 0.5 5.0 0.8 10.0 30.0
Tomato LOD (ppb) 40.0 10.0 5.0 5.0 1.0 0.5 0.5 0.2 6.0 1.0 1.0 1.0 0.1 0.7 0.6 0.6 1.0 0.1 10.0 1.0 10.0 10.0
r2 0.520 0.999 0.999 1.000 0.998 0.989 0.989 0.980 0.959 0.999 0.991 0.993 0.987 0.994 0.998 0.998 0.988 0.985 0.991 0.996 0.983 0.991
Continued 6
Compound Epoxiconazole Ethiofencarb Ethiofencarb sulfone Ethiofencarb sulfoxide Ethion Ethirimol Ethofumesate Ethoprophos Etrimfos Famoxadone Fenamiphos Fenarimol Fenazaquin Fenbuconazole Fenfuram Fenhexamid Fenobucarb Fenoxaprop-ethyl Fenoxycarb Fenpiclonil Fenpropathrin Fenpropimorph
Precursor ion 330 226 258 242 385 210 287 243 293 331 304 331 307 337 202 302 208 362 302 237 350 304
Product ions 121 141 107 164 107 201 107 185 171 199 98 140 121 161 173 215 125 265 195 238 217 234 81 268 57 161 125 70 109 83 97 55 95 152 288 244 88 116 202 166 125 97 147 130
Pepper LOD (ppb) 20.0 0.9 6.0 0.5 1.0 50.0 7.0 2.0 0.7 5.0 1.0 5.0 1.0 0.5 5.0 3.0 1.0 0.6 1.0 1.0 50.0 10.0
Tomato LOD (ppb) *** 1.0 6.0 1.0 5.0 45.0 10.0 2.0 5.0 5.0 0.5 6.0 5.0 2.0 2.0 9.0 1.0 2.0 1.0 10.0 5.0 10.0
r2 *** 0.995 0.998 0.990 1.000 0.990 0.974 0.987 0.999 *** 0.989 0.984 0.999 1.000 0.999 1.000 1.000 0.999 0.999 0.998 0.967 1.000 Continued
Compound Fenthion 1 Fenthion 2 Fenuron Fipronil Flamprop-isopropyl Flamprop-methyl Flazasulfuron Fluazifop-P-butyl Flufenacet Flufenoxuron Fluodioxinil Fluometuron Fluoroglycofene-ethyl Fluoroxypyr Flurtamone Flusilazole Flutriafol Folpet Fonofos Formetanate Fuberidazole Furathiocarb
Precursor ion 279 279 165 437 364 336 408 384 364 489 229 233 344 255 334 316 302 260 247 222 185 383
Product ions 247 169 169 247 72 120 263 368 105 304 105 304 182 301 282 328 152 194 158 141 158 185 72 160 223 300 209 181 247 303 247 165 70 123 130 232 109 137 165 120 157 156 195 252
Pepper LOD (ppb) 5.0 5.0 1.0 10.0 5.0 5.0 8.0 1.0 0.5 9.0 5.0 2.0 25.0 10.0 0.5 0.1 1.0 5.0 1.0 10.0 1.1 0.5
Tomato LOD (ppb) 9.0 10.0 2.0 50.0 10.0 9.0 5.0 0.5 1.0 5.0 10.0 1.0 10.0 5.0 0.4 0.1 1.0 50.0 5.0 9.0 1.0 0.1
r2 0.998 0.999 0.999 0.994 0.974 0.986 0.898 0.991 0.991 0.990 0.987 0.988 0.999 0.997 0.996 0.991 0.998 *** *** 0.995 1.000 0.991
Continued 8
Compound Haloxyfop-methyl Heptenophos Hexaconazole Hexaflumuron Hexazinone Hexythiazox Hydroxyatrazine Imazalil Imazapyr Imazaquin Imidacloprid Indoxacarb Ioxynil Iprodione Iprovalicarb Irgarol Irgarol metabolite Isazofos Isofenphos Isoproturon Isoxaflutole
Precursor ion 376 251 314 461 253 353 198 297 262 312 256 528 372 330 321 254 214 314 346 207 360
Product ions 316 288 127 125 70 159 158 141 171 71 168 228 156 86 159 255 217 234 199 267 175 209 249 293 118 245 245 288 119 203 198 156 158 110 120 162 217 245 72 165 251 69
Pepper LOD (ppb) 1.0 2.0 1.0 5.0 0.5 2.0 4.0 10.0 1.0 0.6 5.0 5.0 30.0 10.0 1.0 1.0 5.0 1.0 3.0 1.0 1.0
Tomato LOD (ppb) 5.0 1.0 0.6 10.0 1.0 1.0 4.0 10.0 1.0 1.0 1.0 1.0 10.0 30.0 1.0 1.0 4.0 0.5 0.5 1.0 20.0
r2 1.000 0.999 0.993 0.984 0.994 0.996 0.998 0.850 0.996 0.998 0.993 0.997 0.998 0.999 0.997 0.999 1.000 0.991 0.996 0.999 1.000
Continued 9
Compound Kresoxim-methyl Lenacil Linuron Lufenuron Malaoxon Malathion Mebendazole Mecarbam Mepanipyrim Metalaxyl Metamitron Metazachlor Metconazole Methabenzthiazuron Methamidophos Methfuroxam Methidathion Methiocarb Methomyl Metobromuron Metolachlor Metolcarb
Precursor ion 314 235 249 511 315 331 296 330 224 280 203 278 320 222 142 230 303 226 163 259 284 166
Product ions 206 267 153 136 160 182 158 141 99 127 99 127 264 105 199 227 77 106 192 220 175 104 134 210 70 125 165 150 94 125 137 111 85 145 121 169 88 106 170 148 252 176 109 91
Pepper LOD (ppb) 5.0 5.0 1.0 5.0 1.0 0.8 0.6 2.0 1.0 1.0 1.0 1.0 1.0 0.5 1.0 0.5 1.0 0.9 1.0 5.0 0.5 3.0
Tomato LOD (ppb) 5.0 25.0 5.0 10.0 0.5 2.0 0.6 1.0 1.0 2.0 1.0 1.0 0.5 1.0 1.0 0.4 0.5 1.0 1.0 10.0 1.0 1.0
r2 0.997 0.998 1.000 0.995 0.996 0.998 0.990 0.989 0.997 0.994 1.000 0.986 0.979 0.999 0.999 1.000 0.987 0.996 0.990 0.999 0.988 1.000 Continued
10
Compound Metosulam Metoxuron Metribuzin Metsulfuron-methyl Molinate Monolinuron Monuron Myclobutanil Naled Napropamide Neburon Nicosulfuron Nuarimol Ofurace Omethoate Oxadixyl Oxamyl Oxydemethon-methyl Paclobutrazol Paraoxon-methyl Parathion-ethyl Parathion-methyl
Precursor ion 418 229 215 382 188 215 199 289 379 272 275 411 315 282 214 279 242 263 294 248 292 264
Product ions 175 354 72 156 187 131 167 141 126 83 126 148 72 126 70 125 127 297 129 171 57 88 182 213 81 252 160 254 125 183 219 102 72 121 169 109 70 165 90 202 236 264 125 232
Pepper LOD (ppb) 5.0 0.5 2.0 1.0 2.0 1.0 1.0 0.5 10.0 1.0 1.0 1.0 1.0 0.5 1.1 5.3 45.0 1.1 0.5 3.0 5.0 10.0
Tomato LOD (ppb) 1.0 1.0 0.5 0.5 5.0 2.0 5.0 1.0 10.0 1.0 0.5 0.6 1.0 1.0 1.0 5.0 50.0 1.0 2.0 5.0 10.0 10.0
r2 0.993 0.996 0.999 0.996 0.999 1.000 0.995 0.997 0.997 1.000 0.989 0.937 *** 0.981 1.000 0.989 0.990 0.990 0.998 0.994 0.999 0.986
Continued 11
Compound Penconazole Pencycuron Pendimethalin Phenmedipham Phenthoate Phorate Phosalone Phosmet Phoxim Picolinafen Picoxystrobin Pirimicarb Pirimiphos-ethyl Pirimiphos-methyl Pirimisulfuron-methyl Prochloraz Procymidone Profenofos Promecarb Prometon Prometryn Propachlor
Precursor ion 284 329 282 301 321 261 368 318 299 377 368 239 334 306 469 376 284 373 208 226 242 212
Product ions 70 159 125 218 212 194 136 168 163 247 75 199 182 322 160 133 77 129 238 359 145 205 72 182 198 182 164 108 254 199 308 266 256 67 303 345 109 151 142 184 158 200 170 152
Pepper LOD (ppb) 0.5 0.5 5.0 1.0 5.0 25.0 5.0 5.0 5.0 1.0 0.5 2.0 0.1 0.1 1.0 5.0 30.0 5.0 1.0 3.0 1.0 1.0
Tomato LOD (ppb) 1.0 5.0 2.0 1.0 9.0 8.0 1.0 7.0 5.0 1.0 0.5 8.0 0.1 0.1 1.0 10.0 8.0 5.0 1.0 3.0 5.0 1.0
r2 0.986 0.987 0.990 0.999 0.975 0.999 1.000 0.988 0.998 0.999 0.991 0.999 0.995 0.934 0.980 0.999 0.998 1.000 0.998 1.000 0.999 0.998
Continued 12
Compound Propamocarb Propanil Propazine Propetamphos Propham Propiconazole-1 Propiconazole-2 Propoxur Propyzamide Prosulfocarb Prosulfuron Pymetrozin Pyraclostrobin Pyrazophos Pyridaben Pyrimethanil Pyriproxyfen Quinalphos Quinmerac Quinomethionate Quinoxyfen
Precursor ion 189 218 230 282 180 342 342 210 256 252 420 218 388 374 365 200 322 299 222 235 308
Product ions 102 144 127 162 146 188 138 156 120 138 159 69 159 69 111 168 190 173 91 128 141 167 105 79 163 194 222 194 147 309 107 183 96 227 147 163 204 176 207 163 197 272
Pepper LOD (ppb) 8.7 1.0 1.0 5.0 5.0 0.5 0.5 0.5 3.0 0.8 1.0 30.0 0.5 0.5 0.1 8.0 0.6 1.0 4.7 5.0 1.0
Tomato LOD (ppb) 9.0 4.0 0.5 0.5 2.0 1.0 1.0 0.5 1.0 1.0 1.0 50.0 0.1 0.1 0.5 5.0 0.5 1.0 5.0 10.0 0.5
r2 0.999 0.963 1.000 0.972 1.000 0.997 0.995 0.999 0.998 1.000 0.989 0.980 0.976 0.996 0.997 0.999 0.999 0.999 0.999 0.994 0.998
Continued 13
Compound Quizalofop-ethyl Rimsulfuron Rotenone Simazine Simetryn Spiromesifen Sulfosulfuron Sulfotep Sulprofos Tebuconazole Tebufenozide Tebutam Tebuthiuron Teflubenzuron Terbufos Terbumeton Terbuthylazine Terbutryn Tetrachlorvinphos Thiabendazole Thiacloprid Thiamethoxam
Precursor ion 373 432 395 202 214 371 471 323 323 308 353 234 229 381 289 226 230 242 365 202 253 292
Product ions 299 271 182 325 213 192 132 124 124 144 273 255 211 261 171 143 219 247 70 151 133 297 91 192 172 116 158 141 57 103 170 114 174 132 186 71 127 239 175 131 126 186 211 181
Pepper LOD (ppb) 0.5 5.0 1.0 1.0 10.0 7.0 1.0 1.0 30.0 1.0 0.5 0.5 0.9 10.0 12.0 5.0 0.1 5.0 1.0 6.0 3.0 5.0
Tomato LOD (ppb) 5.0 1.0 1.0 1.0 10.0 10.0 1.0 5.0 30.0 0.5 0.1 0.5 1.0 9.0 10.0 5.0 0.5 2.0 5.0 1.0 1.0 8.0
r2 1.000 0.810 0.999 0.998 0.998 0.968 1.000 0.999 0.990 0.983 0.983 0.987 1.000 0.981 0.976 1.000 0.999 1.000 0.999 0.999 0.998 0.990
Continued 14
Compound Thifensulfuron-methyl Thiocyclam Thiodicarb Thiofanox Thiophanate-methyl Tolclofos-methyl Tolyfluanid Triadimefon Triadimenol Triasulfuron Triazophos Tribenuron-methyl Trichlorfon Triclocarban Tricyclazole Trietazine Trifloxystrobin Triflumizole Triflumuron Triflusulfuron-methyl Vamidothion
Precursor ion 388 182 355 241 343 301 347 294 296 402 314 396 257 315 190 230 409 346 359 493 288
Product ions 167 141 137 73 88 163 184 57 151 311 125 269 137 238 69 197 70 227 167 141 162 286 155 181 109 221 162 128 163 136 132 202 186 206 278 73 156 139 264 238 146 118
Pepper LOD (ppb) 0.9 54.0 5.0 5.0 10.0 6.0 10.0 1.0 10.0 2.0 0.7 100.0 10.0 1.0 1.0 2.0 0.5 4.0 4.0 4.0 0.8
Tomato LOD (ppb) 1.0 54.0 1.0 10.0 1.0 10.0 1.0 1.0 1.0 0.5 1.0 10.0 10.0 1.0 1.0 0.8 0.5 1.0 1.0 1.0 2.0
r2 0.998 0.986 0.980 0.987 0.999 0.999 0.999 0.998 0.976 0.991 0.984 0.981 0.997 0.998 0.998 1.000 0.997 0.974 0.999 0.990 0.998
15
Table 1B. Compounds Tested With Low Sensitivity or Poor Chromatography on SB C-18
Precursor Acetochlor Acibenzolar-S-methyl Alachlor Aldoxycarb Anilazine Azinphos-ethyl Benfuracarb Bromoxynil Chlorotoluron Ethoxyquin Fenpropidin Itraconazole Methiocarb sulfone Nicotine Phosphamidon Propargite Pyrifenox Spinosad A 270 211 270 223 275 346 411 278 213 218 274 705 258 163 300 373 Na 295 732.4
Product ions 148 224 136 91 162 238 149 177 153 178 160 132 195 252 199 223 72 140 174 148 147 57 450 404 122 217 130 132 129 153 81 57 93 263 142 99 558 142 144 100 161 144
Solvent LOD (pg) 10 50 10 50 1 1 1.0 >100 20 >100 100 *** *** *** 100 >100 >100 20
Reason Poor chromatography on SB C-18 Outside window Poor chromatography on SB C-18 Interference Outside window Outside window Outside window Poor sensitivity Outside window Poor chromatography Poor chromatography No qualifier ion No qualifier ion Poor ionization Outside window Poor sensitivity Poor sensitivity Not eluted from SB C-18 Not eluted from SB C-18 Degraded standard Poor sensitivity
Spinosad D
746.4
100
Spiroxamine Terbacil
298 217
>100 >100
16
The MRM transitions used a dwell time of 10 milliseconds (msec). Eight different time segments were recorded in the chromatographic run, each segment containing approximately 50 pesticides. It was necessary to overlap pesticides at each boundary of the time segment in order to monitor compounds that may elute at the exact moment of the time segment boundary. Figure 1 shows the chromatogram corresponding to 100 parts per billion (ppb) standard on column for all the 301 compounds studied. Extracted ion chromatograms are overlaid for each one of the target analytes according to their respective protonated molecule and product-ion MRM transitions. Application to Vegetable Matrices To confirm the suitability of the method for analysis of real samples, matrix-matched standards were analyzed in two dif-
ferent matrices (green pepper and tomato) and compared with solvent at six concentrations (0.1, 0.5, 1.0, 10.0, 50.0, and 100 ng/mL or ppb concentrations). Figure 2 shows an example standard curve for diazinon in the pepper matrix. The r2 values are shown in Table 1A for the pepper matrix; similar values were found for solvent and tomato matrix. The compounds gave linear results with excellent sensitivity over more than two orders of magnitude, with r2 values of 0.99 or greater for the majority of compounds and LODs of 1 to 10 picograms (pg) for 150 of the compounds and from 10 to 100 pg for 140 compounds. There were 11 compounds that poorly ionized or gave poor chromatography and did not respond with sufficient signal to reach the 0.010 mg/kg level. These compounds are shown in Table 1B.
Figure 1.
Product ion chromatogram (MRM) for 301 pesticides with a concentration of 100 ppb standard, which also shows the eight time segments.
17
10 7 2 1.75 Responses 1.50 1.25 1 0.75 0.50 0.25 0 _10 _5
Diazinon - 6 Levels, 6 Levels Used, 6 Points, 6 Points Used, 0 QCs y = 208335.0516 * x _ 152717.0728 R 2 = 0.99956076
10
15
20
25
30
35
40
45 50 55 60 65 Concentration (ng/mL)
70
75
80
85
90
95
100 105 110 115
Figure 2.
Calibration curve for diazinon in pepper using a six-point curve from 0.1 to 100 ng/mL (ppb) using a linear fit with no origin treatment.
Figure 3 shows the ion ratios qualifying for diazinon in an extract of green pepper spiked with the pesticide mix at 0.010 g/g (100 pg on column). The m/z 169 ion was used for quantification and the m/z 153 ion was used as the qualifier ion, with a window set at 20% for the ion ratios. As shown in Figure 3 in the two ion profiles, diazinon was easily identified in this complex matrix due to the selectivity of the MRM transitions and instrument sensitivity. In general, the LODs for the 301-pesticide mix met the requirements regarding the MRLs imposed by the existing European regulations.
Furthermore, the use of 1.8-micron packing resulted in sharp chromatographic peaks of 5 to 10 seconds in width. Thus, it was important to use fast dwell times of 10 msec in order to keep the quantitation results shown in Table 1A. Finally, the analysis for repeatability of the instrument for the quantifying ion gave a relative standard deviation for five repeats of 6% (median RSD) and a mean RSD of 6.7%. These values were determined at the 0.1 mg/Kg level (100 ppb).
10 6 1.8 1.6 1.4 1.2 Abundance 1.0 0.8 0.6 0.4 0.2 0
305
169.0, 153.0
10 5 5
169.0
Ratio=52.2
4.5 4 3.5 Abundance 3 2.5 2 1.5 1 0.5
153.0
24.4
24.6 24.8 25 25.2 Acquisition time (min)
25.4
150
175
200
225
250
275
300
325
Mass-to-charge (m/z)
Figure 3.
Shows the ion ratios for qualifier ion and the quantifying ion for diazinon in the pepper matrix.
18
Conclusions
The results of this study show that the Agilent 6410 Triple Quadrupole is a robust, sensitive, and repeatable instrument for the study of pesticides in food, such as vegetable extracts, using high-throughput methods. The LOD for the instrument was in the 1 to 10 pg range for 50% of the compounds and 100 pg for 90% of the compounds studied. These LODs included both the quantifying ion and the qualifying ion, which is quite important for identification. These LODs are sensitive given that the segments contained 100 transitions, which is sufficient to analyze approximately 30 to 40 compounds per segment (with an overlap of 5 to 10 compounds per segment, a minimum if good reproducibility of the method is to be obtained). The Agilent 6410 Triple Quadrupole was capable of reaching a LOD of 0.010 mg/kg (ppm) for at least 90% of the pesticides monitored in this study using the two product ion criteria for confirmation, and a 10-L injection, which is a typical injection volume. This MRL is the baby food limit, the limit for banned pesticides, and the typical requirement of a newly purchased LC/MS/MS system by environmental and food scientists who work on real food matrices.
References
1. Imma Ferrer, E. M. Thurman, Y. Fang, P. Zavitsanos, and J. A. Zweigenbaum, Multiresidue Analysis of 100 Pesticides in Food Samples by LC/Triple Quadrupole Mass Spectrometry, Agilent Technologies publication 59895469EN, August 2006 2. M. Anastassiades, S. J. Lehotay, D. Stajnbaher, and F. J. Schenck, J. AOAC International, 86 (2003) 412431.

19
Agilent Technologies, Inc., 2008 Published in the USA November 5, 2008 5989-8614EN
November 2008
Comprehensive Screening, Confirmation, and Quantification of Organic Pesticides in Foods by GCMS and LCMS
The global focus on keeping food safe has intensified, resulting in major changes in the number of pesticides that are being regulated and monitored, as well as the allowable levels of those pesticides in food. These new regulations drive a need for a wide range of comprehensive and complementary analytical methods and instrumentation that possess the throughput, sensitivity, and breadth of pesticide compound coverage to meet the demand of testing laboratories all over the world. This article provides an overview of the instrument platforms, tools, and workflow for analyzing pesticides.
Philip Wylie, Jerry Zweigenbaum, Melissa Churley, Chin-Kai Meng, and Cao Zhe
ith the globalization of trade, food production and distribution have become truly international businesses. When we dine out, the fish might come from Japan, the rice from Australia, the spices from China, and the strawberries from Mexico. We take it for granted that the food we eat is safe and free from contamination that could make us seriously ill. One of the 20th-century advancements that made this cornucopia possible was the development of pesticides. Crops that were previously devastated by pest damage can now be produced at very high yields to meet worldwide demand. However, all pesticide compounds are potentially harmful to humans, and safeguards must be in place to assure that consumers are not exposed to dangerous levels of these pesticides. Many countries have had regulations regarding the use of pesticides, and their allowable levels in foods, for many years.
With the increasing globalization of the food industry, there is greater scrutiny on food safety, resulting in major changes in the number of pesticides that are being regulated and monitored, as well as the allowable levels of those pesticides in food.
A Global Effort to Keep Food Safe

The Food Quality Protection Act (FQPA) of 1996 updated federal regulation of pesticides in the United States to implement a single health-based standard for all pesticides in foods. It provides special protection for infants and children, expedites approval of safe pesticides, and requires periodic evaluation of pesticide registrations and tolerances to ensure that the scientific data supporting registrations will remain up to date in the future. Under this mandate, the US Food and Drug Administration (FDA) performs regulatory monitoring by sampling individual lots of domestically produced and imported foods
their products to assure compliance with FQPA, the EU regulation, and the Japan Positive List, if they hope to sell their products in the U.S., Europe, and Japan, which generate much of the worlds demand for imported food. These new regulations drive a global need for analytical methods that possess the throughput, sensitivity, and breadth of pesticide compound coverage to meet the demand of testing laboratories all over the world that must ensure compliance.
A Broad Range of Complementary Analytical Methods Is Needed

Food testing laboratories need the ability to detect and quantify hundreds of pesticides, in myriad foodstuffs, at very low levels of contamination. They might be screening for a set of hundreds of known target compounds, using a defined test method for each target in the set. Alternatively, a testing laboratory might wish to look for the presence of any pesticide (unknowns), without the need for defining a method for each possible compound. They are often called on to do both types of testing, and no single analytical approach can provide the flexibility required to meet this need. Wide variations in the chemical properties of pesticide contaminants and the necessity to detect a very large number of compounds require a range of chromatography and mass spectrometry (MS) systems. For pesticides that can be vaporized easily without degradation, gas chromatography (GC)MS (single quadrupole) is an ideal analytical tool, due to the availability of large libraries of pesticide spectra and deconvolution software. This approach can be used to screen, confirm, and quantify both target and unknown compounds. Many foodstuffs, such as garlic and ginger, are very complex, or dirty, due to the presence of a very large number of background compounds. GC triple quadrupole MS (QQQ) is very useful for screening, confirming, and quantifying trace-level target compounds in these complex matrices. Pesticides that are quite polar, not easily vaporized, thermally labile, or not easily derivatized are best analyzed using liquid chromatography (LC) methods. LCQQQ is particularly useful for analyzing sets of known target compounds.
Figure 1: GCMS analysis of p,p-DDE in spinach. Data were generated using a J&W DB-5ms column, a model 7890A GC system, and a model 5975C MSD mass spectrometer (Agilent Technologies).
and analyzes them for pesticide residues to enforce established tolerances (http://www.cfsan.fda.gov/~dms/pes0406.html). Domestic samples are collected at the point of production, and import samples are collected at the point of entry into U.S. commerce. Products typically are analyzed as the unwashed, whole, raw commodity. The FDA also performs a total diet study (TDS) on four regional market baskets per year, each basket comprising about 300 foods that represent the average U.S. consumers diet. The foods are prepared as they would be consumed (table-ready) and tested for regulated pesticides. The FDA also performs focused sampling that is generally used to follow up on suspected problem areas or acquire residue data on food commodities that are not usually covered during regulatory monitoring. There are more than 1000 registered pesticides in the U.S., and approximately 400 with tolerances established by the Environmental Protection Agency and enforced by FDA. The European Union (EU) has also recently updated a harmonized regulation of pesticide residues in food, with the new standard taking effect on September 8, 2008 (http://ec.europa.eu/food/plant/protection/pesticides/index_en.htm). This regulation sets maximum residue levels (MRL) for food and animal feed. The
MRL is the highest level of a pesticide residue that is legally tolerated. The regulation covers around 1100 pesticides currently or formerly used in agriculture in or outside the EU. For pesticides that are not specifically mentioned, a general default MRL of 0.01 mg/kg (10 ppb) applies. The EU designates the minimum number of samples to be taken by each member state, and EU reference laboratories coordinate, train staff, develop methods of analysis, and organize tests to evaluate the skills of the various national control laboratories. Japan also adopted a stricter pesticide regulation in 2006, referred to as the Positive List system, which establishes MRLs for about 500 pesticides (http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/introduction.html). Produce with pesticide residues exceeding these MRLs or the default tolerance of the uniform limit of 0.01 ppm (10 ppb) for those pesticides for which there is no Japanese MRL cannot be marketed in Japan. The regulation also lists 16 pesticides that must not be detected in food. Japan tested more than 200,000 food samples in 2007. The increased international focus on pesticide contamination in food impacts food producers in virtually any country, since the food market is a global one. Producers in Thailand, for example, must test
m/z
m/z
m/z
Figure 2: Confirmation of p,p-DDE in spinach using DRS and the Pesticide and Endocrine Disruptor spectral library (Agilent Technologies).
A unique fragment ion for each pesticide is routinely used for screening of each target compound a second precursor to product ion transition can then be used to confirm the identity of the compounds. The transition with the greatest response is typically used to quantify the pesticides. Laboratories may be faced with the challenge of determining all pesticides, if any, present in a food sample. LCtime of flight MS (TOF) screening methods provide extremely accurate molecular mass information (typically < 2 ppm) that can be searched against exact mass databases of thousands of compounds. Compounds found in this type of screen can be confirmed and quantified using LCMS-MS analysis on either QQQ or quadrupole time-of-flight (Q-TOF) instruments.
Analyzing for Target Pesticides

One of the most common activities in a food-testing laboratory is analysis for a set of target pesticides of specific interest. A finite set of pesticides might be used for production of a particular food crop, or a specific set might be especially
toxic to children, necessitating the routine analysis for those compounds. GCMS (single quadrupole) utilizing deconvolution software has broad applicability for this type of analysis. The Automated Mass spectral Deconvolution and Identification System (AMDIS) was developed by the National Institute of Standards and Technology, and is incorporated into deconvolution reporting software (DRS) available from Agilent Technologies. Deconvolution separates unresolved peaks at any given retention time into sets of cleaned component spectra that can be matched to library spectra. The analyst runs the method that was used to generate the Pesticide and Endocrine Disruptor Database Library and locks the method, so that retention times will match those stored in the database. DRS then uses deconvolution and locked retention times to identify pesticides from the database. This is particularly useful in the analysis of extracts from foods such as spinach (Figure 1). In this case, one of the pesticides of interest is p,p-
dichlorodiphenyldichloroethylene (p,pDDE), a breakdown product of DDT. The total ion chromatogram (TIC) does not show a discernible peak at the p,p-DDE retention time (24.07 min). Deconvolution analysis found 768 different components over the whole TIC. After a pesticide library search, 12 of these had a match, including p,p-DDE. The black ions in the bottom window of Figure 1 represent the raw scan, before deconvolution. The white ions are a deconvoluted component that matches the known spectrum of p,p-DDE. Plotting the extracted ion chromatograms for this component (246, 248, 316, and 318) in the top window of Figure 1 also helped to confirm the presence of the p,p-DDE peak in the TIC. Figure 2 illustrates the extraction of the p,pDDE spectrum from the raw spectrum, and its match to the library spectrum, resulting in a positive confirmation of the presence of this pesticide in the spinach sample. The use of powerful integrated deconvolution software such as DRS, in combination with specialized database libraries such as the Japanese Positive List Pesticide Library (1) or the more comprehensive Pesticide and Endocrine Disruptor Library (2) (Agilent Technologies, Palo Alto, California), can increase the efficiency of a food testing laboratory greatly. These GCMS methods can screen for hundreds of compounds and quantify a subset of them in as little as 2 min after the run using DRS. With traditional GCMS data processing, an analyst would need as much as 15 min per sample to review the data and confirm pesticide hits. LCMS methods are also very useful for analysis of target pesticides that are too polar or thermally labile for GCMS. LCQQQ is especially suited to very complex food matrices that can interfere with the analysis.
Abundance
Abundance
Abundance
Dealing with Dirty Sample Matrices

Sample preparation is a significant challenge with food analysis and makes up a considerable portion of the total pesticide analysis time. Robust analysis methods that can tolerate very complex sample matrices without loss of sensitivity or accuracy are a must to reach throughput goals.
y=208335.0516 x - 152717.0728 R2=0.99956076
Figure 3: Calibration curve for diazinon in pepper using a six-point curve from 0.1 to 100 ng/mL with a linear fit and no origin treatment. Data were generated using a ZORBAX SB-C-18 1.8-m LC column with a model G6410A QQQ system in the positive ESI mode (Agilent Technologies).
Area 1 Area 2 Area 3 Area 4 Area 5 Mean RSD %
1197 1280 1347 1186 1291 1260 5.4
Figure 4: Precision of GC-QQQ quantification of 500 fg of cyanophos spiked in garlic. Data were generated using an HP5ms UI (Ultra Inert) 30 m 0.25 mm, 0.25-m film thickness GC column and a model 7000A QQQ MS system (Agilent Technologies). The pesticide cyanophos was spiked into a garlic matrix at 0.5 ppb and replicate analyses performed at 500 fg on-column. The peaks are superimposed for the five runs, with the replicate peak areas, mean peak area, and RSD listed in the table to the left.
Multiple reaction monitoring (MRM) with a QQQ system provides detection of trace amounts of target compounds in complex matrices, providing screening, confirmation, and quantification in one analysis (although a replicate analysis is usually required in most laboratories for confirmation). The precursor ion of interest generated by the source is selected and isolated in the first (Q1) quadrupole. A hexapole collision cell (Q2) fragments the molecule, and the product ions are selected in Q3. By choosing product ions from this transition that are characteristic for the pesticides of
interest, chemical noise is separated from signal, providing very high sensitivity and selectivity, even in very dirty matrices. With careful selection of highly specific product ions for each pesticide, it is possible to create an MRM assay that can be used to identify and quantify hundreds of pesticides in a sample simultaneously. Efficient use of LCQQQ MRM methods requires optimization of a few parameters, including voltages on the fragmentor and the collision cell. Parameters must be optimized to generate the most intense product ions representative for each target compound, and
automated method optimization software simplifies this task enormously. LCQQQ is an excellent platform for screening, confirmation, and quantification of pesticides.A screening method uses one transition per compound to look for a large number of pesticides on the target list. Confirmation of pesticides detected in the screening is done using two transitions per compound. For example, 301 pesticides have been rapidly identified in one analysis with most having a limit of detection (LOD) of 0.01 mg/kg (10 ppb), which is the MRL for baby food in the U.S. and the general default MRL for foods in
the European Union. These levels were reached in a single analysis using positive ion electrospray with 99 transitions per time segment, and a quantifying and qualifying ion for each compound (3). Two types of vegetables (pepper and tomato) were evaluated, and a linearity of response over three orders of magnitude was shown for quantification (r2 > 0.99; Figure 3). GCQQQ also provides sensitive and
reproducible screening, confirmation, and quantification of pesticides that are not highly polar, particularly in very complex matrices such as garlic and ginger. Despite the high background of these matrices, pesticide compounds can be detected at levels as low as 0.5 ppb, with very high signal to noise and peak area precision as good as 6% RSD (Figure 4). For certain active pesticides such as acephate, the
Using oven backflush it took an additional 33 min at 320 C to remove those boilers
Run stopped at 42 min, backflushed at 290 C for 7 min Blank run after backflushed shows the column was clean
Time (min)
Figure 5: Backflushing the GC column for higher signal to noise and shorter cycle times. Comparison of backflushing using capillary flow (red chromatogram) versus column bake-out (blue chromatogram) utilizing a model 7890A GC system and an HP-5ms GC column (Agilent Technologies).
use of chemical protectants can reduce adsorption of analytes during the analysis, increasing sensitivity and improving peak shape. The result can be a marked improvement in accuracy of quantification in complex matrices (4). Backflushing the GC column, rather than using oven bakeout, also removes many high-boiling interfering substances in the sample (Figure 5) without altering retention times in subsequent runs or shortening the life of the column (5). Backflushing improves data quality while shortening cycle time and increasing column life. It also eliminates ghost peaks and reduces ion source contamination by keeping excess column bleed and heavy residues from being introduced into the mass spectrometer. One key to the successful rapid analysis of large numbers of pesticides is the dwell time of the QQQ mass spectrometer the time spent to measure the signal of any one precursor to product ion transition. The shorter the dwell time, the more transitions that can be analyzed per unit time, and the more pesticides that can be quantified per unit time.
Figure 6: Pesticide screening for unknowns using LCQTOF. The top panel shows the full spectrum total ion chromatogram (TIC) using a model 1200 LC system, a ZORBAX Eclipse XDB, 5-m column, and a model 6510 LCMS QTOF system (Agilent Technologies). The Molecular Feature Extractor found 510 compounds in the TIC, 15 of which had hits from the EXACT MASS database search (bottom panel). The three highlighted compounds were further confirmed by MSMS analysis.
Analyzing for Unknown Pesticides

Often, food-testing laboratories want to determine the entire spectrum of pesticides that might be present in a sample, rather than just a limited set of target compounds. While GCMS using deconvolution is an effective tool to screen for over 900 pesticides (2), LCTOF or LCQTOF can be used to screen for any number of LCMS-amenable compounds. TOF and QTOF are extremely accurate MS techniques, routinely achieving <2 ppm mass accuracy.Analysis software such as the Molecular Feature Extractor (MFE) from Agilent Technologies can find all ions representing real compounds in the sample. Noise and other extraneous ions are excluded. The resulting list of masses is then searched against a database of theoretical exact masses of compounds based on their molecular formulas and selected adducts, to identify pesticide compounds (Figure 6). This method has been used to screen food extracts for 600 pesticides and their degradates. The limit of detection (LOD) was determined for 100 of the 600 pesticides, and varied from <0.01 mg/kg (10 ppb) for 34% of the compounds, to <0.5 mg/kg (500 ppb) for 95% of the compounds (6). Using the MFE algorithm streamlines the data analysis for screening of hundreds of compounds at sensitive levels to minutes, compared with hours or even days using a manual approach. TOF can be used to identify an unlimited number of ionizable compounds, and the sensitivity is not impacted by the number of compounds screened.While LCTOF can provide screening and quantification of both unknown and targeted pesticides, LCQTOF or another MS-MS approach is required to confirm identification.
Conclusion
One need only look to recent events to appreciate the importance of pesticide test-
ing in food.A series of poisonings in Japan in December of 2007 and January of 2008 that sickened 10 people was linked to methamidophos pesticide contamination of gyoza dumplings imported from China. The impact on Japanese food buying habits was immediate, as importation of Chinese vegetables fell 40% shortly after the source of the poisoning was determined. In another example, pet food vegetable proteins contaminated with melamine were imported into the United States from China in early 2007, resulting in the deaths of several cats and dogs. More recently, the same contaminant has appeared in baby formula in China. Food-testing laboratories require a broad range of complementary MS-based analytical technologies and methods to screen, confirm, and quantify the presence of as many as 1000 pesticide compounds. Screening methods can be used to detect as many contaminants as possible in a single run, and these methods dont necessarily have to quantify all the compounds. The screening method can be followed by target compound analysis that is very selective, sensitive, and quantitative. For LCamenable pesticides, this means running LCTOF or LCQTOF followed by LCQQQ. For GC-amenable pesticides, this means running GCMS with DRS for screening followed by GCQQQ for target compound analysis at low levels. The method used by any given laboratory is determined by its specific pesticide analysis goals, existing instrumentation, the budget for acquisition of new instrumentation, and required LODs. When choosing new instrumentation, laboratories doing pesticide testing also should consider whether a particular manufacturer has a complete solution that includes all of the above platforms, and whether those platforms share common, interchangeable ion sources, a unified soft-
ware platform, and transferable collision voltages, thus providing a uniform workflow. The robustness of a manufacturers systems is a key factor, including the tolerance of the ion source to contamination and its required frequency of cleaning and maintenance. Utility of the software and its integration with comprehensive pesticide spectra libraries and exact mass databases are also keys to success. Finally, the manufacturers reputation for excellence in GC and LCMS instrumentation should be considered.
References
(1) P.L. Wylie, Screening for pesticides in food using the Japanese Positive List Pesticide Method, Agilent Application Note 59897436 (2007). (2) P.L. Wylie, Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library, Agilent Application Note 5989-5076 (2006). (3) J.A. Zweigenbaum, Multiresidue analysis of 301 pesticides in food samples by LC/Triple Quaduprole mass spectrometry, Agilent Application Note 5989-8614 (2008). (4) M. Anastassiades et al., J. Chromatogr., A 1015, 163184 (2003). (5) C.-K. Meng, Improving Productivity and Extending Column Life with Backflush, Agilent Application Note 5989-6018 (2006). (6) J.A. Zweigenbaum, Automated screening of 600 pesticides in Food by LC/TOF MS using a molecularfeature database search, Agilent Application Note 59895496 (2006).
Philip Wylie, Jerry Zweigenbaum, Melissa Churley, ChinKai Meng, and Cao Zhe are Senior
Applications Chemists, Agilent Technologies, Santa Clara, California.
Printed in U.S.A.
Reprinted from Current Trends In Mass Spectrometry, a Spectroscopy Supplement, November 2008
Determination of Pesticides in Water by SPE and LC/MS/MS in Both Positive and Negative Ion Modes Application Note
Environmental
Author
Chin-Kai Meng Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
Using solid phase extraction (SPE) and liquid chromatography/tandem mass spectrometry (LC/MS/MS), 46 pesticides in positive ion mode and 14 pesticides in negative ion mode were analyzed at low pg level on column without any derivatization. Good linearity was observed for all analytes from 5 pg to 1 ng on column.
Introduction
To monitor trace pesticide residues in surface and ground water, an effective sample preparation and analysis method is required. In 1996, U.S. Geological Survey National Water Quality Laboratory (NWQL) developed and implemented a graphitized carbon-based SPE and high-performance liquid chromatography (HPLC) method to determine polar pesticide concentrations [1]. Subsequently, the NWQL developed an HPLC-mass spectrometry (MS) method to improve the sensitivity and selectivity. This method is capable of quantifying pesticides and pesticide metabolites in filtered water at concentrations as low as 10 ng/L. Taking advantage of the Multiple Reaction Monitoring (MRM) technique, any interference and matrix signal from organic matter in the water can be minimized from the target compound signals for better confirmation and quantitation. In this application, SPE and LC/MS/MS methods are described to analyze 46 pesticides in positive ion mode and 14 pesticides in negative ion mode.
Experimental
Sample Preparation Procedure See reference 1 for more information 1. 2. Filter water samples in the field or laboratory using 0.7-m glass fiber filters. Pump 1 L of the filtered water sample, at a flow rate of 20 mL/min, through a Carbopak-B SPE cartridge containing 0.5 g of graphitized carbon sorbent. Elute the compounds with 1.5 mL methanol, followed by 13 mL of an 80:20 methylene chloride:methanol mixture that has been acidified with trifluoroacetic acid anhydride (0.2%). Reduce the two fractions to near dryness and then combine. The final volume of the extract is 1 mL.
3.
4.
Calibration Samples Separate stock solutions of either positive ion mode or negative ion mode analytes were diluted 1 to 10 to make the calibration standard solutions. The concentrations of the seven calibration solutions were 5, 10, 50, 100, 200, 500, and 1,000 pg/L (ppb).
Instrumentation
Positive Ion Mode LC: Column: Column temperature: Mobile phases: 1200 LC ZORBAX Extend-C-18, RRHT, 2.1 mm 100 mm, 1.8 m 40 C A: 0.1% formic acid in water, add NH4OH buffer to pH 5.5 B: Acetonitrile (ACN) 0.3 mL/min Time 0 15 20 21.5 1.0 L G6410A QQQ ESI (+) 100500 amu 300 ms 3500 V 40 psi 9 L/min 350 C 35 V Injection volume: MS: Ionization: Mass range: Scan time: Capillary: Nebulizer P: Drying gas: Gas temperature: Skimmer: %B 0 100 100 0 Negative Ion Mode LC: Column: Column temperature: Mobile phases: 1200 LC ZORBAX Extend-C-18, RRHT, 2.1 mm 100 mm, 1.8 m 60 C A: 0.04% glacial acetic acid in water B: Acetonitrile (ACN) 0.3 mL/min Time 0 1 2 3 4 8 9 1.0 L G6410A QQQ ESI () 120400 amu 300 ms 3500 V 40 psi 9 L/min 200 C 35 V %B 0 40 52 60 100 100 0
Flow rate: Gradient:
Flow rate: Gradient:
Injection volume: MS: Ionization: Mass range: Scan time: Capillary: Nebulizer P: Drying gas: Gas temperature: Skimmer:
The MRM parameters for positive ion mode and negative ion mode are listed in Tables 1 and 2, respectively.
Table 1. Positive Ion Mode MRM Method Parameters
Name 3(4 chlorophenyl) methyl/urea 3-Keto Carbofuran 3-OH Carbofuran Aldicarb Aldicarb sulfone Aldicarb sulfoxide Atrazine Bendiocarb (Ficam) Benomyl Bensulfuron Bromacil Caffeine Carbaryl Carbofuran Chlorimuron ethyl Cycloate Desethyl atrazine Desisopropyl atrazine
RT 8.05 8.24 6.90 8.20 5.52 4.99 9.86 9.32 6.61 10.86 8.44 5.48 9.69 9.36 11.57 14.52 7.06 5.94
Precursor 185 236 238 116 223 207 216 224 192 411 261 195 202 222 415 216 188 174 142 240 233 165 326 233 198 312 290 256 249 280 226 163 382 275 411 304 347 237 138 342 210 233 233 365 229 161
Quant ion 128 179 163 89 76 89 174 167 160 149 205 138 145 165 186 83 146 68 86 134 72 72 129 72 156 267 177 175 160 220 169 88 167 57 182 284 288 72 120 156 111 137 137 150 127 144
Qual ion 93 151 181 70 86 132 96 109 132 182 162 110 127 123 213 154 79 104 57 167 160 92 262 168 114 252 69 209 182 192 121 106 199 88 213 160 305 90 92 69 168 94 94 199 116 88
Collision V 10 10 10 5 5 5 20 5 30 15 20 15 15 10 10 15 15 30 15 20 20 15 20 20 20 20 30 15 15 10 5 5 15 20 15 30 10 10 10 20 5 15 15 15 15 15
Dwell 30 30 40 30 50 150 40 40 40 60 30 50 40 40 60 60 40 50 150 60 75 40 30 40 40 30 40 40 60 75 60 50 30 60 30 75 60 50 75 60 40 60 60 40 30 30
Segment 4 4 3 4 2 1 5 5 3 7 4 2 5 5 8 8 3 2 1 7 6 3 4 5 3 4 3 3 7 6 7 2 4 8 4 6 8 2 6 8 5 7 7 5 4 4
Desisopropyl desethyl atrazine 1.76 Diphenamid 10.82 Diuron 10.02 Fenuron Flumetsulam Fluometuron Hydroxy-atrazine Imazaquin Imazethapyr Imidacloprid Linuron Metalaxyl (Apron) Methiocarb Methomyl Metsulfuron methyl Neburon Nicosulfuron (Accent) Norflurazon Oryzalin Oxamyl (Vydate) Propham Propiconazole (Tilt) Propoxur (Baygon) Siduron Siduron isomer Sulfometuron, methyl ester Tebuthiuron Terbacil 6.90 7.49 9.70 6.82 7.68 6.99 7.01 11.45 10.15 11.28 5.77 8.43 12.99 7.97 10.51 12.58 5.59 10.68 12.89 9.26 11.12 11.28 9.25 8.14 8.69
Table 2.
Negative Ion Mode MRM Method Parameters
Name Clopyralid Picloram Dicamba DCPA Bentazone 2,4-D Bromoxynil MCPA Triclopyr 2,4-DP 2,4-DB MCPB Acifluorofen Dinoseb
RT 3.47 3.69 4.31 4.49 4.69 5.02 5.06 5.09 5.26 5.42 5.66 5.70 5.89 6.50
MW 191 240 220 330 240 220 275 200 255 234 248 228 361 240
Quant 190 > 146 239 > 195 219 > 175 273 > 215 239 > 132 219 > 161 274 > 79 199 > 141 254 > 196 233 > 161 247 > 161 227 > 141 360 > 316 239 > 193
Qual 192 > 148 241 > 198 219 > 145 271 > 213 239 > 197 221 > 163 274 > 81 201 > 143 256 > 198 235 > 163 249 > 163 229 > 143 360 > 286 239 > 163
Frag V 80 80 60 100 120 80 120 100 80 80 80 80 60 120
Collision V 5 5 0 5 25 15 25 10 10 5 10 5 5 25
Dwell 70 70 50 50 50 25 25 25 25 25 40 40 40 40
Segment 1 1 2 2 2 3 3 3 3 3 4 4 4 4

Figure 1 shows the total ion chromatogram (TIC) for the positive ion mode. As seen in Figure 1, the analysis time is less than 15 minutes for the 46 analytes. Using a 1.8 m
particle size column, the peak widths of these analytes are about 0.1 minute. The narrower peak width helps to achieve a higher signal-to-noise (s/n) ratio. Analysis time in negative ion mode is less than 7 minutes for the 11 analytes, as seen in Figure 2.
107 2
1.5
0.5
9 10 11 12 13 14 15 Abundance vs. Acquisition Time (min)
16
17
18
19
20
21
Figure 1.
Positive ion mode TIC of 46 pesticides.
A few compounds, for example, Dicamba, MCPB and 2,4-DB, are sensitive to heat from the drying gas. Higher drying gas temperature (350 C) will lower the intensity of the precursor ion.
Therefore, in the negative ion mode, the drying gas temperature was set to 200 C. Figure 3 shows the overlaid chromatograms of all 14 pesticides, each at 5 pg on column, from the negative ion mode MRM analysis.
x106 9 8 7 6 5 4 3 2 1 0 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 5.2 5.4 5.6 5.8 6 6.2 6.4 6.6 6.8 7
2,4-D Bromoxynil MCPA
Dinoseb
Bentazone
2,4-DP
Triclopyr Dacthal Dicamba Picloram Clopyralid
2,4-DB MCPB
Acifluorofen
Abundance vs. Acquisition Time (min) Figure 2. Negative ion mode TIC of 14 pesticides.
103 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 3.4 3.6 3.8 4
MCPA 5 pg on column
Dichloroprop 2,4-DP Acifluorofen
2,4-D
MCPB
Bentazone Clopyralid Picloram Dicamba

4.2 4.4
Triclopyr
2,4-DB Dinoseb
Dacthal
Bromoxynil
4.6
4.8
5.2
5.4
5.6
5.8
6.2
6.4
6.6
Abundance vs. Acquisition Time (min)
Figure 3.
Overlay of the MRM results from 14 pesticides in negative ion mode.
Table 3 shows the linearity results for 14 pesticides over the range of 5, 10, 50, 100, 200, 500, and 1,000 pg on column. The calibration model used was a linear model that included origin with no weighting. All analytes showed excellent linearity.
Table 3. Pesticide Linearity: 5, 10, 50, 100, 200, 500, 1,000 pg on Column
Pesticide Clopyralid Picloram Dicamba DCPA Bentazone 2,4-D Bromoxynil MCPA Triclopyr 2,4-DP 2,4-DB MCPB Acifluorofen Dinoseb
R2 (linear fit, include origin, no weighting) 0.9976 0.9993 0.9975 0.9994 0.9975 0.9990 0.9999 0.9980 0.9990 0.9948 0.9887 0.9847 0.9969 0.9905
Conclusions
Using SPE and LC/MS/MS, 46 pesticides in positive ion mode and 14 pesticides in negative ion mode were analyzed at low pg level on column without any derivatization. Good linearity of responses was observed from 5 pg to 1 ng of analytes on column.
Reference
1. U.S. Geological Survey Water-Resources Investigations Report 01-4134, http://nwql.usgs.gov/Public/pubs/WRIR01-4134.html
Acknowledgments
The author gratefully acknowledges the assistance of Stephen Werner and Ed Furlong of the National Water Quality Lab United States Geological Survey (Lakewood, CO) for providing the sample preparation procedures and extracts used in this work.

Agilent Technologies, Inc., 2008 Published in the USA September 18, 2008 5989-5320EN
Automated Screening of 600 Pesticides in Food by LC/TOF MS Using a Molecular-Feature Database Search Application
Food Safety
Authors
E. Michael Thurman and Imma Ferrer Pesticide Residue Research Group University of Almera 04120 Almera, Spain Jerry A. Zweigenbaum Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
the compounds to <0.5 mg/kg for 95% of the compounds. Strengths of the MF algorithm include rapid screening of hundreds of compounds at sensitive levels in minutes compared to a manual approach of hours to days.
Introduction
Contamination of foodstuffs with pesticides always means a risk to the consumer. Fetuses, infants, and children, a particularly sensitive group, are currently protected by a limit of 0.01 mg/kg in baby food by European Union legislation [1]. Maximum residue limits (MRLs) should be set at the lowest possible level with the aim of setting products on the markets without any measurable residues. However, this is often difficult because of the use of fungicides for transport and storage of fruit and vegetables, as well as the use of insecticides for crop protection. The analysis of pesticides in baby food has involved both GC/MS and LC/MS methods. Classically, GC/MS has been used to look for volatile pesticides, especially the chlorinated insecticides and herbicides, and LC/MS has been used to monitor the more polar insecticides, herbicides, and fungicides. GC/MS methods rely on both full scan and selected ion monitoring. Recently the use of reverse search methods for GC/MS has made it possible to search large NIST pesticide libraries in minutes [2] and has made screening quite simple for pesticides amenable to GC/MS methods.
Abstract
Searching a database using a molecular feature (MF) algorithm was developed for the screening of 600 pesticides and degradates in extracts of food by liquid chromatography time-of-flight mass spectrometry in positive ion mode with full-scan accurate mass spectra. The database search works by compiling the accurate mass of the ions detected and identified as real compound chromatographic peaks without ion extraction and compares them with the monoisotopic exact masses of the compounds in the database. The screening criteria consisted of 5 ppm accurate mass window, 0.2 minute retention time window, and a minimum area count of 1,000 counts (signal-to-noise ratio of ~10:1). The limit of detection and retention time was determined for 100 of the 600 compounds and varied from <0.01 mg/kg for 34% of
Unfortunately, similar reverse-search methods have not been available for LC/MS for two reasons. First, the single quadrupole and triple quadrupole methods do not operate in full scan mode for pesticide screening because of a lack of sensitivity. Second, standardization and reproducibility of CID fragmentation energy for broad usage has been difficult, making common spectral libraries unavailable. Recently, the use of LC/TOF MS has shown that it has the capability and sensitivity to obtain full scan spectra with accurate masses of less than 1 ppm [3] suitable for database searching [4].
to 100% B in 30 minutes at a flow rate of 0.6 mL/min Agilent 6210 LC/MS TOF with dual spray electrospray source Positive ESI, Capillary 4000 V Nebulizer 40 psig, drying gas 9 L/min, gas temp 300 C Fragmentor 190 V, skimmer 60 V, Oct DC1 37.5 V, OCT RF V 250 V Reference Masses: mass range (m/z) 121.0509 and 922.0098, resolution: 9500 500 at m/z 922.0098, m/z 50 to 1,000, Reference A Sprayer 2 is constant flow rate during the run
Experimental
Vegetable and Fruit Extraction (QuEChERS) QuEChERS is the acronym for the extractionmethod, which stands for quick, easy, cheap, effective, rugged, and safe. It is a method that is widely receiving acceptance for rapid extraction of pesticides in food [5, 6]. Weigh 15 g of a previously homogenized sample into a 40-mL Teflon centrifuge tube. Add 15 mL of acetonitrile (containing 1% acetic acid), add 6 g of anhydrous MgSO4 and 2.5 g of NaAc63H2O (sodium acetate trihydrate) and shake the sample vigorously for 1 min by using a vortex mixer at maximum speed or hand shaking. Centrifuge for 3 minutes at 3,700 rpm. Take 5 mL of supernatant into a 15-mL tube, add 250 mg of PSA adsorbent and 750 mg of MgSO4, and vortex and shake for 20 s. Then centrifuge again for 3 min at 3700 rpm. Transfer 1.0 mL into LC/MS vial. The 1.0 mL supernatant is then evaporated to dryness and brought back up in 8/92% methanol/water for LC/MSD TOF and ion trap analysis. Direct analysis of the fruit and vegetable extracts were analyzed by injecting 50 L. Nonfortified samples were analyzed directly at this same point by LC/MS TOF. LC/MS TOF Methods LC pumps: Agilent 1100 binary pumps, injection volume 50 L with standard Agilent 1100 ALS Column: ZORBAX Eclipse XDB 4.6 150 mm C-8, 5-micron (p/n 993967-906) Mobile Phase A = 0.1% formic acid in water, and B = acetonitrile, gradient began with 5 minutes isocratic at 10% B followed by a linear gradient

Molecular Feature Database Search Theoretical monoisotopic exact masses of compounds based on their molecular formula were calculated using an Excel spreadsheet and put into csv (comma separated values) format for use by the TOF software of the Agilent LC/TOF MS system for 600 pesticides known to ionize by positive ion electrospray. The csv file is created in the TOF software by an Excel spreadsheet tool called Formula DB Generator. The csv file is then searched automatically by the LC/TOF MS instrument at the completion of the sample run and a report generated on compounds that were found in the database. Search criteria include ppm mass tolerance (5 ppm), retention time window (0.2 minutes) if available, and minimum peak height count, which is called the compound threshold (1,000 counts or a signal-to-noise ratio of ~10:1 or 0.06% relative volume), and adducts and neutral fragment losses. The search routine is called a molecular feature extractor and is software recently available on the Agilent LC/TOF MS (November 2005). The molecular feature extractor finds all ions in an LC/MS TOF data file representing ions of real compounds in the sample analyzed. Noise and other extraneous ions are excluded. The resulting list of ions are then searched against the csv database using the chosen criteria and ions found are then tabulated from the full scan spectrum and checked for accuracy and retention time against the database. The molecular feature approach is more suited to large libraries because of the ease of operation and the quickness for which the search is done. Thus,
each ion of interest in the database is not extracted from the sample file as in a reverse search. This procedure requires much more time with LC/TOF MS data files collected in profile mode. From this point, confirmation is carried out manually by checking the positive screens for retention time match and fragment ions (if present). The sample may be reanalyzed at a higher fragmentor voltage to check for fragment ions and confirmed by authentic standard analysis. Limit of Detection The limit of detection (LOD) was determined in several matrices, including spiked food samples and solvent extracts for 100 compounds (Table 1). These compounds consist of the major classes of pesticides that are commonly used in the United States and Europe for treating crops of fruits and vegetables. The limit of detection was based on an accurate mass of less than 3 ppm and the appearance of the correct accurate mass of A+1 and A+2 isotopic signatures. The LOD was determined at various levels for pesticide work, including the EU regulation of 0.01 mg/kg for baby food, and 0.05, 0.1, and 0.5 mg/kg for various food levels, depending on pesticide and crop type. The LOD was equal to or less than 0.01 ppm for 33 compounds and less than or equal to 0.05 for 60 compounds (60%). The 0.05 mg/kg LOD is also a critical one for food monitoring of banned substances or controlled compounds. The LOD for 95% of the compounds was equal to or less than 0.2 mg/kg for food. Only six compounds were found
to be insensitive with LODs of 0.5 mg/kg for food. The insensitive compounds were promecarb and aldicarb, which are two carbamate pesticides that fragment easily in the electrospray source and give low abundance for the MH+ ion. Likewise, malathion oxon and dimethoate are two organophosphate pesticides that fragment easily. Thus, these compounds could be more sensitively detected by the use of the more abundant fragment ion rather than the MH+. For example, Figure 1 shows the mass spectrum of dimethoate. The MH+ ion is not the major fragment ion of the spectrum, and in fact has an intensity about three to four times less than the m/z 124.9819 ion. Furthermore, it must be taken into consideration that the LOD is affected by the matrix for two reasons. One is suppression of ionization and the second is interfering ions of nearly identical mass. Suppression of ionization has been tested in our previous work for some of these pesticides in food [7], including pepper, broccoli and tomato, melon, orange, and lemon, so we have experience on which matrices are most difficult. For example, Figure 2 shows the matrix chromatogram for pesticide-free pepper, which gives a complicated chromatogram. The MF database search identified ~3,000 compounds of which none were pesticides. These peaks contained signal-to-noise ratio of 10:1 or greater and present an extremely difficult matrix for which to find ions, especially at trace levels. Spiked food extracts of these difficult matrices were used to establish LOD for the compounds shown in Table 1. An example printout is shown in Figure 3. The report contains formula, compound,
SH P H3CO OCH3 O C5H13NO3PS2+ Exact mass: 230.0069 S H N CH3
Figure 1.
Dimethoate mass spectrum showing low intensity of MH+ ion and importance of using characteristic fragment ions to lower LODs on some compounds of low intensity, specifically, m/z 124.9819 ion.
Figure 2.
Blank pepper sample showing complexity of the sample ~3000 accurate mass peaks were detected in this sample at signal to noise of 10:1 or greater.
accurate mass of the neutral molecule, error in mDa and error in ppm, retention time error in minutes, and a description (specifically, fungicide). The mass spectrum of the MH+ and isotope signature of the compound are also shown, which is useful for a quick check and partial confirmation of the formula, especially since most pesticides show an interesting A+2 ion from a halogen or sulfur atom. A+2 Ions and Empirical-Formula Confirmation Because the MF database search is a screening program, a formula is not unequivocally confirmed at this point with a mass accuracy window of 5-ppm. The molecular formula (not molecular identification) may be confirmed by the A+2 isotopic signature of the compound. For example, 70% of the 600-compound database contains either S, Cl, or
Br, which give the isotopic cluster of the A+2 ions. The accurate mass of the isotope along with the intensity profile can then be used as a first level confirmation of the empirical formula. Thus, this adds a great deal of confidence to the screening data of the accurate mass but of course does not yet meet the standards for identification of the molecule, a topic discussed later. For example, let us examine the isotopic signature of imazalil in a pear extract (Figure 4). The measured mass of the MH+ was 297.0564 and the chlorine 37 isotope was 299.0533. Thus the difference in mass is 1.997 mass units, which is the mass defect of a chlorine 37 atom relative to the chlorine 35 atom that has been replaced. Furthermore, the intensity of the A+2 peak is about 2/3 of the A peak, which is consistent with two chlorine atoms in the molecule. This is further established by the presence of an A+4 peak at 301.0503. Thus, these
Limits of Detection for Pesticides in Food Samples with Retention Time and Accurate Mass Accurate mass Compound Ret time Formula molecule Atrazine 21.1 C8H14N5Cl 215.0938 Azoxystrobin 24.0 C22H17N3O5 403.1168 Benalaxyl 26.8 C20H23NO3 325.1678 Buprofezin 27.2 C16H23N3OS 305.1562 Cyanazine 22.0 C9H13N6Cl 240.0890 Diazinon 27.6 C12H21N2O3PS 304.1010 Difenconazole 26.4 C19H17Cl2N3O3 405.0647 Isomer 26.6 C19H17Cl2N3O3 405.0647 Difenoxuron 21.3 C16H18N2O3 286.1317 Dimethomorph 22.2 C21H22NO4Cl 387.1237 Fenamiphos 23.9 C13H22NO3PS 303.1058 Imazalil 18.0 C14H14N2OCl2 296.0483 Imazapyr 20.0 C13H15N3O3 261.1113 Imazaquin 20.0 C17H17N3O3 311.1270 Irgarol 21.2 C11H19N5S 253.1361 Irgarol metabolite 17.0 C8H15N5S 213.1048 Isoproturon 21.3 C12H18N2O 206.1419 Mebendazole 18.2 C16H13N3O3 295.0957 Metolachlor 25.6 C15H22NO2Cl 283.1339 Metribuzin 15.0 C8H14N4OS 214.0888 Nicosulfuron 17.0 C15H18N6O6S 410.1009 Prochloraz 23.0 C15H16Cl3N3O2 375.0308 Prometon 16.6 C10H19N5O 225.1590 Prometryn 19.0 C10H19N5S 241.1361 Propazine 23.0 C9H16N5Cl 229.1094 Propiconazole 25.9 C15H17CI2N3O2 341.0698 Isomer 26.1 C15H17CI2N3O2 341.0698 Simazine 18.8 C7H12N5Cl 201.0781 Spinosyn A 20.9 C41H65NO10 731.4608 Spinosyn D 21.9 C42H67NO10 745.4765 Spiroxamine 19.6 C18H35NO2 297.2668 Isomer 19.7 C18H35NO2 297.2668 Terbuthylazine 23.4 C9H16N5Cl 229.1094 Terbutyrn 20.4 C10H19N5S 241.1361 Triflumazole 25.9 C15H15ClF3N3O 345.0856 Acetamiprid Acetochlor Alachlor Bensultap Bromuconazole Carbaryl Carbendazim Carbofuran Cartap Chlorfenvinphos Cyproconazole Cyromazine Deethylatrazine Deisopropylatrazine Dichlorvos Dimethenamide Dimethoate Diuron 16.3 23.0 23.0 21.4 23.8 21.3 6.2 20.4 3.1 26.5 23.4 2.9 15.3 12.1 20.0 24.0 16.3 21.0 C10H11N4Cl C14H20NO2Cl C14H2ONO2Cl C17H21NO4S4 C13H12N3OCl2Br C12H11NO2 C9H9N3O2 C12H15NO3 C7H15N3O2S2 C12H14Cl3O4P C15H18N3OCl C6H10N6 C6H10N5Cl C5H8N5Cl C4H7Cl2O4P C12H18NO2SCl C5H12NO3PS2 C9H10N2OCl2 222.0672 269.1183 269.1183 431.0353 374.9541 201.0790 191.0695 221.1052 237.0606 357.9695 291.1138 166.0967 187.0625 173.0468 219.9459 275.0747 228.9996 232.0170
Table 1.
LOD food mg/kg 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 5
Table 1.
Limits of Detection for Pesticides in Food with Retention Time and Accurate Mass (Continued) Ret time 21.8 15.7 15.7 19.2 22.5 16.9 21.2 23.5 12.1 18.7 11.9 19.1 28.6 24.0 25.6 29.0 3.7 17.7 4.5 18.5 6.0 20.6 20.4 25.0 11.5 28.6 14.0 24.1 17.4 24.8 27.3 17.0 27.5 11.4 18.5 23.8 15.0 25.0 29.2 18.8 27.2 14.6 25.4 25.0 27.6 24.4 28.2 23.0 3.2 Formula C11H15NO2S C9H12N2O C9H10N5O2Cl C13H18N2O2 C10H19O6PS2 C10H19O6PS2 C15H21NO4 C11H15NO2S C5H10N2O2S C9H11ClN2O C11H15CIN4O2 C14H18N2O4 C11H15BrCIO3PS C12H17NO2 C11H14NOCl C14H21NOS C10H7N3S C10H9N4SCl C5H11NS3 C7H14N2O2S C7H14N2O3S C11H13NO4 C10H13N2OCl C14H13N3O2SF4 C8H15N5O C17H8N2O3Cl2F8 C10H10N4O C6H11N2O4PS3 C11H15NO4S C9H17NOS C10H14NO5PS C9H9NOCl2 C13H9CI3N2O C7H14N2O4S C9H13N2O2Br C13H12N3OCI2Br C11H23NOS C14H9N2O2ClF2 C21H11N2O3ClF6 C7H5N2O3Cl2F C16H8N2O3Cl2F6 C11H10N2OCl2 C13H13N3O3Cl2 C13H19N3O4 C14H6N2O2Cl2F4 C9H8CI3NO2S C7H7CI3NO3PS C23H30O4 C5H13NO6S4 Accurate mass molecule 225.0823 164.0950 255.0523 234.1368 330.0361 330.0361 279.1471 225.0823 162.0463 198.0560 270.0884 278.1267 371.9351 205.1341 211.0764 251.1344 201.0361 252.0236 181.0054 190.0776 206.0725 223.0845 212.0716 363.0665 197.1277 509.9784 202.0855 301.9619 257.0722 187.1031 291.0330 217.0061 313.9780 222.0674 260.0160 374.9541 217.1500 310.0321 488.0362 253.9661 459.9816 256.0170 329.0334 281.1376 379.9742 298.9341 320.8950 370.2144 310.9626 LOD food mg/kg 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.5 0.5 0.5 0.5
Compound Ethiofencarb Fenuron Imidacloprid Lenacil Malathion 1 Malathion 2 Metalaxyl Methiocarb Methomyl Monuron Nitenpyram Oxadixyl Profenfos Promecarb Propachlor Prosulfocarb Thiabendazole Thiacloprid Thiocyclam Aldicarb Aldicarb sulfoxide Bendiocarb Chlorotoluron Flufenacet Hydroxyatrazine Lufenuron Metamytron Methidathion Methiocarb sulfone Molinate Parathion ethyl Propanil Triclocarban Aldicarb sulfone Bromacil Bromuconazole Butylate Diflubenzuron Flufenoxuron Fluroxypyr Hexaflumuron Imazalil degradate Iprodione Pendimethalin Teflubenzuron Captan Chloropyrifos methyl Spiromesifen Thiosultap
Figure 3.
Example of a report from the molecular feature database search.
data are excellent for the confirmation of the molecular formula for imazalil (however not for the identification of imazalil). The mass accuracy was 0.8 mDa off or 2.7 ppm error. Thus, this is an example of how one confirms the database screening of a pesticide molecular formula. The database report prints the isotopic signature and accurate masses to help for the manual screening of the database hits. When the formula does not contain an A+2 atom (that is, consists of only C, H, N, and O), then formula confirmation is not readily possible by this method. The reason being that only an A+1 peak is present and this peak is dominated by the carbon 13 signal of the molecule, which in itself is not enough for formula confirmation. Thus, for these pesticides (about 30% of the database) more data are required (specifically, retention time or fragment ion). Screening of Fruits and Vegetables Table 2 shows the results of screening six fruit and vegetable samples from a nearby grocery store (apple, pear, tomato, potato, pepper, and cucumber) and one commercial brand of olive oil for the 600 pesticides in the MF database search. The MF database search found from 617 to 2,681 accurate mass peaks in the sample chromatograms. The least complicated sample matrix was the tomato with 617 peaks, and the apple was the most complex sample with 2,681 peaks. The sensitivity of the MF database search was set at a signal-to-noise ratio of 10:1. The quantity of peaks found approximately doubles with decreasing the signal-to-noise
ratio from 20:1 to 10:1. The value of 10:1 is chosen in order to obtain good values for the isotopic signature of the compound, which is the A+1 and A+2 isotope signatures with the maximum instrument sensitivity. The accuracy window of the MF database search is set at 5 ppm in order to be well within the mass accuracy of the LC/TOF MS system, which typically operates at less than 3 ppm and often at 1 to 2 ppm or approximately 0.3 mDa [3]. The number of pesticides found in the 5 ppm mass window of these samples varied from 8 to 41 compounds. The only criterion to be included in this match was that the MH+ ion was within 5 ppm of the database value. Thus, as an example, the pepper sample, which had 2,402 peaks, had only 41 of these peaks that met the 5-ppm accuracy window (Table 2). Of these 41 peaks only three formulae were confirmed based on the correct isotope signature and retention time match (for compounds not containing an A+2 isotope), which was checked not only in the printout of the automated database match, but also by manual confirmation of the data file. The confirmation of pesticides varied from no detections in the potato sample, one pesticide in olive oil, three pesticides in pepper and tomato, and five pesticides in the cucumber and apple. The most common compound found in the fruit and vegetable samples was imazalil, which is a postharvest fungicide used for transport and storage of fruits and vegetables before their sale. Other compounds included organophosphate insecticides, such as diazinon, phosmet, and malathion and the
7
Cl
+1.997
H2C
A+2
N NH
Cl
1 37Cl
+1.997
1 35Cl A+4
Figure 4.
Isotope cluster of the m/z 297 ion and imazalil structure in a pear extract.
oxon of malathion, which is a pesticide degradate. The insect growth regulator buprofezin was found in a tomato sample, as was thiophanate methyl and carbendazim, both fungicides. The software originally used for this work did not address saturated peaks and compounds at high concentrations such as imazalil in the pear sample and buprofezin in the tomato sample were incorrectly identified. The latest software handles saturated peaks and correctly identified these compounds. The accuracy of all confirmed samples had an absolute-value average of 0.3 mDa or 1.2 ppm and a standard deviation of 0.25 mDa and 1.0 ppm, respectively (Table 2 ). The absolute-value average for retention time match was 0.07 minutes and standard deviation of 0.09 minutes. Thus, the windows chosen for the database are chosen with enough margin of error to find 99% of the compounds based on two standard deviations of the mean for mass accuracy and retention time.
Screening of Baby-Food Samples Of the 100 tested compounds, 33 met the baby-food screening level of 0.01 mg/kg. The only compound detected by the database search was a trace level of imazalil in the puree of pear, banana, and orange. Furthermore, examination of the label shows that lemon juice was also used in the baby food, which was a possible source for the imazalil. The, concentration though was approximately 0.0005 mg/kg, which is considerably less than the levels for baby-food safety of 0.01 mg/kg. Imazalil is easily screened because it contains the characteristic A+2 chlorine signature with two chlorine atoms. The error in identification was 0.3 mD or 1 ppm. None of the other compounds of the database was detected. Confirmation was not possible on this sample because of the low signal of the fragment ion and the low concentration of the suspected imazalil (0.0005 mg/kg). The sample was screened as safe, though, based on the health limit of 0.01 mg/kg for baby food. Approximately 10 different baby-food samples have been
Table 2.
Screened Pesticides in Food Samples Using the MF Database Peaks screened 2681 Pesticide matches < 5 ppm 12 Pesticides confirmed LC/TOF MS Imazalil Imazalil degradate Iprodione Fluqinconazole Difenoconazole Terbuthylazine (Deisopropylatrazine) Imazalil Diazinon Buprofezin Buprofezin (Benzthiazuron) Carbendazim Thiophanate methyl Thiabendazole Malathion isomer 1 Malathion isomer 2 Malathion oxon Imazalil Imazalil Carbendazim Imazalil degradate Phosmet None Mean Standard deviation 0.30 0.25 1.20 1.00 0.2 0.04 0.22 0.25 0.1 0.21 0.51 0.31 1 0.1 0.7 0.8 0.3 1.1 2 1 0.01 0.05 0.03 0.08 0.05 0.05 0.03 0.04 0 0 Error mDa 1 0.22 0.05 0.74 0.06 0.3 0.11 1 Error ppm 3.9 0.7 0.1 1.8 0.2 1 0.3 3.3 Retention time error (min.) 0.08 0.12 FALSE negative 0 FALSE positive 0
Sample Apple
Olive oil Pepper
1678 2402
10 41
0.09 0.07 0.12 0.1
0 0
0 1
Tomato
617
0 1
Cucumber
1619
17
Pear
1209
14
Potato
1150
11
screened, including a variety of brand-name vegetables and fruits and, fortunately, no positive detections have been found for the pesticides in the MF database search with the exception of imazalil shown above. The baby food samples represent the most difficult samples to screen because of the low LODs required. Weakness and Strengths of MF Database Search The only weakness of the database is the loss of mass accuracy because of interferences in the matrix. This problem is easily solved by the addition of a second ion (fragment ion) or a sodium or ammonium adduct ion for added confidence from matrix interference. The use of accurate mass LC/TOF MS combined with database searching is a powerful example of a new wave of monitoring technology for identification of pesticides in food and water. The use of
classical fragmentation libraries with comparison of fragmentation patterns is not needed in LC/TOF MS. Rather the use of molecular formula and the calculated accurate mass, especially when combined with one or two fragment ions of accurate mass, gives the identity of compounds without the worry of the intensity of the fragment ions and how this may vary from instrument to instrument and matrix to matrix. For example, the Molecular Hunter Software works in conjunction with the Molecular Feature Database using the .mhd files to link ions in groups according to their exact retention time (matching within 0.005 minutes). Therefore, it is possible to find and differentiate the fragment ions of a pesticide, such as imazalil, from the background ions of the matrix. Thus, it is the view of the authors that a large problem in LC/MS libraries is on the verge of being solved with the use of accurate mass
databases for pesticide screening in food with a molecular feature algorithmic approach using the MH+ ion and a major fragment ion.

References
1. PAN Europe position on the European Commission Proposal for a Regulation of the European Parliament and of the Council on maximum residue levels of pesticides in products of plant and animal origin COM(2003) 117 final, 2003/0052 (COD). 2. Phillip L. Wylie, Michael J. Szelewski, Chin-Kai Meng, and Christopher P. Sandy, Comprehensive Pesticide Screening by GC/MSD using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1157EN. 3. Imma Ferrer, E. M. Thurman, 2005, Measuring the mass of an electron by LC/TOF MS: A study of twin ions, Analytical Chemistry, 77, 33943400. 4. E. Michael Thurman, Imma Ferrer, 2005, Identification of unknown pesticides in food using both LC/MSD TOF and ion trap MSn, Agilent Technologies, publication 5989-1924EN. 5. M. Anastassiades, S. J. Lehotay, D. Stajnbaher, and F. J. Schenck, Fast and Easy Multiresidue Method Employing Acetonitrile Extraction/Partitioning and Dispersive Solid-Phase Extraction for the Determination of Pesticide Residues in Produce, (2003) Journal of AOAC International, 86:412431. 6. S.J. Lehotay, K. Matovsk, A.R. Lightfield, Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, (2005) Journal of AOAC International, 88:615629. 7. Imma Ferrer, Juan F. Garcia-Reyes, M. Mezcua, E. M. Thurman, A. R. Fernandez-Alba, 2005, Multiresidue pesticide analysis in fruits and vegetables by liquid chromatography time-offlight mass spectrometry, J. Chromatography A, 1082, 8190.
Multiresidue Analysis of 100 Pesticides in Food Samples by LC/Triple Quadrupole Mass Spectrometry Application
Food Safety
Authors
Imma Ferrer and E. Michael Thurman Pesticide Residue Research Group University of Almeria Almeria, Spain Yanyan Fang, Paul Zavitsanos, and Jerry A. Zweigenbaum Agilent Technologies, Inc. USA
Introduction
In recent years, the established regulations regarding the maximum residue limits (MRLs) in commodities have become more and more stringent. The European Union (EU) has set new directives for pesticides at low levels in vegetables in order to meet health concerns. For fruits and vegetables intended for production of baby food, an MRL of 10 g/kg is applicable for all pesticides, and compounds without a stated regulation also have the lowest MRLs at 10 g/kg. The low MRLs have encouraged the development of more sensitive analytical methods to meet the requirements in complex samples. In this sense, liquidchromatography tandem-mass spectrometry (LC-MS-MS) with triple quadrupole in multiple reaction monitoring (MRM) mode has become, so far, the most widely used technique for the monitoring and quantitation of pesticides in food, as reported extensively in the literature. On the other hand, high-resolving power mass spectrometric techniques, such as time-of-flight mass spectrometry (TOF-MS), have been applied recently for screening purposes as well. Nevertheless, the simplicity of methodologies using triple quadrupole as a detection technique, together with the low limits of detection achieved and the MS/MS capability make this technique a valuable tool for routine
Abstract
An analytical methodology for confirming the presence of a group of 100 pesticides in vegetable and fruit samples was developed using the Agilent G6410AA Triple Quadrupole Mass Spectrometer (QQQ). One transition per parent compound was monitored in a single chromatographic run containing two time segments. The sensitivity obtained meets the maximum residue levels (MRLs) established by the European Union regulation for food monitoring programs. The analytical performance of the method was evaluated for different types of fruit and vegetables + orange, tomato, and green pepper + showing little or no matrix effects. Linearity of response over two orders of magnitude was demonstrated (r > 0.99). This study is a valuable indicator of the potential of the QQQ for routine quantitative multiresidue analysis of pesticides in vegetables and fruits.
monitoring programs established in regulatory official laboratories. The easiness of use is sometimes an essential for these types of regulatory agencies, which lack the high-skilled personnel required for more sophisticated techniques such as TOF-MS. Triple quadrupole technology is not new in the sense that it needs to be validated for monitoring purposes and its basis is already wellestablished for routine analysis. Our study in this report is one of the first of its kind to examine the new Agilent Triple Quad for the analysis of pesticides in fruit and vegetables. This topic was chosen because of the relevance of these compounds and their significant use on food commodities. The sensitivity of the QQQ easily meets the levels required by the regulations on pesticides in food.
LC/MS/MS Instrumentation
LC Conditions Column: Agilent ZORBAX Eclipse XDB C-8, 4.6 mm 150 mm, 5 m, (p/n 993967-906). 25 C A = 0.1% formic acid in water B= Acetonitrile 0.6 mL/min 10% B at 0 min 10% B at 5 min 100% B at 30 min 1-5 L
Column temperature: Mobile phase: Flow-rate: Gradient:
Injection volumes: MS Conditions Mode:
Experimental
Sample preparation Pesticide analytical standards were purchased from Dr. Ehrenstorfer (Ausburg, Germany). Individual pesticide stock solutions (around 1,000 g/mL) were prepared in pure acetonitrile or methanol, depending on the solubility of each individual compound, and stored at 18 C. From these mother solutions, working standard solutions were prepared by dilution with acetonitrile and water. Vegetable samples were obtained from the local markets. Blank vegetable and fruit extracts were used to prepare the matrix-matched standards for validation purposes. In this way, two types of vegetables and one fruit (green peppers, tomatoes, and oranges) were extracted using the QuEChERS method already described in a previous application [1]. The vegetable extracts were spiked with the mix of standards at different concentrations (ranging from 2 to 100 g/kg) and subsequently analyzed by LC/MS/MS.
Positive ESI using the Agilent G6410AA Triple Quadrupole Mass Spectrometer Nebulizer: 40 psig Drying gas flow: 9 L/min V capillary: 4000 V Drying gas temperature: 350 C Q1 resolution: Unit Q2 resolution: Unit Fragmentor voltage: 70 V Collision energy: 525 V MRM: 1 transition for every compound as shown in Table 1 Dwell time: 15 msec

Optimization of LC/MS/MS conditions A preliminary study of the optimal MRM transitions for every compound was carried out by injecting groups of analytes (around 10 analytes in one chromatographic run) at a concentration level of 10 g/mL. Various collision energies (5, 10, 15, 20, and 25 V) were applied to the compounds under study. The optimum energies were those that gave the best sensitivity for the main fragment ion and, as a general rule, left about 10% of parent compound in the spectra, and they were selected as optimum ones. Only one fragment ion was chosen as the most abundant product ion for every target compound. Results are shown in Table 1.
Table 1.
Analytical Conditions and Limits of Detection (LOD) for Each of the Compounds Tested Retention time (min) 2.7 2.7 3 4.5 6.4 6.6 7.9 10.8 11 11.2 11.5 11.9 12.5 13.9 14.5 14.8 14.8 15.4 15.5 15.7 16 16.4 16.9 17 17.2 17.2 17.5 17.8 17.9 17.9 18 18 18.4 18.5 18.5 18.6 18.7 18.9 19.4 19.5 19.5 19.6 19.7 20 20.1 20.3 20.3 20.4 20.4 20.4 20.5 Protonated molecule [M+H]+ 167 312 150 182 207 192 202 223 271 198 163 174 262 203 165 188 256 230 223 226 214 258 411 253 297 296 213 312 279 255 202 199 235 241 166 298 221 215 213 242 242 222 224 732 202 254 216 280 287 207 432 Product ion (m/z) 125 232 105 137 89 160 175 148 225 156 88 132 234 175 72 146 209 199 126 184 158 122 182 126 159 264 89 284 219 209 132 72 153 214 109 144 109 187 72 200 186 165 167 142 145 198 174 248 123 72 290 Collision energy 20 10 15 10 5 15 25 5 10 15 5 15 15 15 15 15 10 5 15 20 15 5 15 15 15 20 10 20 10 10 15 10 10 10 5 15 15 15 15 20 15 10 5 5 5 15 15 10 15 15 15 LOD (pg) 10 90 10 8 9 5 10 50 7 3 4 18 8 8 2 4 7 7 6 4 0.8 6 6 3 7 2 10 15 10 120 5 2 20 70 2 10 10 5 3 2 1 2 2 12 2 0.1 0.3 5 5 1 6 3
Compound name Segment 1 Cyromazine Thiosultap Cartap Thiocyclam Aldicarb sulfoxide Carbendazim Thiabendazole Aldicarb sulfone Nitenpyram Hydroxyatrazine Methomyl Deisopropylatrazine Imazapyr Metamitron Fenuron Deethylatrazine Imidacloprid Dimethoate Acetamiprid Prometon Irgarol metabolite Methiocarb sulfone Nicosulfuron Thiacloprid Imazalil Mebendazole Aldicarb Imazaquin Oxadixyl Fluroxypyr Simazine Monuron Lenacil Cyanazine Metolcarb Spiroxamine Dichlorvos Metribuzin Chlorotoluron Prometryn Terbutryn Carbofuran Bendiocarb Segment 2 Spinosad A Carbaryl Irgarol 1051 Atrazine Metalaxyl Difenoxuron Isoproturon Bensultap
Table 1.
Analytical Conditions and Limits of Detection (LOD) for Each of the Compounds Tested (Continued) Retention time (min) 20.5 20.7 20.7 21.3 21.6 21.7 21.9 22.2 22.5 22.6 22.7 22.8 23 23 23.2 23.2 23.2 23.3 23.3 23.7 23.7 24.1 24.6 24.7 24.8 24.9 24.9 24.9 25 25.1 25.2 25.3 25.4 25.5 25.8 26.2 26.4 26.5 26.7 26.8 26.9 27.1 27.6 27.9 28 28.5 28.7 29.2 29.7 Protonated molecule [M+H]+ 233 746 226 388 212 388 376 218 292 226 230 376 304 303 404 318 300 276 208 376 188 311 330 342 331 342 284 346 270 270 364 406 406 359 326 292 315 461 306 305 381 322 373 511 252 489 218 282 336 Product ion (m/z) 72 558 107 301 170 301 308 162 70 169 174 159 217 145 372 160 264 244 151 159 126 158 245 159 127 159 252 278 238 224 194 251 251 155 294 236 162 158 201 169 158 212 303 158 91 158 57 212 236 Collision energy 15 5 5 20 10 20 10 15 10 5 15 20 15 5 10 5 10 10 10 20 10 10 10 20 5 20 10 10 10 10 5 20 20 10 5 10 15 10 10 15 15 15 10 10 15 10 10 5 15 LOD (pg) 5 100 5 11 1 8 6 10 6 15 0.3 6 0.7 5 0.4 2 50 1 5 6 5 9 8 5 5 5 2 7 8 8 5 4 4 8 5 9 8 7 1 1 22 15 7 10 2 6 2 5 30
Compound name Diuron Spinosad D Ethiofencarb Dimethomorph isomer 1 Propachlor Dimethomorph isomer 2 Prochloraz Propanil Cyproconazole Methiocarb Terbutylazine Bromuconazole isomer 1 Fenamiphos Methidathion Azoxystrobin Phosmet Captan Dimethenamide Promecarb Bromuconazole isomer 2 Molinate Diflubenzuron Iprodione Propiconazole isomer 1 Malathion Propiconazole isomer 2 Metolachlor Triflumizole Alachlor Acetochlor Flufenacet Difenoconazole isomer 1 Difenoconazole isomer 2 Chlorfenvinphos Benalaxyl Parathion ethyl Triclocarban Hexaflumuron Buprofezin Diazinon Teflubenzuron Chlorpyrifos methyl Profenofos Lufenuron Prosulfocarb Flufenoxuron Butylate Pendimethalin Trifluralin
The MRM transitions were included in the method with a dwell time of 15 msec, and two different time segments were recorded in the chromatographic run (each one of them containing about half of the pesticides studied). Figure 1 shows the chromatogram corresponding to 100 pg on column for all the compounds studied. Extracted ion chromatograms are overlaid for each one of the target analytes according to their respective protonated molecule and product ion MRM transition. Linearity and Limits of Detection Linearity was evaluated by analyzing the standards solutions at five different concentration levels in the range 2 to 100 pg on column. As an
example, the calibration curve generated for atrazine is shown in Figure 2. As it can be observed in this figure, the linearity of the analytical response across the studied range is excellent, with a correlation coefficient of 0.998. Similar results were obtained for the rest of the compounds analyzed. The limits of detection (LOD) were estimated from the injection of standard solutions at concentration levels corresponding to a signal-to-noise ratio of about 3. The results obtained are included in Table 1 as well. The best limits of detection were obtained for the triazines (from 100 fg to 2 pg on column) and the highest limits of detection were for fluoroxypyr and spinosad D (above 100 pg).
x105 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
+ MRM (282.0 212.0) mix100_100 pg_5May.d
12
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Abundance vs. acquisition time (min)
Figure 1.
Product ion chromatograms of a mix of 100 pesticides (concentration: 100 pg on column).
500000 400000 Area 300000 200000 100000 0 0
y = 3828.6x % 1790.3 R2 = 0.9988
20
40
60 Amount injected (pg)
80
100
120
Figure 2.
Calibration curve for atrazine using a linear fit with no weighting and no origin treatment.
Application to Vegetable Matrices To confirm the suitability of the method for analysis of real samples, matrix-matched standards were analyzed in three different matrices + green pepper, tomato, and orange + at two different concentration levels (10 and 100 g/kg). Figure 3 shows the analysis of a green pepper spiked with the pesticide mix at 10 g/kg (10 pg on column). As it can be observed in two of the MS/MS extracted product ion chromatograms, for dimethoate and azoxystrobin, compounds can be easily identified in these complex matrices due to the selectivity of the MRM transitions, thus fulfilling the regulation limits imposed by the EU directives. In general, the LOD obtained meet the requirements regarding the MRLs imposed by the existing European regulations.
Reference
1. Imma Ferrer and E. Michael Thurman, Determination of Fungicides in Fruits and Vegetables by Time-of-Flight and Ion Trap LC/MS (2005) Agilent Technologies, publication 5989-2209EN www.agilent.com/chem.

x104 1 1.8 1.4 1 0.6 0.2 0 x103 1 3 2 1 0 x103 5 4 3 2 1 0 1
+ TIC MRM (** **) mix100_10pg_green pepper.d
12
+ MRM (230.0 199.0) mix100_10pg_green pepper.d
12
(a)
+ MRM (404.0 372.0) mix100_10pg_green pepper.d
12
(b)
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Abundance vs. acquisition time (min)
Figure 3.
MRM chromatogram of a spiked green pepper sample at 10 g/kg. Product ion chromatograms for (a) dimethoate and (b) azoxystrobin.
Determination of 44 Pesticides in Foodstuffs by LC/MS/MS Application
Food Safety
Authors
Masahiko Takino Agilent Technologies, Inc. Hachioji, Tokyo Japan Toshitsugu Tanaka Kobe Institute of Health Department of Food Chemistry Chuo-ku, Kobe Japan
Abstract
A sensitive and selective analytical method for the determination of 44 pesticide residues in several foodstuffs using the Agilent G6410AA Triple Quadrupole Mass Spectrometer (QQQ) was developed. This method use two different sample preparation methods followed by LC/MS/MS (liquid chromatography/tandem mass spectrometry). The limits of detection for all pesticides were less than 10 ng/mL in foodstuff. The sensitivity of QQQ easily met the maximum residue limits (MRLs) of all investigated pesticides in Japan Food Hygiene Law.
safety. In recent years, the established regulations regarding MRLs in commodities have been more and more stringent. In Japan, the positive list system was introduced this year, and MRLs have been set for over 500 pesticides in all foodstuffs. This new system sets different MRLs for each pesticide within each food group. Typically, the MRLs range from 0.01 to 3 g/g depending on the commodities and pesticides. The low MRLs fostered the development of more sensitive analytical methods to meet the requirements of complex samples. In this sense, liquid chromatography/tandem mass spectrometry (LC/MS/MS) with QQQ in multiple reaction monitoring (MRM) mode has become so far the most widely used techniques for the quantitation of polar pesticides in food. MRM mode provides for more specific detection in a complex matrix such as food. In this work, 44 pesticides (Tables 1 and 2) are analyzed in two separate runs with sample analytical conditions. The sensitivity requirements set by the positive lost system for these pesticides are easily met.
Experimental
Chemicals The acetonitrile was of LC/MS grade from Wako Pure Chemical Ind (Japan). Toluene, acetone, nhexane, formic acid, sodium chloride, and anhydrous sodium sulfate were of analytical grade from Wako Pure Chemical. All SPE cartridges were purchased from Spelco Japan (Japan). Pesticide standards were obtained from Hayashi Pure Chemical (Japan).
Introduction
Pesticides are widely used in agricultural practices. The main application can be classified in production and post-harvest treatment of agricultural commodities for transport purposes. In this sense, production agriculture comprises the main category of use of pesticides subject to control requirements and, therefore, maximum residue levels (MRLs) have been fixed to assess food
Sample Preparation
Extraction
LC/MS/MS Instrument The LC/MS/MS system used in this work consists of an Agilent 1100-series vacuum degasser, binary pump, well-plate autosampler, thermostatted column compartment, and the Agilent G6410 Triple Quadrupole Mass Spectrometer with an electrospray ionization source (ESI). The objective of the method development was to obtain a fast and sensitive analysis for quantifying pesticides in fruits and vegetables. For chromatographic resolution and sensitivity, different solvents and columns were optimized. It was found that a simple solvent system using water, acetonitrile, formic acid, formic acetate, and a 1.8-m particle size C18 column would work very well.
LC Conditions Instrument: Column: Column temp: Mobile phase: Agilent 1100 HPLC ZORBAX Extend C18, 100 mm 2.1 mm, 1.8 m (p/n 728700-902) 40 C A = 0.1% formic acid +5 mM ammonium formate in water B= Acetonitrile 10% B at 0 min, 80% B at 30 min 0.2 mL/min 5 L Agilent 6410 QQQ Positive ESI 10 L/min 50 psig 350 C 4000 V m/z 100 to 550 Variable 100 V Shown in Tables 1 and 2 Shown in Tables 1 and 2
Vegetable and fruit samples were obtained from the local markets. A sample of 10 to 500 g was chopped in a food processor to obtain thoroughly mixed homogenates. A 20-g portion of sample homogenate was weighed in a 200-mL PTFE centrifuge tube. Then 50 mL of acetonitrile was added and blended in a Polytoron. The extract was then filtered by applying vacuum. The filtrate was collected and the residue was re-extracted with 20 mL of acetonitrile. The filtrates were combined in a 100-mL volumetric flask and made up to volume with acetonitrile. A 20-mL portion of the extract was transferred into a PTFE centrifuge tube, and 10 g of NaCl and 20 mL of 0.5 M phosphate buffer (pH 7.0) were added to the extract followed by shaking for 5 min. Five grams of anhydrous Na2SO4 were added to the acetonitrile layer obtained after salting out. After removing anhydrous Na2SO4, the extract was evaporated to dryness by rotary evaporator (water bath temperature did not exceed 40 C). The residue was dissolved in 2 mL of acetonitrile-toluene (3:1).
Cleanup
Group 1 - The extract was loaded into a GCB/amino propyl SPE cartridge (500 ng/500 mg) preconditioned with 10 mL of acetonitrile-toluene (3:1). The 20 mL of acetonitrile-toluene (3:1) was further added to the SPE cartridge. All eluate was collected and evaporated by rotary evaporator. The residue was dissolved in 4 mL of methanol. Group 2 - The extract was loaded into a silica gel SPE cartridge (500 mg) preconditioned with 10 mL each of methanol, acetone, and n-hexane (10 mL of methanol, 10 mL of acetone, and 10 mL of n-hexane, total volume is 30 mL). The 10 mL of acetone-triethylamine-n-hexane (20:0.5:80) was further added to the SPE cartridge. All eluate was discarded. The 20 mL of acetone-methanol (1:1) was applied and the eluate was collected and evaporated by rotary evaporator. The residue was dissolved in 4 mL of methanol. Standard Preparation Stock solutions of individual pesticides were prepared in methanol at 1 g/mL. Serial dilutions using methanol produced a range of standard mixture solutions at 0.001 g/mL to 1 g/mL. The blank matrix residues were fortified with a mixture of pesticides studied at 10 ng/g.
2
Gradient: Flow rate: Injection vol: MS Conditions Instrument: Source: Drying gas flow: Nebulizer: Drying gas temp: Vcap: Scan: Fragmentor: MRM ions: Collision energy:
LC/MS/MS Method Quantitative analysis was carried out using MRM mode with time program. The parameters of MRM transition are shown in Tables 1 and 2.
Table 1. No 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Data Acquisition Parameters of MRM Transitions of Each Pesticide in Group 1 Pesticides Thiabendazole Thiamethoxam Clothianidin Chloridazon Imidacloprid Dimethirimol Oxycarboxine Thiacloprid Azamethiophos Ferimzone(E) Ferimzone(Z) Phenmedipham Azinphos-methyl Simeconazole Isoxaflutol Pyriftalid Tridemorph Methoxyfenozide Chromafenozide Fenoxycarb Naproanilide Butafenacil Cyazofamide Anilofos Pyrazolate Benzofenap Cyflufenamid Indoxacarb Clomeprop Cloquincet-mexyl Furathiocarb Lactofen Tralkoxydim RT (min) 5.018 6.16 7.83 8.19 8.39 8.8 11.02 11.03 12.87 13.21 13.7 17.77 17.9 18.5 18.7 18.7 19.21 20.06 20.57 20.63 21.27 21.55 21.7 22.5 23.5 24 24.3 24.37 24.78 24.8 25.7 26.3 26.7 Molecular weight 201 291 249 221 255 209 267 252 324 254 254 317 318 293 359 318 297 312 394 301 291 491 324 367 438 430 412 527 372 335 365 478 329 Precursor ion (m/z) 202 292 250 222 256 210 268 253 325 255 255 318 132 294 360 319 298 313 175 302 292 492 325 368 439 431 413 528 373 336 383 479 330 Product ion (m/z) 175 211 169 104 209 171 175 126 183 124 132 136 77 70 251 139 130 149 141 88 171 331 108 199 173 105 241 150 299 238 195 344 284 Collision energy(V) 20 5 10 10 20 20 10 20 10 20 20 20 15 15 15 20 15 20 20 15 10 20 10 10 15 20 20 15 5 15 15 15 10
Table 2. No 1 2 3 4 5 6 7 8 9 10 11
Data Acquisition Parameter of MRM Transitions of Each Pesticide in Group 2 Pesticides Flumetsulam Thidiazuron Imazaquin Thifensulfuron-methyl Florasulam Forchlorfenuron-methyl Clorasulam-methyl Diclosulam Fomesafen Triflusulfuron-methyl Haloxyfop RT (min) 9.96 11.95 12.25 12.89 13.75 14.63 16.41 16.83 18.27 19.29 19.67 Molecular weight 325 220 311 387 359 247 429 405 438 492 361 Precursor ion (m/z) 326 221 312 388 360 248 430 406 456 493 362 Product ion (m/z) 129 102 267 167 129 129 398 161 344 264 316 Collision energy (V) 20 10 20 10 20 10 10 20 10 15 15

Optimization of MRM Transitions Determination of the optimal MRM transitions for each pesticide was carried out using full scan mode followed by product ion scan mode using two pesticide standard mixtures at 1 g/mL. TICs of these standard mixtures in full scan mode and product ion scan mode are shown in Figures 1 and 2. The mass spectrum of each pesticide by full scan mode exhibited protonated molecular ions; [M+H]+ as the base peak ion except azinphos-methyl, furathiocarb, and fomesafen, which exhibited fragment ion and ammonium adduct ion [M+NH4]+. These ions were selected as precursor ions for MRM mode. It was possible to generate individual product ion MS/MS spectrum of each pesticide by using multiple acquisition and time programming mode. As shown in Tables 1 and 2, 10 time segments for 33 pesticides in group 1 and 7 time segments for 11 pesticides in group 2 were used for MRM mode.
Total ion chromatograms of pesticide standard mixture corresponding to the minimum MRL value for pesticides (10 ng/mL) are shown in Figure 3. These show excellent signal-to-noise (S/N) ratios for all pesticides. The limit of detection (LOD) for each pesticide was determined using an S/N ratio of 3 with an MRM chromatogram of each pesticide at 1 ng/mL (see Table 3). To evaluate the linearity of the calibration curves, various concentrations of pesticide standard solutions ranging from 0.001 ng/mL to 1 ng/mL were analyzed. As shown in Table 3, the linearity was very good for all pesticides with correlation coefficients (r2) greater than 0.998 The matrix effect of this method was investigated by using orange, apple, potato, and cabbage extracts spiked with pesticide standards at 10 ng/mL. Typical MRM chromatograms of orange extract are shown in Figures 4 and 5. The other chromatograms of apple, potato, and cabbage extract are shown in Figure 6. There was not additional peak from sample matrix in all food when compared with the pesticide standard mixture. These results indicate that MRM mode has very high selectivity.
x107 1.2
(A)
6 18 19, 20 24 4 2 3 11 9 15,16 13 21 17 22 23
29,30
7,8 0.8 1
10
12
31 27, 28 26 25 32 33
0.4
14
1 x106 6.5
11
13
15
17
19
21
23
25
27
29
31
Abundance versus acquisition time (min) 6
(B)
9 10 12 19, 20 18 21 22 23 21 23 33 25 27, 28 26 25 32 27 29 31 29, 30 24 31 7, 8
4.5
2.5 1 0.5 1 3 5
5 34
11
13
14
15, 16 17
11
13
15
17
19
Abundance versus acquisition time (min)
Figure 1.
TIC of 33 pesticides standard in full scan mode (A) and product ion scan mode (B) at 1 g/mL.
x107 4.0
(A)
6 3.0 2.0 1.0 1 2 4 5 7 8 9
10
11
10
11
12 3
13
14
15
16
17
18
19
20
21
22
23
Abundance versus acquisition time (min) x106 3.0
(B)
4 5 6 10
2.0
1.0
7 8 9
11
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure 2.
TIC of 11 pesticides standard in full scan mode (A) and product ion scan mode (B) at 1 g/mL.
Furthermore, the change on the peak intensity of each pesticide by sample matrix was calculated by comparing with the peak intensity of pesticide standards. As these results show in Table 4, the relative intensity of each pesticide ranged from 91 to 116%. Thus, matrix effect such as ion suppression may be insignificant and it was possible to use external standards instead of matrix matched standards. The repeatability of each pesticide in orange extract is also shown in Table 4, and the RSD of each pesticide was in the range from 1.7 to 5.9%.
7,8 5000 4000 3000 2000 1 1000 2 4 3 5 7 6 10 11 9 11 13 15 17 14 13 15,16 17 19 21 22 23 24 26
(A)
12 9 18 19,20,21 25 29,30
28 27 25
31 32 33 27 29
23
Abundance versus acquisition time (min) 6 16000 12000 1 8000 4000 23 7 9 1 3 5 7 9 11 13 15 17 19 11 21 23 25 27 29 8
(B)
5 4 10
Figure 3.
TIC of 33 pesticide standards (A) and 11 pesticides standard (B) at 10 ng/mL in MRM mode.
Table 3.
Linearity and LOD of 44 Pesticide Standard Solutions r2 0.9999 0.9992 0.9999 0.9993 0.9995 0.9989 0.9993 0.9991 0.9988 0.9993 0.9995 0.9993 0.9997 0.9992 0.9991 0.9988 0.9991 0.9996 0.9994 0.9992 0.9989 0.9969 0.9977 LOD (ng/mL) <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 0.34 0.53 <0.1 <0.1 <0.1 <0.1 <0.1 1.21 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 No 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Pesticides Methoxyfenozide Chromafenozide Fenoxycarb Naproanilide Butafenacil Cyazofamide Anilofos Pyrazolate Benzofenap Cyflufenamid Clomeprop Indoxacarb Quinclorac-methyl Furathiocarb Lactofen Tralkoxydim r2 0.9993 0.9992 0.9988 0.9993 0.9994 0.9987 0.9991 0.9990 0.9982 0.9993 0.9993 0.9991 0.9988 0.9987 0.9987 0.9992 LOD (ng/mL) 0.55 0.49 <0.1 <0.1 <0.1 0.43 <0.1 0.51 0.49 0.43 0.61 1.04 0.63 <0.1 1.10 0.52
No Pesticides Group 1 1 Thiabendazole 2 Thiamethoxam 3 Clothianidin 4 Chloridazon 5 Imidacloprid 6 Dimethirimol 7 Oxycarboxine 8 Thiacloprid 9 Azamethiophos 10 Ferimzone(E) 11 Ferimzone(Z) 12 Phenmedipham 13 Azinphos-methyl 14 Simeconazole 15 Isoxaflutol 16 Pyriftalid 17 Tridemorph Group 2 1 Flumetsulam 2 Thidiazuron 3 Imazaquin 4 Thifensulfuron-methyl 5 Florasulam 6 Forchlorfenuron-methtyl
7 8 9 10 11
Clorasulam-methyl Diclosulam Fomesafen Triflusulfuron-methyl Haloxyfop
0.9987 0.9989 0.9989 0.9992 0.9995
<0.1 <0.1 0.32 <0.1 0.19
25000 15000
1
m/z = 202>>>175
700 300
13
m/z = 132>>>77
550 350
2
m/z = 292>>>211
900 500
14
m/z = 294>>>70
300 200
3
m/z = 250>>>169
2400 1200
15
m/z = 360>>>251
10 11 12 13 14 15 16 17 18 19 20
700 300
4
m/z = 222>>>104
250 150 50
16
m/z = 319>>>139
300 100
24
5 m/z = 256>>>209
17
m/z = 298>>>130
12
800 400
2500
18
m/z = 313>>>149
m/z = 210>>>171
4 5 6 7 8 9 10 11 12 13
1500
7
2500 1500
900
19
m/z = 175>>>141
m/z = 268>>>175
500
4000 2000
8
m/z = 253>>>126
900 600
20
m/z = 302>>>88
3000
9
m/z = 325>>>183
7 8 9 10 11 12 13 14 15 16
1100 500
21
m/z = 292>>>171
1000
300 200 100 400 200
10
m/z = 255>>>124
500 300
22
m/z = 492>>>331
11
m/z = 255>>>132
300 200
23
m/z = 325>>>108
3000 1000
12
m/z = 318>>>136
10 11 12 13 14 15 16 17 Retention time (min) 18 19 20
2200 1000
24
m/z = 368>>199
15 16 17 18 19 20 21 Retention time (min) 22 23 24
Figure 4.
MRM of 33 pesticides in orange extract spiked at 10 ng/mL. (Continued)
400 200
25
m/z = 439>>>173
1000 400
31
m/z = 383>>>195
200 80
26
m/z = 431>>>105
120 60
32
m/z = 479>>>344
180 100
27
m/z = 413>>>241
15 16 17 18 19 20 21 22 23 24
300 100
33
m/z = 330>>>284
21 22 23 24 25 26 Retention time (min) 27 28 29
800 400
28
m/z = 373>>>299
45 25
29
m/z = 528>>>150
3500 1500
30
m/z = 336>>>238
21 22 23 24 25 26 Retention time (min) 27 28 29
Figure 4.
MRM of 33 pesticides in orange extract spiked at 10 ng/mL.
6000 2000
1
m/z = 326>>>129
6000 4000
7
m/z = 430>>>398
7500 2500
2
m/z = 221>>>102
8000 4000
8
m/z = 406>>>161
3
7500 2500
800
9
m/z = 456>>>344
m/z = 312>>>267
400
20000 10000
4
m/z = 388>>>167
20000 10000
10
m/z = 493>>>264
11
7500 2500
5
m/z = 360>>>129
2000 1000
m/z = 362>>>316
16 17 18 19 Retention time (min) 20
6
10000 5000 1
m/z = 248>>>129
2 3 4 5 6 7 8 9 10 11 12 13 14 15
Retention time (min)
Figure 5. 8
MRM of 11 pesticides in orange extract spiked at 10 ng/mL.
7, 8 5000
12 9 13 15,16 10 11 14 17 22 24 29, 30 18 19, 20, 21 23 25 28 26 27 31 32 33
(A)
5 1 2 6 34
3000
1000
5000
(B)
3000
1000
5000
(C)
3000
1000 1 3 5 7 9 11 13 15 17 Retention time (min) 6 19 21 23 25 27 29
15000 10000 5000
(D)
1
10
8 2 3 7 9 11
15000 10000 5000
(E)
15000
(F)
10000
5000
9 11 13 Retention time (min)
15
17
19
Figure 6.
TIC of spiked at 10 ng/mL.
Table 4.
Relative Intensity of Each Pesticide in Sample Extracts Relative intensity(%) Orange* Cabbage 105 (3.2) 101 103 (2.1) 98 106 (2.9) 101 105 (3.3) 102 (1.7) 103 (4.6) 106 (3.7) 104 (3.1) 93 (4.6) 116 (4.1) 96 (5.3) 90 (2.1) 104 (4.4) 102 (2.7) 97 (4.1) 92 (3.1) 96 (2.8) 97 (3.4) 99 (2.1) 91 (4.3) 102 (2.6) 93 (3.5) 102 (2.7) 103 (4.7) 108 (5.2) 108 (3.4) 109 (2.6) 105 (4.2) 105 (4.1) 102 (1.8) 100 (3.7) 101 (3.3) 97 (2.6) 104 (4.8) 105 (3.1) 106 (2.9) 99 (3.1) 101 (4.4) 94 (3.9) 95 (3.3) 99 (5.9) 97 (4.1) 108 (4.8) 106 97 107 102 104 90 109 99 103 102 104 103 99 102 96 105 97 114 92 105 101 111 110 105 107 104 104 109 111 110 101 100 112 106 103 104 102 101 111 114 Apple 116 104 109 101 102 103 104 106 94 102 100 104 106 108 104 104 103 100 102 98 104 87 103 103 98 105 100 106 104 105 105 102 156 102 100 116 103 100 97 96 95 104 110 Potato 107 105 112 109 104 108 106 108 84 112 104 110 110 103 93 97 101 111 101 103 114 95 107 97 108 101 111 104 105 101 112 117 104 113 101 113 109 108 142 107 109 108 124
No Pesticides Group 1 1 Thiabendazole 2 Thiamethoxam 3 Clothianidin 4 5 6 7 8 9 Chloridazon Imidacloprid Dimethirimol Oxycarboxine Thiacloprid Azamethiophos
10, 11 Ferimzone(E,Z) 12 Phenmedipham 13 Azinphos-methyl 14 15 16 17 18 19 20 21 22 23 24 25 26 27 29 28 30 31 32 33 Simeconazole Isoxaflutol Pyriftalid Methoxyfenozide Chromafenozide Tridemorph Fenoxycarb Naproanilide Butafenacil Cyazofamide Anilofos Pyrazolate Benzofenap Cyflufenamid Indoxacarb Clomeprop Cuinclorac-methyl Furathiocarb Lactofen Tralkoxydim
Group 2 1 Flumetsulam 2 Thidiazuron 3 Imazaquin 4 5 6 7 8 9 10 11 Thifensulfuron-methyl Florasulam Forchlorfenuron-methyl Clorasulam-methyl Diclosulam Fomesafen Triflusulfuron-methyl Haloxyfop
*( ): RSD,% calculated based on five replicates within one day
10
Conclusions
The multiresidue method by LC/MS/MS described here was suitable for the determination of 44 pesticides in a variety of food samples due to its high sensitivity and high selectivity. Another advantage of this method is that ion suppression was not observed for all food samples studied. Thus, it may eliminate the need for matrix-matched standards, which make analysis more tedious for samples from different origins. For more details concerning this application, please contact masahiko_takino@agilent.com

11
Determination of Chlorinated Acid Herbicides in Soil by LC/MS/MS Application Note
Environmental
Author
Chin-Kai Meng Agilent Technologies 2850 Centerville Road Wilmington, DE 19808-1601 USA
This application note is based on standards and sample preparation procedures from the Montana Department of Agriculture in Bozeman, Montana.
Experimental
Standard and Sample Preparation A stock solution of each analyte at 200 ppm is prepared in methanol. Intermediate mixed solutions for fortifying soil samples and making calibration standards are made by accurately mixing aliquots of each standard stock solution. The concentrations of each analyte in the intermediate solution used in this study are listed in Table 1.
Table 1. Acid Herbicide Mixed Intermediate Standards in Methanol 5930 (pg/L) 1800 8200 1740 5480 1240 1410 2710 6900 3,6-dichloro-2-pyridinecarboxylic acid 4-amino-3,5,6-trichloropicolinic acid 3,6-dichloro-2-methoxybenzoic acid 2,4-dichlorophenoxyacetic acid 2-methyl-4-chlorophenoxyacetic acid [(3,5,6-trichloro-2-pyridinyl)oxy] acetic acid 2,4-dichlorophenoxypropionic acid or dichloroprop 2-(2-methyl-4-chlorophenoxy) propionic acid 2,4-dichlorophenoxybutyric acid
Abstract
Chlorinated acid herbicides were analyzed at the picogram level on column without any derivatization using liquid chromatography and tandem mass spectrometry (LC/MS/MS). Good linearity was observed for all the selected analytes, from low pg to low ng levels on column.
Introduction
Chlorinated acid herbicides are a popular class of broad-leaf weed killer in lawn and grain crops. Due to their widespread use, environmental contamination in water and soil from run-off is a serious concern. Traditional analytical methods based on gas chromatography (GC) and/or mass spectrometry (MS) require derivatization of the analytes. Combining LC and electrospray ionization (ESI) in negative ion mode, these herbicides can be analyzed without derivatization. The multiple reaction monitoring (MRM) mode in MS/MS operation provides low pg detection limits.
Clopyralid Picloram Dicamba 2,4-D MCPA Triclopyr 2,4-DP MCPP 2,4-DB
Sample extraction and cleanup procedures are shown below. Sample Extraction Procedure 1. Weigh 20 0.1 g of soil. 2. Add 50 mL of 0.5N KOH in 10% KCl extracting solution to each sample. Mix thoroughly by shaking. 3. Place samples in boiling water bath for 15 minutes. 4. Place samples on horizontal shaker for 15 minutes. 5. Centrifuge samples at 1200 to 1500 rpm for 15 minutes. 6. Transfer a 3.0-mL aliquot into a 13-mL conical centrifuge tube and add 150 L of 12 N sulfuric aicd. 7. Vortex and confirm the pH is <1.5. If not, add additional acid solution. Sample Cleanup Procedure 1. Add 2 mL of chloroform to the acidified extract. 2. Vortex 30 seconds and centrifuge at 3000 rpm for 2 minutes. 3. Remove the lower chloroform layer into another centrifuge tube. Repeat these three steps two more times. 4. Evaporate the chloroform extract to just dryness. 5. Immediately add 4.0 mL HPLC-grade water, vortex briefly, sonicate 5 minutes, and briefly vortex again. Fill autosampler vial.
See Reference 1 for more detailed information on sample preparation.
Instrumentation
LC: Column: Column temperature: Mobile phases: 1200 LC ZORBAX Extend-C18, RRHT, 2.1 mm 100 mm, 1.8 m 60 C A: 0.04% Glacial acetic acid in water B: Acetonitrile (ACN) Flow rate: Injection volume: Gradient: 0.3 mL/min 1.0 L Time, min. 0 1 2 3 4 8 9 G6410A QQQ ESI () 120 to 400 amu 3500 V 40 psi 9 L/min 200 C 35 V %B 0 40 52 60 100 100 0
MS: Ionization: Mass range: Capillary: Nebulizer pressure: Drying gas flow: Gas temperature: Skimmer:
MRM parameters are listed in Table 2.
Table 2. Name Clopyralid Picloram Dicamba 2,4-D MCPA Triclopyr 2,4-DP MCPP 2,4-DB
Method Parameters for MRM RT 3.47 3.69 4.31 5.02 5.09 5.26 5.42 5.46 5.66 MW 191 240 220 220 200 255 234 214 248 Quant 190 > 146 239 > 195 219 > 175 219 > 161 199 > 141 254 > 196 233 > 161 213 > 141 247 > 161 Qual 192 > 148 241 > 197 219 > 145 221 > 163 201 > 143 256 > 198 235 > 163 215 > 143 249 > 163 Frag V 80 80 60 80 100 80 80 100 80 Col cell 5 5 0 15 10 10 5 10 10 Dwell 70 70 150 25 25 25 25 25 25 Segment 1 1 2 3 3 3 3 3 3
Due to the concentration differences for these analytes in the intermediate solution (Table 1), Dicambas concentration was used as the concentration level shown in Table 3.
Table 3. Concentration Levels (8000 to 10 pg/L) Used in This Study 800 593 180 820 174 548 124 141 271 690 400 90 410 87 274 62 70.5 135.5 345 200 45 205 43.5 137 31 35.2 67.7 172.5 80 59.3 18 82 17.4 54.8 12.4 14.1 27.1 69 40 9.0 8.7 6.2 7.1 13.6 20 4.5 4.4 3.1 3.5 6.8 10 7.4 2.3 2.2 6.9 1.6 1.8 3.4 8.6
For example, concentration level 8000 is the intermediate solution, and concentration level 20 is a 1:400 dilution of the intermediate solution.

Figure 1 shows the overlaid chromatograms of all nine herbicides from the MRM analysis. The run time was less than 6 minutes. Using a 1.8-m particle size column, the peak widths of these analytes are in the range of 0.1 to 0.2 min. The narrower peak width helps to achieve a higher signal-tonoise ratio. The MRM results of all nine herbicides at 1.6 to 10.3 pg on column are shown in Figure 2. As shown in Table 3, the linearity range, incorporating 10, 20, 40, 80, 200, 400, 800, and 8000 pg on column, mimics the concentration of Dicamba.
Solution concentration level 8000 Clopyralid Picloram Dicamba 2,4-D MCPA Triclopyr 2,4-DP MCPP 2,4-DB 5930 1800 8200 1740 5480 1240 1410 2710 6900
296.5 148.2
29.7 14.8
41.0 20.5 10.3 27.4 13.7
34.5 17.3
106 5
-EIC (161.0 m/z) MRM (247.0 m/z); from mont_neg_8000.d
MCPA
MCPP
4
Dicamba
2
Clopyralid Picloram
2, 4-D
2, 4-DP Triclopyr
2, 4-DB
0 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9
Figure 1.
Overlaid MRM results from the nine selected herbicides.
102 3 1
-EIC (146.0 m/z) MRM (190.0 m/z); from mont_neg_10.d Smooth (1) 3.522 Clopyralid, 7.4 pg -EIC (195.0 m/z) MRM (239.0 m/z); from mont_neg_10.d Smooth (1) 3.722 Picloram, 2.3 pg
102 1.5 0.5 103 1 0 102 4 0 103 3 1
-EIC (175.0 m/z) MRM (219.0 m/z); from mont_neg_10.d Smooth (1) 4.379
Dicamba, 10.3 pg
-EIC (161.0 m/z) MRM (219.0 m/z); from mont_neg_10.d Smooth (1)
2,4 -D, 2.2 pg
5.098
-EIC (141.0 m/z) MRM (199.0 m/z); from mont_neg_10.d Smooth (1)
MCPA, 6.9 pg
5.130
102 -EIC (196.0 m/z) MRM (254.0 m/z); from mont_neg_10.d Smooth (1) 1.5 Triclopyr, 1.6 pg 0.5 102 6 -EIC (161.0 m/z) MRM (233.0 m/z); from mont_neg_10.d Smooth (1) 2 103 -EIC (141.0 m/z) MRM (213.0 m/z); from mont_neg_10.d Smooth (1) 2 1 103 -EIC (161.0 m/z) MRM (247.0 m/z); from mont_neg_10.d Smooth (1) 1 0.5 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5
5.327
2,4 -DP, 1.8 pg
5.481
MCPP, 3.4 pg
5.497
2,4 -DB, 8.6 pg
5.670
6
5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9
Figure 2.
MRM results. Table 4. Linearity (10 to 8000 pg/L) and Repeatability Compound Clopyralid Picloram Dicamba 2,4-D MCPA Triclopyr 2,4-DP MCPP 2,4-DB R2 (linear fit, include origin) 0.9995 0.9991 0.9999 0.9998 0.9999 0.9993 0.9998 0.9999 0.9973 Repeatability, %RSD (n=7) 3.5 13.1 2.4 6.5 2.2 11.9 4.5 5.0 8.4 Amount on column 29.7 pg 9.0 41.0 8.7 27.4 6.2 7.1 13.6 34.5
Therefore, the corresponding on column amounts for Triclopyr are: 1.6, 3.1, 6.2, 12.4, 31, 62, 124, and 1240 pg. The calibration model used was a linear model that included origin with no weighting. All analytes showed excellent linearity. In the repeatability study, seven 1-L injections of the level 40 solution (Table 4) were analyzed to calculate the RSD. At this low concentration, all RSDs were <15%, with the majority in the single digits. The matrix effect from three different matrices was evaluated. Water and soil extracts were spiked with the herbicide standards (50 L of the level 400 standard were added to 950 L of water, silt, clay, or sandy extracts). The resulting concentration of the analytes is equivalent to the level 20 solution (Table 3). Table 5 shows the RSD of eight 1-L injections of the level 20 solutions, that is, <20 pg of each ana4
lyte on column, in three different matrices. As expected, analytes with lower absolute responses showed higher RSD value.
Table 5. RSDs from Eight Injections of <20 pg of Each Analyte in Three Matrices Clay Triclopyr MCPP MCPA Clopyralid 2,4-DP 2,4-DB 2,4-D 27 6 2 14 9 9 11 Sandy 22 5 6 15 8 3 14 Silt 22 7 5 13 13 9 13 On column 3.1 pg 6.8 13.7 14.8 3.5 17.3 4.4
The RSDs of <20 pg analytes in the selected matrices were comparable to that in water (RSDs ~5%), except for 2,4-D, due to the higher response from the silt.
Acknowledgement
The author would like to thank Ms. Heidi Hickes and Ms. Angela Schaner from the Montana Department of Agriculture for valuable discussions, sample preparation procedures, and the samples used in this study.
Reference
1. Determination of Chlorinated Acid Herbicides in Soils by Liquid ChromatographyElectrospray/Mass Spectrometry/Mass Spectrometry, by Angela Schaner and Laura Luckey, Revision 2, April 2, 2004. Montana Department of Agriculture, Laboratory Bureau, McCall Hall, Montana State University, Bozeman, MT 59717.
The repeatabilities of responses in three matrices for all analytes were <15% except for Triclopyr, which was >20%. The responses of ~20 pg analytes among water and matrices are compared in Figure 3. The listed response for each matrix is the average of responses from eight injections. The RSDs for water and the three matrices are shown in Figure 3 and are comparable. In general, the variation of the responses among water and different matrices is less than 5% for all analytes except for 2,4-D, which has an RSD close to 10% due to the higher responses from the silt matrix. This shows that the method described in this application note is free from matrix interferences from clay, sand, and silt.

For more information on our products and services, visit our website at www.agilent.com/chem
Conclusion
Using LC/MS/MS, chlorinated acid herbicides were analyzed at pg levels on column without any derivatization. At <40 pg on column, the repeatability of seven 1-L injections showed RSD <15%. Good linearity was observed for all analytes from low pg to low ng on column.
25000 20000 15000 10000 5000
3.0 Level 20 in water 3.8 2.7 4.4 3.8

MCPP MCPA Clopyralid
Level 20 in clay Level 20 in sandy 5.1

2,4-DP 2,4-DB
9.3
2,4-D
Level 20 in silt
0 Triclopyr
Figure 3.
Analyte RSDs from water and soil extracts are comparable. 5
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2006 Printed in the USA July 25, 2006 5989-5246EN
Analysis of Post-Harvest Fungicides and Their Metabolites in Citrus Fruits and Juices by Time-of-Flight and Ion Trap LC/MS Application
Food Safety
Authors
E. Michael Thurman and Imma Ferrer Pesticide Residue Research Group University of Almera, 04120 Almera, Spain Michael Woodman and Jerry Zweigenbaum* Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
*Agilent contact
Introduction
Thiabendazole and imazalil are widely used for post-harvest treatment of citrus fruits (Figure 1) in order to preserve the fruit during the transport process, which may take from several days to several weeks. The fungicide treatment is of the fruit skin only. The maximum residue limits (MRLs) for imazalil and thiabendazole in Europe and the United States require that these compounds be monitored before consumption of the fruit [1, 2]. Since these compounds are used ubiquitously, they are frequently detected in imported citrus fruits [3]. Thus, many commercial samples must be inspected to ensure food safety. Additionally, imazalil is easily metabolized to 1-(2,4dichlorophenyl)-2-(1H-imidazole-1-yl)-1-ethanol, which is sometimes detected in citrus fruits [4]. In the United States, the sum of imazalil and its metabolite is regulated, so a method for the parent compound should also include its metabolite [5]. While papers have reported analytical LC/MS procedures for imazalil and thiabendazole, less is known about the metabolite of imazalil, and there are, as yet, no published reports using LC/MS accurate-mass analysis [6]. This application note describes a rapid and simultaneous method using electrospray liquid chromatography/time-of-flightmass spectrometry and liquid chromatography/ion trap mass spectrometry (LC/TOFMS and LC/ITMS) for the analysis of these two important postharvest fungicides in citrus. An earlier application note reported on the analysis of carbendazim and thiabendazole in fruits and vegetables and may be of interest as well [7].
Abstract
This application note applies both time-of-flight and ion trap liquid chromatography/mass spectrometry (LC/TOFMS and LC/ITMS) to determine two important post-harvest fungicides (imazalil and thiabendazole) and a metabolite of imazalil in oranges and in their juice. Included is both a detailed rapid procedure for sample preparation of citrus fruits and a solid-phase extraction for the isolation of these fungicides and metabolite from juice. Identification of these fungicides and metabolite is carried out using both accurate mass for elemental composition and fragmentation pathways with LC/ITMS to help elucidate pathways of degradation. The results of the analysis of actual samples from the marketplace for fruit and juice are included.
1 ppm
0 ppm
5. Shake the sample vigorously for 1 minute by using a Vortex mixer at maximum speed or hand shaking. 6. Centrifuge for 3 minutes at 3700 rpm.
Treated
Untreated
7. Add 5 mL of supernate into a 15-mL tube. 8. Add 250 mg of PSA adsorbent and 750 mg of MgSO4. 9. Vortex and shake for 20 s.
Cl O H2C N NH
10. Centrifuge again for 3 minutes at 3700 rpm.

HO Cl
11. Transfer 1.0 mL into a LC/MS vial. 12. Evaporate the 1.0 mL supernate to dryness and bring back up in 8%/92% methanol/water for LC/TOFMS analysis and ion trap analysis. LC/TOFMS analysis of fruit and vegetable extracts is performed by injecting 50 L. Nonfortified samples are analyzed directly at this same point by either LC/TOFMS or LC/ITMS. Fruit-Juice Extraction
Cl
m/z 297
NH
Cl
m/z 257
Imazalil
H2 N N N
Imazalil metabolite
m/z 202
Thiabendazole Figure 1. Structure of imazalil, thiabendazole, and an imazalil metabolite, which are post-harvest fungicides used on citrus (oranges and lemons), showing the [M+H]+ ion. Top panel compares treated and untreated oranges.
Solid-phase extraction (SPE) is used for the extraction of post-harvest fungicides from fruit juices and sodas based on juices. The cartridge of choice is the Accubond C-18 SPE cartridge (part number 188-1356). 1. Prepare the 5-mL SPE cartridge by taking 5 mL of methanol to wet the cartridge followed by 5 mL of deionized water to remove the methanol for good recovery of the fungicides. 2. Allow juice to stand, equilibrate to room temperature and degas prior to SPE, especially if bubbles are present. 3. Centrifuge and filter (0.2-micron TFPE filter to remove remaining solids) 5 mL of juice prior to dilution with 5 mL of deionized water. 4. Pass the juice through the cartridge and rinse with 2 mL of deionized water. 5. Elute with 5 mL of methanol. 6. Evaporate the methanol to 200 L under nitrogen. 7. Dilute 1:3 with deionized water before injection into the LC/TOFMS or LC/ITMS.
Experimental Methods
Fruit Extraction: QuEChERS QuEChERS (quick, easy, cheap, effective, rugged, and safe) is a method that is receiving wide acceptance for rapid extraction of pesticides in food [8]. 1. Weigh 15 g of a previously homogenized sample into a 40-mL Teflon centrifuge tube. 2. Add 15 mL of acetonitrile (containing 1% acetic acid). 3. Add 6 g of anhydrous MgSO4. 4. Add 2.5 g of NaAc3H2O (sodium acetate trihydrate).
LC/TOFMS Method
LC Pumps Column Mobile Phases Gradient Instrument Nebulizer Drying gas Voltages Reference masses Resolution Mass range Flow rate Agilent 1100 binary pumps, injection volume 50 L with standard Agilent 1100 ALS ZORBAX Eclipse XDB-C8, 4.6 mm 150 mm, 5 m, product number 993967-906 A = acetonitrile and B = 0.1% formic acid in water 5 minutes isocratic at 10% A, followed by a linear gradient to 100% A in 30 minutes at a flow rate of 0.6 mL/min Agilent LC/MSD TOF time-of-flight mass spectrometer with dual electrospray source, positive ESI, capillary 4000 V 40 psig 9 L/min, 300 C Fragmentor 190 V, skimmer 60 V, Oct DC1 37.5 V, OCT RF V 250 V m/z 121.0509 and 922.0098 9500500 at m/z 922.0098 501000 m/z Reference A Sprayer 2 is at a constant flow rate during the run
LC/ITMS Method
Chromatography LC Pumps Column Mobile phases Gradient Instrument Detection Identical to LC/MSD TOF methods for direct comparison of peaks Agilent 1100, injection volume 50 L ZORBAX Eclipse XDB-C8, 4.6 mm 150 mm, 5 m, product number 993967-906 A = acetonitrile and B = 0.1% formic acid in water 5 minutes isocratic at 10% A, followed by a linear gradient to 100% A in 30 minutes at a flow rate of 0.6 mL/min Agilent 1100 Series LC/MSD Trap. Positive ESI, Capillary 3200 V, Nebulizer 40 psig, drying gas 9 L/min, gas temp 300 C. Capillary exit 70 V, Skimmer 60 V Trap accumulation, 50,000 counts with maximum accumulation time of 200 ms

Imazalil in Oranges Figure 1 shows two oranges, one with a treatment of the post-harvest fungicides (imazalil and thiabendazole) and the other orange without treatment. Note the intense growth of mold when no treatment of post-harvest fungicides is used. The purpose of the post-harvest fungicides is to prevent serious mold formation during transport and sale of citrus fruits. This treatment is directed at the orange peel itself, rather than the entire fruit; therefore, it poses a lesser risk to the consumer. Figure 2 shows the LC separation and chromatogram of the orange-peel extract that contains two large peaks, which were suspected to be the post-harvest fungicides, imazalil and thiabendazole. They were identified using the following
technique. The accurate mass spectrum of each peak was taken and A+2 isotopes were examined (Figures 2 and 3). For example, the peak at 17.9 minutes was suspected to be a chlorinecontaining species based on the mass spectrum (Figure 3). In fact, the presence and number of chlorine atoms in the compound can be easily attained taking into account both the relative intensity of the 37Cl/35Cl signals and the accurate mass differences between the two masses. As can be seen in Figure 3, the accurate mass of the m/z 297 peak was 297.0556 with a 37Cl isotope signal of 299.0527, with a relative intensity of about two-thirds of the main peak. The mass difference between both signals is 1.9971, which is very near to the exact mass difference between 35Cl (34.9689) and 37Cl (36.9659) (1.9970). This evidence combined with peak area shows that the unknown compound unequivocally contains chlorine atoms.
3
3.9e7 3.8e7 3.6e7 3.4e7 3.2e7 3.0e7 2.8e7 2.6e7 2.4e7 Intensity, cps 2.2e7 2.0e7 1.8e7 1.6e7 1.4e7 1.2e7 1.0e7 8.0e6 6.0e6 4.0e6 2.0e6 0.0 2 4 6 8 10 12 14 2.1 2.4 2.7 7.6 16.7
17.9
Imazalil
Thiabendazole Imazalil degradate

14.9 18.2 21.0 20.4 21.5 24.8 26.0 26.4 22.7 23.9 25.4
16 18 Time, min
20
22
24
26
28
30
Figure 2.
LC/MSD TOF total ion chromatogram (TIC) of orange peel, note the two large peaks in the chromatogram are imazalil and thiabendazole.
Furthermore, the relative abundance of the isotopic signal indicates that the chlorine isotope is present with two atoms (the natural distribution of 35 Cl: 37Cl is nearly 3:1 (75.77% 35Cl; 24.23 % 37Cl), and two chlorine atoms give the characteristic pattern shown in Figure 3. This fact makes much easier the assignment of an elemental composition for the suspected species. Using the calculator tool of the TOF software, we set, as a stringent criterion, the number of chlorine atoms to two. Using an accuracy error threshold of 3 ppm (the maximum error of the LC/MSD TOF), only one formula
was found (C14H15N2OCl2, 0.3 ppm error). This formula was searched against the The Merck Index database. The use of the Merck Index is explained in an application note [9]. It has to be taken into consideration that the additional hydrogen atom present in the measured ions due to the positive ionization mode must be subtracted from the elemental compositions provided by the calculator tool before entering the formula into the database. The Merck Index gave a unique match with the formula for imazalil.
297.0556
A (Cl35)
Imazalil
A+2 (Cl35 and Cl37)

299.0527
Intensity
Accurate mass Accurate mass spectrum spectrum
299.0501
A+4 (Cl37 and Cl37)

301.0498
288
290
292
294
296
298
300 302 m/z, amu
304
306
308
310
312
314
Figure 3.
Imazalil detection in orange peel measured accurate mass, m/z = 297.0556 theoretical mass = 297.0556, accuracy 0.3 ppm, formula is C14H15N2OCl2.
Likewise the mass spectrum was taken for the peak at 7.6 minutes in Figure 2 and the spectrum is shown in Figure 4, m/z 202.0434. The relative abundance of the isotopic signal indicates that sulfur may be present with one atom. The relative intensity of the A+2 peak is 4.2% with a mass depletion of 1.996 daltons when going from 32S (31.9721) to 34S (33.9679). The mass spectrum shown in Figure 4 has a mass difference of 1.997 daltons and a peak height of 4.7%, which is
quite close to the expected values. Thus, these data indicate the presence of 1 sulfur atom. Entering a mass of m/z 202.0434 gives the unique formula of C10H8N3S, which is the formula for a protonated ion of thiabendazole, based on the Merck Index search. The next step is to examine the fragmentation of the molecules using the LC/MS ion trap for further structural confirmation.
1.9e5 1.8e5 1.7e5 1.6e5 1.5e5 1.4e5 Intensity, counts 1.3e5 1.2e5 1.1e5 1.0e5 9.0e4 8.0e4 7.0e4 6.0e4 5.0e4 4.0e4 3.0e4 2.0e4 1.0e4 186 188 190 192 194 196 198 200 202 m/z, amu 203.0458 204.0406 204 206 208 210 212 214 216 202.0434
Figure 4.
LC/MSD TOF mass spectrum of thiabendazole.
Fragmentation of Imazalil The next step in the discovery process was to search for characteristic fragment ions of the pesticide to confirm (or refute) its identity as imazalil. The fragmentation pathway of the large peak at 17.9 minutes was determined with LC/ITMS at MS3. The pathway is examined to see if reasonable structures could be drawn for the imazalil fragments based on the detection by LC/TOFMS. Figure 5 shows the structure for the proposed ions from the fragmentation of the unknown peak. The m/z 255 ion results from the loss of 42. MS/MS of the m/z 255 ion gives the m/z 187 and 159 ions, whose proposed structures are shown in Figure 5.
The m/z 159 ion is proposed as a doubly chlorinated tropylium ion. All the structures are consistent with the proposed identification as imazalil. Next is to examine the accurate mass spectrum from the LC/TOFMS. Here we found two fragment ions (m/z 255 and 159) with a relative abundance of respectively, 5% and 10% of that of the protonated molecule. The accurate mass of fragment 1 was 255.0084 (Table 1), and a 37Cl signal m/z of 257.0055 (data not shown). From both the relative intensity of these signals and the difference between the two masses, it can be deduced that the two chlorine atoms present in imazalil remain in this fragment.
5 6 4 2 0 100 200 300
297.0
Cl O H2C N
400 500 m/z 600 700 800 900
Cl Neutral loss of 42 H2C HO
NH
Cl
NH
Cl
m/z 255 at MS2
5000 4000 Intensity 3000 2000 1000 109.0 0 100 200 200.8 158.7
254.8
Imazalil, EE ion
EE ion fragment
MS/MS (297 255, 201,159)
Neutral loss of 68
NH
Cl
300 400 500 m/z 600 700 800 900
Cl _CO of 28
+
1.0 0.8 Intensity x104 0.6 0.4 0.2 0.0 100 200 300 400 500 m/z 600 700 800 900 158.7
m/z 159 at MS4
Double Cl trophilium Cl
Cl
m/z 187 at MS3
MS3 (255 187,159)

186.7
EE ion fragment
EE ion fragment
82.0
219.8
Figure 5.
Ion trap MS2 and MS3 to show fragmentation pathway of imazalil.
As can be seen in Table 1, the accurate mass of this fragment gave also three possible elemental compositions in the calculator tool. The first formula (C11H9N2OCl2), (0.8 ppm error) fitted with the proposed structure, involving a loss of C3H6 (accurate mass loss of 42.0469 versus 42.0465 u) in relation with the proposed parent species. Moreover, the accurate mass spectrum (relative intensity and mass differences) evidenced the presence of two chlorine atoms on fragment 2. The measured mass (m/z 158.9762) gave a unique elemental composition (C7H5Cl2), that corresponds to the formation of a doubly chlorinated tropylium fragment-ion. At this point, we combined the fragmentation information from the LC/MSD Trap, as shown in Figure 5 for the proposed structure of
imazalil, with the accurate mass information, that all show a perfect match for the detection of imazalil. These two lines of evidence (LC/TOFMS and LC/ITMS) provide fragment ions and information to confirm the identity of the proposed species based on fragmentation of the parent structure as imazalil, without the use of standards! Likewise, thiabendazole was examined by ion trap and gives a m/z 175 fragment ion, that fits with the structure of thiabendazole. The data are not shown here but were shown in a previous application note for thiabendazole [9]. We verified these determinations by standards for final confirmation.
Table 1.
LC/MSD TOF Fragment Ions of Imazalil Elemental compositions C14H15N2OCl2 C3H15N8O4Cl2 C11H19N2OSCl2 m/z calculated 297.0556 297.0588 297.0590 255.0086 255.0073 255.0060 158.9763 Error, ppm <0.1 8.3 9.0 0.8 4.3 9.4 0.6 Comments Imazalil Fragment 1 Fragment 2
m/z experimental 297.0556
255.0084
C11H9N2OCl2 C9H7N5Cl2 C8H11NO4Cl2
158.9762
C7H5Cl2
Imazalil Metabolite In the orange extract, we found a peak (at a retention time of 14.9 minutes, see Figure 2) with an ion with the same isotopic pattern as imazalil (two chlorine atoms). Taking into account that it had the same number of chlorine atoms, and it also appeared before imazalil, we considered that it could be related to imazalil, perhaps a metabolite. The accurate mass of the ion was 257.0245 with a 37 Cl signal of 259.0216 (see Figure 6). It gave a unique elemental composition in the calculator tool: C11H10N2OCl2 (0.8 ppm error, also see Table 2). For confirmation purposes, we searched for additional fragments but we did not find any characteristic fragment ion of that compound by TOF LC/TOFMS. The obtained elemental composition was then searched against two databases (The Merck Index and ChemIndex) with no positive
results. Then, we search the elemental composition against the Sigma-Aldrich commercial electronic catalogue, and we found a unique match: 1-(2,4dichloro-phenyl)-2-imidazol-1-yl-ethanol. The structure of this compound, shown in Figure 1, is compatible with the degradation of imazalil (see Figure 5 the m/z 255 ion). This suggests that the degradation product is the neutral species corresponding to the degradation of imazalil at the same site that cleaves to yield the m/z 255 fragment. A literature search on imazalils degradation products produced data and reports that agreed with our results (Imazalil-metabolite, R14832) [4]. In fact, in the United States, the sum of imazalil and imazalil metabolite is regulated, so the survey of residual imazalil metabolite is also required [4]. Finally, we confirmed the identity of the degradate by standard matching.
Table 2.
Accurate Mass and Elemental Composition of Imazalil, Imazalil Degradate, and Thiabendazole in Orange Peel Extracted with Methanol Formula C14H14Cl2N2O C11H10Cl2N2O C10H7N3S Selected ion [M+H]+ [M+H]+ [M+H]+ m/z experimental 297.0556 257.0245 202.0434 m/z calculated 297.0556 257.0243 202.0434 Error mDa +0.1 +0.2 +0.2 ppm 0.3 0.8 <0.1
Compound Imazalil Imazalil Degradate Thiabendazole
2.4e5 2.3e5 2.2e5 2.1e5 2.0e5 1.9e5 1.8e5 1.7e5 1.6e5 1.5e5 Intensity, counts 1.4e5 1.3e5 1.2e5 1.1e5 1.0e5 9.0e4 8.0e4 7.0e4 6.0e4 5.0e4 4.0e4 3.0e4 2.0e4 1.0e4 0.0 235 240 245 250
257.0245
C11H11N2OCl2
HO N N Cl
1 (2,4 dichloro-phenyl)-2-imidazol-1-YL ethanol (Imazalil degradate)
[M+H]+ Protonated imazalil degradate Cl
259.0216
261.1115
255
260 m/z, amu
265
270
275
280
285
Figure 6.
LC/MSD TOF spectrum of imazalil degradate.
Identification of Imazalil in Orange Juice Figures 7 and 8 show the LC/TOFMS chromatogram (see peak at 18.0 minutes) and identification of imazalil (m/z 297.0556) in a juice product made from oranges. Several samples were analyzed and only this sample contained imazalil. No identifications of thiabendazole were observed. It is assumed that the production of juice usually does not involve the squeezing of the peel. However, in some types of juice machines the entire orange is squeezed without removing the peel. In this case, the opportunity for including fungicides in the juice may occur. Note that the LC/TOFMS analysis gave good results in spite of the complex matrix and complex chromatogram of the juice extract.
3.7e7 3.6e7 3.4e7 3.2e7 3.0e7 2.8e7 2.6e7 2.4e7 Intensity, cps 2.2e7 2.0e7 1.8e7 1.6e7 1.4e7 1.2e7 1.0e7 8.0e6 6.0e6 4.0e6 2.0e6 0.0 2 2.1 3.0
3.8
25.4 3.4 16.7 12.2 17.5 24.3 23.4 6.2 11.5 4.9 14.4 14.8 15.3 13.8 21.1 23.1 18.0 19.5 20.3 21.5 22.1 28.5 27.9 26.1
24.7
10
12
14
16 Time, min
18
20
22
24
26
28
30
Figure 7.
LC/MSD TOF TIC of orange juice extract.
9.2e4 9.0e4 8.5e4 8.0e4 7.5e4 7.0e4 6.5e4 6.0e4 Intensity, counts 5.5e4 5.0e4 4.5e4 4.0e4 3.5e4 3.0e4 2.5e4 2.0e4 1.5e4 1.0e4 5000.0 0.0 260 265 270 275 280 285 290
297.0556
299.0528
298.0585
295 m/z, amu
300
305
310
315
320
325
330
Figure 8. 10
LC/MSD TOF mass spectrum of imazalil in orange juice made from peel.
Conclusions
LC/TOFMS analysis is a powerful tool for identification of fungicides in fruits and juices and their metabolites and is a new tool for environmental food chemistry. Elemental composition and structural fragmentation of standards and fragment ions are possible, especially with ion trap MS/MS confirmation. The identification of metabolites is possible, in the case of imazalil, without standards, when both LC/TOFMS and LC/ITMS are used. These concentrations are equal to or lower than the EU directives for controlled fungicides in fruits and juices.

References
1. EU Food Directives 2002, 91/414/EEC. 2. Food Quality Protection Act, 1998 (FQPA). 3. M. Fernandez, R. Rodriguez, Y. Pico, J. Manes, (2001) J. Chromatogr. A 912, 301310. 4. N. Yoshioka, Y. Akiyama, K. Teranishi, (2004) J. Chromatogr. A, 1022, 145150. 5. http://www.epa.gov./pesticides/registration/ imazalil/ImazProductChem.pdf. 6. E. M. Thurman, Imma Ferrer, Jerry A. Zweigenbaum, Juan F. Garca-Reyes, Michael Woodman, and Amadeo R. Fernndez-Alba, (2005) Discovering Metabolites of Post-Harvest Fungicides in Citrus with LC/MS TOF and Ion Trap MS/MS, J. Chromatogr., in press. 7. Determination of Fungicides in Fruits and Vegetables by LC/MS/ESI/TOF and LC/MS Ion Trap, Agilent Technologies, publication 5989-2209EN www.agilent.com/chem 8. M. Anastassiades, S.J. Lehotay, D. Stajnbaher, and F.J. Schenck, (2003) J. AOAC Internat., 86, 412431. 9. Identification of Unknown Pesticides in Food Using Both LC/MSD TOF and Ion Trap MSn, Agilent Technologies, publication 5989-1924EN www.agilent.com/chem
11
Acknowledgments We acknowledge Amadeo Fernandez-Alba and Juan-Francisco Garcia-Reyes, Department of Hydrogeology and Analytical Chemistry, University of Almera, for help, ideas, and analytical support. Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. ZORBAX is a registered trademark of E.I. duPont de Nemours Co. Inc. Agilent Technologies, Inc. 2005 Printed in the USA August 24, 2005 5989-2728EN
Analysis of Terbuthylazine in Olive Oil by LC/Ion Trap MS and LC/Time-of-Flight MS Application
Food safety
Authors
Imma Ferrer and E. Michael Thurman University of Almera, 04120 Almera, Spain
Introduction
Olive oil is one of the most-used food products in Mediterranean countries. The positive effects of olive oil on health have prompted a demand for this product world-wide. Virgin olive oil is obtained from the fruit of the olive tree (Olea Europaea) exclusively by mechanical and physical processes without any further treatment, which may alter the olive oil quality. The most extensively applied agrochemicals in olive plantations of Mediterranean countries (Greece, Spain, and Italy) are herbicides and insecticides. Although herbicides are mainly applied to soils, some residues can persist to the harvest stage, thus contaminating the olives picked up from soil. This can result in the presence of trace amounts of these pesticides in olive oil. Consequently, both the European Union and the Codex Alimentarius Commission of the Food and Agriculture Organization (FAO) of the United Nations have established maximum pesticide residue levels in olives and olive oil. Currently, various olive oil pesticide residue regulatory programs are being carried out to update and to establish new and more stringent regulations concerning the maximum residue levels in these commodities [1]. This fact has fostered the development of more powerful analytical tools in order to provide enough sensitivity and selectivity to meet these requirements in food samples such as edible oils, which have a complex matrix due to the high fat content of the extracts obtained after the sample treatment step.
Abstract
This application note describes the use of liquid chromatography/ion trap mass spectrometry (LC/ITMS) and liquid chromatography/time-of-flight mass spectrometry (LC/TOFMS) for the identification and quantitation of terbuthylazine in olive oil samples. The method includes a sample treatment step based on a preliminary liquidliquid extraction, followed by matrix solid-phase dispersion (MSPD) using an aminopropyl-bonded silica as a sorbent material. A final clean-up step is performed with Florisil using acetonitrile as an eluting solvent. The analysis by ion trap was achieved in MS/MS mode, monitoring the characteristic fragment ion at m/z 174. The identification by LC/TOFMS was accomplished with the accurate mass (and the subsequently generated empirical formula) of the protonated molecule ([M+H]+ m/z 230), along with the accurate mass of the fragment ion and the characteristic chlorine isotope cluster present in terbuthylazine. The accuracy typically obtained was routinely better than 2 ppm. The method sensitivity, linearity, precision, accuracy, matrix effects, and limit of detection were also studied; they supported the potential of LC/TOFMS and LC/ITMS for the routine quantitative analyses of terbuthylazine in olive oil.
Many multiresidue procedures employing different clean-up techniques and a variety of detection methods are reported on the determination of pesticide residues in olive oil. The most commonly used methodology is based on GC after a comprehensive clean-up step, in most cases based on liquid-liquid partitioning or gel permeation chromatography to separate the low molecular mass pesticides from the higher molecular mass fat constituents of the oil, such as triglycerides [2, 3]. The preparation of oil samples for the determination of pesticides by GC requires the complete removal of the high-molecular-mass fat from the sample to maintain the chromatographic system in working order. Recently, a multiresidue method for the determination of triazines and organophosphorous pesticides using MSPD followed by GC/MS and ion trap MS techniques was reported [4]. In this work, we further explore the capabilities of LC/ITMS and LC/TOFMS techniques for the identification of terbuthylazine, one of the most common pesticides found in olive oil.
- Another 10 mL of acetonitrile saturated with petroleum ether was added to the petroleum ether extract. The mixture was shaken again for 3 minutes and the acetonitrile phase was collected and added to the previous one. - Finally, a 7-mL aliquot of the acetonitrile extract was transferred to a 10-mL glass test tube. The extract was then carefully evaporated down to a final volume of about 2 mL. This remaining extract was transferred to a glass mortar to be subjected to matrix solid-phase dispersion. 2. The MSPD: - The extract obtained in the liquid-liquid extraction step was homogenized with 2 g of aminopropyl-bonded sorbent (Bondesil-NH2, 40-m particle size, Varian Inc.) until a fine powder was obtained. - The mixture was transferred to a commercially available minicolumn containing 2 g of Florisil (12-mL Bond-Elut Varian minicolumn, Varian Inc.). This minicolumn was connected to a vacuum system for solid phase extraction adjusting the flow to 3 mL/min. - An elution step was carried out with 2 5 mL of acetonitrile. The final extract was evaporated until near dryness, then dissolved with 1:1 acetonitrile:water. - Prior to LC/ITMS and LC/TOFMS analysis, the extract was filtered through a 0.45-m PTFE filter (Millex FG, Millipore, Milford, MA, USA). Using the MSPD method, recoveries for terbuthylazine from olive oil samples were higher than 96% with a 6% relative standard deviation (RSD) (n = 5).
Olive Oil Extraction MSPD was used for the extraction of terbuthylazine from olive oil after a preliminary liquidliquid extraction with organic solvents. 1. The liquid-liquid extraction: - An aliquot of approximately 5 g (ca. 5.5 mL) of olive oil sample was mixed with 15 mL of petroleum ether, saturated with acetonitrile, in a 100-mL separatory funnel, in which a two-step liquid-liquid extraction was performed. - A solution of 25 mL of acetonitrile, saturated with petroleum ether, was added to the olive oil mix obtained previously. The funnel was shaken vigorously for 3 minutes, and the acetonitrile phase was separated from the petroleum ether phase.
Agilent 1100 Series LC/MSD TOF Methods

LC conditions LC Pumps were Agilent 1100 binary pumps Injection volume: 50 L with standard Agilent 1100 ALS Column: ZORBAX Eclipse XDB-C8, 4.6 mm 150 mm, 5 m, p/n 993967-906 Mobile phases: A = acetonitrile and B = 0.1% formic acid in water Gradient: 5 minutes isocratic at 10% A, followed by a linear gradient to 100% A in 25 minutes at a flow rate of 0.6 mL/min MS conditions LC/MSD TOF Source: Capillary: Nebulizer: Drying gas: Gas temp: Fragmentor: Skimmer: Oct DC1: Oct RF V: Reference masses: Resolution: Mass range: Reference A sprayer 2:
Positive ESI (electrospray ionization) 4000 V 40 psig 9 L/min 300 C 190 V 60 V 37.5 V 250 V m/z 121.0509 and 922.0098 9500 500 @ m/z 922.0098 501000 m/z Constant flow during the run
Agilent 1100 Series LC/MSD Trap Methods LC/MSD Trap Methods identical to LC/MSD TOF for direct comparison of peaks. LC Pumps: HP 1100 Inject vol: 50 L Column: ZORBAX Eclipse XDB-C-8, 4.6 mm 150 mm, 5 m, p/n 993967-906 Mobile phases: A = acetonitrile (ACN) B = 0.1% formic acid in water Gradient: 10% A, isocratic, for 5 minutes followed by linear gradient to 100% A in 25 min at a flow rate of 0.6 mL/min LC/MSD Trap Positive ESI: Nebulizer: Drying gas: Capillary exit: Skimmer: Trap accumulation: Capillary 3200 V 40 psig 9 L/min, gas temp 300 C 70 V 60 V 50,000 counts with maximum accumulation time of 200 ms

Identification of Terbuthylazine by LC/ITMS and LC/TOFMS Olive oil is one of the most difficult food matrices due to the presence of numerous interferences that show up in full-scan mode. For this reason, the ion trap method was optimized in MS/MS mode by isolating the precursor ion at m/z 230, which corresponds to [M+H]+. An isolation mass window of m/z 2 and optimized fragmentation amplitude was used in order to enhance the signal-to-noise (S/N) ratio. An amplitude voltage of 0.8 V was used to
obtain good fragmentation for terbuthylazine, which fragments by losing the terbutyl group (56) forming the m/z 174 fragment ion. Figure 1 shows the analysis of an olive oil sample (spiking level 0.025 mg/kg) by ion trap MS/MS. The extracted ion chromatogram (EIC) for m/z 174, the main fragment of terbuthylazine, as well as its mass spectrum is shown. The fragmentation occurs at the CN bond via the cleavage of the terbutyl group. This represents a characteristic fragmentation for this analyte, allowing for the structural confirmation of terbuthylazine in a relative complex matrix such as olive oil.
1.25
1.00 Intensity 105
0.75
0.50
0.25
0.00 0 5 10 15 20 25 Time [min]
MSPAC25.D: EIC 174 All 1.50
173.8
1.25 Cl Intensity 105 1.00 0.75 0.50 0.25 0.00 100 N N H N N N H _56
+MS2(230.0), 23.7 min #556
210.8
200 300 400 500 600 700 800 900 m/z
Figure 1.
Ion Trap MS/MS chromatogram corresponding to the analysis of a spiked olive oil sample with terbuthylazine (0.025 mg/kg). The EIC for m/z 174 and its corresponding spectrum are shown.
LC/TOFMS analyses were optimized in terms of fragmentation. The in-source collisionally induced dissociation (CID) fragmentation is greatly enhanced at high fragmentor voltages. This provides highly valuable structural information since the accurate mass of the characteristic fragment ion can be used along with that of the molecular ion for confirmation purposes. The relative abundances for both the main fragment and the protonated molecule of terbuthylazine are summarized in Table 1 at three different voltages: 160 V (low),
Table 1. m/z ion 230 [M+H]+ 174 [M+HC4H8]+ Effect of the Fragmentor Voltage on CID Fragmentation for LC/TOFMS Relative abundance 160 V 190 V 230 V 100 5 100 30 20 100
190 V (medium), and 230 V (high). In order to obtain sufficient sensitivity for quantitative purposes (using the protonated molecule) and additional qualitative spectral information provided by the fragment ions generated by in-source fragmentation, a value of 190 V was chosen for further analyses. The accurate masses for both the main fragment and the protonated molecule of terbuthylazine in a spiked olive oil matrix are shown in Table 2.
Table 2.
LC/TOFMS Accurate Mass Measurements for Terbuthylazine and its Main Fragment Ion in Olive Oil Matrix-Matched Standard (Fragmentor Voltage 190 V). Spiking Level: 0.025 mg/Kg Ion [M+H]+ Theoretical mass 230.1167 232.1137 174.0541 176.0511 Measured mass 230.1168 232.1134 174.0542 176.0511 Error mDa 0.1 0.3 0.1 0.05 ppm 0.4 1.5 0.6 0.3
Elemental composition C9H16N535Cl C9H16N537Cl C5H9N535Cl C5H9N537Cl
[M+HC4H8]+
Along with the accurate mass of the protonated molecule and the information provided by the fragments obtained with an optimized in-source fragmentation, terbuthylazine presents another feature which enables its identification; the presence of a chlorine atom. The accurate mass pattern of the 37 Cl isotope signal confirms that the peak unequivocally contains only one chlorine atom (Figure 2). Therefore, the accurate mass of each ion as well as the presence of the chlorine signal, together with the characteristic retention time represent sufficient information to unequivocally identify and confirm this specie in such complex matrices.
230.1168 9.50e4 8.50e4 7.50e4 6.50e4 5.50e4 4.50e4 3.50e4 2.50e4 1.50e4 5000.00
Intensity, counts
174.0542 176.0511
232.1134
60 80 100 120 140 160 180 200 220 240 260 280 m/z, amu
Figure 2.
LC/TOFMS total ion chromatogram (TIC) corresponding to the analysis of a spiked olive oil sample with terbuthylazine (0.025 mg/kg). The EIC for m/z 230 and its corresponding spectrum are also shown.
Analytical Performance The analytical performance of the proposed methods was studied in order to explore the ability to carry out quantitative analyses of terbuthylazine in these complex matrices with a high content of fat. The calibration was carried out using matrixmatched standards prepared by the extraction method based on MSPD described in the experimental section. Linearity was evaluated by analyzing solutions of matrix-matched standards at seven different concentration levels in the range 0.0050.5 mg/kg. Using ion trap the quantitation was performed in MS/MS mode of the m/z 230 ion. Using LC/TOFMS the quantitation was carried out using the peak area from the EIC of the protonated molecule with a mass window of 0.1 Da. As an example, Figure 3 shows the linear calibration curve obtained by LC/TOFMS for terbuthylazine in an olive oil matrix compared to the curve obtained in pure solvent.
The limits of detection (LOD) were estimated from the injection of matrix-matched standard solutions with concentration levels giving a S/N ratio of about 3. The results obtained are shown in Table 3. The LOD obtained are remarkable since they are far below the maximum residue level regulations established for these pesticides. In this sense, LC/TOFMS analyses benefits of the use of narrow mass windows for quantitation purposes, which results in enhanced S/N ratio, thus providing lower detection limits. This fact illustrates the analytical potential of the proposed method based on MSPD and LC/TOFMS for the analyses of pesticides in complex matrices with high content of fat.
40
Pure solvent
y = 70.697x + 0.5666 R2 = 0.9977
30
Spiked oil solvent
y = 60.442x + 0.7386 R2 = 0.9977
Area 10 E+06
20
10
0 0.0 0.1 0.2 0.3 Concentration (mg/kg) 0.4 0.5 0.6
Figure 3.
Calibration plot obtained from spiked olive oil samples versus solvent based samples by LC/TOFMS.
Table 3.
Analytical Parameters for the Analysis of Terbuthylazine in Olive Oil Samples by Ion trap MS/MS and LC/TOFMS Concentration Linearity LOD RSD (%) Method range (mg/kg) (R2) (g/kg) n = 5 LC/ITMS LC/TOFMS 0.0050.5 0.0050.5 0.991 0.995 0.2 1 5.5 2
overcome using the isotopic chlorine signal or a characteristic fragment not affected by other interfering species.
Conclusion
LC/ITMS and LC/TOFMS can be used for the identification and quantitation of terbuthylazine in olive oil samples after performing several preliminary sample treatments. The analysis by ion trap was achieved in MS/MS mode, monitoring the characteristic fragment ion at m/z 174. The identification by LC/TOFMS was accomplished with the accurate mass (and the subsequently generated empirical formula) of the protonated molecule ([M+H]+ m/z 230), along with the accurate mass of the fragment ion and the characteristic chlorine isotope cluster present in terbuthylazine. The accuracy typically obtained was routinely better than 2 ppm. The method sensitivity, linearity, precision, accuracy, matrix effects, and LOD were also studied and they supported the potential of LC/TOFMS and LC/ITMS for the routine quantitative analyses of terbuthylazine in olive oil.
Analysis of Olive Oil Samples To evaluate the effectiveness of the proposed method, it was applied to the analysis of olive oil samples. An example is shown in Figure 4 where traces of terbuthylazine were found in an olive oil extract; the EIC for m/z 230 is also shown. This is a real example illustrating the usefulness of the routine accurate mass measurement capabilities of LC/TOFMS, especially when analyzing traces of pesticides in complex samples such as olive oil. In this sense, the selectivity of LC/TOFMS relies on the resolving power of the instrument on the m/z axis, enabling discrimination between the target species and an isobaric interference within 0.05 Da of the mass difference (using 230 m/z, as example). In the case of terbuthylazine, it is easily
Figure 4.
Upper: LC/TOFMS TIC of a "positive" for terbuthylazine in an olive oil sample. Lower: EIC of terbuthylazine using a 20 mDa mass window.
References
1. Proposal for a Regulation of the European Parliament and of the Council on maximum residue levels of pesticides in products of plant and animal origin, COM/2003/0117/final COD2003/0052*/; www.europa.eu.int/pol/food/indexes.htm 2. Ch. Lentza-Rizos, E. J. Avramides, and F. Cherasco, (2001) J. Chromatogr. A 912: 135142. 3. J. J. Vreuls, R. J .J. Swen, V. P. Goudriaan, M. A. T. Kerkhoff, G. A. Jongenotter, and U. A. Th. Brinkman, (1996) J. Chromatogr. A 750: 275286. 4. C. Ferrer, M. J. Gmez, J. F. Garca-Reyes, I. Ferrer, E. M. Thurman, and A. R. FernndezAlba, (2005) J. Chromatogr. A 1069: 183194.
Acknowledgments
We acknowledge Amadeo Fernandez-Alba, JuanFrancisco Garca-Reyes and Carmen Ferrer from the Department of Hydrogeology and Analytical Chemistry, University of Almera, for help, ideas, and analytical support. Agilent contact: Jerry Zweigenbaum

Agilent Technologies, Inc. 2005 Printed in the USA August 23, 2005 5989-3573EN
Determination of Fungicides in Fruits and Vegetables by Time-of Flight and Ion Trap LC/MS Application
Food Safety
Author
Imma Ferrer and E. Michael Thurman Pesticide Residue Research Group University of Almera, 04120 Almera, Spain
Introduction
The importance of the identification and quantitation of fungicides in fruits and vegetables was described [1, 2]. Increasingly, LC/MS is being used for the analysis of many polar fungicides of the European Union (EU) Directives and is rapidly becoming an accepted technique in pesticide residue analysis for regulatory monitoring. In particular LC/MS works well on the various families of fungicides that are polar and labile. Thus, often LC/MS methods are preferred over the older gas chromatography/mass spectrometry (GC/MS) methods. For example, the use of MS and MS/MS for pesticides (including fungicides) in food was reviewed by Pico et al. [3], and there are currently about 100 published papers dealing with the analysis of pesticides in food. Of these, there are approximately 25 papers that deal with fungicides by LC/MS, the majority published in the early 2000s. Thus, LC/MS is an emerging technology for the analysis of food products. Interestingly though, none of the reported papers in this most recent review deal with the use of LC/TOFMS and accurate mass analysis for fungicides in food. Thus, there is an important need for research studies and methods development on the analysis of pesticides in food by accurate mass using LC/ TOFMS [1, 2]. Our study in this report is one of the first of its kind to examine the new Agilent LC/MSD TOF time-of-flight mass spectrometer for the analysis of fungicides in food. This topic was chosen because of the relevance of these fungicides and their significant use on apples, oranges, lemons, melons, tomatoes, broccoli, and peppers.
Abstract
This application examines the feasibility of the new instrumentation of electrospray and liquid chromatography/ time-of-flight mass spectrometry (LC/TOFMS) in conjunction with liquid chromatography/ ion trap mass spectrometry (LC/ITMS) to analyze five major fungicides in fruit extracts (apple, lemon, melon, and orange) and in salad vegetables (tomato, broccoli, and pepper). Included are the LC/TOFMS and LC/ITMS MS/MS spectra of five important fungicides (carbendazim, thiabendazole, azoxystrobin, dimethomorph, and triflumizole), the TOF empirical formula and MS/MS fragmentation and diagnostic ion(s) for these fungicides in the matrices of various important fruits and vegetables. A detailed rapid procedure for sample preparation and extraction of these fungicides from the fruit and vegetables is given. Spectral quality at the limit of detection (LOD), linearity, and quantitation of the fungicides in pure solvent and in the fruit and vegetable extracts using TOF and ion trap are provided. Last are the results of the analysis of real samples from the marketplace for these fungicides in fruits and vegetables: apples, oranges, lemons, and melons.
Furthermore, LC/ITMS will be demonstrated as a companion technique to LC/TOFMS for the analysis of fungicides in fruits and vegetables. The companion use of LC/ITMS was described in more detail [2]. The molecular structures of the common fungicides of this study (carbendazim, thiabendazole, azoxystrobin, dimethomorph, and triflumizole) are shown in Figure 1.
2. Take 5 mL of the supernatant into a 15-mL tube, add 250 mg of PSA adsorbent and 750 mg of MgSO4, and vortex for 20 s. Then microcentrifuge again for 3 min at 3700 rpm. Transfer 1.0 mL into an autosampler vial. Then evaporate the fruit and vegetable residues to dryness and redissolve in 8/92% methanol/ water. Standards were prepared by fortifying with a mixture of carbendazim, thiabendazole, azoxystrobin, dimethomorph, and triflumizole at concentrations ranging from 0.01 to 0.5 mg/kg for analysis using the Agilent LC/MSD TOF and the Agilent 1100 Series LC/MSD Trap. Nonfortified samples were analyzed directly at this same point by either the LC/MSD TOF or the LC/MSD Trap by injecting 50 L. The extraction method is summarized in Figure 2.
O H N NH N Carbendazim [M+H]+ = m/z 192 OCH3
H N
S N
Thiabendazole [M+H]+ = m/z 202
N O CN Azoxystrobin [M+H]+
N O H3CO = m/z 404 O OCH3 Cl F
Sample 1 kg of fruit or vegetable. F
H3CO
OCH3 O N O H3C O N
N Cl Dimethomorph [M+H]+ = m/z 388 Triflumizole [M+H]+ = m/z 346
Combine 15 g of sample, 15 mL of acetonitrile (1% acetic acid), extract and shake for 60 s with some Vortex to mix. Add 6 g of MgSO4 and 2.5 g of NaAc3H2O and shake. Centrifuge for 3 min at 3700 rpm.
Put 5 mL of sample into a 15-mL tube, add 250 mg of PSA adsorbent and 750 mg of MgSO4, and vortex and shake for 20 s. Then microcentrifuge for 180 s at 3700 rpm.
Figure 1.
Structures of major fungicides used on fruits and vegetables, and the corresponding [M+H]+ ion. Figure 2.
Transfer 1.0 mL into LC/MS vials
Fruit and Vegetable Extraction (QuEChERS) QuEChERS is the acronym for the method of extraction, which stands for quick, easy, cheap, effective, rugged, and safe [4]. It is a method that is receiving wide acceptance for rapid extraction of pesticides in food. 1. Weigh 15 g of a previously homogenized sample into a 40-mL Teflon centrifuge tube. Add 15 mL of acetonitrile (containing 1% acetic acid), 6 g of anhydrous MgSO4 and 2.5 g of NaAc3H2O (sodium acetate trihydrate) and shake the sample vigorously for 1 min by using a Vortex mixer at maximum speed or by hand shaking. Afterwards, centrifuge for 3 min at 3700 rpm.
2
QuEChERS Extraction method for fungicides in fruits and vegetables.
Agilent LC/MSD TOF Methods LC Pumps were Agilent 1100 binary pumps, injection volume 50 L with standard Agilent 1100 ALS. Column: ZORBAX Eclipse XDB-C8, 4.6 mm 150 mm, 5 m, part number 993967-906 Mobile Phase A=acetonitrile and B=0.1% formic acid in water, gradient was 10% A for 5 min, then to 100% A in 25 min at a flow rate of 0.6 mL/min Agilent LC/MSD TOF with electrospray source Positive ESI, Capillary 4000 V
Nebulizer 40 psig, drying gas 9 L/min, gas temp 300 C Fragmentor 190 V, Skimmer 60 V, Oct DC1 37.5 V, OCT RF V 250 V Reference Masses: m/z 121.0509 and 922.0098, resolution: 9500500 @ m/z 922.0098, Reference A Sprayer 2 is constant flow rate during the run Agilent 1100 Series LC/MSD Trap Methods Chromatographic methods were identical to Agilent LC/MSD TOF for direct comparison of peaks LC Pumps were Agilent 1100, injection volume 50 L Column: ZORBAX Eclipse XDB-C8, 4.6 mm 150 mm, 5 m, part number 993967-906 Mobile Phase A=acetonitrile and B=0.1% formic acid in water, gradient was 10% A for 5 min, then to 100% A in 25 min at a flow rate of 0.6 mL/min. Agilent 1100 Series LC/MSD Trap Positive electrospray ionization (ESI), Capillary 3200 V Nebulizer 40 psig, drying gas 9 L/min, gas temp 300 C Capillary exit 70 V Trap accumulation, 50,000 counts
Intensity, cps
17.6 5.0e7 4.0e7 3.0e7 2.0e7 1.0e7 0.0 2 1.6e6 1.4e6 1.2e6 1.0e6 8.0e5 6.0e5 4.0e5 2.0e5 0.0 2 4 4 6 8 2.1 3.6 4.1 13.4 17.4 15.1 18.0 20.8 22.7 24.0 25.1
(a) 28.9
10 12 14 16 18 20 22 24 26 28 30 Time, min 4 (b)
Intensity, cps
3 5
10 12 14 16 18 20 22 24 26 28 30 Time, min
Figure 3.
(a) TIC of a lemon standard solution (0.125 ppm); (b) EICs of carbendazim (1), thiabendazole (2), dimethomorph (3), azoxystrobin (4), and triflumizole (5)
of the compounds are shown in Figure 3b. The window of extraction was clean, which is commonly a feature of accurate mass extraction of ions, where the width of the window of extraction may be narrowed to (0.02 amu) or 100 ppm. In Figure 3, the window of extraction was 0.1 amu. At concentrations as low as 0.05 mg/kg, the extracted ions (m/z 192, 202, 346, 388, and 404) still yielded clean chromatograms testifying to the importance of accurate mass and its ability to give clean extracted ion chromatograms (EICs) with a narrow mass window. This accurate mass window is reflected in the determinations of the accurate mass of the protonated molecule of each of the fungicides. Table 1 shows the elemental composition, the exact mass, and errors, in mDa and ppm, for each of the fungicides at the 0.50 mg/kg concentration. Mass accuracy was always better than 2 ppm in all the fruit and vegetable matrices, except for dimethomorph.

Fragmentation and Mass Accuracy Figure 3 shows the LC separation of five fungicides: carbendazim, thiabendazole, azoxystrobin, dimethomorph, and triflumizole. The sample contained the fungicides at 0.125 mg/kg (ppm) in a fortified lemon extract. The extracted ions for each
Table 1.
Accurate Mass and Elemental Composition of Five Fungicides at 0.50 mg/kg in Lemon Matrix Error mDa -0.05 -0.34 +1.1 0.20 -0.35 Error ppm 0.27 1.7 2.5 0.50 1.0 3
Compound Carbendazim Thiabendazole Dimethomorph Azoxystrobin Triflumizole
Formula C9H9N3O2 C10H7N3S C21H22NO4Cl C22H17N3O5 C15H15N3OF3Cl
Selected ion [M + H]+ [M + H]

+
m/zexperimental 192.0767 202.0430 388.1321 404.1243 346.0925
m/zcalculated 192.07675 202.04334 388.13101 404.1242 346.09285
[M + H]+ [M + H]+ [M + H]+
These results show that the use of continuous calibration is effective for accurate mass across an order of magnitude concentration range in complex fruit and vegetable matrices. Figures 4a4e show the fragmentation pathway of each of the five fungicides, based on accurate mass spectra with pure standards in a methanol solution. The fragmentation ions of the LC/MSD TOF were verified with the LC/MSD Trap spectra at MS2 with identical chromatographic conditions. Beginning with carbendazim, the MS2 shows that the protonated molecule loses methanol (32 Da) to give the m/z 160 fragment ion (Figure 4a). This fragment ion of m/z 160 is also seen in the LC/MSD TOF. The fragmentation pathway for thiabendazole is shown in Figure 4b with the protonated molecule, m/z 202, fragmenting to give the m/z 175, with the neutral loss of 27 Da or HCN. The fragmentation pathway for azoxystrobin is shown in Figure 4c with the protonated molecule, m/z 404, fragmenting to m/z 372 (loss of methanol of 32 Da) at MS2, and to m/z 342 at MS3. The fragmentation pathway for dimethomorph is shown in Figure 4d with the protonated molecule, m/z 388, fragmenting to give the m/z 301 at MS2, then m/z 165 at MS3. The fragmentation pathway for triflumizole, m/z 346, is to m/z 278 at MS2, then to m/z 250 at MS3 (Figure 4e). The combination of the LC/MSD Trap and LC/MSD TOF is extremely valuable for interpretation of spectra in that both instruments work well for these compounds and give complementary information on the structure and identity.
MSD Ion Trap of carbendazim

+H+ NH O
Fragment ion at MS2

+H+ O N
N H
OCH3
Carbendazim, [M+H]+ = m/z 192
Fragment ion, m/z 160
4.7e4 4.0e4
N N O
192.0766
H2 N N
NH
Int ensity, counts
N H
OCH3
3.0e4
2.0e4
160.0502
1.0e4
193.0796
0.0 145 155 165 175 m/z, amu 185 195 205 215
Figure 4a. Structure and fragmentation pathways for carbendazim using the LC/MSD TOF and LC/MSD Trap. The spectrum is from the LC/MSD TOF.
MSD Ion Trap of thiabendazole

H N +H+ S
Fragment ion at MS2

+H+ N S
N Thiabendazole, [M+H]+
N N = m/z 202 Fragment ion, m/z 175
1.00e5 9.00e4 8.00e4 Intensity, counts 7.00e4 6.00e4
202.0433
S 5.00e4 4.00e4 3.00e4 202.2072 2.00e4 203.0459 1.00e4 200.0 202.0 204.0404 204.0 206.0 m/z, amu 208.0 210.0 N H
Figure 4b. Structure and fragmentation pathways for thiabendazole using the LC/MSD TOF and LC/MSD Trap. The spectrum is from the LC/MSD TOF. 4
N O
N Cl O H3CO +H+ OCH3 O
Ion Trap MS/MS of dimethomorph
Fragment ion at MS2

H3CO OCH3
Ion Trap MS/MS of azoxystrobin, [M+H]+ = m/z 404
CN
Azoxystrobin, m/z 404
+H+ N H3CO H3CO O Dimethomorph, [M+H+]+ = m/z 388 O
N O +H+
Fragment ion, m/z 372, at MS2

CN
Cl Fragment ion, m/z 301
Fragment ion, m/z 372 H3CO O
H3CO
Fragment ion at MS3 Fragment ion, m/z 342, at MS3

CN N O N O +H+ O 8.0e4 7.0e4 Intensity, counts Intensity, counts 4.0e5 1 ppm accuracy 3.0e5 O 2.0e5 CN H3CO 1.0e5 372.0982 OH CN H3CO OH 405.1275 426.1061 427.1092 420 430 N N O 1 ppm accuracy N O N O OCH3 2.0e4 1.0e4 404.1246 6.0e4 5.0e4 4.0e4 3.0e4 389.1341 Cl 390.1280 O O 388.1310 O NH
O +H+ Fragment ion, m/z 165 MS3
CH3 O CH3
391.1317 386.0 387.0 388.0 389.0 390.0 391.0 392.0 393.0 394.0 m/z, amu
373.1008 370 380 390
406.1301 400 410 m/z, amu
Figure 4d. Structure and fragmentation pathways for dimethomorph using the LC/MSD TOF and LC/MSD Trap. The spectrum is from the LC/MSD TOF.
Figure 4c. Structure and fragmentation pathways for azoxystrobin using the LC/MSD TOF and LC/MSD Trap. The spectrum is from the LC/MSD TOF.
Ion Trap MS/MS of triflumizole
Fragment ion at MS2
+H+ O+ O
N N N N
CF3
CF3
Cl
Triflumizole [M+H]+ = m/z 346
Cl
Cl
Fragment ion at MS3

N O + CF3 Fragment ion, m/z 250 at MS3
broccoli, and pepper). Results showed the similarity among matrices with the LC/MSD TOF and LC/MSD Trap, given the variability in fruit and vegetable matrices. For example, Figures 5a5b show the standard curves for two representative fungicides, carbendazim and azoxystrobin. These figures demonstrate that there was little or no matrix suppression in the LC/MSD TOF system for the fruit extracts up to 0.5 ppm of parent compound for carbendazim, thiabendazole, and azoxystrobin. However, triflumizole and dimethomorph showed both enhanced signal and suppressed signal (data not shown here). The enhancement occurred for the fruit extracts of melon, orange, and lemon. The vegetable extracts of pepper showed no enhancement and some suppression for broccoli was observed. One possible explanation for these results for triflumizole has to do with the late retention time of the analyte (26 min), which means that the mobile phase is approximately 80% acetonitrile. The ESI signal is susceptible to matrix effects at these high concentrations of organic solvent [5], which indicates the importance of using matrix matched standards for unknown analysis of food extracts. Thus, best accuracy for reporting concentrations was with the standard curve made up in matrix for each fruit or vegetable. This was the procedure that was used for unknown analysis in real fruit and vegetable samples.
(a)
F
Cl 278.0558 2.5e5
Carbendazim
30 F Area 10 E+06
O+ H3C N CF3 O N
2.0e5 Intensity, counts
20
1.5e5
10
Solvent Pepper Lemon Orange Broccoli Melon
1.0e5
Cl 280.0529 279.0586 281.0556 280 290 300 310 320 m/z, amu 330
N 346.0930
0.1
0.2
5.0e4 1.0e4
0.3 0.4 0.5 Concentration (mg/kg)
0.6
347.0958 348.0912 340 350 Area 10 E+06
50
Azoxystrobin
(b)
40
Figure 4e. Structure and fragmentation pathways for triflumizole using the LC/MSD TOF and LC/MSD Trap. The spectrum is from the LC/MSD TOF.
30 Solvent Pepper Broccoli Lemon Orange Melon
20
Linearity and Detection Limits

Calibration curves were established for the five fungicides using both the LC/MSD TOF and LC/MSD Trap over the analyte concentration range of interest, which was from 0.01 mg/kg to 0.5 mg/kg (ppm) in solvent and in each of the fruit and vegetable matrices (orange, lemon, melon,
6
10
0 0 0.1 0.2 0.3 0.4 0.5 Concentration (mg/kg) 0.6
Figure 5. Overlay of standard curves for the LC/MSD TOF analysis of carbendazim (5a) and azoxystrobin (5b) in five matrices.
Furthermore, Figures 5ab also show that the standard curves were linear across the range of concentration tested with correlation coefficients of 0.990 to 0.999 for all matrices tested and for all four of the fungicides (Table 2). Similar results for standard curves were seen with the LC/MSD Trap, with correlation coefficients typically of 0.9900.001.
Table 2. Correlation Coefficients for Standard Curves for Various Pesticides in Representative Fruit Matrices Orange 0.998 0.999 0.990 0.996 Melon 0.997 0.999 0.990 0.999 Lemon 0.999 0.999 0.997 0.996
For example, carbendazim is regulated at 0.10 mg/kg in tomatoes, thiabendazole is 0.50 mg/kg for tomatoes, while azoxystrobin is not regulated in tomatoes in Spain; therefore, the maximum residue limit is automatically 0.01 mg/kg. The LC/MSD TOF is capable of these LODs for each of the fungicides in the various fruit and vegetable matrices. The quantitation for real fruit extracts from actual grocery store samples are shown in Table 5 for LC/MSD TOF analysis of fungicides with LC/MSD Trap confirmation. Thus, these data show that the LC/MSD TOF is capable of accurate mass measurements of the fungicides in fruit and vegetable extracts with LODs needed for their monitoring in the EU.
Compound Carbendazim Thiabendazole Azoxystrobin Trifluomizole
Table 3 shows the limits of detection (LODs) of the fungicides for various matrices. Typically, the values of Table 3 vary from 0.001 ppm to 0.010 ppm. The European standard for pesticides with no regulatory standard is 0.010 ppm. Thus, the LC/MSD TOF is sufficiently sensitive to detect these compounds in all matrices. A similar result was achieved for the LC/MSD Trap for LODs (Table 4) using single MS. The accurate mass capabilities of the LC/MSD TOF appear capable of seeing through the complexity of the fruit and vegetable matrices without interferences to the LODs. These LOD results of the various fungicides in fruits and vegetables are capable of meeting the regulation limits of Spain and at those of the EU.
Table 3.
Table 4.
LOD in mg/kg for Fungicides by LC/MSD Trap in Three Matrices in Full-Scan Mode Orange 0.005 0.005 0.005 0.002 0.002 Melon 0.005 0.010 0.003 0.005 0.002 Lemon 0.010 0.005 0.010 0.010 0.005
Compound Carbendazim Thiabendazole Dimethomorph Azoxystrobin Triflumizole
Table 5.
Concentration (in mg/kg) of Fungicides by LC/MSD TOF and Confirmed by LC/MSD Trap in Samples of Fruits from Grocery Store Produce Orange 0.03 <LOD 0 0 0 Melon <LOD 0 0 0 0 Apple 0.1 0 0 0 0 Lemon 0.5 0.1 0 0 0
Compound LOD in mg/kg for Fungicides by LC/MSD TOF in Three Carbendazim Matrices Thiabendazole Compound Orange Melon Lemon Dimethomorph Carbendazim 0.004 0.005 0.008 Azoxystrobin Thiabendazole 0.005 0.010 0.005 Triflumizole Dimethomorph 0.005 0.002 0.008 Azoxystrobin Triflumizole 0.001 0.001 0.001 0.001 0.001 0.001
Conclusions
LC/MSD TOF analysis is a powerful tool for identification of fungicides in fruits and vegetables and is a new tool for environmental food chemistry. Quantitation is easily possible over 2 orders of magnitude with accuracy better than 3 ppm, typically less than 2 ppm, and in this work was an amazing 1 ppm in ESI+ ion for most compounds! Elemental composition of fungicides and fragment ions are possible with LC/MSD TOF, and further confirmed with LC/MSD Trap. LODs of the five fungicides in fruit and vegetables were from 0.001 to 0.010 g/g. These concentrations are equal to or better than the EU directives for controlled fungicides in fruits and vegetables.
References
1. Imma Ferrer and E. Michael Thurman, (2004) "Determination of chloronicotinyl insecticides in salad vegetables by LC/MSD TOF and LC/MSD ion trap", Agilent Technologies, publication 5989-1842EN www.agilent.com/chem 2. E. Michael Thurman and Imma Ferrer, (2005) "Identification of unknown pesticides in food using both LC/MSD TOF and Ion Trap MSn", Agilent Technologies, publication 5989-1924EN www.agilent.com/chem 3. Y. Pico, C. Blasco, and G. Font, (2004) Mass Spectrometry Reviews, 23, 45-85. 4. M. Anastassiades, S.J. Lehotay, D. Stajnbaher, and F.J. Schenck, (2003) Journal of AOAC International, 86: 412-431. 5. E.M. Thurman, I. Ferrer, and A.R. FernndezAlba, Chapter 8: LC/MS I. Basic Principles and Technical Aspects of LC/MS for Pesticide Analysis, in Chromatographic-Mass Spectrometric Food Analysis for Trace Determination of Pesticide Residues, Ed. A.R. Fernndez-Alba, Elsevier, Amsterdam, 2005.
Acknowledgements
We acknowledge Amadeo Fernndez-Alba, Department of Hydrogeology and Analytical Chemistry, University of Almera for analytical support.

For more information on our products and services, visit our Web site at www.agilent.com/chem. Agilent Contact: Jerry Zweigenbaum
ZORBAX is a registered trademark of E.I. duPont de Nemours Co. Inc. Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2005 Printed in the USA March 15, 2005 5989-2209EN
Identification of Unknown Pesticides in Food Using Both LC/MSD TOF and Ion Trap MSn Application
Food Safety
Author
E. Michael Thurman and Imma Ferrer Pesticide Residue Research Group University of Almera 04120 Almera Spain Agilent Contact: Jerry Zweigenbaum 2850 Centerville Road Wilmington, DE 19808-1610
searching of ions. The procedure here consists of multiple steps. First is the initial detection of a possible unknown pesticide in actual market-place vegetable extracts (tomato or another salad vegetable) using either the LC/MSD TOF or the LC/MSD ion trap. Second is the application of the "Dual Approach." This consists of accurate mass identification of an empirical formula or molecular mass (usually containing an A+2 isotope signature, such as from S or Cl), followed by MS/MS ion trap confirmation of the chlorinated and/or sulfur-containing pesticides and their major fragment ions and chlorine or sulfur signatures. Third is the database analysis of the empirical formula, which is carried out with either the Merck Index CD, containing 10,000 compounds, or the ChemINDEX CD, containing 77,000 compounds, and a laptop computer. In the examples given, the unknowns are identified as the organophosphate insecticide, malathion, in cucumbers; and the insect growth regulator, buprofezin in tomato extracts. The fourth and final step involves confirmation with authentic standards.
Abstract
Traditionally, the screening of unknown pesticides in food was accomplished by gas chromatography/mass spectrometry (GC/MS) methods using conventional library searching routines. However, many of the new polar and thermally labile pesticides are more readily and easily analyzed using liquid chromatography/mass spectrometry (LC/MS) methods, even though no searchable libraries currently exist (with the exception of some limited user libraries). There is, therefore, a need for LC/MS methods to detect true pesticide unknowns. This application develops an identification scheme for unknown pesticides using a combination of liquid chromatography/mass selective detector (LC/MSD) with time-of-flight (TOF) and ion trap (MSn) options, and searching for the empirical formula using the accurate mass and the ChemINDEX or Merck Index databases. The approach is different from the conventional library
Introduction
The identification and quantitation of insecticides in vegetables is of great importance to individuals and health organizations around the world. The European Union (EU), has set new Directives for pesticides at low levels in vegetables in order to
meet these health concerns. For example, new laws such as the European Directive 91/414/EEC [1], or the Food Quality Protection Act (FQPA) [2] in the US have increased the standards for human health, workers, and environmental protection. They also require re-registration for older pesticides. Furthermore, the review programs have withdrawn authorizations for many of the crop protection products currently on the market, 177 compounds in the US and 320 in Europe. Moreover, it was announced in Europe that a total of 110 products will be withdrawn in the near future [1]. The revision of the review program in Europe dictates that over 450 existing active substances will be taken off the market by 2008 with about 400 pesticides remaining in use. Next, the quality standards within the new regulations include the re-assessment of the maximum residue limits (MRLs) for vegetables, and EU directives are setting different MRLs for each pesticide within each food group. Typically, the MRLs are lower than the previous ones. Furthermore, the new Directive leads to different MRLs for each EU country, which are still being decided. The EU Directive states that individual country MRLs will be maintained in the new program. Finally, banned compounds have the lowest MRLs, which is set now at 0.01 mg/kg (ppm). With the planned program to remove so many compounds from the market, it is important, even necessary, that screening for unknown pesticides may be done by LC/MS on vegetable extracts. Because it is not always possible to know which banned substances may be used, it is of vital importance to environmental food monitoring that there be a system to give fast and accurate screening of unknown substances in food and food products. Thus, there is an important need for research studies and methods development on the analysis of unknown pesticides in food by new LC/MS methods, such as the combination of accurate mass using LC/MSD TOF, and MS/MS using LC/MSD ion trap. Our study in this report is one of the first of its kind to examine the new Agilent LC/MSD TOF combined with LC/MSD ion trap, and the use of commercial databases, such as the Merck Index and the ChemINDEX database to identify unknown pesticides in food. Several advantages of the combination of LC/MSD TOF and ion trap are that accurate mass and empirical formulas may be combined with the MS/MS spectra or even MSn for spectral information. The identification of unknowns using the
LC/MSD TOF and ion trap consists of six steps, which are outlined below. 1. Analyze the vegetable extract with ion trap in full-scan mode looking for important unknown peaks or alternatively, using in-source CID on LC/MSD TOF. 2. Next check for A+2 isotope patterns to see if Cl, Br, or S is present. 3. Search Merck Index or ChemINDEX for possible unknowns using molecular mass (nominal mass weight), A+2 isotopes, and molecular formula. 4. Proceed to ion trap MS/MS with proposed ideas for structure and do MS2 or MS3. Identify ion fragments, accurate mass fragments, and possible structures. 5. Combine with LC/MSD TOF data of accurate fragment ions and protonated molecule to make tentative identification. 6. Obtain and analyze appropriate standard for final confirmation. Given are two detailed examples of this process using store-purchased salad vegetables, cucumbers and tomatoes, both of which contained "unknown white powders" that were subsequently identified by the above process for various "unknown pesticides."
Standard Vegetable Extraction [3] 1. Tomato, lettuce, pepper, and cucumber, in 2-kilogram portions, were obtained from the market place and homogenized by chopping. 2. Then 15 g of the homogenized vegetable was accurately weighed and placed into a 200-mL PTFE centrifuge tube. 3. Ethyl acetate (45 mL) and 6.5 M NaOH (1 mL) were added and blended in a high-speed blender (Polytron) for 30 s at 21,000 rpm [4]. After this time, 13 g of anhydrous Na2SO4 was added and the extraction repeated for 30 s. 4. The combined extracts were then filtered through a thin layer of 20 g of anhydrous Na2SO4. The solid was washed with 50 mL of ethyl acetate and the combined extracts were evaporated to dryness on a vacuum rotary evaporator using a water bath at 45 5 C.
5. The dried residue was dissolved by sonication in 15 mL of methanol. The extracts, which contained 1 g of sample per mL, were filtered through 0.2-m PTFE filters (Millex FG, Millipore) prior to LC/MS analysis. Rapid Vegetable Extraction 1. Select tomato or cucumber containing white powder from a commercial market place. 2. Carefully wash the vegetable skin three times with methanol to remove the white powder, recovering a total of 2 to 5 mL of wash, (depending on the size of the vegetable) into a 150-mL Pyrex beaker. 3. After mixing, transfer the methanol wash to a 5-mL syringe and filter through a Millex-FH PTFE filter and aliquot 0.3 mL. 4. Dilute with 0.6 mL of deionized water. 5. Analyze by either LC/MSD TOF or LC/MSD ion trap. LC/MSD TOF Methods Instrument: Agilent model LC/MSD TOF with electrospray source LC pumps were Agilent 1100 binary pumps, injection volume 50 L with standard Agilent ALS Column: ZORBAX Eclipse XDB-C8, 4.6 mm 150 mm 5 m, part number 993967-906 Mobile phases: A: ACN, B: 0.1% formic acid in water Gradient: 15% A to 100% A over 30 minutes, at 0.6 mL/min Positive ESI, capillary: 4000 V Nebulizer 40 psig, drying gas 9 L/min, gas temp 300 C Fragmentor: 190 V, skimmer: 60 V Oct DC1: 37.5 V, OCT RF V: 250 V Reference masses: m/z 121.0509 and 922.0098, resolution: 9500 500 @ m/z 922.0098 Reference A Sprayer 2 set to constant flow rate during the run
LC/MSD Ion Trap Methods Chromatographic conditions identical to those using LC/MSD TOF for direct comparison of peaks. Instrument: LC/MSD Ion Trap Classic LC pumps were Agilent 1100 binary pumps, injection 50 L with standard Agilent ALS Column: ZORBAX Eclipse XDB-C8, 4.6 mm 150 mm 5 m, part number 993967-906 Mobile phases: A: ACN, B: 0.1% formic acid in water Gradient: 15% A to 100% A over 30 minutes, flow rate: 0.6 mL/min Positive ESI, Capillary 3000 V Nebulizer 40 psig, drying gas 9 L/min, gas temp 350 C Skimmer 1: 20 V, Skimmer 2: 10 V, Capillary exit offset: 50 V, Capillary exit: 70 V Octopole 3 V, Oct RF 100 V, Octopole : 2 V, Lens 1: -3 V, Trap drive: 50, Lens 2: -50 V Target: 50,000, max. accumulation time: 200 ms, scan m/z 50 to 800, averages: 5 Rolling average: on, MS/MS m/z 2.0 isolation width, amplitude: 1.2 V, fragmentation cutoff m/z 83

Cucumber Extract Figure 1 shows the ion trap total ion chromatogram (TIC) for the rapid extract of the white powder present on a store-purchased cucumber. The largest peak in the chromatogram, deemed the Star*TIC, has a spectrum shown in Figure 1 (bottom panel) with ions at m/z 331 and 353. The spacing of m/z 22 indicates a sodium adduct. Because the analysis was carried out in positive ion, the [M+H]+ is the m/z 331, and the [M+Na]+ is the m/z 353. Furthermore, note that the m/z 331 ion also contains an A+2 ion at m/z 333, which is 8% of the m/z 331 ion. The two important A+2 ions to consider in unknown ID are chlorine and sulfur (occasionally bromine). Sulfur-34 is present in natural abundance of 4.8%; thus, the 8% peak is most likely due to two sulfur-34 atoms present in the unknown at m/z 331.
Intens. 106 2.5 2.0 1.5 1.0 0.5 0.0 0 5 TIC EIC 301 Intens. 106 0.8 0.6 0.4 0.2 0.0 98.9 170.0 100 200 285.1 300 400 500 Possible fragment ions 353.1 575.5 600 331.1 10 15 20 EIC 331 m/z 331 25 m/z 301 Time [min]
Intens. 106 2.0 1.5 1.0 0.5 0.0 20.5 Intens. 3000 2000 1000 0 Fragment ions 126.9 172.9 240.1 98.7 100 200 300 21.0 21.5 22.0 22.5 23.0 23.5 24.0 24.5 Time [min] TIC
284.9
MS/MS of 331 No fragmentation of 353 verifying it is a Na adduct The m/z 285, 240, and 99 ions are present 400 500 600 700 m/z
Figure 2.
Ion-Trap MS/MS and corresponding product mass spectrum for m/z 331.
TIC of cucumber skin
Star*TIC
The next step is the search of the Merck Index for the molecular weight and sulfur atom content. Figure 3 shows an example search sheet on a laptop computer. Because the ion trap gives the protonated molecule (m/z 331) the search is done by subtracting a proton to give the molecular weight of 330. The molecular weight is then searched across one mass unit to 331. The molecular formula has only the S2, which is a requirement now of the database search.
1. [M+H]+ = m/z 331 2. m/z 353 = [M+Na]+ 3. S-34 ~8% = 2 S atoms 682.8 709.5 700 m/z
Figure 1.
Upper: cucumber skin ion-trap TIC; lower: extracted ion chromatogram (EIC) for m/z 331. The lower panel is the spectrum of the Star*TIC peak. Figure 3. Database search for MW 330-331.
Figure 2 shows the LC/MS/MS of m/z 331. The ions present include: m/z 285, 240, 127, and 99. These ions will be used later for structural identification after a database search on the Merck Index finds a possible structure. These ions will then be compared with that structure.
Figures 4a and 4b show the two results of the Merck Index database search. The compounds found were penicillin O and malathion, with molecular weights of 330.43 and 330.36, respectively.
Results of library search for MWt = 330-331 Molecular Formula = S2*
Monograph Number: 7170 Title: Penicillin O CAS Registry Number: 87-09-2

H2C S N H N O COOH S CH3 CH3 O H H
CAS Name: (2S,5R,6R)-3,3-Dimethyl-7-oxo-6-[[(2-propenylthio)acetyl]amino]-4-thia-1-azabicyclo[3.2.0]heptane-2carboxylic acid Additional Names: [(allylthio)methyl] penicillin; allylmercaptomethylpenicillin; allylmercaptomethylpenicillinic acid; penicillin AT Molecular Formula: C13H18N2O4S2
Molecular Weight: 330.43. Percent Composition: C 47.25%, H 5.49%, N 8.48%, O 19.37%, S 19.41% Literature References: Antibiotic produced by Penicillium chrysogenum. Biosynthesis of salts: Behrens et al., J. Biol. Chem. 175, 793 (1948); Rhodehamel, Behrens et al., US 2528175 and US 2623876 (1950, 1952, both to Lilly); Ford et al., Antibiot. & Chemother. 3, 1149 (1953); Ford, US 2647894 (1953 to Upjohn); Paleckov, Slechta, C.A. 50, 17309g (1956).
Derivative Type: 2-Chloroprocaine salt monohydrate Additional Names: Chloroprocaine penicillin O; penicillin O 2-chloroprocaine Trademarks: Depo-Cer-O-Cillin Chloroprocaine (Upjohn) Molecular Formula: C H ClN O S .H O 26 37 4 6 2 2 Molecular Weight: 619.20. Percent Composition: C 50.43%, H 6.35%, Cl 5.73%, N 9.05%, O 18.09%, S 10.36% Properties: Slender needles from hot water, mp 79-81. Practically insol in cold water. Stable in dry form at room temp. Aq suspensions are stable at room temp for 1 week, at refrigerator temps for 3 weeks. Calculated activity: 949 units/mg. Solubilities: Weiss et al., Antibiot. & Chemother. 7, 374 (1957). Melting point: mp 79-81
Figure 4a. First result for database search for MW 330-331.
Monograph Number: 5723 Title: Malathion CAS Registry Number: 121-75-5 CAS Name: [(Dimethoxyphosphinothioyl)thio]butanedioic acid diethyl ester
H3CO H3CO S P S
O O
CH3 CH3
Additional Names: diethyl mercaptosuccinate S-ester with O,O-dimethyl phosphorothioate; S-(1,2dicarbethoxyethyl) O,O-dimethyldithiophosphate; insecticide no. 4049; carbofos; malathon (obsolete); mercaptothion; phosphothion Manufacturers' Codes: ENT-17034 Trademarks: Cythion; Derbac-M (International); Malamar 50; Malaspray; Organoderm (Mundipharma); Prioderm (Coates & Cooper); Suleo-M (International) Molecular Formula: C10H19O6PS2
Molecular Weight: 330.36. Percent Composition: C 36.36%, H 5.80%, O 29.06%, P 9.38%, S 19.41% Literature References: Cholinesterase inhibitor. Prepn: Johnson et al., J. Econ. Entomol. 45, 279 (1952); Cassaday, US 2578652 (1951 to Am. Cyanamid). Purification: Usui, US 2962521 (1960 to Sumitomo). Treatment of Pediculus humanus (head lice) infestation: D. Taplin et al., J. Am. Med. Assoc. 247, 3103 (1982). Toxicity data: T. B. Gaines, Toxicol. Appl. Pharmacol. 14, 515 (1969). Review of distribution, transport and fate in the environment: M. S. Mulla et al., Residue Rev. 81, 1-159 (1981); of carcinogenic risk: IARC Monographs 30, 103-129 (1983). Properties: Deep brown to yellow liq, mp 2.9. bp 156-157. Characteristic odor. d 25 1.23. n 25 1.4985. Vapor pressure at 30: 4 10-5 mm Hg. Slightly sol in 4 D and alkylated aromatic hydrocarbons and vegetable oils. water (145 ppm). Misc with many organic solvents 0.7 including alcohols, esters, ketones, ethers, aromatic Limited soly in certain paraffin hydrocarbons. Petroleum ether is sol to about 35% in malathion. Hydrolyzed at pH >7.0 or <5.0. Stable in an aq soln buffered to pH 5.26. LD in female, male rats (mg/kg): 1000, 1375 orally (Gaines). 50 Melting point: mp 2.9 Boiling point: bp 156-157 0.7 Index of refraction: n 25 1.4985 D Density: d 25 1.23 4 Toxicity data: LD in female, male rats (mg/kg): 1000, 1375 orally (Gaines) 50 CAUTION: Potential symptoms of overexposure are miosis, aching eyes, blurred vision and lacrimation; eye and skin irritation; salivation; anorexia, nausea, vomiting, abdominal cramps, diarrhea, giddiness, confusion and ataxia; rhinorrhea, headache; tightening of chest, wheezing and laryngeal spasms. See NIOSH Pocket Guide to Chemical Hazards (DHHS/NIOSH 97-140, 1997) p 188. Use: Insecticide. Therap-Cat: Pediculicide. Therap-Cat-Vet: Ectoparasiticide.
Figure 4b. Second result for database search for MW 330-331. 5
The fragmentation ions shown in Figure 5 fit the ions from the MS/MS experiment of the m/z 331 (Figure 2). Therefore, malathion is a good possibility of an identification from the Database search of the Merck Index.
Intensity, cps OH S H3CO H3CO P S O C10H20O6PS2 Exact Mass: 331.0433 Error = 0.8 ppm
+
2.1 1.00e7 8.00e6 2.8 6.00e6 4.00e6 2.00e6 4.0 18.6 2.4 0.7 3.5 Star*TIC 27.5 22.5 28.6 24.9 26.0 27.8 30.5 21.1 23.3
O O
CH3 CH3 S H3CO H3CO P S Na
O O O O CH3 CH3
0.00
10 12 14 16 18 20 22 24 26 28 30 32 Time, min
Figure 6a. LC/MS TOF analysis of a cucumber extract.
-46
HO
C10H19NaO6PS2+ Exact Mass: 353.0253 Error = 1 ppm
H3CO H3CO
S P S O
When this formula is entered into the Merck Index database the only match is malathion (Figure 7).
OH
H3CO H3CO
S P S
O C8H14O5PS2+ Exact Mass: 285.0015 Error = 0.8 ppm HO
O C7H14O4PS2+ Exact Mass: 257.0066 Error = 0 ppm O
O C6H7O3+ Exact Mass: 127.0390 Error = 0 ppm
O H
C4H3O3+ Exact Mass: 99.0077 Error = 3.0 ppm
Thus, the accurate mass gives the same formula match that is consistent with the identification by ion trap. Furthermore, Figure 6b also shows the accurate mass ions for the possible fragmentation of the unknown at masses of m/z 353.0256, 285.0017, 127.0389, and 99.0081. These accurate mass ions are the same ones measured with the ion trap in MS/MS mode of the m/z 331 at nominal mass (with the exception of the m/z 353 adduct). The masses of these ions match the formulas shown in Figure 5 including the sodium adduct of the m/z 331. Thus, the probability of this being the correct identification is quite high. The final step is the identification by standard matching. This was done by both ion trap MS/MS and by TOF; both gave this compound as the correct identification. Malathion is a banned substance for cucumbers so this identification represents an important finding. Based on the identification point scheme for unknowns, this compound receives 2.0 points for the protonated molecule, 2.0 points for the m/z 285 ion, and 1.5 points for the MS/MS transition from m/z 331 to 285 for a total of 5.5 points. Four identification points are required for banned substances in food (see reference by Stolker et al. [5] in book by Ferrer and Thurman [6]). The complementary nature of the two instruments is also shown with the next unknown example.
Figure 5.
Fragmentation possibilities for malathion.
The next step is obtaining the accurate mass with the LC/MSD TOF. Figure 6a shows the TIC and the nearly identical retention times between the two instruments for the unknown Star*TIC because of using the same HPLC column. Figure 6b shows the accurate mass for the m/z 331.0435 and only one formula match of C10H19O6PS2.
OH 1.6e5 1.5e5 1.4e5 1.3e5 1.2e5 1.1e5 1.0e5 Intensity, counts 9.0e4 8.0e4 7.0e4 6.0e4 5.0e4 4.0e4 3.0e4 2.0e4 1.0e4 333.0404 O 127.0389 O H3CO H3CO S P S O OH 257.0066 353.0256 99.0081 HO H3CO O HO S P S O H3CO H3CO O S P S O Na O O O CH3 CH3 331.0435
Accurate Mass of 331 m/z 331.0435 Formula matches...C10H19O6PS2
S H3CO H3CO P S O
O O
CH3 CH3
285.0017
O H H3CO
60
80
100
120
140
160
180
200
220
240 m/z, amu
260
280
300
320
340
360
380
400
Figure 6b. LC/MS TOF spectrum of peak at 22.5 min (m/z 331).
Monograph Number: 5723 Title: Malathion CAS Registry Number: 121-75-5 CAS Name: [(Dimethoxyphosphinothioyl)thio]butanedioic acid diethyl ester
H3CO H3CO S P S
O O
CH3 CH3
Additional Names: diethyl mercaptosuccinate S-ester with O,O-dimethyl phosphorothioate; S-(1,2dicarbethoxyethyl) O,O-dimethyldithiophosphate; insecticide no. 4049; carbofos; malathon (obsolete); mercaptothion; phosphothion Manufacturers' Codes: ENT-17034 Trademarks: Cythion; Derbac-M (International); Malamar 50; Malaspray; Organoderm (Mundipharma); Prioderm (Coates & Cooper); Suleo-M (International) Molecular Formula: C10H19O6PS2
Molecular Weight: 330.36. Percent Composition: C 36.36%, H 5.80%, O 29.06%, P 9.38%, S 19.41% Literature References: Cholinesterase inhibitor. Prepn: Johnson et al., J. Econ. Entomol. 45, 279 (1952); Cassaday, US 2578652 (1951 to Am. Cyanamid). Purification: Usui, US 2962521 (1960 to Sumitomo). Treatment of Pediculus humanus (head lice) infestation: D. Taplin et al., J. Am. Med. Assoc. 247, 3103 (1982). Toxicity data: T. B. Gaines, Toxicol. Appl. Pharmacol. 14, 515 (1969). Review of distribution, transport and fate in the environment: M. S. Mulla et al., Residue Rev. 81, 1-159 (1981); of carcinogenic risk: IARC Monographs 30, 103-129 (1983). Properties: Deep brown to yellow liq, mp 2.9. bp 156-157. Characteristic odor. d 25 1.23. n 25 1.4985. Vapor pressure at 30: 4 10-5 mm Hg. Slightly sol in 4 D and alkylated aromatic hydrocarbons and vegetable oils. water (145 ppm). Misc with many organic solvents 0.7 including alcohols, esters, ketones, ethers, aromatic Limited soly in certain paraffin hydrocarbons. Petroleum ether is sol to about 35% in malathion. Hydrolyzed at pH >7.0 or <5.0. Stable in an aq soln buffered to pH 5.26. LD in female, male rats (mg/kg): 1000, 1375 orally (Gaines). 50 Melting point: mp 2.9 Boiling point: bp 156-157 0.7 Index of refraction: n 25 1.4985 D Density: d 25 1.23 4 Toxicity data: LD in female, male rats (mg/kg): 1000, 1375 orally (Gaines) 50 CAUTION: Potential symptoms of overexposure are miosis, aching eyes, blurred vision and lacrimation; eye and skin irritation; salivation; anorexia, nausea, vomiting, abdominal cramps, diarrhea, giddiness, confusion and ataxia; rhinorrhea, headache; tightening of chest, wheezing and laryngeal spasms. See NIOSH Pocket Guide to Chemical Hazards (DHHS/NIOSH 97-140, 1997) p 188. Use: Insecticide. Therap-Cat: Pediculicide. Therap-Cat-Vet: Ectoparasiticide.
Figure 7.
Database search for C10H19O6PS2. 7
Tomato Extract In this next example, a similar protocol is followed except that this time we use the LC/MSD TOF first to obtain an in-source CID spectrum. Figure 8a shows the Star*TIC for the rapid extraction of the tomato white powder.
23.9 3.0e7 2.5e7 2.0e7 Intensity, cps 14.7 1.5e7 1.0e7 2.2 5.0e6 0.0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Time, min 8.3 14.5 11.3 18.5 24.9 3.2 2.0 26.0 26.7 30.4
The accurate mass is m/z 306.1642, which results in two possible formulae after examination of the A+2 isotope pattern, which shows one sulfur atom in the molecule. Furthermore, the A+1 has an area relative to the m/z 306 of 17%, which indicates 15 carbon atoms. See Figure 8b.
4.0e5 3.0e5 Intensity, counts 2.0e5 1.0e5 306.6200 0.0 280 290 300 310 m/z, amu 307.1664 A+117%~C15 308.1624 A+25%S1 320 330 306.1642 m/z 306.1642 C16H23N3OS C14H28N O2PS 2.4 ppm -3.0 ppm
Star*TIC
Figure 8b. LC/MS TOF spectrum of peak at 23.9 min (m/z 306).
Figure 8a. LC/MS TOF analysis of a tomato extract.
The two formulae were both then searched in the Merck Index and only one formula gave a database hit. That compound was C16H23N3OS, the insect growth regulator buprofezin. Buprofezin is used extensively on white flies according to the Merck Index (Figure 9). Thus, this compound was a good candidate for a positive identification.
Monograph Number: 1486 Title: Buprofezin

N N O CH3 H3C S N C(CH3)3
CAS Registry Number: 69327-76-0
CAS Name: 2-[(1,1-Dimethylethyl)imino]tetrahydro-3-(1-methylethyl)-5-phenyl-4H-1,3,5-thiadiazin-4-one Additional Names: 2-tert-butylimino-3-isopropyl-5-phenylperhydro-1,3,5-thiadiazin-4-one Manufacturers' Codes: NNI-750; NNK-758; NN-29285; PP-618 Trademarks: Applaud (Nihon Nohyaku) Molecular Formula: C16H23N3OS
Molecular Weight: 305.45. Percent Composition: C 62.92%, H 7.59%, N 13.76%, O 5.24%, S 10.50% Literature References: Insect growth regulator which inhibits chitin synthesis. Prepn: Z. Grnecker et al., DE 2824126; K. Ikeda et al., US 4159328 (1978, 1979 both to Nihon Nohyaku); H. Kanno, Pure Appl. Chem. 59, 1027 (1987). Mode of action study: T. Asai et al., Appl. Entomol. Zool. 20, 111 (1985). Control of whiteflies and scale insects: I. Ishaaya et al., Meded. Fac. Landbouwwet., Univ. Gent 54, 1003 (1989). GC-MS determn in clementine citrus: P. Cabras et al., J. Agr. Food Chem. 46, 4255 (1998). Persistence in olives and olive oil: idem et al., Food Addit. Contam. 17, 855 (2000). Review of physical properties, activity and field trials: H. Kanno et al., Proc. Br. Crop Prot. Conf. - Pests Dis. 1981, 59-66. Properties: Crystals from isopropyl alcohol, mp 106.1. Soly at 25 (g/l): acetone 240, chloroform 520, ethanol 80, toluene 320; water 0.9 mg/l. Vapor pressure at 25: 9.4 10-6 mmHg. LD in mice, rats (mg/kg): 10000, 8740 orally. LC (48 hr) in carp: 2-10 mg/l (Kanno, 1981). 50 50 Melting point: mp 106.1 Toxicity data: LD Use: Insecticide. 50 in mice, rats (mg/kg): 10000, 8740 orally; LC 50 (48 hr) in carp: 2-10 mg/l (Kanno, 1981)
Results of Library SearchC14H28NO2PSNo hit. Molecular Formula = C16H23N3OS

Figure 9. Database search for C14H28NO2PS and C16H23N3OS. 8
Note in the spectrum the presence of the ion at m/z 201.1059 (Figure 10). LC/MSD ion trap MS/MS of the m/z 306 ion gave the m/z 201 and the further MS3 yielded the m/z 116 ion.
Intens. 107 1.0 306.1656 0.5 0.0 23.00 Intens. 6 10 3 2 1 115.9 1.0e5 201.1059 116 0.0 100 200 300 400 0 MS3 gives 116
4.0e5
23.50
24.00
Intensity, counts
3.0e5
24.50 25.00 Time [min]
200.9 MS/MS gives 201
2.0e5
Furthermore, the expected mass for the 201.1056 fragment ion matched the value from the LC/MSD TOF (201.1059) quite closely (0.0003 u), which gave a high certainty for identification. After obtaining the buprofezin standard, the final data show a perfect match, which further shows the ability of the LC/MSD TOF and LC/MSD ion trap to identify unknowns. Buprofezin is allowed for tomatoes; therefore, it is not a banned substance. The quantitative tolerance for buprofezin may then be measured by the standard extraction and analysis for vegetable acceptance to European Union standards [1].
100 200 300 400 500 600 700 m/z Formula = C9H17N2OS 500 600 m/z, amu 700 800 900 1000
Conclusions
LC/MSD TOF and LC/MSD Trap are complementary and powerful tools for identification of pesticides in vegetables and represent a new approach for environmental food chemistry using LC/MS. The combination of these two tools, the twin mass spectrometer(s), with a pesticide database, Merck Index or ChemINDEX, works often for identification of unknown pesticides. The use of identification of fragment ions with MS2 and MS3 is also powerful when combined with accurate mass of LC/MSD TOF CID spectra. Combining TOF and Trap for unknown ID works [4, 7]! Future work should include chromatographic data to help in the identification of unknowns and to provide a simple pesticide library generated through calculation of empirical formula and isotope ratios.
Figure 10. LC/MS TOF spectrum and LC/MS/MS Ion Trap spectrum for m/z 306.
It was possible to draw reasonable chemical structures for the m/z 201 and 116 fragment ions (Figure 11).
Buprofezin m/z 306 S N N O H3C CH3 N C(CH3)3
MS2
S N N O H3C 201 m/z fragment ion C9H17N2OS Exact mass: 201.1056 CH3 C(CH3)3
MS3
H N m/z 116 C(CH3)3
Conclusion: Buprofezin Standard confirmationyes.
Figure 11. Reasonable structures of fragment ions consistent with the structure of buprofezin.
References
1. EU Food Directives, 2002, 91/414/EEC 2. Food Quality Protection Act, 1998 (FQPA) 3. Aguera, A.; Lopez, S.; Fernandez-Alba, A.R.; Contreras, M.; Crespo, J.; Piedra, W., 2004, J. Chromatogr. A., 1045: 125-135. 4. Imma Ferrer and E. Michael Thurman, "Determination of Chloronicotinyl Insecticides in Salad Vegetables by LC/MS/ESI/TOF and LC/MS Ion Trap", Agilent Technologies publication 5989-1842EN, www.agilent.com/chem 5. A.A.M Stolker, E. Dijkman, W. Niesing, E.A. Hogendoorn, 2003, "Identification of residues by LC/MS/MS" In: Imma Ferrer and E. Michael Thurman, editors Liquid Chromatography/Mass Spectrometry/Mass Spectrometry and Time-ofFlight Mass Spectrometry for the analysis of emerging contaminants, (2003) American Chemical Society Symposium Volume 850. 6. Imma Ferrer and E.M. Thurman, (2003), Liquid Chromatography/Mass Spectrometry/Mass Spectrometry and Time-of-Flight Mass Spectrometry for the analysis of emerging contaminants, American Chemical Society Symposium, 850. 7. E. Michael Thurman, Imma Ferrer, A. R. Fernandez-Alba, (2005), "Matching Unknown Empirical Formulas to Chemical Structure Using LC/MS TOF Accurate Mass and Database Searching: Example of Unknown Pesticides on Tomato Skins" Journal of Chromatography, Special Issue on Mass Spectrometry, In press.
Acknowledgments Amadeo Fernandez-Alba, Department of Hydrogeology and Analytical Chemistry, University of Almera for analytical support. ChemOffice with Chemfinder, the Merck Index, and ChemIndex are products of CambridgeSoft. ZORBAX is a registered trademark of E.I. duPont de Nemours Co. Inc. Millex is a registered trademark of the Millipore Corp. Merck Index is a registered trademark of Merck. Pyrex is a registered trademark of Corning Corporation. Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2005 Printed in the USA February 18, 2005 5989-1924EN
Analysis of Components, Contaminants, and Impurities in Fungicide Formulations by GC/MS and LC/MS Application
Agriculture, Specialty Chemical, Environmental, Ag Chem
Author
James D. McCurry Paul Zavitsanos Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Abstract
A commercially available fungicide formulation was analyzed by both gas chromatography/mass spectrometry (GC/MS) and electrospray ionization liquid chromatography/mass spectrometry (ESI-LC/MS). The GC/MS analysis provided a detailed look at the volatile components in the formulation, but did not yield any results for the active ingredient, triforine. The ESI-LC/MS provided information on the stereoisomers of triforine as well as the nonvolatile surfactants and contaminants in the formulation. This paper demonstrates the complementary nature of these two analytical techniques when trying to fully characterize a complex chemical formulation containing a broad range of components.
unexpected impurities and breakdown products that can affect product quality. However GC/MS can only provide meaningful information for compounds that are volatile, nonionic, thermally stable, and have relatively low molecular weight. Liquid chromatography is much better suited to analyzing compounds that are nonvolatile, ionic, polar, thermally labile, or have high molecular weight. This includes about 80% of all known organic compounds [1]. When coupled with a modern atmospheric pressure ionization (API) mass spectrometer, LC/MS offers a complementary tool to GC/MS in the chemical diagnostic laboratory. Commercial pest control formulations contain one or more active compounds along with a recipe of ingredients that can play an important role in the products efficacy. These inactive ingredients are often a combination of solvents and surfactants that allow for easy application and dispersal of the active ingredient onto the target substrate. For this work, an over-the-counter fungicide formulation was purchased at a local home products store. The active ingredient in this product is 6.5 % (wt) of N,N-[1,4-piperazinediylbis(2,2,2-trichloroethylidene)] bisformamide. This is also known as triforine (CAS registry number 26644-46-2), and the structure is shown in Figure 1. The inactive ingredients in this formulation are listed as cyclohexanone, N-methyl pyrrolidone, and Atlox 3406-F. The Atlox 3406-F is an agricultural dispersant that contains ionic and nonionic surfactants and mixed aromatic solvents.
Introduction
Gas chromatography/mass spectrometry (GC/MS) is an indispensable tool for solving complex problems in the chemical industry. This fast and powerful technique yields detailed information about the expected compounds in the mixture along with any
Cl Cl * Cl NH O
Table I.
GC/MS Analysis Conditions
Gas chromatograph conditions Column: Carrier gas: Flow rate: Inlet: 30 m 0.25 mm HP5-MS, 0.25 m (p/n 19091S-433) Helium at 13.00 psi 1.6 mL/min., constant flow mode Cool on-column at 50 C, oven track mode
N O HN Cl Cl N * Cl
*Optically active
Oven temperature program: 50 C for 3 min 10 C/min to 275 C 275 C for 4 min MS Transfer line: 280 C 1 L 1400 V 3 min 30 to 800 m/z 50 counts 2 1.95 scans/s Injection volume: Electron multiplier: Solvent delay: Scan range: Scan threshold: A/D Samples: Scan rate
Figure 1.
Chemical structure of triforine, the active ingredient in some commercial fungicides. The nominal molecular weight is 432, and the structure contains two optically active carbons.
A complete analysis of this formulation requires GC/MS to separate and identify the volatile components and LC/MS for the surfactants and polar components. Analysis of the active ingredient, triforine, presents a separate challenge. References for triforine analysis cite gas chromatography as the method of choice when analyzing environmental residues [2]. However, the melting point is reported to be 155 oC with decomposition, indicating that gas chromatography may only be possible with on-column injection.
Electrospray Ionization Liquid Chromatography/Mass Spectrometry (ESI-LC/MS) The same fungicide sample was run on the Agilent 1100 Series LC/MSD. This system included a vacuum degasser, a binary pump, an autoinjector, a thermostatted column compartment, and the LC/MSD SL quadrupole mass spectrometer. LC/MS instrument conditions for this analysis are shown in Table 2.
Experimental
Gas Chromatography/Mass Spectrometry (GC/MS) A 1% (v/v) solution of the triforine formulation was made in acetonitrile and the GC/MS analysis was performed with an Agilent 5973 GC/MS system. The components in this system were a 6890N gas chromatograph, a 7683 autoinjector, and a 5973 mass spectrometer. A cool-on-column inlet in the Agilent 6890 GC was used to avoid decomposition of the triforine. Instrument conditions for the GC/MS analysis are listed in Table 1.

Gas Chromatography/Mass Spectrometry (GC/MS) The complex nature of this fungicide formulation is revealed when one looks at the GC/MS data. Figure 2 shows the total ion chromatogram (TIC) of the fungicide sample. The volatile components
Table 2.
LC/MS Analysis Conditions
Liquid chromatograph conditions Column: Mobile phase A: Mobile phase B: Mobile phase gradient: Flow rate: Column temperature: Injection volume: Source: Drying gas flow: Nebulizer: Drying gas temperature: Vcap: Stepsize: Peak width: Time filter: Scan range Fragmentor 150 4.6 mm Zorbax XDB-C8, 5 m (p/n 993967-906) 0.1% Formic acid in water 0.1% Formic acid in acetonitrile 30% B at 0 min; 50% B at 7 min; 95% B at 10 min 1.0 mL/min 30 C 1 L Electrospray 12 L/min 40 psig 350 C 3500 V (positive) and 3000 V (negative) 0.1 amu 0.1 min On 120 to 1200 m/z Fixed at 60 V
in the formulation are easily identified from the mass spectral data. The major solvents, cyclohexanone and N-methyl-2-pyrrolidone, dominate the chromatogram while smaller amounts of C9 aromatics, C10 aromatics, and substituted napthalenes are easily separated and identified. There were no peaks in the TIC whose spectra matched the triforine reference spectra from the Wiley mass spectral library. An extracted ion profile using the triforine base peak of 203 m/z did not produce any chromatographic peak indicating the presence of triforine. From this data, it appears that the triforine did not elute from the column into the mass spectrometer. However, a spectral average of the large hump between 18 and 20 minutes shows an isotope pattern indicating one chlorine atom (Figure 3A). Since no chlorinecontaining species other than triforine are components in the formulation, the presence of chlorine and the broad peak shape indicates triforine decomposition in the gas chromatograph. The peak at approximately 20-minute retention time also has a mass spectrum containing an isotope pattern indicating the presence of two chlorine atoms in the structure (Figure 3B). This peak could be a decomposition product or a contaminant in the formulation.
5 2 4 1 2 3 4 5 6 7 1-hexanol Cyclohexanone C9-aromatics N-methyl-2-pyrrolidone C10-aromatics Naphthalenes Triforine decomposition
6 3
10
12
14
16
18
20
Figure 2.
GC/MS TIC showing the complex volatile components in the commercial fungicide formulation.
145 158 187 152
215
140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320 m/z
288
144
140 150
157
160 170 180
189
190 200 210
225
220 230 240 250
259
260 270 280 290 300 310 320 m/z
Figure 3.
(A) Average mass spectrum of broad hump between 18 and 20 minutes of TIC. Isotope patterns of the peaks at m/z 145, 158 and 187 indicate the presence of one chlorine atom. (B) Mass spectrum of the peak at 20 minutes shows the presence of two chlorine atoms in the structure.
Electrospray Ionization Liquid Chromatography/Mass Spectrometry (ESI-LC/MS) The positive ion ESI-LC/MS chromatogram is shown in Figure 4. Several major peaks are observed along with several minor components. The spectra of the three peaks eluting between 0 and 2 minutes are shown in Figure 5. Since electrospray is a soft ionization technique, these spectra do not exhibit the detailed fragmentation
needed to interpret structures for these three compounds. However, peak number 2 does have an isotopic pattern indicating the presence of two chlorine atoms in the structure. This compound could be a contaminant related to triforine production or a triforine decomposition product. Figure 6 shows the spectra of the three peaks between 10.5 and 13 minutes. These compounds are the various surfactants that make up the agricultural dispersant used in the formulation.
3 8
2 6
10
12
Figure 4.
TIC from positive ion ESI-LC/MS of fungicide formulation.
122
167
Peak 1
126 146 185 208
200 m/z
120
140
160
180
260 262
Peak 2
264
250 275
300
325
350
375
m/z
199
Peak 3
200
180 190 200
210
220
m/z
Figure 5.
Electrospray spectra from LC/MS peaks 1, 2, and 3. The spectra from peak 2 shows an isotope pattern indicating two chlorine atoms in this structure. This compound may be a contaminant in the formulation from the active ingredient triforine.
Peak 6
600
700
800
900
m/z
Peak 7
400
600
800
1000
m/z
Peak 8
400
500
600
700
800
m/z
Figure 6.
Electrospray mass spectra of LC/MS peaks 6, 7, and 8 from Figure 4. These compounds are the surfactants used in the formulation.
The spectra of LC/MS peaks 4 and 5 (Figure 7) are identical and correspond to the active ingredient, triforine. The protonated molecular ion is observed at m/z 433 along a sodium adduct at m/z 455. The multiplets for m/z 433 to 439 and m/z 455 to 461 exhibit an isotopic pattern consistent with six chlorine atoms. The ion at m/z 388 is due to a rearrangement and subsequent loss of a formamide group from the protonated molecular ion (m/z 433). This is also confirmed by the isotopic pattern indicating six chlorine atoms (m/z 388 to 396). The presence of two triforine peaks in Figure 4 can be explained by the stereochemistry of the structure. Triforine contains two optically active carbons that give rise to four stereoisomers. Figure 8 shows the four configurations that can be grouped into two pairs of mirror images that are diastereomers. The S,R and R,S configurations are mirror
images that are superimposable, resulting in a meso compound that exhibits no optical activity or differences in physical properties. Therefore, because the S,R and R,S configurations are identical, they will elute as one chromatographic peak. The second pair of mirror images are the R,R and S,S configurations. These are not superimposable and are, therefore, enatiomers that will exhibit different optical activity, but identical physical properties. Conventional reverse-phase liquid chromatography cannot separate these enantiomers, and they will co-elute as a single peak. However, these enantiomers are not mirror images of the meso compound and can be chromatographically separated from the meso compound. This is why there are two triforine peaks, one for the meso compound and one for the enatiomers. Without pure standards of the stereoisomers, it is not possible to determine which configurations can be attributed to the observed chromatographic peaks.
(M+H)+ -43
Peak 4
7.554 min
(M+H)+ (M+Na)+
390 392
Peak 5
7.757 min
388
394
433 435 437 439
400
420
440
457 459 461

460 m/z
Figure 7.
Electrospray mass spectra of peaks 4 and 5 from Figure 4. Both spectra show a protonated molecular ion at m/z 433 representing the active ingredient triforine. There is also a sodium adduct (m/z 455) of triforine observed for both peaks. A rearrangement and loss of a formamide group from the protonated molecular ions give rise to the multiplet at m/z 388 to 396. Mirror Mirror
O HN H
(S)
396
O NH HN H Cl3C
(R)
O NH
CCl3
Cl3C
(R)
(S)
CCl3
N H
(R)
N CCl3 Cl3C
(S)
N H H
(R)
N CCl3 Cl3C
(S)
NH O (S,R)
HN O (R,S) O
NH
HN O
(R,R)
(S,S)
Meso structure no optical activity Figure 8.
Enantiomers optically active
The four triforine stereoisomers arising from the two chiral carbons in the structure. These two pairs of mirror images account for the two triforine peaks observed in the chromatogram (Figure 4). 7
The fungicide formulation was also run by ESI-LC/MS in the negative ion mode. The results of this analysis are shown in Figure 9. The negative ion mass spectra for these two peaks are shown in Figure 10. For both triforine peaks, the most stable negative ion species is the chloride adduct (m/z 467). However, the spectra for the first peak contains a
7.554
deprotonated molecular ion (m/z 431) and a formate adduct (m/z 477) that is not observed in the spectra of the later eluting peak. This selective adduct formation is likely related to the stereochemstry of the triforine, but again, without pure standards, the correct configurations cannot be assigned to the chromatographic peaks.
7.757
10
12
Figure 9.
TIC from negative ion ESI-LC/MS of fungicide formulation.

_ (M+Cl)
469 471
RT = 7.554 min
(M+HCO2)
479 _
(M+TFA)
_ 545 547
467
473
477
483
(M_H)
(M+NH4CO2)
494 494 494 494
431 433 435 437
RT = 7.757 min
425
450
475
500
525
550
551 m/z
Figure 10. Negative ion electrospray mass spectra of the two triforine peaks. The spectra from peak at 7.554 minutes shows a deprotonated molecular ion (m/z 431) and a formate adduct (m/z 477) that is not seen in the later eluting peak (7.757 minutes).
549
481
Conclusions
This paper demonstrates the complimentary nature of GC/MS and LC/MS when trying to characterize a formulation that is composed of many different chemical species. The volatile compounds in the formulation can be easily separated and identified by GC/MS. In this case, polar solvents such as cyclohexanone and N-methyl-2-pyrrolidone were the major components while 1-hexanone, C9 aromatics, C10 aromatics, and substituted naphthalenes were present as minor components or contaminants. However, GC/MS did not yield any information on the active fungicidal ingredient, triforine, a hexachlorinated compound. This was most likely due to thermal decomposition during GC/MS analysis. Evidence for this was seen in a broad chromatographic hump containing chlorine-containing constituents. The nonvolatile components in this fungicide were quickly analyzed by ESI-LC/MS. This analysis yielded information on several polar contaminants, some containing chlorine, which may be by-products of triforine production or triforine breakdown products. Also observed were several surfactants that are used in agricultural products as dispersants. The LC/MS analysis did yield significant information on the triforine active ingredient, showing a distribution of stereoisomers in the formulation.
References
1. Willoughby, R., Sheehan, E, and Mitrovich, S. A., Global View of LC/MS: How to Solve your Most Challenging Problems, p. 80, Global View Publishing, 1998. 2. The Pesticide Manual 9th Edition, Worthing, C.R. and Hance R.J. eds., p. 853, The British Crop Protection Council, 1991.

Determination of Acidic Herbicides in Groundwater and Potable Water by LC/MSD Using Selective Ion Monitoring
Application
Environmental
Authors
Neil Cullum Anglian Water Laboratories Huntingdon UK Paul Stephens and Stan Evans Agilent Technologies, Ltd. Bracknell UK
The Prescribed Concentration or Value (PCV) for an individual herbicide in drinking water in the UK, as defined by the Water Supply (Water Quality) Regulations, is set at 0.1 g/L. Ideally, the method of analysis is capable of detecting 10 to 20% of the PVC, that is, 0.01 to 0.02 g/L. This method clearly attains these lower limits of detection. Standard analytical methods for these herbicides in water matrices involve either liquid-liquid extraction or solid phase extraction (SPE), followed by a derivatisation using diazomethane or pentafluorobenzyl bromide and analysis by gas chromatography (GC) or gas chromatography/ mass spectrometry (GC/MS). The derivatisation step involves use of potentially hazardous chemicals. This application note gives details of a validated liquid chromatography/mass spectrometry (LC/MS) method requiring no derivatisation step and which uses a relatively small volume of sample.
Abstract
A validated liquid chromatography/mass spectrometry method for acidic herbicides in groundwater and potable water is described using atmospheric pressure electrospray ionisation in negative ion mode and selective ion monitoring. No derivatisation step is required and only small volumes of sample are used. Detection levels of 10 to 20% of the Prescribed Concentration or Value for an individual herbicide in drinking water in the UK are attained.
Experimental
All analysis was performed using an Agilent 1100 series LC/MS quadrupole system coupled to an Agilent 1100 series LC system consisting of a quaternary pump, autosampler, thermostated column compartment and vacuum degasser. A diode array detector, in-line before the mass spectrometer was used as a troubleshooting tool. The quadrupole mass spectrometer was operated with the Agilent atmospheric pressure electrospray ionisation (API-ES) source in negative ion mode.
Introduction
A validated method is described for the analysis of 19 acidic herbicides and one amide herbicide, Propyzamide, as a single analytical suite. The various herbicides cannot be characterized as a single class but belong to several different groups: chlorophenoxy acids such as mecoprop, are well represented.
Sample Preparation 1. Fully automated SPE used Oasis HLB, 60 mg, 3 mL cartridges and a series of solutions: #1) 90:10 TMBE: methanol (0.01% formic acid), #2) methanol (0.01% formic acid), #3) HPLC grade water, and #4) HPLC grade water (0.25% hydrochloric acid). 2. Each cartridge was initially conditioned with sequential additions of 5 mL of solution #1, 5 mL of solution #2, 3 mL of solution #3, and 3 mL of solution #4. 3. An aqueous sample of 50 mL was diluted to 200 mL with deionised water and acidified with 0.75 mL of hydrochloric acid. 4. A diluted sample of 100 mL was pumped through the conditioned cartridge at 10 mL/min. 5. Cartridge was dried 10 minutes, using forced air. 6. Cartridge was eluted with solution #1 twice using 1.5 mL and once using 1 mL. 7. The extract was evaporated to dryness in a 45 C heated block, using a gentle air stream. 8. Residue was dissolved in 250 L of solution #1.
LC conditions
Column: Zorbax Eclipse XDB-C18, 150 mm long, 2.1 mm id, 3.5 m particles, 60 C Pre-column quaternary pump Mobile phase A: Mobile phase B: Gradient program: Initial 37.5 min 38.0 min 48.0 min Pre-column flow rate: 0.3 mL/min Sample size injected: 50 L A 90% 43% 90% 90% B 10% 57% 10% 10% 0.01% Formic acid in water Acetonitrile
Binary pump with diode array detector (DAD)
Instrument: Agilent 1100 LC/MS with API-ES in Negative Ion Mode

Drying gas temperature and flow: Nebulizer gas pressure: Vcap: Fragmentor voltage: SIM ions: 300 C, 11 L/min 30 psig 2500 Volts Variable, see Table 1 See Table 1
Table 1.
SIM Parameters SIM ions Quantitation, Qualification (q) 178.9, 180.9 q 190.1, 192.0 q 238.9, 240.9 q 260.1, 261.1 q 174.9, 177.0 q 169.9, 171.9 q 195.0, 196.9 q 239.0, 240.1 q 259.0, 261.0 q 199.0, 201.0 q 219.0, 220.9 q 275.9, 278.0 q 160.9, 162.9 q 195.0, 196.9 q 195.9, 198.0 q 213.0, 215.0 q 369.9. 371.0 q 160.9, 163.0 q 227.1, 229.1 q 254.1, 256.0 q
Compound 2,3,6-TBA Clopyralid Picloram Imazapyr Dicamba Benazolin Fluroxypyr Bentazone Bromacil MCPA 2,4-D Bromoxynil Dichloroprop 2,4,5-T Triclopyr Mecoprop Ioxynil 2,4-DB MCPB Propyzamide
Time 1.00
Group number 1
Fragmentor voltage 90 75 120 75 85 140 145 140 135 150 150 140 150 150 150 150 145 150 70 120
12.0 17.0
2 3
24.5
28.0
33.5 36.25
6 7
The SIM ions and fragmentor voltages listed in Table 1, page 2, were all optimized using Flow Injection Analysis (FIA). Standard solutions of 10 mg/L of each herbicide were injected using scan mode range 150 to 400 amu, and fragmentor voltage ramped from 70 to 150 V in 5 V steps. A typical FIA pattern for dicamba appears in Figure 1, and the peak area relationship appears in Figure 1A.
5 4 Signal 3 2 1 0
70 V
80 V
90 V
120 V
110 V
120 V
130 V
140 V
150 V
Figure 1.
Dicamba TIC versus fragmentor voltage.
14.0
12.0
10.0 EIC Peak area, 10-6
8.0
6.0
4.0
2.0
0.0 70 80 90 100 110 Fragmentor voltage 120 130 140 150
Figure 1A. Dicamba 174.9 m/z area vs. fragmentor voltage.
Using the FIA data, the fragmentor voltage generating the maximum response was selected. For dicamba, this corresponded to 8085 V. The mass spectrum of each herbicide was obtained at the optimum fragmentor voltage. The mass spectrum of dicamba, which has a molecular weight of 220, appears in Figure 2. This shows the M-H ion at 219 with the expected two chlorine isotope pattern, but with a much stronger 174.9 ion, corresponding to the additional loss of CO2. For maximum sensitivity the 174.9 ion was chosen for quantitation and the 177.0 ion as the qualifier ion (q). This process was repeated for all 20 herbicides.
MHCO2
177.0
174.9
219.0 178.9
MH
176.1
220.9 220.0 223.1
200
250
m /z
Figure 2.
Mass spectrum of dicamba at 85 fragmentor voltage.
Results
A typical chromatogram for the low level standard is shown in Figure 3. This is equivalent to a concentration of 0.1 g/L.
14
250000
200000
150000
1 2 3 4 5 6 7 8 9 10
Clopyralid Picloram Imazapyr 2,3,6-TBA Dicamba Benazolin Fluroxypyr Bromacil Bentazone Bromoxynil
11 12 13 14 15 16 17 18 19 20
2,4-D MCPA Triclopyr Ioxynil Dichloroprop 2,4,5-T Mecoprop 2,4-DB MCPB Propyzamide
15, 16
10 13
100000
8 1
17 18 19
50000
3 2
4 5
6 7 11 12
20
0 5 10 15 20 25 30 35 min
Figure 3.
Low level standard chromatogram equivalent to 0.1 g/L.
Figure 4 shows an extracted ion chromatogram from a real sample found to contain bentazone at a concentration of 0.08 g/L. The mass spectrum is also shown.
140000
SIM ion: 239.0 m/z

100000
80000 240.1 180 200 220 240
60000
239.1
120000
22.852 - Bentazone
40000
m /z
20000
0 5 10 15 20 25 30 35 40 45 min
Figure 4.
Borehole sample containing 0.08 g/L bentazone.
Calibration curves were also produced using three calibration levels at 0.1, 0.3 and 0.5 g/L. The calibration curves are all produced using a quadratic fit and forced through the origin. Typical correlation values are 0.9999 or better for all the herbicides in the suite. A typical calibration curve is shown in Figure 5.
Area
1600000
1
1400000
2,4 - DB 161 m/z
1200000
1000000
2
800000
600000
400000
Correlation: 0.99999 3
200000
0 0 0.1 0.3 Amount [g/L] 0.5
Figure 5.
Typical calibration plot.
Validation of the method was carried out on 11 batches of samples. The borehole groundwater was spiked at three levels: 0.01 g/L, 0.10 g/L, and 0.40 g/L. The potable tap water (which was from a surface water source) was spiked at two levels: 0.01 g/L and 0.10 g/L. Each batch of samples was analysed in duplicate and in a random order. The limit of detection (LOD) for each herbicide was calculated from the within-batch standard deviation of the borehole sample spiked at 0.01 g/L. Recovery for both the groundwater and potable water samples was calculated from the 0.10 g/L spike after subtraction of the 0.01 g/L standard. Hence the recovery value is based on 0.09 g/L. Experimental results are shown in Table 2.
Table 2.
Comparison of Spike Recoveries from Groundwater and Potable Water Samples Groundwater samples Recovery % RSD % 89.4 10.1 67.1 9.4 73.1 11.0 93.4 6.4 92.6 6.5 83.7 7.4 84.8 7.6 89.2 7.9 106.8 4.1 99.5 4.2 85.1 8.8 91.5 8.1 97.7 7.4 100.2 6.3 95.8 6.1 86.7 7.3 96.6 5.9 90.7 6.3 93.2 5.8 86.8 5.9 90.2 7.1 Potable water samples Recovery % RSD % 61.1 12.9 67.7 9.3 47.5 13.6 83.5 9.0 83.0 7.5 67.6 7.5 72.7 7.6 61.6 9.2 98.6 4.9 94.0 5.1 76.7 6.1 83.2 4.9 85.0 6.8 99.8 4.5 89.8 4.4 83.0 5.7 93.9 4.6 81.6 5.7 84.2 7.9 87.5 5.3 80.1 7.1 LOD g/L 0.00658 0.00477 0.00631 0.00335 0.00882 0.00629 0.00694 0.00702 0.01034 0.00666 0.00603 0.00526 0.00704 0.00653 0.00545 0.00562 0.00593 0.00488 0.00647 0.00574 0.00630
Compound Clopyralid Picloram Imazapyr 2,3,6-TBA Dicamba Benazolin Fluroxypyr Bromacil Bentazone Bromoxynil 2,4-D MCPA Triclopyr Ioxynil Dichloroprop 2,4,5-T Mecoprop 2,4-DB MCPB Propyzamide Average
Discussion
The data produced clearly show that there is a difference in recovery between the two sample matrices tested. This is in part due to different extraction efficiencies from the two different water types, with the potable water in general showing a lower recovery. This potable water sample typically has a Total Organic Carbon (TOC) content of about 4 mg/L, compared to the borehole sample with a typical TOC of about 0.5 mg/L. The higher organic content of the potable water may well lead to slightly lower spiked recoveries. The other factor involved, especially with the earlier eluting compounds (for example, clopyralid and imazapyr), is suppression of the ionisation. This occurs where there is competition for ionisation in the spray chamber from other compounds in the sample. In the case of potable waters, in particular waters derived from surface water sources, these are likely to be due to humic and fulvic acids, which elute very early in the analytical run. Hence ionisation suppression will lead to an apparent reduction in recovery values obtained. Consequently, when analyzing samples routinely, the
recovery correction should be made from reference values obtained from an appropriate matrix, or adjusted by the use of an internal standard(s) where available.
Conclusion
The data shows that the method presented is capable of quantitative analysis for the 20 herbicides in single analytical suite. The performance requirements set by the Drinking Water Inspectorate (DWI) for standard deviation, bias and total error are all met. Although spiked recovery targets of 90 to 110% are not achieved in all cases, this can be compensated for by the application of recovery factors which are calculated from the performance data. The method was granted UKAS accreditation at the laboratory where the method was validated.

Analysis of Glyphosate and Aminomethyl Phosphonic Acid by Liquid Chromatography/Mass Spectrometry
Application
Environmental
Authors
Paul Zavitsanos* and Chin-Kai Meng Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA Lorna Grey, Bick Nguyen, and Paul Yang Ontario Ministry of Environment Ontario Canada
over its toxicity, which are documented in a recent review [2]. Judging from the scope and quantity of glyphosate in use worldwide, it is important that a rugged, routine method be available for the accurate determination of both glyphosate and its metabolite, aminomethyl phosphonic acid (AMPA), (HO)2P(O)-CH2-NH2, in environmental matrices. Any analytical method must analyze for both, since an absence of glyphosate may have been caused by its conversion to AMPA. Because of the highly polar nature of these compounds, organic solvents cannot extract them from environmental matrices. This has made their determination a difficult challenge. Many analytical methods for these molecules were reported by Stalikas and Konidari [3] listing diverse techniques including gas chromatography (GC), high performance liquid chromatography (HPLC), ion chromatography (IC), enzyme-linked immunosorbent assays, and capillary electrophoresis (CE). Although, gas chromatography/mass spectrometry (GC/MS) based methods are economical and have good selectivity and sensitivity, they use complicated derivatization processes and require highly specialized personnel. HPLC methods suffer from the relatively poor response of these molecules to ultraviolet (UV) detection. The present method, offering a liquid chromatography/mass spectrometry (LC/MS) approach with pre-column derivatization, combines simplicity and sensitivity, adequate for a 1050 ppb determination, given a 100-mL water sample.
Abstract
A liquid chromatography/mass spectrometry method using an electrospray ionization source in positive ion mode was developed for the analysis of glyphosate (N-phosphonomethyl glycine) and its metabolite, aminomethyl phosphonic acid in water. Both glyphosate and its metabolite were derivatized using 9-fluorenylmethyl chloroformate in buffer solution prior to reversed phase high performance liquid chromatography separation. The method cleanly resolved both target molecules with excellent sensitivity in both positive and negative ion modes.
Background
Glyphosate (N-phosphonomethyl glycine), (HO)2P(O)-CH2-NH-CH2CO2H, is a global herbicide widely used in forest management, agricultural applications, and urban landscape management. Glyphosate is recognized as a benign, environmentally friendly herbicide with low toxicity [1]. However, there are health and environmental concerns
*Corresponding author
Method
Sample Preparation 1. Prepare reagent solution: 10 mg/mL 9-fluorenylmethyl chloroformate (FMOC) in acetonitrile. 2. Prepare buffer solution: 5% sodium tetraborate decahydrate (Na2B4O710 H2O), pH = 9.1 3. To 50 L of sample extract, add 50 L of buffer solution and mix. 4. Add 50 L of reagent solution. 5. Let stand 4 hrs. Instrument: Agilent 1100 LC/MS with electrospray ionization (ESI) in Positive Ion Mode Drying gas: 12 L/min, 350 C Nebulizer gas: 60 psi Vcap: 3500 V Fragmentor: 100 V for both scan and SIM runs SIM ions: 334 and 392 m/z Scan range: 120 to 1000 m/z
LC conditions Column: ZORBAX XDB-C8, 4.6 mm id 50 mm long, 5 m particles, 40 C Precolumn pump: Agilent 1100 binary Mobile phase A: 50 mM ammonium acetate, aqueous Mobile phase B: acetonitrile, 0 to 95% in 5 min, hold 3 min Pre-column flow rate: 0.7 mL/min Sample size injected: 1 L Post-column pump: Agilent 1100 isocratic. Flow: 0.3 mL/min of 0.6% formic acid Binary pump with diode array detector (DAD) and well-plate sampler
Results
Figure 1 shows molecular structures for the starting compounds and their derivatized products. It is these derivatives that are analyzed and depicted in subsequent figures.
O OH H2C NH H 2C P HO O OH + Cl
O O
O OH H 2C N H 2C P HO
FMOC (MW = 259)
O O O OH
Gly-FMOC (MW = 391)
HCl
Glyphosate (MW = 169)
O NH2 H 2C P HO O OH HO
AMPA (MW = 111) FMOC (MW = 259)
Cl +
O H 2C
H N P O OH
O O + HCl
AMPA-FMOC (MW = 333)
Figure 1. Derivatization reactions for glyphosate and AMPA with FMOC.
Figure 2 displays the mass spectra of both derivatized molecules. Positive ions 392 and 334, representing glyphosate and AMPA respectively, were chosen for further analysis.
[M+H]+ 392.0
Glyphosate
80 60
393.0
40 20 0 300
400
500
600
700
783.1
800
[2M+H]+
m/z
[M+H]+ 334.0
35 30 25 20
351.0
[M+NH4]+
335.0
15 10 5 0 300
352.0
400
500
600
668.1 [2M+H]+
700
667.0
AMPA
800
m/z
Figure 2. Mass spectra of target molecules.
Figure 3 is a stacked extracted ion chromatogram of the derivatized target molecules at low concentration.
160000 140000 120000 100000 80000 60000 40000 20000 0 1 120000 100000 80000 60000 40000 20000 0 1 1.5 2 2.5 3 3.5 4 4.5 min 1.5 2 2.5 3 3.5 4 4.5 min
Glyphosate 20 ng on column
AMPA 10 ng on column
Figure 3. Low level extracted ion chromatograms for both target molecules.
Figure 4 is a similar stacked plot, but at a higher concentration.

Glyphosate ~300 ng on column
2500000 2000000 1500000 1000000 500000 0 1 1400000 1200000 1000000 800000 600000 400000 200000 0 1 1.5 2 2.5 1.5 2 2.5
3.5
4.5
min
AMPA ~200 ng on column
3.5
4.5
min
Figure 4. Higher level extracted ion chromatograms for both target molecules.
Figures 5 and 6 are abundance vs. concentration plots for the derivatized target molecules, as the positive ions for glyphosate and AMPA, respectively.
Area 14000000 12000000 10000000 8000000 4000000 6000000 3000000 4000000
Glyphosate
Area 8000000 7000000 6000000 5000000
AMPA
Correlation: 0.99339
2000000 0 0 300 Amount [ng on column]
2000000 1000000 0 0
Correlation: 0.99427
200
Amount [ng on column]
Figure 5. Abundance vs. concentration for derivatized glyphosate, detected as [M+H]+, 392 m/z.]
Figure 6. Abundance vs. concentration for derivatized AMPA, detected as [M+H]+, 334 m/z.]
The non-linearity of the positive ion experiments at the higher concentrations is believed due to insufficient FMOC used in this initial experiment. This belief is strengthened by the work of Yang, et al [4] where linearity was achieved over a comparable concentration range with an optimized derivatization, also using FMOC, but analyzed in negative ion mode. Figures 7 and 8 show abundance vs. concentration data, per Yang [4], for the derivatized target molecules as the negative ions, 390 and 332 m/z for glyphosate and AMPA, respectively.
3000000 2500000 2000000 Area 1500000 1000000 500000 0
Chromatography is excellent with good separation between parent and breakdown product. Instrument can easily measure 1020 ng on column. Sensitivity is adequate for a 1050 ppb determination given a 100 mL water sample size. In positive ion mode, linearity is good below 200 ng on column. Non-linearity at higher concentrations is caused by depletion of FMOC. A drop in FMOCs UV response at higher target concentrations supports this, even though absorbencies were well within the UV detector linear range. The present positive ion work can be a starting point for further method development to increase both linear range and applicable matrices. The negative ion method shows what a fully engineered method can accomplish. Full sample extraction procedures can be found in Reference 4. This work represents an example of how derivatization can enhance the power of LC/MS. Generally, derivatization is not thought of as an LC/MS option. Other derivatization strategies can be studied for similar compounds.
Glyphosate Correlation: 0.9996
100
200 300 Amount [ng on column]
400
Figure 7. Abundance vs. concentration for derivatized glyphosate, detected as [M-H]-, 390 m/z.
350000 300000 250000 200000 150000 100000 50000 0
References
1. Glyphosate: A unique global herbicide. ACS Monograph 189, Washington D.C., American Chemical Society.
AMPA Correlation: 0.984
Area
2. Cox, C., Herbicide Fact Sheet. Glyphosate (Roundup). Journal of Pesticide Reform, 1989. 18, 3-17. 3. Stalikas, C. D. and C. N. Konidari, Analytical methods to determine phosphonic and amino acid group-containing pesticides. J. Chromatography, 2001 A 907 1-19.
100
200 300 Amount [ng on column]
400
Figure 8. Abundance vs. concentration for derivatized AMPA, detected as [M-H]-, 332 m/z.
4. L. Grey, B. Nguyen, and P. Yang High Performance/Electrospray Ionization/Isotopic Dilution Mass Spectrometry Analysis of N-phosphonomethyl glycine and mass spectrometry analysis of aminomethyl phosphonic acid in environmental water and vegetation matrices, J. AOAC, 2001, 84(6), 1770-1780.
Conclusions
Both glyphosate and AMPA, as FMOC derivatives, can be readily and sensitively detected using LC/MS with electrospray ion source in positive or negative ion mode.

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance or use of this material. Information, descriptions and specifications in this publication are subject to change without notice. Agilent Technologies Inc. 2002 Printed in the U.S.A. February 20, 2002 5988-4981EN
Analysis of Simazine, Thiobencarb, and Thiuram by Liquid Chromatography/Mass Spectrometry Application

Environmental
Author
Hiroki Kumagai
Abstract
A liquid chromatography/mass spectrometry method using electrospray ionization in positive ion mode was successfully applied to the sensitive and simultaneous determination of the pesticides Simazine, Thiobencarb, and Thiuram.
LC Conditions: Mobile phase A: CH3OH/30 mM CH3COONH4 (50/50) Mobile phase B: CH3OH Gradient: 0 % to100 % B in 20 min Flow rate: 0.2 mL/min Oven temperature: 40 C Injection volume: 50 L Column: Inertsil ODS3, 3.1 mm id 250 mm long 5 m
Sample Analysis
All three pesticides were determined simultaneously using the Agilent 1100 LC/MS. The following figures illustrate both the sensitivity and applicability of this method.
Background
In recent years, the potential contamination of water supplies by runoff of many kinds of pesticides from golf courses and agricultural fields has become a societal problem. Many governments have established guidelines for pesticide use and water quality standards to limit such contamination. In Japan, the concentration limits in drinking water for the pesticides Simazine, Thiobencarb, and Thiuram are 3, 20, and 6 ppb, respectively. Typically, gas chromatography-mass spectrometry (GC/MS) is used to determine Simazine and Thiobencarb in drinking water, while Thiuram is determined by high-performance liquid chromatography (HPLC) with ultraviolet (UV) detection. However, the Thiuram method used to date has problems with both selectivity and sensitivity. A better method of analysis is needed for this chemical. Such a method is described below.
1
140000 120000 100000 80000 60000 40000 20000 0 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00
1 Simazine 2 Thiuram 3 Thiobencarb
Method
Instrument: Agilent 1100 Liquid Chromatograph/Mass Spectrometer (LC/MS) with electrospray ionization (ESI) positive ion mode Drying gas: N2 (8 L/min, 350 C) Nebulizer: N2 40 psi Fragmentor: 40 V (Thiuram), 70 V Mass range: 100500 amu
Figure 1. Total ion chromatogram of target pesticides, each at 5 ng.
C2H5NH N C1
Abundance 90 60 30 0 140 180 220 260 300 340 m/z
N N
NHC2H5
O CH2SCN(C2H5)2
(CH3)2NCSSCN(CH3)2
Abundance
202 (MH)+
Abundance 90 60 30 0 140 180 220
241 (MH)+
90 60 30 0
258 (MH)+ Thiobencarb
Simazine
Thiuram
260
300
340
m/z
120
160
200
240
280
320
360
m/z
Figure 2. Mass spectra of target pesticides.
Abundance 2400 2000 1600 1200 800 400 0 2.00 4.00 6.00 8.00 10.00 12.00 min
Abundance 5500 4500
Abundance dance
1700 1500
Simazine (5 pg)
3500 2500 1500
Thiuram (0.25 ng)
1300 1100 900 700
Thiobencarb (50 pg)
500 4.00
6.00
8.00
10.00
12.00
min
4.00
9.00
11.00
13.00min
Figure 3. SIM chromatograms of target pesticides.
Conclusion
The LC/MS method described above is suitable for the simultaneous determination of the pesticides Simazine, Thiuram, and Thiobencarb. The peaks are well separated with detection limits of 0.02, 2.5, and 1 ppb, respectively, approximately 1/10 of the Japanese concentration limits.
Hiroki Kumagai is an application chemist at Agilent/Yokogawa Analytical Systems, Tokyo, Japan.
Agilent Technologies shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Copyright 2001 Agilent Technologies, Inc. www.agilent.com/chem Printed in the USA October 18, 2001 5988-4233EN
The Analysis of Organophosphate Pesticides by LC/MS Application
LC-MS
Authors
Paul Zavitsanos Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA Paul Yang MOE Ontario Canada Lorna Grey MOE Ontario Canada
Additionally, LC-MS provides unequivocal identification of each pesticide, even if the pesticide was not completely resolved from neighboring eluants. Traditional UV detection cannot provide the required specificity because many of the pesticides within the same class exhibit similar UV spectra.
Sample case
A mixture of organophosphate pesticides and an internal standard were analyzed using an Agilent 1100 LC/MS with an ESI source (Table 1).
Table 1. Mixture of Organophosphate Pesticides
Abstract
Organophosphate pesticides were readily analyzed using liquid chromatography-mass spectrometry with electrospray ion source. Sensitivity and selectivity were significantly better than using a diode-array UV detector.
Elution order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Compound Mevinphos isomer 1 Dimethoate Mevinphos isomer 2 Dichlorvos Azinphos methyl Parathion methyl Malathion Diazinon Triphenyl orthophosphate* Parathion ethyl Phorate Reldan Ronnel Terbuphos Dursban Ethion Temephos
[M+H]+ 225 230 225 221 318 264 331 305 327 292 261 322 321 289 350 385 467
Concentration g/mL 0.2 0.5 0.5 0.5 0.05 0.2 0.5 0.2 1.0 0.1 0.1 0.5 0.1 0.2 0.1 0.2 0.1
Overview
Liquid chromatography-mass spectrometry (LC-MS) is rapidly becoming a routine technique for efficient trace analysis of polar pesticides in various types of samples. In comparison to existing methodologies, such as gas chromatography-mass spectrometry (GC-MS) and ultraviolet (UV) detection, LC-MS considerably simplifies cleanup procedures, reducing both time of analysis and method development time.1 LC-MS with an electrospray ion (ESI) source avoids the thermal degradation of labile pesticides encountered with GC and eliminates the need for preliminary derivatization to increase compound volatility.
* Internal standard
Method summary
Column 2.1 mm id 5 cm long, filled with
Results
Simultaneous UV (220 nm) and MS detector outputs are compared in Figure 1. The MS plot is a composite of all the individual extracted ion chromatograms. Each was obtained at the [M+H]+ value given in Table 1, and are separated and stacked in Figure 2 for easy comparison.
3.5 m particles, C18 chemistry

20 mM ammonium acetate vs. acetonitrile
mobile phase gradient 5 % to 95 % acetonitrile in 4 minutes Hold 2 minutes

Splitless 400 L/min flow 3 L injection volume Scan data 120 to 600 m/z
SIM data as per Table 1. 95 msec dwell/ion in two groups
5 0 -5 -10 -15 -20 -25 0 2 4 6 8 Normalized UV Plot 10 12 14 min.
7000000 6000000 5000000 4000000 3000000 2000000 1000000 0 2 4 6 8 10 12 14 min.
MS selected ion mode (SIM), overlayed and normalized
Figure 1. Comparison of UV and MS chromatograms.
17 16
m/z 467 385
15 350 14 289 13 321 12 322 11 261 10 292 9 327 8 305 7 331 6 264 5 348 4 221 2 230 1
2
3 225
4 6 Minutes 8 10
Figure 2. Stacked normalized extracted ion chromatograms for compounds 1 through 17.
Figures 3 through 6 show the resulting normalized mass ion spectra for each compound included in Table 1.
Mevinphos isomer 1
193.0
[M+H]+
225.0
Dimethoate
[M+H]+
232.0
230.0
226.0
[M+H]+
192.9
222.9
Dichlorvos
[M+H]+
220.9
237.9
100
150
200
224.8
250
[M+NH4]+
239.9
242.9
[M+NH4]+
Mevinphos isomer 2
225.0
242.1
283.0
300
m/z
Figure 3. Stacked normalized ion mass spectra for compounds 1 through 4.
Diazinon
Malathion
Azinphos methyl
Parathion methyl
100
118.9 133.1 150.0 160.0
132.0
150
200
235.4
250

[M+H]+ 265.1 285.0 [M+H]+ 305.1 291.4 Background 263.9 260.9 290.1 291.5 290.4 [M+H]+ [M+H]+ 332.0 348.1 331.0 340.1 318.8 317.9
300
306.1
m/z
Triphenyl orthophosphate
[M+H]+
327.0
Parathion ethyl
[M+H]+
292.1
235.1
[M+H]+
Phorate
263.0
261.0
293.0
304.9
[M+H]+
Reldan
325.8
100 150 200 250 300
321.9 323.9
324.9
328.0
m/z
Ronnel
[M+H]+
324.9 101.9
322.9 232.9 253.3 244.8 186.9 198.9 234.9 [M+H]+

150 200 250
320.8
Terbuphos
289.0
100
291.0
98.0
300
324.0
339.1
m/z
350.9
Dursban
[M+H]+
353.9
351.9
[M+H]+
Ethion
Temephos
[M+NH4]+ 386.9 467.0
400 425 450
325
350
375
[M+H]+
475
485.0
484.0
m/z
Figure 6. Stacked normalized ion mass spectra for compounds 13 through 17. 7
www.agilent.com Conclusions
When determining organophosphate pesticides using LC-MS with an ESI source:
All the tested organophosphate pesticides
ionized well and gave definite [M+H]+ ions

Sensitivity and selectivity are significantly
better than using diode-array UV detector

Overall chromatography and analysis is simple
and straightforward
Positive identification and quantification are
performed using integrated software For more information, contact your local Agilent sales representative, or visit www.agilent.com.
Reference
1. Elbert Hogendoorn and Peit van Zoonen, Journal of Chromatography A, 892 (2000) 435-453.
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Copyright 2001 Agilent Technologies, Inc. Printed in the USA September 13, 2001 5988-3774EN
Development of an LC/MS Method for the Analysis of Rodenticides Application Note

Environmental
Friedrich Mandel and Juergen Wendt Agilent Technologies Raffaele Vistocco and Christian Bachmann Chromatography Department, A.P.P.A. Bolzano (Italy)
Introduction
Rodents such as rats and mice must be controlled because they destroy food supplies and serve as vector hosts for human diseases such as hantavirus. Although individual animals or small groups can be removed by trapping, rodenticides are frequently used in rodent control. Most rodenticides are also toxic to humans and domesticated animals such as dogs. In this work, an LC/MS method was developed to monitor several coumarin rodenticides in complex matrices. Anticoagulant poisons work by interfering with the blood clotting mechanism. After repeated ingestion of relatively small doses, a lethal dose accumulates and causes death due to internal bleeding. Anticoagulant poisoning can also cause spontaneous bleeding from the nose, gums and the gastrointestinal and urinary tracts.
Development of an LC/MS Method for the Analysis of Rodenticides
Agilent Technologies
Experimental
All experiments were done on an Agilent 1100 Series LC/MSD system that was comprised of a binary pump, vacuum degasser, autosampler, thermostatted column compartment with column-switching valve, diode-array detector, and an LC/MSD. The LC/MSD was used with either the electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI) source. Complete system control and data evaluation was done on the Agilent ChemStation for LC/MS. Reagent grade chemicals and HPLC grade solvents were used for preparing mobile phases and standards. The compounds used for this study are all classified as coumarin rodenticides due to the common coumarin moiety in the structures (Figure 1). Method development included assessing ionization technique (ESI and APCI), ionization polarity and fragmentor voltage to determine the best conditions for the three analytes. In order to test the performance of the method, spiked sausage extract and dog stomach extract were analyzed. The samples were spiked at either 0.5 or 5 ppm, then subjected to a general extraction and sample cleanup.1 The spiked samples were extracted in acetone for four hours. The acetone was then evaporated to a volume of approximately 50 ml and extracted with 100 ml of methylene chloride. The acetone layer was dried with anhydrous sodium sulfate, filtered, and evaporated to dryness. After redissolving in 2 ml of acetone, the extract was applied to a column containing 15 g silica gel and 1 g charcoal. After elution with methylene chloride/benzene/acetone (10/2/2 v/v), the eluate was evaporated to dryness and redissolved in 5 ml of methanol. The sample extracts were then analyzed by ESI-LC/MS.
O O CH OH H2C OH O C CH3 O O O
OH
Difenacoum C31H24O3 mass 444.173

O O
Coumatetralyl C19H16O3 mass 292.11

Figure 1. Coumarin rodenticides.
Coumafuryl C17H14O5 mass 298.084

Both coumafuryl and coumatetralyl showed poor response in positive mode ESI compared to negative mode ESI (Figure 2). The positive mode response for these compounds was better in APCI than ESI, but the negative mode APCI response for coumatetralyl was weak compared to ESI. Based on these results, ESI negative mode was selected for the optimized method.
10000 0
Positive m/z 299
ANALYSIS METHOD Chromatographic Conditions Column: 150 2.1 mm Zorbax XDB-C18, 5 m (p/n 993700-902) Mobile phase: A = 2 mM ammonium acetate in water B = methanol Gradient: start with 30% B at 2 min 50% B at 4 min 100% B Flow rate: 0.4 ml/min from 0 to 2 min, then 0.5 ml/min Column temp: 50C Injection vol: 2 l Diode-array detector: signal: 280, 16 nm; reference: 550, 100 nm
Negative m/z 297 40000 0 Positive m/z 293 1000 0 10000 0 Positive m/z 445 20000 0 20000 0 0 1 2 3 4 5 6 7 min Negative m/z 443 Negative m/z 291
coumafuryl
coumatetralyl
difenacoum
ESI-MS Conditions Source: Drying gas flow: Nebulizer: Drying gas temp: Vcap: Stepsize: Peakwidth: Time filter: Scan: SIM ions (negative mode):
Fragmentor:
ESI 10 l/min 40 psig 350C 2500 V (positive and negative) 0.1 0.2 min On m/z 150500 296, 297, 298 Coumafuryl 290, 291, 292 Coumatetralyl 442, 443, 444 Difenacoum variable 180 V (50275) 100 V (280) 120 V (400)
Figure 2. Comparison of positive and negative ionization modes for electrospray LC/MS.
Typically, positive ions will fragment more easily than negative ions, so the fragmentor voltage was optimized for both modes. For example, difenacoum showed extensive fragmentation at 160 V in positive ion mode but no fragmentation at the same voltage in negative ion mode. Optimized fragmentor voltages were used for each ion in the final method. For samples extracted from a complex matrix, the deprotonated molecule as well as the ions 1 m/z above and below can be monitored to check for coeluting artifacts. In the final method, this strategy was employed so a set of three ions was used for each analyte. The ions for each analyte were time-programmed as a SIM ion group in order to maximize the dwell time and thus maintain maximum sensitivity. Figure 3 shows the results for the
spiked sausage extract. For this analysis, data was collected in scan mode because the coumafuryl and difenacoum were present in sufficient concentration. The mass spectra from the spiked analytes show good quality at the 5 ppm level. The dog stomach extract was spiked at a lower level (0.5 ppm) so SIM analysis was done for this sample (Figure 4). Only the spiked analyte, coumafuryl, shows a significant signal. This work demonstrates that anticoagulant rodenticides can be detected by APCI and ESI in both positive and negative ion modes. Negative mode ESI-LC/MS was found to be the best choice for the three target compounds. This LC/MS method was capable of detecting the rodenticides in complex matrices.
coumafuryl
250000 150000 50000 m/z 297
coumafuryl
60000 m/z 297
2.833
2.882
40000 443.1 20000 0
coumafuryl
Max: 153123 311.2 250000 150000 265.1 343.2 50000 m/z 291
60000 40000 20000 1 2 3 4 5 minutes 6 200 300 400 m/z 0
m/z 291
difenacoum
difenacoum
297.0 60000 Max: 60723 5.374 40000 20000 0 211.1 0 1 2 3 4 5 6 min m/z 443
250000 150000 50000
m/z 443
3 4 5 minutes
200 300 400 m/z
Figure 4. Extracted ion chromatograms from the analysis of dog stomach extract spiked with 0.5 ppm of coumafuryl. Data was collected in negative ion using SIM mode.
Figure 3. Extracted ion chromatograms (left) and mass spectra (right) from the analysis of sausage extract spiked with 5 ppm of coumafuryl and difenacoum. Data was collected in negative ion using scan mode.
References
1. Faucannet, V., Pouliquen, H., and Pinault, L., Journal of Analytical Toxicolocy, 27, 1997.
For more information on our products and services, you can visit our site on the World Wide Web at: http://www.agilent.com/chem. Agilent Technologies shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance or use of this material. Information, descriptions and specifications in this publication are subject to change without notice. Copyright 2000 Agilent Technologies All rights reserved. Reproduction and adaptation is prohibited. Printed in the USA June 2000 (23) 5968-0561E
Authors
Friedrich Mandel and Juergen Wendt are application chemists at Agilent Technologies in Waldbronn, Germany. Raffaele Vistocco and Christian Bachmann are in the Chromatography Department of A.P.P.A. Bolzano in Italy.
Productivity Tools
Can "Deconvolution" Improve GC/MS Detectability?
Application Note
All Industries
Authors
Chin-Kai Meng and Mike Szelewski Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
This study uses 35 pesticides spiked in spinach extracts at the 50 ppb level to find the optimal AMDIS deconvolution settings. Additional advantages of using deconvolution versus MSD ChemStation, to find more compounds in an extract are also discussed. The detectability of compounds in a complex matrix is significantly improved with deconvolution. This can also be viewed as better or increased sensitivity through improved selectivity versus the background. Agilents MSD ChemStation add-on - Deconvolution Reporting Software (DRS) runs AMDIS automatically to generate an easy-to-read quantitation report.
Introduction
Instrument detectability is usually determined by the amount of sample injected, the responses from the detector and matrix interferences. The signal-to-noise ratio (S/N) can be used to gauge the sensitivity of an instrument in a clean sample. The presence of matrix alters this sensitivity due to a lack of selectivity between compounds of interest and background. In a multiresidue analysis, the data reviewing process is also very important in confirming the hits found by the software and reviewing the integration and quantitation for accuracy. Agilent Deconvolution Reporting Software (DRS) has been proven as a powerful data processing tool for finding trace compounds in complex matrices [1]. In this study, results from the Automated Mass spectral Deconvolution and Identification System (AMDIS), part of DRS is closely studied and compared to the results from ChemStation. The goal is to determine if deconvolution (DRS) can provide better results (detectability) than routine ChemStation data processing.
Instrument parameters
GC: Autoinjector: Retention gap: Column: Oven ramp: Initial Ramp 1 Ramp 2 Ramp 1 Run time: Inlet: RT locking: Liner: Injection mode: Inlet temp. ramp: Initial Ramp 1 Septum purge: Purged Union: Split vent: Gas saver: Cryo on: Backflush Postrun: Oven: Purged Union: MMI: Restrictor: MSD: Solvent delay: EMV mode: Mass Range: Threshold: Sample number: Transfer Line: Source: Quad: 5 min 280 C 70 psi 2 psi 0.7 m 0.15 mm deactivated fused silica tubing (from Purged Union to MSD) 5975C 2.5 min Gain Factor = 2 Full scan, 45-550 0 2 A/D Samples 4 280 C 300 C 200 C 7890A 7693A 2 m 0.25 mm id Siltek capillary tubing HP-5MS UI (ultra inert), 15 m 0.25 mm, 0.25 m (from inlet to Purged Union) Agilent p/n 19091S-431 UI Rate (C/min) 50 6 16 Temp (C) 100 150 200 280 Time (min) 1.6 0 0 5
20.933 min Multimode Inlet (MMI) at 17.73 psi (Retention Time Locked), constant pressure mode Chlorpyrifos-methyl locked to 8.297 min Helix double taper, deactivated (Agilent p/n 5188-5398) 2-L cold splitless (fast injection) Rate C/min 720 Temp C 50 300 Time min 0.01 hold
3 mL/min 4 psi (PCM) 50 mL/min at 0.75 min 20 mL/min after 4 min Cryo use temperature 150 C; time out at 15 min
Experimental
Spinach extracts (see Acknowledgement) were prepared using the QuEChERS [2, 3] protocol shown below:
15 g homogenized sample + 15 mL ACN + internal standard
Add 1.5 g NaCl and 6.0 g MgSO4
Shake and centrifuge
Transfer 9 mL extract to tube containing 0.4 g PSA + 0.2 g GCB + 1.2 g MgSO4 and vortex
Add 3 mL toluene
Deconvolution
Deconvolution is a process for extracting ions from a complex total ion chromatogram (TIC), even with the target compound signal at trace levels. The software used for this technique is AMDIS developed by NIST (National Institute of Standards and Technology) [4].
Shake and centrifuge
Reduce 6 mL to ~100 L Add 1.0 mL toluene + QC standard + MgSO4 and centrifuge
Transfer to ALS vials for GC-MS analysis
Thirty-five pesticides were spiked into spinach extract at 50 ppb (pg/L).

2
As a review, let's look at the deconvolution process. AMDIS considers the peak shapes of all extracted ions and their apex retention times (RT). In this example, only some of the extracted ion chromatograms (EICs) are overlaid for clarity with the apex spectrum (Figure 1A).
Ion 160 EIC has the same RT as ions 50, 170 and 280, but has a different peak shape. Ion 185 has a different peak shape and an earlier RT. Ions 75 and 310 have similar peak shapes but they have different RTs.
Figure 1A
170 50 75 160 185 280 310
160 shape
Extracted Ion Chromatograms (EIC) After de-skewing
50 170 280 185 shape & early retention time 75 late retention time 310 early retention time Same shape and same retention time
Figure 1B shows the EICs after the different peak shapes or RTs are eliminated from Figure 1A. Ions 50, 170, 280 and a few others remain.
Figure 1B
170 50 75 160 185 280 310
Extracted Ion Chromatograms (EIC)
50 170 280
Only the ions in black have the same shape and retention time as shown by 50, 170, 280plus others
Figure 1A-1C. Simplified deconvolution process (continued).
Figure 1C shows all of the ions in black that have similar peak shapes and RTs, within the criteria set earlier by the analyst. These are grouped together and referred to as a component by AMDIS.
Figure 1C
170 50
280
50
Extracted Ion Chromatograms (EIC)
170
These deconvoluted ions are grouped together as a component
280
Figure 1A-1C.
Simplified deconvolution process (continued).
Deconvolution finds the components from a complex TIC. Each component is searched against a retention time locking (RTL) library in AMDIS format. In addition to spectral matching, the locked RT can also be used as a criterion for hits. Depending on the match factor from the search, target compounds can be identified or flagged in a complex TIC. The power of deconvolution is appreciated while comparing the top two spectra in Figure 2. The raw scan or original nondeconvoluted scan is shown on top. The clean scan, that is the
deconvoluted component, is shown in the middle. The bottom scan is the identified compound in the AMDIS library. Without deconvolution, the analyst would visually compare the background subtracted raw scan and library scans for confirmation. It would be very difficult, if not impossible, to say that Fenbuconazole, the target compound in this example, is present using that type of comparison.
Scan at 10.776 min
Deconvoluted/extracted spectrum A component in the scan above.
Library spectrum
Fenbuconazole
Figure 2.
Comparison of raw, deconvoluted, and library spectra.
AMDIS Settings
Previous publications that discussed the power of using deconvolution to screen complex matrices, did not discuss specific AMDIS settings to define components [1, 5, 6]. In this study, several settings (that is, resolution, sensitivity, and
shape requirements) are compared to find the maximum number of spiked compounds. The minimum match factor is set to 30 and the retention time window is limited to 30 seconds (RI window is set to 30) to qualify the hits from the retention time library search (Figure 3). The expected retention times of the compounds in the library database are obtained in acetone solvent without a retention gap. The samples in this study are in toluene solvent with a retention gap. Therefore, the retention time window is set wider than the normal 10 or 15 seconds, at 30 seconds.
Figure 3.
AMDIS identification settings.
Figures 4 and 5 describe some of the parameters in the AMDIS deconvolution tab. In this article, "1 M H M" means: adjacent peak subtraction = 1, resolution = medium, sensitivity = high, shape requirements = medium.
Settings can be optimized for chromatographic resolution, peak shape, retention time windows, acceptance criteria, and so forth. Settings can be saved to "ini" files. The chemist has control over the deconvolution and identification process by varying numerous AMDIS settings. Most of these parameter settings are not independent; so changing one parameter can affect another.
Assumed component width in scans. Increase this if all peaks are wider. If the box is checked, masses entered here will not be used as models but can still be included in a component. A closely eluting large ion will be subtracted to allow more models to be considered. None yields the fastest processing and Two the slowest.
Figure 4.
AMDIS deconvolution settings.
Higher Resolution will separate closer eluting peaks to find more components and thus runs slower Higher Sensitivity will find smaller, noisier components but may result in more false positives and runs slower Higher Shape requirements requires that EICs have exactly the same shape, thus resulting in fewer components found and more uncertain peaks present.
Figure 5.
AMDIS deconvolution settings.

Deconvolution Settings
Figure 6 shows effects on match factors (y-axis) due to variation of adjacent peak subtraction and sensitivity across 35 pesticides (x-axis). This figure shows two things:
The adjacent peak subtraction (1 or 2) makes little difference in match factor The sensitivity setting (very high and high) makes little difference in match factor In the next few figures, the AMDIS setting is varied one at a time to observe the number of pesticides found. The reference point is the optimal setting (HHM) where the maximum number of hits were obtained.
100
90
80
70
60
Match Factor
50
40
1 H VH M 2 H VH M 1HHM 2HHM
30
20
10
Pesticide
Figure 6.
Comparison of match factors with four AMDIS settings.
Figure 7 shows that keeping the sensitivity and peak requirements the same, and lowering the resolution from H to M will find fewer targets. The number of targets found is in the yellow circle. A resolution setting of "low" yields even fewer targets.
Changing resolution only
M H M
31
35
H H M
Figure 7.
Number of compounds found by varying resolution.
Figure 8 shows that while keeping the resolution and peak requirement constant, lowering the sensitivity from H to M will find fewer targets. However, increasing the sensitivity from H to VH does not affect the number of targets found, similar to that in Figure 6. Figure 9 shows that while keeping the resolution and sensitivity the same, lowering or increasing the peak shape requirement from M to L or H will find less targets.
Changing sensitivity only
35
Changing resolution only
M H M
31
62.3
35
61.6
H VH M
H H M
62.0
H M M
33
61.9
58.5
33
H H H H H Changing shape requirement only L
32
63.6
35
Changing sensitivity only
Figure 10. Comparison of average match factors with AMDIS settings.
35
H VH M
H H M
H M M
33
ChemStation Quant settings

Figure 11 shows part of the "Edit Compound" screen in the MSD ChemStation. This shows the quant database for locating and confirming compounds using three ion ratios of each target analyte. The RT window is specified in the upper box and the ions and ion ratios are specified in the lower box. As shown in Figure 11, the Extraction RT window is set to 0.5 min and the Qualifier Ion (Q1, Q2, and Q3), % Uncertainty is set to Absolute 50%. In ChemStation, the
Figure 8.
Number of compounds found by varying sensitivity.
35
H H M
H H H
33
Changing shape requirement only
H H L
32
Figure 9.
Number of compounds found by varying peak shape.
In addition to the number of targets found, we should look at the Average Match Factor (AMF) of all the targets found. The AMF is the number in the green triangle. Figure 10 shows that there is no significant variation in AMFs except in HHH mode (58.5) which is much lower than others (>61.6). This supports that HHM is still the optimal setting, considering processing speed and number of false positives.
Figure 11. Target compound RT and ion setup.
target compound identification is based on four ions and three qualifier ion ratios. However, the target compound identification in AMDIS (Figure 2) was based on the full spectral library match which is more dependable. Another key parameter in quantitation is the "Quantitation subtraction method" which is set to "Avg first and last" and not shown here. Figure 12 is an overlay of four ions (Quant and Qualifiers) from ChemStation and the quant ion from AMDIS (in magenta).
Due to the chemical background, the four ions from ChemStation have offset and noisy baselines, which will affect the peak integration and proper quantitation results. In comparison, the magenta trace is the deconvoluted quant ion from AMDIS. The chemical noise had been removed in the deconvolution process. It shows a flat baseline and accurate integration. There are other advantages of using deconvolution in GC/MS analysis as discussed below.
14000 12000 10000 8000 6000 4000 2000 0 13.60 13.70 13.80 13.90 14.00 14.10 14.20
Ion 123
14.079
Ion 171 Ion 128 Ion 143 AMDIS
|
| | | |
14.078
14.30
Deconvolution shows a flat and accurate integration baseline
Figure 12. Target, qualifier and AMDIS deconvoluted EIC overlay.
10
Additional Advantages of Using Deconvolution

Finds more compounds than ChemStation does In Figure 13, ChemStation did not integrate ion 109 (ChemStation target ion) at the expected RT, therefore, the compound was not found. AMDIS found Fonofos correctly, at 6.898 min. The qualifier ion ratios at this RT also match that required by ChemStation for identification.
9000 8000 7000 6000 5000 4000 3000 2000 1000 0 6.76 6.78 6.80 6.82 6.84 6.86 6.88 6.90 6.92 6.94 6.96 6.98
Ion 109 Ion 246 Ion 137 Ion 110 6.898 AMDIS
(242) Fonofos 6.944 min (-6.944) 0.00 response 0 Ion 109.00 246.00 137.00 110.00 Exp% 100 59.00 54.60 24.20
AMDIS: 0.08 AMDIS: 70868 Act% 0.00 0.00# 0.00# 0.00
11
Finds the correct peak In Figure 14, from the size and location of the three qualifier ions, it is obvious that ChemStation picked the wrong peak (at RT = 4.067) to quantitate. However, AMDIS found a peak (at RT = 3.873) whose ion ratios are in agreement with the ChemStation qualifier ions. Again, this demonstrates that the AMDIS full-spectrum matching process is a more robust approach for identifing a compound in a complex matrix.
3500 3000 2500 2000 1500 1000 500 0
Ion 147 Ion 76 Ion 104 Ion 103 AMDIS
|
3.873
4.067
| | | |
|
3.50 3.60 3.70 3.80 3.90
|
4.00 4.10 4.20
(79) Phthalimide 4.069 min (+0.079) 0.07 AMDIS: 0.04 response 62142 AMDIS: 36450 Ion 147.00 76.00 104.00 103.00 Exp% 100 60.50 57.30 28.80 Act% 100 48.95 14.64 35.45
12
Higher discrimination power than ChemStation In Figure 15, the target ion (ion 235) is overwhelmed by the matrix background (shown as a large fronting peak). ChemStation was not able to differentiate the ion 235 contribution from the background or the compound; therefore it
integrated the distorted peak. Due to the rising baseline, ChemStation integrated a large area of chemical background as the "target compound signal". On the other hand, AMDIS was able to deconvolute the compound signal away from the background ion and remove noise properly before the integration. This provides a more reliable quant result.
70000 60000 50000 40000 30000 20000 10000 0 11.60
Ion 235 Ion 237 12.234 Ion 165 Ion 199 AMDIS
|
12.234
| | | |
| | |
11.80 12.00 12.20 12.40 12.60
| | |
12.80 13.00
Deconvoluted ion is noise-free, thus easier to integrate for more reliable quantitation results In Figure 16, ChemStation and AMDIS found the same peak. Due to the noisy baseline, ChemStation drew the integration
baseline (red dash line) incorrectly. Again, deconvolution removes chemical noise first, and can therefore, integrate the peak easily and reliably.
3000 2500 2000 1500 1000
Ion 269 Ion 325 Ion 271 AMDIS
|
13.129 13.130
ChemStation
500 0 12.90 12.95 13.00 13.05
|
AMDIS
13.10 13.15 13.20 13.25 13.30
13
Agilents ChemStation add-on - Deconvolution Reporting Software (DRS) incorporates AMDIS deconvolution. Therefore, the above AMDIS advantages are automatically captured in DRS data processing which combines results from ChemStation, AMDIS, and NIST MS Search into one report.
Comparing number of compounds found between ChemStation and AMDIS

Figure 17 is a summary of the hits from ChemStation and AMDIS under four different settings, respectively. The blue bars represent the number of false positives and the red bars represent the number of actual target compounds found. On the left side of the graph, the settings of ChemStation are Ion
Ratio Uncertainty. Although the absolute 30% and 50% increase the total number of compounds found, only about half of the 35 targets are found. The analyst is forced to review more hits and does not gain any additional information. The entire target list of 900+ compounds must be reviewed for false negatives. The right side of the graph shows that the four AMDIS settings gave similar results. In each case, all 35 targets were found with a reasonable number of false positives. There were no false negatives. The analyst must only review the positives, which is a significant time savings. This shows that AMDIS (DRS) is much more capable than ChemStation in finding target compounds in a complex matrix. AMDIS (DRS) provides better detectability and faster data processing.
120
ChemStation Results
110
AMDIS Results
False Positive Actual Targets Found
88 72 83 73
100
80
60 49 40 11 17 6 0 50% Relative 30% Relative 50% Absolute 12 35 20 19 35 35
20
35
30% Absolute
1 H VH M
2 H VH M
1HHM
ChemStation Settings
AMDIS Settings
2HHM
Figure 17. Overall comparison of AMDIS and MSD ChemStation compounds found.
14
Conclusions
AMDIS finds more target compounds than ChemStation in a complex matrix. Deconvolution (DRS) provides a cleaned peak to integrate properly giving more reliable results. AMDIS did not miss any target compounds at the 50 ppb level using scan data. This minimizes the time an analyst must spend reviewing results. Confirmation of compounds is done in significantly less time with deconvoluted component spectra available. The detectability of compounds in a complex matrix is significantly improved with deconvolution. This can also be viewed as better or increased sensitivity through improved selectivity versus the background. Deconvolution Reporting Software (DRS) automates the deconvolution (AMDIS) process to produce an easy-toread quantitation report.
3.
S. J. Lehotay, K. Matovsk, and A.R. Lightfield, "Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables," 2005, J. AOAC Int, 88:615-629 http://chemdata.nist.gov/mass-spc/amdis/overview. html Philip L. Wylie, "Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library," Agilent Technologies publication, 5989-5076EN, April 2006 Chin-Kai Meng and Mike Szelewski, "Replacing Multiple 50-Minute GC and GC-MS/SIM Analyses with One 15Minute Full-Scan GC-MS Analysis for Nontargeted Pesticides Screening and >10x Productivity Gain" Agilent Technologies publication, 5989-7670EN, December 2007
4. 5.
6.

Acknowledgement
The authors would like to thank Dr. Jon Wong (FDA-CFSAN, College Park, Maryland) for graciously provided samples for this study.
References
1. Christopher P. Sandy, "A Blind Study of Pesticide Residues in Spiked and Unspiked Fruit Extracts Using Deconvolution Reporting Software," Agilent Technologies publication, 5989-1654EN, October 2006 M. Anastassiades, S. J. Lehotay, D. Stajnbaher, and F. J. Schenck, "Fast and Easy Multiresidue Method Employing Acetonitrile Extraction/Partitioning and 'Dispersive Solid-Phase Extraction' for the Determination of Pesticide Residues in Produce," 2003, J. AOAC Int, 86:412-431
2.
15
Agilent Technologies, Inc., 2010 Printed in the USA January 22, 2010 5990-5052EN
Maintaining Compound Retention Times with the Backflush Enabled Pressure Controlled Tee Configuration for Agilent 7890A GCs with Agilent 5975 Series MSD and Agilent 7000 Series Triple Quadrupole MS Systems
Application Note
Environmental
Authors
Harry Prest Agilent Technologies, Inc. 5310 Stevens Creek Blvd. Santa Clara, CA 95051 USA
Abstract
The Pressure Controlled Tee configuration provides a simple approach to rapid backflushing and GC-serving on all MS systems. To further improve acquisition and analysis, this note demonstrates how to generate compound retention times that are permanent and universal through Retention Time Locking. If retention times can be permanent on any system, then complex MS acquisition methods such as selected ion monitoring and multiple reaction monitoring modes need not be changed when columns are trimmed or replaced. Similarly, compound retention times are unchanged in data analysis which simplifies identification, quantitation and the use of databases of compound retention times. This note applies Retention Time Locking and the pressure Controlled Tee to analysis of the Japan Positive (Pesticide) List of about 430 compounds to add the feature of rapid backflushing without additional analytical time over the published method. The results show retention times may be reproduced from the cited database values to better than an average of 1-sec, which is a very good match. Rapid servicing of the column and inlet are possible without venting the MS system and applying simple "mechanical Retention Time Locking" quickly returns the GC/MS to acquisition without method changes or additional complexity.
Introduction
Agilent's GC Capillary Flow Technologies (CFTs) have greatly expanded the capability and flexibility of gas chromatographic analysis. CFT can provide rapid column and inlet maintenance without MS venting as well as backflushing of the accumulated sample matrix preventing carryover, which compromises the column performance. One highly flexible CFT arrangement, the Pressure Controlled Tee (PCT) [1], has been shown to provide rapid backflushing and enhance analytical robustness in GC/MSD single quadrupole analysis in a very challenging sample matrix [2]. Other work has shown that introducing the PCT configuration maintains signal and therefore compound detection limits [3]. Another concern of analysts involves method maintenance, both in acquisition and data processing. After GC column maintenance or replacement, compound retention times change. This affects scheduled events in MS acquisition such as selected ion monitoring (SIM) or multiple reaction monitoring (MRM) modes as well as compound retention times in quantitation or library databases. One powerful approach to maintaining compound retention times is Retention Time Locking (RTL). Retention Time Locking provides a very simple idea: RTL makes compound retention times permanent and universal. This can greatly simplify methods and their development. An example is the Retention Time Locked compound database for the Japanese Positive List for pesticide analysis [4]. In this case, the retention times for 430 pesticides are "locked" to a particular column and oven program. Using this method of given compound retention times as an example, this note explores and addresses tools provided by the PCT for maintaining compound retention times in quantitation databases as well as for the SIM and selected reaction acquisition modes used by the Agilent 5973, 5975 and 7000 series Mass Spectrometers.
described in the Rapid Universal GC/MS Backflushing Kit (G1472A) and included documents (Manual G1472-90001). An Agilent 5975 Series MSD was used but results are independent of the MS system and so apply to the 7000 Series Triple Quadrupole. Details of the Japan Positive List method are provided [4] and only those altered to accommodate the PCT and backflushing are cited here (Table 1). Note the original method has an extended oven hold at 300 C. This has been shortened by 5 min, and 3 min of that time has been incorporated for backflushing overall the method is slightly shorter.
Table 1. GC Parameters for RTL with Backflush (the actual arrangement is two 15-m columns but the GC is configured as follows) Front Inlet to MSD Outlet: DB-5ms 30-m 0.25-mm id 0.25 m film AUX EPC Channel 4 to MSD Outlet: DB-5ms 15-m 0.25-mm id 0.25 m film. Helium in constant flow mode As required = Column 1 + 0.04 mL/min 50 C for 1 min, 25 C/min to 125 C for 0 min, 10 C/min to 300 C for 5 min 300 C for 3 minutes 0.25-mL/min 3-mL/min (or as allowed by the pumping system)
Configuration: Column 1: Column 2: Carrier and mode: Column 1 flow: Column 2 flow: GC oven: Post run: Column 1: Column 2:
Results
RTL Experiments
In order to lock the method, the helium carrier flow must be set to produce a retention time of 13.443 min for the compound chlorpyrifos-methyl. To do this, the column flows for both Column 1 and Column 2 are manually incremented by 0.1 mL/min and the elution of the chlorpyrifos-methyl "locking compound" is determined by injection of a standard at each flow setting. The data and resulting regression are given in Table 2 and Figure 2, respectively. Note that more flow situations were acquired than necessary and usually only 5 flows would be required. The regression correlation is very high and, using the equation for the regression, it was calculated that a Column 1 flow of 1.47 mL/min and Column 2 flow of 1.51 mL/min should produce a retention time of 13.443 min for chlorpyrifos-methyl. These values were applied to the GC acquisition method and several standards containing chlorpyrifos-methyl and other compounds were acquired. The results showed a high degree of agreement not only for the locking compound but also other compounds in the Japan Positive List screening database (Table 3). In general,
2
Experimental
Hardware and method parameters
Schematically the experimental arrangement is shown in Figure 1. The commonly used 30-m analytical column configuration is replaced by two 15-m columns joined by the Purged Ultimate Union. For these experiments, two 15-m DB-5ms (0.25-mm id 0.25-m film) were used (122-5512 UI). Makeup gas can be supplied by either an Auxillary Electronic Pneumatic Control (Aux EPC) or Pressure Control module (PCM) but here an Aux EPC was used. The Column 2 flow is always set slightly higher than the Column 1 flow by 0.04 mL/min to prevent back-diffusion in the makeup gas line. More details and the installation and use of the PCT are
Figure 1.
Schematic of the Pressure Controlled Tee configuration illustrating the arrangement. Upper panel shows the forward flow during analysis and the lower panel shows the column flows during backflushing.
retention times agreed within better than 1 sec of their expected time. The worst case was on the order of 2 or 3 sec but considering the peak widths are on the order of 3 to 6 sec, this must still be considered within the analytical identification window. (The data also suggest a slight systematic error
Table 2. Column 1 (30 m) and Column 2 (15 m) Flow Settings and the Measured Retention Time of the Chlorpyrifos-methyl Locking Compound. Column 2 flow 1.04 1.14 1.24 1.34 1.44 1.54 1.64 1.74 Time (min) 14.164 13.978 13.814 13.665 13.529 13.406 13.292 13.185
across the elution order which could be corrected to improve the fit on a permanent basis.)
Mechanical Retention Time Locking

Pesticide analysis typically involves intensive column and inlet maintenance. Setting the PCT in Backflush mode and cooling the injection port (and oven), allows the septum, liner and column to be rapidly serviced [5]. Usually a column "loop" of about a half-meter is cut off the column head. One approach to restoring retention times would be to check the new retention time of the locking compound via an injection and correct accordingly. However, Agilent Capillary Flow Technology (CFT) enables a more rapid and convenient approach. If the removed section of column is simply replaced with an equal length of capillary column, here DB-5ms 0.25-mm 0.25-m, then the retention times will remain unchanged. This is possible using the Ultimate Union as the column connector which is included in the G1472A kit. The Ultimate Union conveniently mounts inside the bracket next
Column 1 flow 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70
Figure 2.
Regression of locking compound retention time versus column 1 flow.
to the Purged Ultimate Union (Figure 3). Employing this approach, a column loop of the DB-5ms column at the injection port was removed and replaced with the Ultimate Union connection. Standards were rerun with no other change to the GC acquisition method and the results are shown in Table 4. Again, deviations from expected retention times are very small, less than 1 sec and much less than a peak width, suggesting MRM and SIM tables would not require amendment.
Figure 3.
Schematic illustration of the PCT configuration and mounting bracket: The Purged Ultimate Union is mounted behind the Ultimate Union. The Purged Ultimate Union is at the mid-point of the columns (gold lines) while the Ultimate Union connects a short pre-column to the injection port. Not shown is the retaining pin which secures the union.
Table 3.
Agreement between Measured Compound Retention Times and the Expected Retention Times for Selected Compounds in the Japan Positive List Database. Columns A, B, and C Refer to Multiple Measurements for Compounds Present in Different Analyzed Standards. RT (measured-expected) sec. a b c 0.66 0.66 0.66 0.48 0.36 0.36 1.14 0.72 0.12 0.00 0.06 0.72 0.24 0.12 0.06 0.24 0.48 0.36 0.24 0.42 0.36 0.18 0.6 0.42 0.36 1.08 0.6 0.3 1.02 1.02 0.24 0.96 1.08 0.84 2.28 1.14 1.02 1.5 0.84 1.56 1.02 1.86 2.7 2.34 0.36 0.18 -0.06 0.48 0.18 0.06 0.12 0.24 0.36 0.06 -0.42 0.78 0.48 0 0.72 0.96 1.02 0.66 0.66 0.9
Table 4. Compound Dichlorvos Mevinphos
Results for "Mechanical Retention Time Locking" by Use of the Ultimate Union Connector RT Error (s) 0.66 0.66 0.9 0.36 0.96 0.54 0.48 0.48 0.06 0.78 0.42 0.42 0.12 0.12 0.36 0.12 0 0.06 0.12 0.48 0.18 0.18 0.42 0.36 0.72 2.34
Compound Dichlorvos Mevinphos Ethoprophos Naled Sulfotep Phorate BHC alpha isomer Lindane Fonofos Diazinon Disulfoton BHC delta isomer Chlorpyrifos Methyl Methyl parathion Heptachlor Fenchlorphos Fenitrothion Malathion Chlorpyrifos Fenthion Aldrin 4,4'-Dichlorobenzophenone Heptachlor exo-epoxide Tetrachlorvinphos Endosulfan (alpha isomer) Fenamiphos Prothiofos p,p'-DDE S,S,S-Tributylphosphorotr Dieldrin Endrin Fensulfothion Ethion p,p'-DDD o,p'-DDT Sulprofos Carbophenothion Endosulfan sulfate p,p'-DDT EPN Azinphos-methyl Azinphos-ethyl Coumaphos
Ethoprophos Sulfotep Phorate Fonofos Diazinon Disulfoton Chlorpyrifos Methyl Methyl parathion Fenchlorphos Fenitrothion Malathion Chlorpyrifos Fenthion Tetrachlorvinphos Fenamiphos Prothiofos S,S,S-Tributylphosphoro Ethion Sulprofos Carbophenothion EPN Azinphos-methyl Azinphos-ethyl Coumaphos
Conclusions
Maintaining compound retention times can greatly simplify acquisition and data analysis methods in GC/MS or GC/MS/MS. The PCT configuration can easily be used to generate permanent and universal retention times for compounds through Retention Time Locking. To show the "universal" attribute, it has been demonstrated that given the initial Japan Positive List method, RTL with the PCT can replicate the 430 compound retention times to a very high degree. By extension, any method and database can be replicated. Using a replaceable section of column which can be rapidly exchanged during servicing (without venting or cooling the MS), compound retention times can be conserved and can be considered permanent. This was demonstrated in constant carrier gas flow mode as this provides the best chromatographic conditions for GC/MS and is the normal approach of MS users. The Ultimate Union, like the Purged Ultimate Union, has very low dead volume, low activity, and minimal influence on chromatography. Overall, the Pressure Controlled Tee configuration provides a number of simple strategies for improving method robustness and method maintenance without loss in sensitivity.
Acknowledgement
The author is very grateful for a past discussion on this topic with an outstanding colleague, Satoshi Ito.

References
1 Capillary Flow Technology for GC/MS: A Simple Tee Configuration for Analysis at Trace Concentrations with Rapid Backflushing for Matrix Elimination, Agilent Technologies publication 5989-8664EN Capillary Flow Technology for GC/MS: Efficacy of the Simple Tee Configuration for Robust Analysis Using Rapid Backflushing for Matrix Elimination, Agilent Technologies publication 5989-9359EN Implementation of the Pressure Controlled Tee for Backflushing for the 7000 Series Triple Quadrupole Mass Spectrometer: Implications for Sensitivity, Agilent Technologies publication 5990-4504EN Screening for Pesticides in Food Using the Japanese Positive List Pesticide Method: Benefits of Using GC/MS with Deconvolution Reporting Software and a Retention Time Locked Mass Spectral Database, 5989-7436EN The Rapid Universal GC/MS Backflushing Kit, part number G1472A and contains the manual (G1472-90001) which describes this configuration and operation for 5973 or 5975 or 7000 series Mass Selective Detectors.
Agilent Technologies, Inc., 2009 Printed in the USA September 30, 2009 5990-4643EN
Achieving Lower Detection Limits Easily with the Agilent Multimode Inlet (MMI)
Application Note
All Industries
Authors
Bill Wilson and Chin-Kai Meng Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
This application note discusses three injection techniques: hot splitless, cold splitless, and solvent vent mode available on the Multimode Inlet. The cold splitless and solvent vent mode injections allow analysts to achieve a lower detection limit by making large volume injections (LVI). A total ion chromatogram overlay of 40-ppb pesticide standards from 2-L hot splitless, 10-L cold splitless and 25-L solvent vent illustrates the improvement in signal-to-noise ratios using LVI.
Introduction
A growing number of analysts are exploring large volume injection (LVI) techniques to improve existing analyses. With traditional liquid injection techniques in capillary gas chromatography, most inlets and columns can only handle 1 2 L at a time. Attempts to increase the injection volume can lead to broadened and distorted analyte peaks, large and long solvent peak tails, and saturated or damaged detectors. The purpose of increasing the injection volume is normally to improve detection limits in trace analysis. By introducing more of the sample to the system, the mass of analyte reaching the detector will be proportionally increased, resulting in larger peak areas and peak heights. If the baseline noise is constant, larger peak heights mean greater signal to noise ratios and lower system detection limits. An additional benefit of LVI is the ability to reduce the amount of sample originally processed. By injecting 10 100 times more volume of processed sample and concentrating it in the inlet, the sample preparation can start with 10 100 times smaller sample volume and still achieve the same mass of analyte on column. Another advantage of using LVI (solvent vent) is the decrease in solvent that actually reaches the detector. Usually, only 10 30% of the injection solvent actually enters the column and makes it to the detector. LVI can be applied to injection volumes ranging from a few microliters up to 1 mL or more. In most LVI approaches, the sample solvent is evaporated and removed from the inlet system before the analytes are transferred to the separation column. In this way, LVI is similar to nitrogen evaporation or rotary evaporation of the solvent, with the added benefit of being performed in the GC inlet rather than in a fume hood. Analytes that would be lost during nitrogen evaporation may be retained in the inlet and successfully analyzed via LVI. Furthermore, the LVI process can be automated and is reproducible. As in the other evaporation techniques, the LVI approach is a function of the solvent type, the inlet temperature, the vent flow of evaporation gas, and the analyte boiling point. In addition, the inlet pressure during evaporation and the inlet liner have an impact on the rate of solvent removal and analyte recovery. These parameters will be discussed in this application note.
mode), Cold Split/Splitless (also in pulsed mode), Solvent Vent and Direct mode. Hot Splitless (for 1 3 L injections) For most analysts considering LVI, their current methods are using hot splitless injection. This proven and reliable sample introduction technique has worked well for almost 40 years; however, it does present some challenges to the sample integrity and to the method developer. First, the inlet must be hot enough to flash vaporize the solvent and analytes so that the resulting vapor cloud can be transferred to the column. The inlet liner volume must be sufficiently large to contain this vapor cloud. If the liner volume is too small, the vaporized sample can overflow the liner and reach reactive surfaces, leading to analyte loss. In addition, the pressure wave generated by the vaporized sample can push back against the incoming carrier gas and enter sensitive pressure and flow control systems. Using the Agilent pressure/flow calculator [1], a 1-L injection of acetone into an inlet at 240 C and 14.5 psig expands to 288 L of gas. Most inlet liners for standard split/splitless inlets have a nominal volume of 1 mL. An increase of injection volume to only 3.5 L under these conditions creates a vapor cloud of 1 mL which could easily overflow the inlet liner. Hot splitless injection also creates a challenging environment for thermally unstable or labile analytes. Compounds such as the organochlorine pesticides DDT and endrin can rearrange to form breakdown compounds. This process is accelerated with the inlet temperatures normally used to analyze them. Effective chemical deactivation of the liner can minimize analyte breakdown. However, high inlet temperatures can decrease the lifetime of deactivated liners. Another challenge created by hot splitless injection is the opportunity for needle fractionation or analyte discrimination. The needle temperature increases as the sample is being transferred from the syringe to the inlet because the needle is in contact with the septum. The rise in needle temperature can cause the solvent to "boil" away and deposit high boiling analytes inside the needle. To avoid this fractionation problem, some analysts load a solvent plug into the syringe first and then draw up the desired sample volume (available in 7693A Automatic Liquid Sampler). The thought is that the solvent plug will wash any deposits into the inlet. An effective way to address this problem is to make a high speed injection. This minimizes the time the needle is in contact with the septum and the time the sample touches the needle. Even with these issues, hot splitless injection is a well-accepted technique. An alternative technique, such as cold splitless can address these concerns and improve the analysis results.
Experimental
MMI Operational Modes
The Agilent Multimode Inlet (MMI) uses the same liners and consumables as a standard split/splitless inlet, making it compatible with existing hot split and splitless methods. Its operational modes include: Hot Split/Splitless (also in pulsed
2
Cold Splitless (for 1 10 L injections) MMI's versatile temperature programmability allows it to perform cold split and splitless analyses. In cold splitless mode, the MMI is cooled to a temperature below the normal boiling point of the sample solvent so that when the sample is injected, no vaporization takes place. The injection is simply a liquid transfer from the syringe to the inlet. Once the syringe is removed from the inlet, the inlet is heated to vaporize the sample and transfer it to the column. The solvent vaporizes first and moves to column, allowing analyte focusing to take place as in normal hot splitless injections. The analytes subsequently vaporize and move to the column. The main advantage is that the analytes vaporize at the lowest possible inlet temperature, rather than at a constant high temperature. This minimizes thermal degradation while still allowing a wide range of analytes to vaporize. Cold splitless operations also do not thermally stress the liner as harshly as hot splitless does, prolonging its usable life. Cold splitless can also extend the amount of sample that can be injected in some cases. If a slow inlet temperature program is used, the solvent can be vaporized slowly and will not overflow the liner volume. As long as the analytes can be refocused on the column, slow inlet temperature programs cause no detrimental effects to the chromatography. Solvent Vent (for 5 1000 L injections) The solvent vent mode is the method which enables MMI to do LVI of more than 5 L. In solvent vent mode, the inlet is kept at a low initial temperature during sample injection. Pneumatically, the inlet is in split mode with a low inlet pressure. The flow of gas through the inlet liner and out to vent removes the evaporating solvent. The sample is injected slowly so that the incoming liquid is deposited on the liner wall and the solvent evaporates at a similar rate. Once the entire sample has been injected, the inlet switches to a splitless mode for analyte transfer. The inlet is then heated to vaporize the concentrated sample and any remaining solvent and the vapor is transferred to the column. After a sufficient period to ensure the sample transfer, the inlet is then switched to a purge mode to allow any remaining material in the inlet liner to be vented. During the sample injection and solvent venting period, the GC oven has been held at an appropriate temperature to allow the solvent to refocus the analytes on the column. When this refocusing is complete, the oven is then programmed to perform the separation.
LVI Method Development

An effective procedure for developing an LVI method on a MMI is to run the existing method first to determine peak areas for a small volume injection. Such results serve as a baseline for evaluating the LVI method performance. The next step is to switch to the solvent vent mode with a slightly larger injection volume (for example, 2 to 5 times larger). By comparing the resulting peak areas and accounting for the increased injection volume, the analyte recovery can be calculated and conditions can be further optimized.
Backflush
A traditional bakeout step for removing late eluters can be very time consuming for samples with complicated matrices, even as long as the analysis time. Capillary flow devices (in this case, a purged ultimate union) provide backflush [2, 3] capability. "Backflush" is a term used for the reversal of flow through a column such that sample components in the column are forced back out the inlet end of the column. By reversing column flow immediately after the last compound of interest has eluted, the long bake-out time for highly retained components can be eliminated. Therefore, the column bleed and ghost peaks are minimized, the column will last longer, and the MS ion source will require less frequent cleaning. The split vent trap may require replacement more frequently than usual.
Instrument Parameters GC MS Column MMI MMI liner Septum purge Purged Union Restrictor Syringes ALS MS parameters Solvent delay Gain factor Mass range Threshold Samples Tune file Agilent 7890A Agilent 5975C MSD HP-5MS UI, 15 m 0.25 mm 0.25 m (19091S-431UI), from inlet to purged union Constant pressure (~18 psi), chlorpyrifos-methyl RT locked to 8.297 min, 2 psi at post run for backflush Double taper deactivated, Helix (5188-5398) 3 mL/min 4 psi; 70 psi at post run for backflush 0.7 m 0.15 mm deactivated fused silica tubing (from purged union to MSD) 10 L, for splitless injections (5181-3354) 50 L, for solvent vent mode (5183-0318) Agilent 7693A 2.5 min 1 44550 0 2 atune.u
Oven Initial temperature Initial hold time Rate 1 Temperature 1 Hold time Rate 2 Temperature 2 Hold time Rate 3 Temperature 3 Hold time Total runtime Post run Oven post run temp 70 C 1 min 50 C/min 150 C 0 min 6 C/min 200 C 0 min 16 C/min 280 C 5 min 20.933 min 5 min (for backflush) 280 C
The parameters for the 25-L Solvent Vent injection were determined with the Solvent Elimination Calculator integrated in the ChemStation. This calculator was designed to help determine reasonable starting conditions for LVI methods. When the MMI is put into the PTV Solvent Vent mode, an additional button appears in the inlet screen, shown in Figure 1. In the first screen of the Solvent Elimination Calculator (Figure 2), the sample solvent and desired injection volume are selected and entered. The calculator "knows" the syringe currently installed and will only allow 50% of that volume to be injected at once. Larger injection volumes can be entered into the calculator but the injection volume will not be downloadable. The calculator also requests the boiling point of the earliest eluting analyte, as this allows the initial inlet temperature to be selected. If the boiling point is unknown, the temperature should be left at 150 C as this will work for a wide range of analytes.
Sample: 40-ppb pesticide standards in acetone (for a list of compounds, see Figure 5). Multimode Inlet (MMI) Parameter Initial temperature Initial time Rate 1 Final temperature Vent flow Vent pressure Vent time Purge time Purge flow Injection volume Injection speed Cryo Cryo fault detection Cryo use temperature Time out detection Hot Splitless 280 C 0.75 min 50 mL/min 2 L Fast Cold Splitless 30 C 0.01 min 700 C/min 320 C 1.25 min 50 mL/min 10 L Fast On 125 C On (15 min) Solvent Vent 35 C 0.35 min 700 C/min 320 C 150 mL/min 5 psig 0.33 min (from calculator, Figure 3) 1.5 min 50 mL/min 25 L 75 L/min (from calculator, Figure 3) On 125 C On (15 min)
On (liquid CO2) On (liquid CO2)
Figure 1.
Multimode Inlet "Solvent Elimination Calculator" imbedded in ChemStation for easy method development.
Figure 2.
Select solvent of choice and enter the injection volume to start the calculation.
Figure 3 shows the calculation screen. The calculator uses an initial set of inlet conditions to determine the solvent elimination rate according to fundamental theory [4]. This "Elimination Rate" does not account for other factors (for example, local cooling due to solvent evaporation) specific to LVI and is normally faster than that determined from practical experience. The "Suggested Injection Rate" does consider these factors and is designed to leave a small amount of solvent in the liner at the end of the venting period. This solvent serves as a liquid "trap" for the more volatile analytes and promotes their recovery. The "Suggested Vent Time" is determined by dividing the injection volume by the "Suggested Injection Rate."
Several variables for determining elimination rate can be set by the user in the lower portion of the window. A small change in inlet temperature has a significant impact on elimination rate. Vent flow has a linear effect such that a decrease by a factor of two in vent flow gives an equal decrease in elimination rate. As the vent pressure decreases, the elimination rate increases. Bear in mind that the vent pressure also impacts the amount of solvent that reaches the column during venting. As the vent pressure is increased, more solvent is loaded onto the column before the analytes are transferred. Finally, the type of solvent, specifically its normal boiling point, has a substantial impact on the elimination rate.
Figure 3.
The calculator calculates the injection rate and vent time according to the selected inlet temperature and vent flow.
The download screen in Figure 4 shows all of the method changes that are downloaded to the edit parameters screen. The check boxes allow the user to accept (by checking) or reject any of these parameters. The oven initial temperature and hold times are not automatically checked in case the current method requires these values to be unchanged (for example, a Retention Time Locked method).
Figure 4.
Confirm values suggested by the Calculator and download to ChemStation.

Figure 5 compares the responses of a 40-ppb standard solution from three injection modes. The bottom total ion chromatogram (TIC) is a typical 2-L hot splitless injection. Some of the 40-ppb pesticides are barely visible (80 pg each on column). The middle TIC is from a 10-L cold splitless injection. The MMI starting temperature was
30 C. In this TIC, the on column amount for each analyte is 400 pg. Lastly, the top TIC is from a 25-L solvent vent injection with MMI starting temperature at 35 C. In this TIC, the signal-to-noise ratio is significantly better than the TIC from hot splitless injection (bottom TIC), as noted in the Introduction section. The peak shape and resolution are maintained, even with the 25-L injection volume. This implies that the solvent was mostly eliminated during the injection.
Significant Improvement in Responses from LVI

Mirex Methyl parathion Heptachlor Chlorpyrifos Methyl Bromacil Malathion Chlorpyrifos Dieldrin p,p-DDE Hexazinone Ethalfluralin Trifluralin Prometon b-BHC atrazine Lindane Leptophos
25-L Solvent vent (35 C)

Cypermethrin I
16.00
Mevinphos
Dichlorvos
Propargite
Chlorothalonil
4.00
6.00
8.00
10.00
12.00
14.00
Fenvalerate I
Vernolate
10-L Cold splitless (30 C)
2-L Hot splitless (280 C)

18.00 20.00
Figure 5.
Overlay of total ion chromatograms (TICs) from three injection modes, plotted on the same scale.
Conclusion
The new Agilent Multimode Inlet (MMI) has the same form factor and uses the same consumables (for example, liners, o-rings and septa) as the existing split/splitless inlet, allowing existing hot splitless methods to be replicated. In addition, the temperature programmability permits both cold splitless and large volume injection (LVI) methods for improved detection limits. An integrated Solvent Elimination Calculator provides a complete set of initial conditions for easy LVI method development. The application results show a significant signal-to-noise improvement (lower detection limits) comparing the 25-L solvent vent injection to the 2-L hot splitless injection.
References
1. Agilent Pressure/Flow Calculator Included in the Instrument Utility DVD, available with each gas chromatograph and MMI accessory kit. Chin-Kai Meng, "Improving Productivity and Extending Column Life with Backflush, "Agilent Technologies publication, 5989-6018EN, December 2006. Matthew Klee, "Simplified Backflush Using Agilent 6890 GC Post Run Command," Agilent Technologies publication, 5989-5111EN, June 2006. J. Stanieski and J. Rijks, Journal of Chromatography 623 (1992) 105-113.
2.
3.
4.

Agilent Technologies, Inc., 2009 Printed in the USA June 18, 2009 5990-4169EN
Capillary Flow Technology for GC/MS: A Simple Tee Configuration for Analysis at Trace Concentrations with Rapid Backflushing for Matrix Elimination Application
Environmental, Drug Testing, and Forensics
Author
Harry Prest Agilent Technologies Inc. 5301 Stevens Creek Blvd. Santa Clara, CA 95051 USA
Abstract
Capillary Flow Technology devices offer the potential to enhance GC/MSD operation and robustness. In operation, they can allow rapid service of the GC column and inlet, including liner and septum, without venting or subjecting the MSD to air. In terms of robustness, late eluting compoundscan be removed from the column by "backflushing," which forces components to retreat through the column into the injection port before they damage the MSD source or compromise the next analysis. This leads to higher analytical integrity as both the column phase and the MSD can be protected. This application describes a simple arrangement for Capillary Flow Technology devices that provides ventless maintenance features with highly accelerated backflushing and minimal losses in the MSD signal. This solution supports GC analysis in constant flow mode with pressure pulsed injections and is recommended for all MSD users (in both electron impact or chemical ionization modes), including those with diffusion pump systems.
chromatographic peaks of consistent width (time) and allows optimization of MS cycle times to meet either qualitative or quantitative requirements. Also, splitless injections gained pressure pulsing or ramped flow modes, which lowered the analytes residence time in the hot injection port and confined the expansion of the injection solvent (avoiding overfilling of the liner). The power of this approach lead to continued evolution of EPC technology with the present state of the art represented in the new 7890A GC. The recent addition of Capillary Flow Technology (CFT) devices has reinvigorated and recast Deans switching and other pressure control approaches to GC analysis. One such CFT device, the QuickSwap [13], provides two important capabilities to GC/MS: 1) The ability to service and/or replace the entire analytical column or the injection port liner and septum without venting the MSD (yet still retaining high vacuum integrity) 2) The ability to remove from the column late-eluting, highly retained components that elute after the target compounds of analytical interest by reversing the carrier flow direction through the column in what is called backflushing. With the oven temperature elevated and the flow reversed, these very high boiling interferences can be pushed off the column into the split vent and thereby prevent degradation of the column phase or the detector. A schematic representation of the arrangement that makes this possible is shown in Figure 1.
Introduction
The introduction of Electronic Pressure Control (EPC) was a major advance for GC and especially GC/MS analysis. EPC allowed development of the constant flow mode of analysis, which generates
1 to 4 psi
Aux Aux EPC EPC
During GC run
Split Vent
0.8 to 2.5 mL/min
Z mL/min
S/SL Inlet
MSD
17.1 cm
30 m
Quickswap
10 to 75 psi
MSD Optimum < 1.5mL/min
After GC run
Split Vent
Aux EPC
10 to 25 mL/min
S/SL Inlet
2 psi
Figure 1.
30 m
17.1 cm
MSD
Quickswap
Schematic of QuickSwap arrangement.
Every new approach has a downside and for QuickSwap it is the additional makeup flow required to purge the QuickSwap device during analysis which dilutes the signal in the GC/MSD. This is not an issue for many users since the sensitivity of the MSD is usually more than adequate. However, analysis at trace concentrations has more stringent requirements and maintaining a signal closely comparable to that of a single continuous column is essential. Another CFT configuration for GC/MSD applications designed specifically for trace GC/MS analysis where customers do not wish to surrender signal is possible using the QuickSwap or any of several other CFT devices. In this alternate configuration, the CFT device is located in the middle of the analytical column, essentially splitting the column in half. For example, a 15-m column preceeds and follows a CFT tee. Schematically this arrangement is illustrated in Figure 2. The auxiliary EPC device adds just enough pressure (flow) to match the flow (pressure) from the first column, so there is little flow addition and therefore less dilution and loss in the GC/MSD signal. Backflushing is similarly simple; the pressure or flow is dropped in the first column section while the second section column flow is increased.
Advantages of this pressure controlled tee (PCT) approach are similar to those of QuickSwap, such as: Service of injection port liner and septum without venting the MSD Column cutback or replacement of the front or first column without venting the MSD But additional advantages of the PCT arrangement over QuickSwap are: Minimal or no signal loss (in EI- or CI-MS) is obtained because of the very small additional makeup gas flow. Constant flow mode and pressure pulsed injections are straightforward. This configuration is suitable for diffusion pumped systems and allows backflushing in diffusion pumped systems. Backflushing is more rapid and can be initiated earlier. This application details some configurations and provides an example of backflushing.
Split/splitless injection port
Pressure/flow controller
Vent
7890A GC
Z mL/min
CFT Device
Z mL/min
5975C MSD EI mode
Z mL/min
15-m HP-5 ms (0.25 mm id 0.25 m)
15-m HP-5 ms (0.25 mm id 0.25 m)
Figure 2.
Schematic of pressure controlled tee arrangement for the GC-MSD: solid lines indicate the forward flow during GC/MSD analysis and the dashed lines indicate backflushing flows.
Experimental
A number of devices can be used in this approach and those arrangements will be cited later, but for these experiments the instrument configuration was as follows: 7890A GC with split/splitless ports in front and back and a 7683B ALS 5975C MSD with performance turbomolecular pump 2 HP-5ms 15 m 0.25 mm id 0.25 m film columns (19091S-431) CFT device: 2-way unpurged splitter (G318160500) with SilTite ferrules and nuts CFT GC mounting hardware: dual-wide mounting bracket (G2855-00140) or single-wide mounting bracket kit (G2855-00120) Deactivated 0.25 mm id column approximately 1 m long 2 CFT blanking plugs (G2855-60570 or as G2855-20550 with G2855-20593)
As an overview of the configuration, the 1-m column was connected to the back injection port and to the first position on the CFT splitter using the appropriate SilTite fittings. (This CFT device has three connection points and is really best thought of as a simple tee reminiscent of glass Yor T-connectors and will be referred to as a CFT device or CFT tee from here forward). One of the 15-m HP-5ms columns was connected at the uppermost position on the CFT tee and the other end through the transfer line into the MSD as usual. The other 15-m HP-5ms column was connected to the midpoint of the CFT device and the front injection port. In detail, the arrangements were as follows. The CFT tee was attached to the forward position on the mounting hardware on the right side in the GC oven. The 1-m long section of guard column was wound on a spare column cage and hung on the column hanger in the back of the oven. (This could simply be added to one of the 15-m HP-5ms column
cages to avoid the extra cage.) Using a Vespel/ graphite ferrule, one end was connected to the back injection port and the other end to the lowest connection of the CFT device with a SilTite ferrule and nut. The other two CFT tee connections were sealed with CFT blanking plugs and the back injection port was pressure tested as described in the 7890A Advanced User Guide (part number G343090015). One of the 15-m columns was then hung on the cage carrying the 1-m column and installed with one end through the MSD transfer line. Since this column (column #2) can be expected to have a rather long life as it will be protected by the upstream column, a SilTite ferrule is recommended for the transfer-line seal. These ferrules do not develop leaks as the transfer-line temperature is cycled; however, the Vespel/graphite ferrules can shrink and develop leaks. (Note that if the surface of the transfer line is very worn it may fail to seal well, in which case the Restek Agilent interface cleaner [P/N 113450] can be used to resurface the sealing surface if very carefully employed). The
other end of this GC column was connected to the uppermost connection on the CFT tee with the SilTite ferrule. The upstream 15-m GC column (column #1) was hung on the other 15-m column cage and installed in the front split/splitless injection port with a Vespel/graphite ferrule, liner, and BTO septum, as usual. The other end was connected to the CFT tee middle post and, after temporarily removing the other connected columns, blanked off and pressure tested as above. All connections were then re-established to the CFT tee with the 1-m column in the lower position; the front, first column (#1) connected in the middle position; and the rear, second or MSD column (#2) in the uppermost connection. Helium was supplied to both the front and back ports, and a helium leak detector was used to check for any leaks. A picture of the arrangement is shown in Figure 3.
To MSD second column
From front inlet first column From back inlet flow control
Figure 3.
Picture of the installed pressure controlled tee arrangement for the GC/MSD.
GC Configuration The GC can be configured in several ways. However, for instructional purposes and those of these experiments, the GC was configured as follows: Column #1: Inlet: 30 m 0.25 mm id 0.25 m column Front injection port: pulsed splitless mode, split flow 15 mL/min MSD (vacuum) Constant flow 15 m 0.25 mm id 0.25 m column Back injection port: split mode, split flow 15-mL/min MSD (vacuum) Constant flow
Operating with Pressure Pulsed-Splitless Injection Figures 4A and 4B show screen captures of the 7890A GC configuration for a standard pressurepulsed splitless injection with constant flow mode operation; they show the front and back injection port parameters. Remember, the arrangement is set up such that the front port, into which the sample will be injected, is configured as if a 30-m column were installed into the MSD. Typical pressure-pulse conditions are set for these parameters: a 25 psi pulse for 0.5 minutes; split flow on at 0.75 minutes at 50-mL/min; with gas saver on at 2 minutes at 15-mL/min. The general rules apply for pressure-pulsed splitless injections: given a particular liner, inlet temperature, injection volume, and solvent, the expansion of the solvent is confined to a fraction of the interior volume (< 0.75) of the liner by the pressure applied. Figure 4B shows that the back injection port is in split mode, at 120 C (to remove water background), with split flow and gas saver set at 15-mL/min flow.
Outlet: Mode: Column #2: Inlet: Outlet: Mode:
The flows were set to 1.2-mL/min, all zones were left cold, and the MSD power was turned on. With the MSD and GC zones still cold, the MSD background was checked to be sure m/z 28 was decreasing, indicating that the system was tight. Only after there was confidence that there was no leak were other zones brought up to temperature.
Figure 4A (upper panel).
Typical pressure-pulsed splitless injection parameters for constant flow: front injection port.
Figure 4B (lower panel).
Typical pressure-pulsed splitless injection parameters for constant flow: back injection port (not used for injection but for column control).
Figures 5A and 5B show the constant flow mode settings for the two columns. The front column flow is the typical 1.20 mL/min, but the back
column flow is slightly higher at 1.25 mL/min to prevent any backflow. Essentially the additional flow is equivalent to an extra meter of column length.
Figure 5A (upper panel). 6
Typical pressure-pulsed splitless injection parameters for constant flow: First column section (configured as a 30-m column).
Figure 5B (lower panel).
Typical pressure-pulsed splitless injection parameters for constant flow: Second column section (configured as a 15-m column).
column configuration. Both peak height and area remain the same, indicating that there is no loss in Figure 6 shows the results for pressure-pulsed signal. This is as expected since no signal dilution splitless injections of octafluoronaphthalene (OFN) is taking place. There is a slight degradation in S/N at 1-pg/L acquired in selected ion monitoring for the CFT tee results as the background noise is (SIM) with the two 15-m column and CFT tee conraised by about 35% due to the additional flow configuration and the standard 30-m continuous troller. The important point is that the signal is preserved at trace levels.
CFT tee
Abundance
Standard 30-m column
5.00
5.05
5.10
5.15
5.20
5.25
5.30
5.35
5.40
5.45
5.50
5.55
5.60
5.65
5.70
5.75
5.80
5.85
Time
Figure 6.
Reconstructed total ion chromatogram (RTIC) of three replicate SIM acquisitions of octafluoronaphthalene using pulsed splitless injection with CFT tee (left profiles) and with a standard 30-m continuous column configuration (right profiles). 7
Chromatographic Character Beyond preserving signal, the CFT device should exhibit reasonable chromatographic performance. One indication of chromatographic integrity is the peak shape profiles of the fatty acid methyl esters (FAMEs). The result for GC/MS analysis of a FAMEs standard acquired using a metabolomics method is shown in Figure 7 and suggests very little degradation of chromatography using this PCT. This can be expected as the path is deactivated and the path length in the channels in the
PCT relative to the linear velocity suggests a relatively rapid transit through the device. Another common chromatographic test used in organochlorine pesticide analysis (as in USEPA method 8081) examines degradation of 4,4'-DDT and Endrin. This degradation test was developed to indicate the degree of activity of the injection port by examining the amounts of DDD and DDE products of DDT and the ketone and aldehyde products of Endrin. The situation is complicated here as the degradation products can be generated in both the injection port and the CFT tee. How-
Abundance 6.00
8.00
10.00
12.00
14.00 Time
16.00
18.00
20.00
22.00
24.00
18.30
18.50
18.70
18.90
Abundance 6.00
8.00
10.00
12.00
14.00 Time
16.00
18.00
20.00
22.00
24.00
Figure 7.
Reconstructed total ion chromatogram (RTIC) of a multicomponent FAMEs standard using pulsed splitless injection with CFT tee (upper) and the reconstructed extracted ion chromatogram (REIC) for m/z 74. The enlarged panel is for octadecanoic methyl ester.
ever, because those products formed in the injection port and those formed at the CFT device will have different retention times due to differing lengths of column, the degradation contributions from the two origins should be discernable. By analyzing these known breakdown products in the PCT and then injecting the DDT and Endrin agents themselves, an estimate of the activity contributed by the CFT device can be calculated. The upper panel of Figure 8 presents the reconstructed total ion current (RTIC) for the selected ion monitoring (SIM) signals of the four breakdown products. These were acquired in SIM-scan mode with a single SIM group composed of one or two major ions for each compound so there was no time
selection for the compounds appearance. On the basis of summed areas, the total breakdown for Endrin is less than 13% with the CFT device contributing less than 10% of the total breakdown area or less than 1.2% to the area total. The DDT breakdown is less than 4% for the system; however, the CFT device contributes about 46% of the total observed breakdown and is about double the breakdown generated by the port. It is possible some DDD breakdown is hidden under the DDT peak. On the basis of the DDT to DDE contribution from the CFT tee, however, it is likely to increase the breakdown perhaps less than about another 2%. A better study would use on-column injection
DDD
Abundance
DDE EA
EK
20.00
20.20
20.40
20.60
20.80
21.00
21.20
21.40 Time
21.60
21.80
22.00
22.20
22.40
22.60
22.80
Endrin
4,4-DDT
Abundance
EK*
DDD* DDE*
EA*
20.00
20.20
20.40
20.60
20.80
21.00
21.20
21.40 Time
21.60
21.80
22.00
22.20
22.40
22.60
22.80
Figure 8.
CFT tee activity. A: the REIC of a GC-MS SIM acquisition using pulsed splitless injection with the PCT configuration of the expected degradation products of DDT and Endrin at 0.2 ng on column : 4,4-DDE (DDE), 4,4-DDD (DDD), Endrin aldehyde (EA), and ketone (EK). B: REIC for an injection of 2.0 ng of 4,4-DDT and Endrin identifying degradation products. Those with an asterix (*) are attributed to the injection port and due to the CFT device activity such as; from Endrin (5 as ketone) and from 4,4-DDT (6 as DDE). Note 7 is tenatively identified as DDMU, source unknown.
of all components, but the verdict is likely the same: the CFT device has some activity but is comparable to that of other elements (for example, in the inlet and liner). It is worth noting that this CFT device has a very long path compared to others (see the Alternative Configurations section) and that air intrusion in any part of the system will be a major issue in considering activity problems. Adding Backflush Figures 9A, 9B, and 9C show the GC parameters for adding backflush. They are quite simple. The oven temperature can remain the same as the temperature at the end of the oven program or can be raised to the isothermal or programmed temperature limits in Post Run for backflushing. Raising the column temperature during Post Run helps condition the column and removes some column bleed but is not necessary. The front column (column #1) flow is dropped to 0.3 mL/min and the back column (column #2) flow is raised to 4 mL/min. To quickly estimate the duration of the Post-Run time parameter, notice that the back column (column #2) in Figure 9C cites the column Holdup Time at a given flow. At the 1.25-mL/min shown, the Holdup Time is roughly 0.4 minutes. When the
column #2 flow is raised to 4 mL/min, the Holdup Time for back flow through column #1 will be less than this (actually around 0.26 min). But estimating that every 0.4 minute the front 15-m column section would be flushed at least once is very conservative and an adequate approximation. Five to 10 column volumes will flush this front 15-m section in less than 2 to 4 minutes, which is relatively rapid. Choose a time in this range (for example, 3 minutes) and test the effectiveness of the backflush method by injecting a sample and follow this with a solvent blank injected under the nonbackflush GC/MSD method. There should be no sign of carryover. Extend this Post-Run time if there is carryover or further raise the Post-Run temperature or both. This is a very conservative approach. Column or Inlet Servicing and Maintenance To change the liner, septum, cutback the column, or replace the front 15-m column, simply cool the inlet(s) and increase the flow on the back column (column #2) to 4 mL/min and set the front injection port pressure to OFF. It is worth saving this method (such as SERVICE-Front.M). When the head of the column is removed from the injection port, one can confirm that the carrier is flowing back up the column by immersing the tip in liquid.
Figure 9A (upper panel). 10
Adding backflushing in Post Run: oven parameters.
Figure 9B (middle panel).
Adding backflushing in Post Run: front column (column #1) parameters.
Figure 9C (lower panel).
Adding backflushing in Post Run: back column (column #2) parameters.
11
This backflow also prevents fines from the column cutting from entering the column. Make the necessary service and reattach and reload the analytical method. If a completely new 15-m column (#1) is installed, it can be conditioned in situ by setting up the backflow condition with the oven at the conditioning column temperature. Advanced Techniques: Concurrent Backflushing If the fastest possible total analytical time is the highest priority, one will realize that backflush can begin earlier than the elution of the last component. In other words, backflushing can occur
during the analytical acquisition, thereby increasing productivity. After the last compound of interest has passed the CFT tee and entered the back 15-m column, the pressure or flow through the earlier 15-m column can be dropped and compounds will cease moving forward and actually begin to retreat. When the last compound elutes, then the flow in the back column can be raised to complete backflushing. This is demonstrated in Figure 10. The calculations are also very simple. To calculate when the flow (pressure) in the front column (column #1) is to be reduced, simply subtract the Holdup Time (Figure 9C) from the last compounds
9.00
9.50
10.00
10.50
11.00
11.50
12.00
a
9.00 9.50 10.00
b
10.50 11.00 11.50 12.00
9.00
9.50
10.00
10.50
11.00
11.50
12.00
Figure 10. Example of backflushing with flow or pressure control. Upper panel: RTIC of original six-component standard. The third peak is considered the last analyte and the fourth peak the beginning of the lateeluting interferences. Middle panel: RTIC of the same standard with backflushing beginning at 10.1 min (a), where the first 15-m column (column #1) flow is dropped and at (b) where column #2 flow is increased to 4 mL/min. Note that the last analyte is retained but the late eluters never enter the MSD. Lower panel: solvent blank run without backflush after the backflush method which shows no carryover.
12
elution time. After this last compound has eluted, go into Post Run and set the second 15-m column (#2) flow to 4-mL/min (or the pumping system maximum) with the front column (#1) pressure remaining low and the oven at the final programmed temperature. This can best be accomplished in ramped flow mode or in pressure programming. Do this for two to three column volumes and test with a sample followed by a solvent blank to see if this is sufficient. Experimentation with particular samples will enable setting these requirements more efficiently.
However, the best CFT tee device appears to be the new Purged Ultimate Union (G3186-60580), Figure 11. As the name describes, this is essentially a union with a gas purging line, making it a very low dead volume tee. It occupies very little space and can be suspended from the column cage, the oven wall, or through the upper GC wall. Preliminary tests of this Purged Ultimate Union using DDT and Endrin have shown very little breakdown. Chromatographic behavior is also very good. Similarly, the carrier control need not be the back injection port split/splitless module; a Pressure Control Module (PCM) or EPC module can be used. Of the two, the Pressure Control Module may be more convenient. Most importantly, the CFT tee position itself does not need to be exactly in the middle. The best arrangements can be considered on the basis of selection against components and the rapidity of backflushing. In other words, rapid backflushing suggests a shorter upstream column #1. So another arrangement is at the two-thirds mark or a 10-m column, then the CFT tee, and then a 20-m column to create a 30-m analytical column. Here
Conclusions
Alternative Configurations The CFT is very rich and allows many possible arrangements; these are only a few suggestions or alternatives. The CFT tee used here can be replaced by a purged two-way splitter with one channel plugged (G3180-61500) or even the QuickSwap itself can be moved back from the MSD interface and suspended in the oven.
Purged Ultimate Union
Figure 11. Purged Ultimate Union.
13
backflushing would be nearly 10 times faster than the arrangement with QuickSwap and more than twice as fast as the 15-m column for the same pressure. This would be the best arrangement for the MSD with a diffusion pump. Also, in terms of analytical time, this approach would provide even higher efficiency since 10 column volumes could be flushed in about 2 minutes. If backflushing begins before the analytical run ends (as shown in Advance Techniques and in Figure 10), then in many cases the Post-Run time would be very short or entirely unnecessary, yet still provide sufficient backflushing. This would further reduce total cycle times. The joined columns need not match in many aspects. For example, a 0.32-mm id may be the first column and a 0.25-mm id the second column. In this situation it will be better to have the columns configured and described as they actually exist in the 7890A. For example, column #1 inlet is the splitless port and the outlet is the PCM module A; column #2 inlet is the PCM module A and the outlet is the MSD. Considerations of capacity, resolution, robustness, etc., can be entertained in several innovative ways to enhance productivity and data quality. This solution can also be implemented on the Agilent 6890 GC. Of course, the PCT tee configuration is not confined to the Agilent GC/MS detector, but is suitable for other detection schemes as well. Future software releases will contain a key command that will allow more functionality and greater ease of use: it will allow the user to apply the IGNORE READY = TRUE condition to the EPC device controlling the CFT tee. This will prevent the pressure pulse or other flow conditions from producing a not ready condition for the instrument.
References
1. The 5975C Series GC/MSD, Agilent Technologies publication 5989-7827EN 2. Frank David and Matthew S. Klee, Analysis of Suspected Flavor and Fragrance Allergens in Cosmetics Using the 7890A GC with Column Backflush, Agilent Technologies publication 5989-6460EN 3. Frank David and Matthew S. Klee, GC/MS Analysis of PCBs in Waste Oil Using the Backflush Capability of Agilent QuickSwap Accessory, Agilent Technologies publication 5989-7601EN (These references are available in the Literature Library at www.chem.agilent.com.)
Acknowledgements
The author is very grateful to Bruce Quimby, Wes Norman, and Matthew Klee for several informative and encouraging discussions. Also, a special thanks to Wes Norman for providing superb CFT devices tailored to the needs of the GC/MSD and an advance example of the Purged Ultimate Union.

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2008 Printed in the USA June 24, 2008 5989-8664EN
Retention Time Locking with the MSD Productivity ChemStation
Technical Overview
Introduction
A retention time is the fundamental qualitative measurement of chromatography. Most peak identification is performed by comparing the retention time of the unknown peak to that of a standard. It is much easier to identify peaks and validate methods if there is no variation in the retention time of each analyte. However, shifts in retention time occur frequently. Routine maintenance procedures such as column trimming alter retention times. In a multi-instrument laboratory running duplicate methods, the retention times for each instrument will differ from each other, even when run under nominally identical conditions. These differences in retention times mean that each instrument must have a separate calibration and integration event table, making it time-consuming to transfer methods from one instrument to another. Differences in retention time also complicate comparison of data between instruments over time.
mance. The ability to correct for degrading chromatographic performance, optimize lab resources, and still provide the correct answer saves time, money, and results in significant productivity gains. Using GC/MS, it is also possible to screen samples for the presence of target compounds using a mass spectral database of RTL spectra. The RTL mass spectral database provides additional confirmatory information in spite of changes to the chromatographic system. One such database is the G1672AA Pesticide RTL Library. This database allows for quick and easy screening of pesticide samples.
When Should I Lock My Methods?

Locking or relocking your methods should be done whenever you make any changes to the chromatographic system or move methods from one system to another (GC or GC/MS). To establish and maintain a locked method, RTL should be performed whenever: The column is changed or trimmed The method is installed on a new instrument A detector of different outlet pressure is used (GC vs. MS) System performance is validated Troubleshooting chromatographic problems
What Is Retention Time Locking?

Retention time locking (RTL) allows a close match of retention times on one Agilent Technologies GC/MSD or GC system to those on another like system with the same nominal column. By making an adjustment to the inlet pressure, the retention times on one system can be closely matched to those on another system using the same nominal column. The ability to very closely match retention times from one system to another can greatly reduce the time it takes to develop and transfer methods. RTLocked methods can also compensate for degradations in chromatographic perfor-
How Does It Work?

The process of RTL is to determine what adjustment in inlet pressure is necessary to achieve the
desired match in retention times. To lock a given method for the first time or for the reasons above, you must first develop a Retention Time vs. Pressure (RT vs. P) calibration curve. Using an established method, multiple (five) injections of the standard are used to calculate the retention times at predefined inlet pressures. The use of an automatic sampler simplifies this process. The RT vs. P calibration data for each of the five injections are saved and used to correct locked methods. The five defined pressures are: Target pressure 20% Target pressure 10% Target pressure (nominal method pressure) Target pressure +10% Target pressure +20% Even when using columns with the same part number (same id, stationary phase type, phase ratio, and same nominal length), separate and different locking calibration curves may be needed. Other examples of when separate locking calibration curves are required include: Systems with different column outlet pressures (MSD/vacuum, FID/atmospheric) Columns differing from the nominal length by more than 15% (for example, due to trimming) Systems where the predicted locking pressure falls outside the range of the current calibration Selecting the Standard Compound for RTLocking A specific compound (usually one found in the normal method calibration standard) must be
chosen and then used for both developing the locking calibration and locking all future systems. The compound, or target peak, should be easily identifiable, symmetrical, and should elute in the most critical part of the chromatogram. Compounds that are very polar or subject to degradation should be avoided. Once the target compound has been chosen and all other chromatographic parameters of the method have been determined, the five calibration runs are performed. The resulting RT vs. P calibration curve data are saved. The software is then used to select and integrate the peak used for locking.
Acquire RTLock Calibration Data.... Agilent Technologies GCs are the only ones supported. See Figure 1. If you are using an automatic sampler, the five sample injections will be made automatically. This example illustrates how RTL works on a GC/MS system. Upon completion of the fivesample analysis, Data Analysis will begin and display the nominal MS total ion chromatogram (TIC). If the system was configured as a stand-alone GC, all five chromatograms will be displayed. From these displays, you will select the compound that will be used for RTLocking of the method. Selecting the RTLock Compound Use the mouse to select the compound or peak that you would like used for locking. For GConly mode, you must select one peak from each of the five chromatograms for RTLocking. Once you have made your selection, you will be asked to allow the software to automatically find the remaining peaks. You may choose to zoom the display for better visibility. See Figure 2.
Creating an RTL Method

The following is an overview of the actual steps taken to acquire the RT vs. P calibration curve data, selecting the compound/ peak to use and locking the method. Acquiring the RTLock Data From Instrument Control, select the Instrument menu and
Figure 1.
From instrument control, you select the mode of data acquisition for RTL.
Calculating the RTLock Curve and Saving a Method Once the RTLock compound has been selected, the new RT vs. P curve for each compound will be displayed. To select a new compound and calculate new RT vs. P curves, reset the nominal MSTIC from the RTLock menu item. Select a new compound as the RTL compound and from the RTLock menu item, select Calculate New Curve from Selected Peaks. This will generate new RT vs. P calibration curves. The new RT vs. P calibration curve equation will be displayed on the screen along with the correlation coefficient. Select Yes to either create a new, or update an existing, RT vs. P calibration file. See Figure 3. Next, you can enter the name and retention time of the RTLock compound. See Figure 4 and Figure 5.
Figure 2.
From this panel, the user selects the peak used for RTL. In this example the second peak has been selected. The spectrum of the selected compound is also displayed and is used to confirm the RTL compound.
Figure 4.
Enter or confirm the name of the RTLock compound.
Figure 3.
When the RT vs. P calibration curve equation is calculated, the correlation coefficient is determined for the RTLocked compound in each of the five calibration samples. The resulting coefficient is displayed at each peak. The nominal, or no change to pressure calibration sample, has a correlation of 1.
Figure 5.
Enter or confirm the retention time of the RTLock compound.
You will then be asked to confirm the RTLock pressure that has been calculated and will be used for that method. Select Yes to confirm the new RTLock pressure and save the method. See Figure 6.
A report is also available that provides detailed information regarding the RTLock method. See Figure 8. The report includes: Method name Calibration date Instrument name Operator name Status of method (on or off)
RTLock compound name Tabular retention time/ pressure calibration table Maximum deviation RTLock curve equation and correlation coefficient data Locked retention time information (file name, acquisition date, instrument name, and operator name) Report date
Figure 6.
By selecting Yes, you RTLock and save the method.
View Current RTLock Method Setpoints and Report Once the system is calibrated and locked, you can view and confirm the RTLock setpoints by selecting View Current Method Setpoints. See Figure 7.
Figure 7.
Viewing the RTLock retention times and corresponding pressures for the RTLock method.
Figure 8.
Example RTLock Report
Running an RTLock Method Once the RTL method is created, you can analyze and process new samples. This requires that you unlock and relock the methods without editing the quantitation calibration data. The following example demonstrates how retention times might change when column maintenance is performed or a method is moved to a new GC/MS system. See Figure 9. Once the method is RTLocked, new samples are analyzed and retention times corrected. See Figure 10. Unlocking and Relocking a Method Once a method is locked you may unlock or relock it using the same or different compounds or after additional maintenance. To define a new compound for RTLocking, select RTLock Setup from the View menu in Data Analysis. See Figure 11. The nominal RTLock sample that represents the method you are working on will be displayed. Once again, use the mouse to identify a new RTLock compound.
After cut Locked
Figure 9.
Overlay of an RTLocked data file (labeled Locked) and the resulting data file after clipping one meter off the column (labeled After Cut).
Figure 10. Offset overlay of RTLocked data file and the RTCorrected datafile show peaks overlapped and corrected for column maintainence.
Figure 11. To define a new compound for RTLocking, select RTLock Setup from the View menu in data analysis.
After selecting the compound to RTLock, from the RTLock menu item, select Calculate New Curve from Selected Peaks. This will generate a new RT vs. P calibration curve. The new RT vs. P calibration curve equation will be displayed on the screen along with the correlation coefficient. Select Yes to create a new or update an existing RTLock calibration. See Figure 12. From the RTLock menu you can also: View current setpoints Calculate the RT vs. P curve Restore the original chromatograms Report the RTLock calibration Unlock the method Relock the method
Summary
RTLocking provides an easy and flexible tool that can be used to reduce the time and complexity often associated with routine chromatographic maintenance. It allows methods to be transferred between like GC/MS systems without time-intensive edits to the quant database and reacquisition of standards. It also simplifies the process for executing routine chromatographic maintenance. RTLocking can minimize mistakes and provide a productivity improvement for most applications by reducing the time and setpoint changes required to update a method.

Figure 12. RTLock view and menu choices.
Screening for Pesticides in Food Using the Japanese Positive List Pesticide Method: Benefits of Using GC/MS with Deconvolution Reporting Software and a Retention Time Locked Mass Spectral Database Application
Food
Author
Philip L. Wylie Agilent Technologies, Inc. 2850 Centerville Rd. Wilmington, DE 19808-1610 USA
Abstract
In 2006, the Japanese Ministry of Health, Labour and Welfare introduced a new system for the regulation of pesticides, feed additives, and veterinary drugs. This Positive List system stipulates that only compounds on the approved list can be used in food production and provides the framework for regulation of these compounds. These new regulations have increased the need for analytical methods capable of detecting residues of these chemicals in a wide variety of food products. The majority of the chemicals under regulation are pesticides, and their residues are most often measured by gas or liquid chromatography with mass spectral detection (GC/MS or LC/MS). To address the need for rapid and comprehensive analysis of food samples in the Japanese market, Agilent has introduced a new Japanese Positive List Pesticide Database for use with its Deconvolution Reporting Software (DRS). With this new database and DRS, analysts can screen their GC/MS data files for the 430 GCamenable pesticides that are being regulated by the Japanese government. The process is fully automated and takes about two minutes per sample.
Positive List system for the regulation of pesticides, feed additives, and veterinary drugs [1, 2]. At that time it published provisional maximum residue limits (MRLs) for 758 chemicals and designated 65 others that would be exempted from regulation. There are 15 substances that must have no detectable residues in food products because of their high risk to humans [35]. This Positive List system implies that only chemicals listed can be used in agricultural production and any residues must comply with the MRLs set by the Japanese government. Other agricultural chemicals not mentioned have a uniform MRL of 0.01 ppm. This new regulation took effect on May 29, 2006. Since its introduction, all Japanese agricultural products and imports to Japan have had to comply with the Positive List system. This has led to an increased need for screening agricultural commodities for the pesticides, feed additives, and veterinary drugs on the list. This application describes a rapid method to screen food extracts for all the GC-amenable pesticides listed in the Japanese Positive List system, together with other pesticides that are monitored by the Japanese Quarantine Stations. In all, the method can screen samples for 430 different pesticide residues. The method uses an Agilent 7890A/ 5975C GC/MS system running the MHLW GC method with the MSD operating in the scan mode. A new retention time locked (RTL) mass spectral library has been developed specifically for this method. When combined with Agilents Deconvolution Reporting Software (DRS), it is possible to screen GC/MS data files for all 430 pesticides in about two minutes per sample.
Introduction
On November 29, 2005, the Japanese Ministry of Health, Labour and Welfare (MHLW) published a
Experimental
The MHLW has published methods for the extraction and analysis of plant- and animal-based foods [6]. Depending upon the target compounds, the extracts are analyzed by GC/MS or LC/MS. The GC/MS method specifies the column phase and dimensions, oven temperature program, inlet temperature, carrier gas, ionization mode and energy, and the ions to be monitored for each compound in a SIM analysis. The method parameters used here are the same, except that the analysis is usually run in the scan
mode so that all ions are monitored for each compound. This facilitates the use of Agilents DRS, which has significant advantages over other methods. The instrumentation, software, and instrumental parameters are listed in Table 1. Samples The samples analyzed by this method were extracted using the QuEChERS method developed by Lehotay, et al. [7, 8].
Table 1.
Instrumentation and Parameters for Analyzing Pesticides According to the Japanese Positive List System Agilent 7890A Agilent J&W 30-m 0.25-mm 0.25-m DB-5MS (P/N 122-5532) 5-m 0.25-mm Siltek Deactivated Fused Silica Tubing [Restek (Bellefonte, PA USA) P/N 10026] Helium at a nominal flow rate of 1.0 mL/min with no retention gap. About 1.1 mL/min with 5-m 0.25-mm retention gap connected to head of GC column. About 1.7 mL/min with 5-m 0.25-mm retention gap connected to the head of the GC column and a QuickSwap installed and set to a pressure of 5.2 psi. Chlorpyrifos-methyl locked to 13.443 min 50 C (1 min), 25 C/min to 125 C (0 min), 10 C/min to 300 C (10 min) Split/splitless 250 C Helix Double Taper Deactivated (P/N 5188-5398) Agilent 7683B Series Injector and Tray 2 L Agilent 5975C MSD Scan 45 550 u 0 (or set according to noise level) 70 eV n=2 280 C 3.5 min 230 C 150 C Atune.u On Agilent GC/MS ChemStation (P/N G1701EA, Ver. E.01.00 or higher) Agilent Deconvolution Reporting Software (P/N G1716AA, Ver. A.03.00 or higher) NIST MS Search (Ver. 2.0d or greater) (comes with NIST05 mass spectral library Agilent P/N G1033A) Automated Mass Spectral Deconvolution and Identification Software (AMDIS_32 version 2.64; comes with NIST05 mass spectral library Agilent P/N G1033A) NIST05 mass spectral library (Agilent P/N G1033A) Agilent Japanese Positive List Pesticide Library in Agilent and NIST formats (P/N G1675AA)
Gas Chromatograph Column Retention gap (optional) Carrier gas
Retention time locking Oven temperature program Inlet Inlet temperature Inlet liner Automatic Sampler Injection volume Mass Spectrometer Acquisition mode Scan range Threshold Ionization energy Sampling rate Transfer line temperature Solvent delay Source temperature Quadrupole temperature Tune file Trace ion detection Software GC/MS instrument control Deconvolution Reporting Software Library searching software Deconvolution software
MS libraries

Several years ago Agilent Technologies introduced Retention Time Locking (RTL) for gas chromatography (GC) and GC with mass spectral detection (GC/MS). RTL software makes it possible to reproduce retention times from run to run on any Agilent GC or GC/MS, in any laboratory in the world, as long as the same nominal method and GC column are used [9]. Since any laboratory can reproduce retention times generated in another, it is possible to create mass spectral libraries that contain locked retention times. By locking their method to the published database, users can screen GC/MS files for all of the librarys compounds. Hits are required to have the correct retention time as well as the correct spectrum, which eliminates many false positives and gives more confidence in compound identifications [9, 10]. More recently, Agilent introduced Deconvolution Reporting Software (DRS) that incorporates mass spectral deconvolution with conventional library searching and quantification. DRS results from a marriage of three different GC/MS software packages: 1) the Agilent GC/MS ChemStation, 2) the National Institute of Standards and Technology (NIST) Mass Spectral Search Program with the NIST05 MS Library, and 3) the Automated Mass Spectral Deconvolution and Identification System (AMDIS) software, also from NIST. DRS performs a normal quantitative analysis for all target compounds that have been calibrated. It then sends the data file to AMDIS, which deconvolutes all the spectra in the total ion chromatogram. (A discussion of deconvolution principles follows.) The deconvoluted spectrum of each peak is then
searched against a target compound library in this case Agilents Japanese Positive List Pesticide Library, containing 430 compounds. Hits are identified by spectral matching and by comparing their locked retention times to values stored with the library. Because retention times are locked to the library, very narrow retention time windows are used typically 10 seconds around the librarys value. For confirmation, the deconvoluted spectra of all hits are compared to the entire NIST05 mass spectral library. The results are summarized in a simple report. Deconvolution While a thorough discussion of deconvolution is beyond the scope of this application, the basic principles are illustrated in Figure 1. The chromatographic peak shown in black looks Gaussian, but it is actually the result of at least three compounds that were only partially resolved. The spectrum at the apex of this peak is composed of ions from all three compounds, some of which are common to two or three of the overlapping analytes. AMDIS deconvolutes the chromatogram and pulls out cleaned spectra from the overlapping peaks. In most cases AMDIS is very successful at isolating a compounds spectrum from column bleed, other analytes, and coextracted interferences, even when interference abundances are far greater than the target analyte. Using the deconvoluted full spectrum, AMDIS searches each peak against Agilents Japanese Positive List Pesticide Library and reports a hit if the match quality exceeds a user-settable threshold. Since compounds are also required to have the
Figure 1.
An illustration of the mass spectral deconvolution process. 3
correct retention time false positives are virtually eliminated. At the same time, false negatives are minimized because AMDIS deconvolution is able to generate library searchable spectra, even for traces of pesticides that are obscured by heavy matrix. As a confirmation step, the deconvoluted spectra of all AMDIS hits are searched against the 163,000compound NIST mass spectral library; for this step, there is no retention time requirement. More details about DRS can be found in earlier publications [11-17]. Experience in this laboratory and others [14, 15] has shown that DRS is the fastest, most comprehensive method for pesticide screening and that it produces the fewest false positives and false negatives. Pesticides Included in the Japanese Positive List Pesticide Database Pesticides included in the Japanese Positive List Pesticide Database were derived from three sources. First are the pesticides listed for analysis by GC/MS under the current MHLW regulations. Some additional pesticides were added because they were published in a recent paper written by analysts from the Kobe and Yokahama Quarantine Stations[18]. A third source of pesticides came from the Japanese Office of Food Safetys Imported Foods Monitoring Plan for FY 2006. This plan says that imported foods and agricultural products must conform with Schedule 6 of this document, which is a list of 447 pesticides [19]. Of the 447 pesticides, some are better ana-
lyzed by GC and others by LC. Some are amenable to either method, but the document does not specify the methods to be used. Agilents Japanese Positive List Pesticide Database includes all of the compounds in this list that can realistically be analyzed by GC/MS. The actual MRL values appear in two lists one containing the finalized MRLs [4] and another with provisional MRLs [5]. On February 5, 2007, the Japanese MHLW published revised versions of both MRL lists. Sixty-seven new drugs and pesticides were added to the revised provisional list. Of the pesticides in the original Japanese Positive List, 265 were to be analyzed by GC/MS. Many of the 59 newly added pesticides are also amenable to GC/MS analysis. The resultant Japanese Positive List Pesticide Database contains the mass spectra and locked retention times for 430 pesticides and one internal standard (phenanthrene-d10). Analyzing Samples Figure 2 shows a chromatogram of a strawberry extract that was spiked with eight pesticides (500 ng/g each) and analyzed using the Japanese Positive List method. The first 18 minutes of the chromatogram are very dirty, with many large peaks from endogenous compounds that were extracted along with the pesticides. DRS was run on this sample but the GC/MS was not calibrated for the target compounds. Instead,
Figure 2.
Chromatogram of a strawberry extract that was spiked with eight pesticides each at 500 ng/g. Seven of the eight pesticides are in the Japanese Positive List Pesticide Database and were easily identified by DRS.
an average response factor was used for all pesticides in the database. Figure 3 shows the report that was generated by DRS in about 90 seconds. Temephos was spiked into the sample but was not found by DRS because this compound is not in the database. Pyridaben was identified by the Agilent ChemStation but was not confirmed by AMDIS. The ChemStation uses four ions for identification while AMDIS uses the whole deconvoluted spectrum. It is rare that the ChemStation identifies a target compound that AMDIS doesnt find. A quick review of the data showed that pyridaben was not present in the sample. As with methiocarb, it is much more common for AMDIS to find compounds that are not reported by the ChemStation. This example shows how DRS helps to eliminate both false positives and false negatives. For each of the seven hits, DRS provided the retention time, CAS number, and name of the pesticide. Column four shows the amount of each hit as determined by the ChemStation software. Agilent supplies four methods for use with the Japanese Positive List Pesticide Database that can be used with different instrument configurations. Each of these methods comes with a quant database for all 430 target pesticides. The response factors proMSD Deconvolution Report Sample Name: Data File: Date/Time: Strawberry extract C:\msdchem\1\DATA\Strawberry_TID_2uL.D 04:44 PM Thursday, Sep 6 2007
vided with these methods were derived by taking the average response factor for about 25 representative pesticides. These response factors are not correct, but allow the analyst to estimate concentrations of the target compounds. The estimated values are usually within an order of magnitude of the actual concentrations. For accurate quantitative analysis, laboratories must do their own calibrations. Since it is unlikely that a laboratory would calibrate for all 430 pesticides, they can use these average values for compounds that are not calibrated. When a new compound is identified by DRS, it can be added to the calibration solutions. Column five shows how well the deconvoluted spectrum matches the target pesticide library spectrum (100 = perfect match). Also under the AMDIS heading is the retention time difference between the library and observed values (column 6). Because RTL was used in the creation of the library and for the strawberry analysis, the observed values are extremely close to the library. Remarkably, only one compound, fenamiphos, deviated from its library retention time by more than one second. The last step in the DRS analysis is to compare the
The NIST library was searched for the components that were found in the AMDIS target library. Agilent ChemStation Amount (ng/L) 0.99 1.4 1.09 0.91 2.57 3.32 0.04 5
R.T. 11.4914 12.3647 12.4698 12.5726 12.7135 14.0851 14.4553 16.0453 20.630 12.575 Figure 3.
Cas # 298022 13071799 333415 1517222 298044 2032657 55389 22224926 96489713
Compound name Phorate Terbufos Diazinon Phenanthrene-d10 Disulfoton Methiocarb Fenthion Fenamiphos Pyridaben Phenanthrene-d10
AMDIS R.T. Diff Match sec. 95 97 95 98 87 85 99 96 0.4 0.5 0.3 0.6 0.3 0.5 0.3 2.4
NIST Reverse Hit match number 90 89 80 84 84 81 90 88 1 1 1 1 1 1 1 1
DRS report for the strawberry sample whose chromatogram is shown in Figure 2. Amounts shown in column 4 are only approximations, derived using an average response factor rather than individual pesticide calibrations.
deconvoluted spectrum for each hit to the entire NIST05 mass spectral library. If it finds the pesticide among the top 100 library hits, it prints the match value and the hit number in the last two columns. This step provides further verification of the compounds identity. Figure 4 shows the chromatogram of a mixed fruit extract that was not spiked. DRS found two fungicides diphenylamine and thiabendazole, along with one organophosphorus pesticide azinphosmethyl (Figure 5). In the initial analysis, dipheny-
lamine was not reported by the ChemStation even though it was an excellent match in AMDIS and to the NIST05 library. In this case, the diphenylamine peak was integrated manually using the Q-Edit feature of the ChemStation and DRS was rerun using the existing quant results. Figure 6 shows the new DRS report for the mixed fruit extract with quant results for diphenylamine. The reported values are only estimates because an average response factor was used for quantification. Azinphos-methyl was identified in the mixed fruit
Figure 4.
Chromatogram of an unspiked mixed fruit extract.
MSD Deconvolution Report Sample Name: Data File: Date/Time: Mixed fruit C:\MSData\Sept 04_07 Lehotay samples using TID & Japanese method\Mixed Fruit_TID_2uL.D 04:01 PM Monday, Sep 10 2007
The NIST library was searched for the components that were found in the AMDIS target library. Agilent ChemStation Amount (ng/L)
R.T. 10.7853 12.5733 15.3882 19.4523 12.575 Figure 5.
Cas # 122394 1517222 148798 86500
Compound name Diphenylamine Phenanthrene-d10 Thiabendazole Azinphos-methyl Phenanthrene-d10
AMDIS R.T. Diff Match sec. 97 99 0.1 0.6 2.6 0.6
NIST Reverse Hit match number 91 84 92 76 1 2 1 2
0.66 0.02 5
99 63
Initial DRS report for the mixed vegetable sample whose chromatogram is shown in Figure 4.
MSD Deconvolution Report Sample Name: Data File: Date/Time: Mixed fruit C:\MSData\Sept 04_07 Lehotay samples using TID & Japanese method\Mixed Fruit_TID_2uL.D 04:11 PM Monday, Sep 10 2007
The NIST library was searched for the components that were found in the AMDIS target library. Agilent ChemStation Amount (ng/L) 0.92 0.66 0.02 5
R.T. 10.783 12.5733 15.3882 19.4523 12.575 Figure 6.
Cas # 122394 1517222 148798 86500
Compound name Diphenylamine Phenanthrene-d10 Thiabendazole Azinphos-methyl Phenanthrene-d10
AMDIS R.T. Diff Match sec. 97 99 99 63 0.1 0.6 2.6 0.6
NIST Reverse Hit match number 91 84 92 76 1 2 1 2
DRS report after using Q-Edit to integrate the diphenylamine peak. DRS was run a second time using the existing quant results.
extract even though it was buried by coeluting compounds. The benefits of deconvolution are apparent in Figure 7, which shows a screen capture from the AMDIS software. The arrow in Figure 7a shows where azinphos-methyl elutes. Figure 7b shows the spectrum at that point while Figure 7c juxtaposes the deconvoluted spectrum (white) with the library spectrum (black). Without
deconvolution it would have been impossible to identify azinphos-methyl by library searching. Even knowing that azinphos-methyl was present, it was impossible to create a library-searchable spectrum by standard background subtraction techniques. After deconvolution, the spectrum is a reasonable match to the library.
Figure 7.
Screen capture from AMDIS showing: a) the total ion chromatogram of a mixed fruit extract, b) the spectrum where Azinphos-methyl elutes, and c) the deconvoluted spectrum (in white) juxtaposed to the library spectrum for Azinphos-methyl (black). 7
Several studies have shown that DRS is capable of identifying pesticides that are not found by conventional pesticide analysis [14, 15]. Most GC and GC/MS pesticide methods target a fixed number of compounds and generally do not identify compounds unless they are on the target list. Moreover, it can easily take a skilled analyst 30 minutes or more per sample to verify the analytical results. By contrast, Agilents DRS method can screen for all of the GC-amenable pesticides of interest to the Japanese government (430 pesticides) in about two minutes.
4. Table in item 6 (1), Section A General Compositional Standards for Food, Part I Food: The maximum residue limits of substances used as ingredients of agricultural chemicals in foods http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/dl/index-1a.pdf 5. Table in item 7 (1), Section A General Compositional Standards for Food, Part I Food: The maximum residue limits of substances used as ingredients of agricultural chemicals in foods (Provisional MRLs List) http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/dl/index-1b.pdf 6. Analytical Methods for Residual Compositional Substances of Agricultural Chemicals, Feed Additives, and Veterinary Drugs in Food, Department of Food Safety, Ministry of Health, Labour and Welfare, http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/dl/060526-1a.pdf 7. S. J. Lehotay, A. de Kok, M. Hiemstra, and P. van Bodegraven (2005) J. AOAC Int. 88, 595614 8. M. Anastassiades, S. Lehotay, D. tajnbaher, and F. J. Schenck (2003) J. AOAC Int., 86, 412-431 9. V. Giarocco, B. Quimby, and M. Klee, Retention Time Locking: Concepts and Applications, Agilent Technologies, publication 5966-2469E, www.agilent.com/chem 10. H. Prest, P. Wylie, K. Weiner, and D. Agnew, Efficient Screening for Pesticides and Endocrine Disrupters Using the 6890/5973 GC/MSD System, Agilent Technologies, publication 5968-4884E, www.agilent.com/chem 11. Philip L. Wylie, Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library, Agilent Technologies, publication 5989-5076EN, www.agilent.com/chem 12. Bruce Quimby, Screening for Hazardous Chemicals in Homeland Security and Environmental Samples Using a GC/MS/ECD/FPD with a 731 Compound DRS Database, Agilent Technologies, publication 5989-4834EN, www.agilent.com/chem 13. Mike Szelewski and Bruce Quimby, New Tools for Rapid Pesticide Analysis in High Matrix Samples, Agilent Technologies, publication 5989-1716EN, www.agilent.com/chem
Conclusions
Agilent Technologies has introduced a new retention time locked mass spectral database to address the needs of laboratories that must comply with the Japanese Positive List pesticide regulations. This database contains mass spectra in Agilent, NIST, and AMDIS formats along with locked retention times for 430 pesticides. The list of pesticides was derived from the most up-to-date publications of the Japanese Ministry of Health, Labour and Welfare and its agencies. The locked retention times were obtained using the GC/MS conditions recommended by the MHLW. Together with Agilents Deconvolution Reporting Software, analysts can screen data files for all 430 pesticides in about two minutes per sample. The method is rapid, comprehensive, accurate, and automated so it depends less upon the skill of the individual analyst.
References
1. Introduction of the Positive List System for Agricultural Chemical Residues in Foods, Department of Food Safety, Ministry of Health, Labour and Welfare, June 2006 http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/introduction.html 2. Positive List System for Agricultural Chemical Residues in Foods, Ministry of Health, Labour and Welfare http://www.mhlw.go.jp/english/topics/foodsafety/positivelist060228/index.html 3. Maximum Residue Limits (MRLs) List of Agricultural Chemicals in Foods, The Japan Food Chemical Research Foundation http://www.m5.ws001.squarestart.ne.jp/ foundation/search.html
14. Chris Sandy, A Blind Study of Pesticide Residues in Spiked and Unspiked Fruit Extracts Using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1654EN, www.agilent.com/chem 15. Philip Wylie, Michael Szelewski, Chin-Kai Meng, Christopher Sandy, Comprehensive Pesticide Screening by GC/MSD Using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1157EN, www.agilent.com/chem 16. Xiaofei Ping, Chin-Kai Meng, Michael Szelewski, Building Agilent GC/MSD Deconvolution Reporting Libraries for Any Application, Agilent Technologies, publication 5989-2249EN, www.agilent.com/chem 17. C. Lesueur and M. Gartner, Routine Identification and Quantification of Pesticide Multiresidues in Fruit and Vegetable Samples with Full Scan, SIM, and Deconvolution Reporting Software, Ernahrung/Nutrition 29 (11), 466471 (2005) 18. Y. Hirahara, et Al., Validation of Multiresidue Screening Methods for the Determination of 186 Pesticides in 11 Agricultural Products Using Gas Chromatography (GC), J. Health Science 51 (5) 617-627 (2005) 19. Imported Foods Monitoring Plan for 2006 Notices Nos. 0331006 and 0525001, Office of Imported Food Safety, Ministry of Health, Labour and Welfare, May 26 2006 http://www.mhlw.go.jp/english/topics/ importedfoods/dl/4.pdf
Acknowledgements
The author would like to thank Drs. Steven Lehotay and Katerina Mastovska of the U.S. Department of Agriculture for providing the samples of spiked and unspiked extracts.

Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2007 Printed in the USA September 21, 2007 5989-7436EN
Reducing Analysis Time Using GC/MSD and Deconvolution Reporting Software
Application
Food and Flavors
Authors
Mike Grady, Steve Morrison, and Bob Deets Campbell Soup Company Campbell Place Camden, NJ 08103 USA Mike Szelewski Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
Analyzing a complex matrix can be accomplished using multiple specific detectors, but a significant time savings is realized using GC/MSD. Using an available Retention Time Locked database with Deconvolution Reporting Software (DRS) adds a second expert opinion. GC/MSD with DRS provides the fastest methodology to the fewest number of false positives/negatives with the greatest confidence in the results.
dual flame photometric for phosphorus and sulfur (DFPD). Most of the analytes, depending on their molecular formula, could be run on two of these specific detectors for confirmation. However, that would take twice the time. A second choice is to run each sample on unlike columns of differing polarities to the same type detector. Both of these approaches, using ECD, NPD, and/or DFPD, involve more sample handling/tracking, usually twice the number of runs and subsequent data interpretation, and still only have retention time as an identifier. Some methodologies require final confirmation by GC/MSD. Also, it is very difficult to analyze for hundreds of compounds using a nonMS method due to peak overlaps. Campbell Soup Company has been moving from GC-specific detector analyses to GC/MSD analyses for the above-mentioned reasons. A significant time savings can be realized using GC/MSD. It is imperative, however, that data quality be maintained while increasing productivity. Agilent Technologies recently introduced Deconvolution Reporting Software (DRS) for use with a GC/MSD system [1]. DRS automatically combines the results from the MSD ChemStation Quantitation with AMDIS Deconvolution and NIST Search into an easy-to-read report. Using an available Retention Time Locked database reduces methods development for DRS and speeds data comparison among labs. Positives found by normal quantitation are confirmed, and the analyst is directed to further target compound hits to verify. The analyst
Introduction
Analyzing a complex matrix can be accomplished using multiple specific detectors such as electron capture (ECD), nitrogen-phosphorus (NPD), and
need not spend the time to review hundreds of possible compounds, which could take hours per sample. GC/MSD with DRS provides the fastest methodology to the fewest number of false positives/negatives with the greatest confidence in the results. It typically takes less than 2 minutes to process a DRS Report. The purpose of this application note is to show the use of GC/MSD with DRS on complex samples in Campbells Research Lab. It is also intended to show a time savings while maintaining data quality.
The HP-5MS column was used by Agilent to develop the original method and is run in constant pressure mode. Constant pressure methods can be precisely scaled, or sped up, for faster analyses. The retention times of 927 analytes have been recorded on this column. The system is Retention Time Locked to methyl chlorpyrifos at 16.596 min. The primary benefit of RTL for a food laboratory is the ability to maintain retention times after clipping or changing the column. Other benefits include: 1) constant quantitation database and integration events times; 2) switching group times remain constant for laboratories performing SIM analyses; 3) multi-site laboratories can easily compare data; 4) commercially available RTL databases can be used. Additional information is available at www.agilent.com/chem, with application notes detailing the numerous benefits of RTL. The injector parameters are modified from the default Fast Injection to Variable. This allows matching the injector parameters to the PTV-SV requirements. Most importantly, the injection speed is slowed to accommodate the evaporation of the solvent during injection. The 5973N MSD had been upgraded to an inert source and Performance Electronics. The inert source has shown improved response and less degradation for active compounds. Performance electronics minimize noise at faster Scan sampling rates and allow shorter dwell times and more ions/group for SIM acquisitions. A sampling rate of 2 and a scan range of 35-500 yields 3.12 scans/sec. This results in at least 10 data points across the earliest narrow peaks that are 0.055 min wide. This is a good number of points for both quantitation and deconvolution while minimizing noise.
Experimental
Instrument Operating Parameters The instrument operating parameters are listed in Table 1. These conditions may have to be optimized for use in another laboratory. A Programmable Temperature Vaporizing inlet (PTV) is used in the solvent vent mode (SV). The sample is injected at or below the boiling point of the solvent, in this case 50 C. Solvent is evaporated through the split vent line with helium at 200 mL/min for 0.3 min. At 0.5 min, the PTV is rapidly heated to 320 C, transferring the analytes, with minimum solvent, onto the column. The column is held at the initial temperature of 70 C during this process. The PTV-SV allows larger volumes of sample to be injected, 10 L for this study, versus the typical 1 L for this column. The PTV inlet liner, 5183-2037, is multi-baffled and deactivated. It does not contain glass wool, which could contribute to active compound degradation. This liner has sufficient capacity to accommodate the 10 L injection volume
Table1. GC Inlet Mode
Gas Chromatograph and Mass Spectrometer Conditions Agilent Technologies 6890N EPC PTV Solvent vent C/min 600 50 On 50 C 10.00 min (On) On 17.98 psi (On) 0.30 min 200.0 mL/min 0.0 psi 400.0 mL/min 1.00 min 403.9 mL/min On 20.0 mL/min 2.00 min Helium Agilent PTV Liner part# 5183-2037 120 V C/min 25 3 8 41.87 min 0.5 min 325 C Agilent Technologies HP 5 MS, part# 19091S-433 30.0 m 0.25 mm 0.25 m Constant Pressure 17.98 psi 1.9 mL/min Front MSD Vacuum Next C 70 150 200 280 Hold min 2.00 0.00 0.00 10.00 Next C 50 320 200 Hold min 0.50 3.00 0.00
RTL Front injector Sample washes Sample pumps Injection volume Syringe size PreInj Solv A washes PreInj Solv B washes PostInj Solv A washes PostInj Solv B washes Viscosity delay Plunger speed Injection speed Draw speed Dispense speed PreInjection dwell PostInjection dwell MSD Upgrades Solvent delay Low mass High mass Threshold Sampling Scans/sec Quad temperature Source temperature Transfer line temp Tune Type Calibration Standards
System Retention Time Locked to methyl chlorpyrifos at 16.596 min
Temperature ramp Initial Ramp 1 Ramp 2 Cryo Cryo use temperature Cryo timeout Cryo fault Pressure Vent time Vent flow Vent pressure Purge flow Purge time Total flow Gas saver Saver flow Saver time Gas type PTV liner Oven Oven ramp Initial Ramp 1 Ramp 2 Ramp 3 Total run time Equilibration time Oven max temperature Column Length Diameter Film thickness Mode Pressure Nominal initial flow Inlet Outlet Outlet pressure
0 3 10.00 L 25.0 L 0 1 2 2 1s Variable 50.00 L/min 600.00 L/min 1000.00 L/min 0.00 min 0.00 min Agilent Technologies 5973N Inert source and Performance Electronics 4.00 min 35 amu 500 amu 50 2 3.12 150 C 230 C 280 C Autotune
Prepared from certified reference standards available from ChemServe and Crescent Chemical Company. All standards were corrected for purity.
Extraction Procedure An appropriate amount of commodity is weighed, typically 10-15 grams. Surrogates and, if necessary, fortification standards (spike) are added. The commodity is extracted with 1% acetic acid in acetonitrile, centrifuged [2], and passed through an SPE cartridge [3]. Analytes are eluted from the cartridge using acetonitrile/toluene. A 1-gram volume equivalent is taken from the eluant and internal standard(s) is added. The extract is brought to near dryness and solvent exchanged into ethyl acetate for GC/MSD analysis.
trace quantities (< 0.02 ppm) to 0.48 ppm. In most cases, the trace quantities were not found by the GC/MSD standard quantitation using 3 qualifier ion identification. As a verification of the methodology, a second sample of strawberries and tomatoes were each fortified with six analytes at the 0.1 ppm level, together with the Aldrin SS/IS. The analyses results for these spiked samples are shown in Table 3. There are excellent recoveries for most analytes. Responses for two analytes in tomato, atrazine, and permethrin were higher than expected and could be due to matrix enhancement during injection. Time constraints did not allow for further investigation. Duplicate results are also shown in Table 3. The chlorothalonil in tomato at 0.08 ppm compares favorably with the 0.09 ppm found in the first sample. The same is true for captan at 0.47 ppm in the duplicate versus 0.48 ppm in the original sample. The GC/MSD system was calibrated for more than 50 compounds. Using DRS, the analyst can get a verification of the presence of those 50 compounds together with an automated expert second opinion of other compounds that may be present. This second opinion is in two distinct parts. First, the deconvolution of the complex TIC with subsequent matching of clean spectra to a database is provided. Second, the matching of these clean spectra to an independent database, in this case the NIST05a library of > 163,000 compounds.
Celery Gr pepper Strawberry Tomato
Results
Six commodities - apples, lettuce, carrots, celery, green peppers, strawberries, and tomatoes - were purchased at local supermarkets. They were extracted and analyzed using the GC/MSD conditions described earlier. Aldrin was added to each during the extraction process and acts as both a surrogate standard (SS) and an internal standard (IS). The datafiles were processed using Agilents Deconvolution Reporting Software. DRS automatically combines the results from the MSD ChemStation Quantitation with AMDIS Deconvolution and NIST Search into an easy-to-read report. The results are shown in Table 2. Each of the samples showed at least one residue, ranging from
Table 2. Market Basket Commodity Results Commodity Apple Lettuce Compound Table R.T. Methamidopho Acephate Diphenylamine Chlorothalonil Carbaryl Metalaxyl Malathion Isodrin Thiabendazole Captan Phosmet Permethrin Cyfluthrin Cypermethrin Fenvalerate Addional DRSonly compounds
* First R.T. of multiple isomers.
Carrot
5.655 7.690 10.516 14.784 16.806 17.337 18.800 20.031 20.939 21.227 28.504 31.369* 32.218* 32.690* 34.271*
t t 0.02 0.02
0.04 0.23 t t t t 0.09
0.02 0.48 0.03 t t t t 8 0 1 0.03 t
13
10
t = Trace quantity
Table 3.
Spiked and Duplicate Commodity Results, ppm Spikes Strawberry 0.15 0.11 0.11 0.13 0.09 0.14 0.15 Tomato 0.25 0.11 0.15 0.19 0.13 0.26 0.25 Duplicates 0.08
Compound Table R.T. Atrazine 13.159 Lindane 13.461 Carbaryl 16.806 Linuron 18.187 Parathion 19.275 Permethrin I 31.369 Permethrin II 31.550 Chlorthalonil Captan 14.784 21.227
0.47
A portion of the DRS Report for celery is shown in Figure 1. Chlorothalonil was found at 14.823 minutes at 0.02 ppm. AMDIS verified chlorothalonil with a 97 match eluting only 2.7 seconds from its expected RTL time. NIST search further verified chlorothalonil with a 93 match, as the third hit out of the top 100 hits. Aldrin and permethrin are similarly verified with permethrin below the normal reporting limit. Malathion was found by AMDIS and verified by NIST. It was not found by ChemStation because one of three qualifier ions was out of range. AMDIS mitigates this problem because deconvolved spectra are cleaned of interferences and full spectrum matching is used. It would be nearly impossible to identify the analytes of interest in the presence of > 650 individual components in the celery without deconvolution. The TIC for celery is shown in Figure 2.
For all of the commodities tested, DRS verified the presence of all the calibrated peaks found by the ChemStation. Most laboratories only calibrate a fixed number of compounds, say 50-100, as it is not practical to calibrate for all 927. At the same time, these laboratories are interested in identifying other compounds that may be present in samples. Numerous uncalibrated compounds were identified using the 927 compound DRS database. The number of these additional compounds is shown on the last line in Table 2, excluding phthalates, cresols, and sulfur. When these are important to the laboratory, the GC/MSD system can, of course, be calibrated using additional standards. If an estimate of the amount is needed, or if a standard is unavailable, an average response factor can be used. The DRS database provides an average response factor for all 927 compounds. In contrast to the above methodology is the use of multiple element specific detectors such as ECD, NPD, and DFPD. Most of the analytes could be run on two of these specific detectors, but that would take twice the time. Also, it is very difficult to analyze for 927 compounds using a non-MS method due to peak overlaps. A second choice is to run each sample on unlike columns of differing polarities to the same type detector. Both of these approaches, using ECD, NPD, and/or DFPD, involve more sample handling/tracking, usually twice the number of runs, and still only have retention time as an identifier.
MSD Deconvolution Report Sample Name: celery Data File: C:\msdchem\1\Data\sjm03.D Date/Time: 02:30 PM Friday, Jan 19 2007 The NIST library was searched for the components that were found in the AMDIS target library.
R.T. 14.823 18.5992 18.7799 31.3134 Figure 1. CAS # 1897456 309002 121755 52645531 Compound Name Chlorothalonil Aldrin Malathion Permethrin I Agilent ChemStation Amount (ppm) 0.02 1 0.03 Match 97 99 59 71 AMDIS R.T. Diff Sec 2.7 4.3 -1.2 -3.3 NIST Reverse Match Hit Num. 91 93 47 53 3 1 1 5
DRS report for celery.
Aldrin
Chlorothalonil Malathion
Permethrin
10
12
14
16
18
20
22
24
26
28
30
32
Figure 2.
Market basket celery total ion chromatogram.
Conclusions
Analyzing a complex matrix can be accomplished using multiple specific detectors, but a significant time savings is realized using GC/MSD. Many more analytes can be determined simultaneously using mass spectra for confirmation. Using an available Retention Time Locked database reduces methods development and speeds data comparison among labs. Deconvolution Reporting Software adds a second expert opinion. Positives found by normal quantitation are confirmed, and the analyst is directed to further target compound hits to verify. GC/MSD with DRS provides the fastest methodology to the fewest number of false positives/negatives with the greatest confidence in the results.
2. Fast and Easy Multiresidue Method Employing Acetonitrile Extraction/Partitioning and Dispersive Solid-Phase Extraction for the Determination of Pesticide Residues in Produce, M. Anastassiades, S. J. Lehotay, D. Stajnbaher, F. J. Schenck. (2003) J.AOAC Int. 86, 412-431. 3. Analytical Methods for Residual Compositional Substances of Agricultural Chemicals, Feed Additives, and Veterinary Drugs in Food, Japan Department of Food Safety, Ministry of Health, Labour and Welfare.

References
1. Comprehensive Pesticide Screening by GC/MSD using Deconvolution Reporting Software, Philip L. Wylie, Michael J. Szelewski, and Chin-Kai Meng, Agilent Technologies Pub # 5989-1157EN.
Direct Injection of Fish Oil for the GC-ECD Analysis of PCBs: Results Using a Deans Switch With Backflushing Application
Environmental and Pharmaceutical
Author
Philip L. Wylie Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
A Deans switch, employing Agilent's Capillary Flow Technology, was configured on an Agilent 7890A gas chromatograph (GC) equipped with dual electron capture detectors (ECDs). A method was developed for the analysis of fish oil for polychlorinated biphenyl (PCB) contamination. The Deans switch was used to heart cut 7 indicator PCBs (IUPAC congeners 28, 52, 101, 118, 138, 153, and 180) from the primary DB-XLB column onto a DB200 column for further separation. Fish oil from a supplement capsule was simply diluted 1:10 in isooctane and injected directly. In a separate experiment, the fish oil was analyzed by GC with a flame ionization detector (GC/FID) without backflushing. From these analyses, it was estimated that about two-thirds of the fish oil components would remain on the column after the 17.4-minute GC/ECD run. To prevent carryover, contamination, and retention time shifts, the Deans switch was used to backflush the primary column at the end of each run. Evidence shows that backflushing removed the fish oil residue, which otherwise would quickly degrade the chromatography.
beneficial health affects. In addition to eating fish, many people take fish oil as a supplement to their daily diet. However, fish, especially those high on the aquatic food chain, can bioaccumulate fat-soluble pollutants. Among these are polychlorinated dibenzodioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and polychlorinated biphenyls (PCBs). Therefore, fish oil used in supplements undergoes a variety of analyses, including tests for halogenated pollutants. One of the quality assurance tests is to analyze fish oil for PCB contamination. This is complicated by the fact that fish oil is a very complex mixture containing high-boiling fatty acids and triglycerides of fatty acids; chain lengths are mostly between 14 and 22 carbons. They also contain varying amounts of phospholipids, glycerol ethers, wax esters, and fatty alcohols. PCB analysis is complex by itself, with 209 possible congeners. Of these, 140 to 150 have been observed in commercial PCB mixtures called Aroclors. PCB analysis usually focuses on the 12 planar, dioxin-like PCBs and/or on seven indicator PCBs (IUPAC Numbers 28, 52, 101, 118, 138, 153, and 180). To obtain sufficient sensitivity and selectivity for these compounds, analysts have traditionally employed very expensive techniques such as highresolution mass spectrometry (HR/MS) or HR/MS/MS. Analysis of the fish oil generally follows a series of extraction and cleanup steps. This paper focuses on the analysis of the seven indicator PCBs in fish oil using an Agilent 7890A GC configured with a Deans switch, two columns of differing selectivity, and dual electron capture
Introduction
Fish oils contain high levels of eicosapentanoic acid (EPA) and docosahexanoic acid (DHA), omega-3 fatty acids that are thought to have
detectors (ECDs). Fish oil from a commercially available supplement was simply diluted 10:1 in isooctane and injected into the GC. No cleanup steps were employed.
Instrumental Conditions for Analysis Inlet Split/splitless at 330 oC Oven temperature program 80 oC (1 min), 50 oC/min to 200 oC (0 min), 10 oC/min to 290 oC (5 min) Detectors Dual ECD at 340 oC N2 at 60 mL/min H2 at 41.040 psig (constant pressure mode) H2 at 20.610 psig (constant pressure mode) ECD make-up gas Inlet pressure PCM pressure to Deans switch
Experimental
The fish oil supplement was obtained from a local grocery store. According to the bottles label, each gelatin capsule contains 1.0 g of fish oil of which 180 mg is EPA and 120 mg is DHA. Oil was removed from a capsule and diluted with isooctane (pesticide grade from Sigma-Aldrich, Milwaukee, WI, USA) to make a 10% solution. This solution was spiked with various Aroclors (Supelco, Bellefonte, PA, USA) or with individual PCB congeners (AccuStandard, New Haven, CT, USA). Table 1 lists the instrumentation and experimental conditions for the analysis.
Table 1. Instrumentation and Experimental Conditions
Post-Run Backflush Conditions Post-run duration 2.4 min Inlet pressure PCM pressure Oven temperature during backflush H2 at 0 psig H2 at 80 psig 290 oC for 2.4 min

Without backflushing, the high-boiling components of fish oil can be retained by the GC column, causing severe carryover problems from one run to the next. After a few injections, so much of the fish oil residue builds up on the column that it causes PCB retention times to shift by a minute or more. Such dramatic retention time shifts would prevent the use of the Deans switch, where heart cuts are just a few seconds wide. Deans SwitchHeart Cutting The Deans switch is one of Agilents new devices that employ Capillary Flow Technology. These devices have extremely low dead volumes, are inert, and do not leak, even with large cycles in oven temperature. Columns are easy to install into the Deans switch, which is mounted on the side of the oven wall (Figure 1).
Instrumentation and Software Gas chromatograph Agilent 7890A Automatic sampler Agilent 7683B Series injector and tray Primary column Primary column connections Secondary column Secondary column connections Restrictor Restrictor connections Inlet liner Auxiliary pressure control device Deans switch calculator software J&W 30-m 0.18-mm 0.18-m DB-XLB (P/N 121-1232) Split/splitless inlet to Deans switch J&W 30-m 0.25-mm 0.50-m DB-200 (P/N 122-2033) Deans switch to back ECD 76.8-cm 0.100-mm deactivated fused silica tubing Deans switch to front ECD Agilent deactivated single taper with glass wool (P/N 5062-3587) Agilent 7890A Pneumatic Control Module (PCM) Option # 309 Agilent Technologies Deans Switch Calculator (Rev. A.01.01)
Software for data acquisition Agilent GC ChemStation and analysis (Rev. B.03.01)
Restrictor to back detector
Primary column DB-XLB
Pneumatic connections to solenoid valve
Secondary column DB-200
Figure 1.
Photograph of the Deans switch installed on the side of the 7890A GC oven. The column and restrictor connections are indicated by an * in Figure 2a.
As shown in Figure 2a, the 30-m 0.18-mm 0.18-m DB-XLB column is connected between the split/splitless inlet and the Deans switch. A short length of deactivated fused silica tubing (76.8 cm 0.100 mm) connects the Deans switch to the front ECD. The secondary column (30-m 0.25-mm 0.5-m DB-200) was chosen because it is more polar than the DB-XLB column and has a different selectivity for PCBs. It has an upper temperature limit of 300 C, which is high enough to elute the PCBs of interest. Figure 2a shows the Deans switch in the normal mode with the solenoid valve in the off position.
In this mode, the effluent from the primary DB-XLB column is directed through the restrictor to the front ECD. When the solenoid valve is switched, the effluent is directed through the secondary DB-200 column to the back ECD (Figure 2b). The retention times for the seven indicator PCBs were initially determined with the valve in the off position. Using the timed events table in the ChemStation, the valve was switched to on just before each PCB peak and off immediately after. This produced seven heart cuts that were directed through the DB-200 column to the back ECD.
Front ECD
*
Restrictor Solenoid valve (off)
S/S Inlet
41.040 psig
*
DB-XLB
PCM
20.610 psig
Back ECD
DB-200
Figure 2a. Deans switch in the no cut position. The effluent from the DB-XLB column goes directly to the front ECD through the short restrictor. The intersections marked with an * are column and restrictor connections to the Deans switch plate (Figure 1).
Front ECD
Restrictor
Solenoid valve (on)
S/S Inlet
41.040 psig
PCM
20.610 psig
DB-XLB
Back ECD
DB-200 Figure 2b. Deans switch in the cut position. The effluent from the DB-XLB column goes to the DB-200 column and then to the back ECD.
In some Deans switch systems, the second column is placed in a separate GC oven or cryogenic cooling is used to trap the heart cut components at the head of the second column. In this case, both columns were mounted inside of the 7890A oven and cooling was not used to focus compounds at the head of the DB-200 column. Figure 3a shows the chromatogram for a fish oil sample spiked with Aroclor 1260. PCBs 28, 52, 101,
units 120000
118, 138, 153, and 180 were cut out of the primary chromatogram (Figure 3b) and sent to the second column (Figure 3c). The purpose of the DB-200 column is to resolve the target PCBs from other PCBs and matrix components that co-elute with them on the DB-XLB column. Six of the 7 PCBs appear to be well resolved on the DB-200 column. PCB 118 is only partially resolved by this method.
100000
80000
60000
40000
20000
0 8 10 12 14 16 min
Figure 3a. GC/ECD chromatogram of Aroclor 1260 spiked into fish oil. This is the effluent from the primary DB-XLB column with seven heart cuts.
units
700
600
500
400
300
200
28
100 8
52
101
118 153
138
180
10
12
14
16
min
Figure 3b. Enlargement of the chromatogram in Figure 3a showing where heart cuts were made for the seven target PCBs. 5
units
153 138
80000 70000 60000 50000 40000 30000 20000 10000
180 101
28
0 8
52
10 12
118
14
16
min
Figure 3c. GC/ECD chromatogram from the DB-200 column. The peaks in this chromatogram were heart cut from the DB-XLB column. Except for congener 118, the target PCBs were separated from co-eluting interferences by the DB-200 column.
Deans SwitchBackflushing Data collection with the Deans switch system ended at 17.4 min with the oven at 290 C. While it was assumed that a lot of the fish oil components remained on the column at this point, it was impossible to tell because the ECD responds poorly to these compounds. The fish oil does contribute some small peaks (both positive and negative) to the chromatogram, but it is impossible to see the full contribution of the fish oil. So a sample of the fish oil was analyzed on an identical DB-XLB column using a flame ionization detector (FID) with no Deans switch installed. The temperature was held at 290 C for an extra 25 minutes to determine if high boiling compounds were still eluting.
Figure 4 shows that a great deal of the fish oil continued to elute after 17.4 minutes (arrow in figure). When a blank run was made with a final oven temperature of 310 C, much more of the fish oil eluted from the column (Figure 4, middle chromatogram). A second blank run (Figure 4, top chromatogram) showed that fish oil components were still eluting from the column. In actuality, only about a third of the fish oil comes off the column under the Deans switch conditions. This is why other fish oil methods begin with a solvent extraction followed by solid phase extraction for sample cleanup.
Response 9000000 8500000 8000000 7500000 7000000 6500000 6000000 5500000 5000000 4500000 4000000 3500000 3000000 2500000 2000000 1500000 1000000 500000 0 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 Time
2nd Bakeout
1st Bakeout
1 L 10% fish oil splitless injection
Figure 4.
GC/FID chromatogram from a 1 L splitless injection of 10% fish oil using a 30-m 0.18-mm 0.18 m DB-XLB column. The arrow indicates where the GC/ECD method ends and the post-run backflush begins. In this case, there was no backflushing so the oven was held at 290 C for an extra 25 min. The run was repeated two more times without injection but with the oven held at 310 C for 30 minutes at the end of the run. Residue from the fish oil injection continued to elute, even during a second bakeout step.
The 7890A has been designed to make column backflushing a routine process. It has been shown empirically that backflushing should continue for about five times the holdup time. In this case the column was held at 290 C during the post run backflush. At the same time, the inlet pressure was dropped to 0 psig while the PCM pressure was increased to 80 psig. Using Agilents GC Pressure/ Flow Calculator software, the H2 flow rate backwards through the column was 3.81 mL/min and the holdup time was 0.466 min. Backflushing was, therefore, continued for 2.4 minutes, which is slightly more than five times the calculated holdup value. Figure 5 shows the Deans switch in the backflush mode.
Front ECD
Restrictor
Solenoid valve (off)
S/S Inlet
0 psig
PCM
80.000 psig
DB-XLB
Back ECD
DB-200 Figure 5. Deans switch in the backflush mode. The inlet pressure is dropped to 0 (or 1) psig while the PCM pressure is raised to 80 psig. This causes the carrier gas to flow backwards through the DB-XLB column. The reverse flow sweeps highboiling fish oil components off the head of the column and out the split vent.
As mentioned earlier, just a few injections of fish oil can cause dramatic shifts in PCB retention times. Backflushing forces the remaining fish oil components backwards through the primary column and out through the split vent. This prevents fish oil buildup on the column, thus eliminating carryover and retention time shifts. Figure 6a compares the first and last chromatograms in a six-run sequence. One-L splitless injections were made of 10% fish oil spiked with Aroclor 1260. This sequence was run after many previous injections of fish oil using this method, and it is clear that the retention times did not shift. Figure 6b shows the seven PCBs that were heart cut from the two analyses shown in Figure 6a. Figure 6b shows no differences in the first and last heart cut chromatograms, providing further proof that there were not even subtle shifts in the PCB retention times. Each heart cut was just 4 to 5 seconds wide, so very small RT shifts in the first column would dramatically alter the results in the second.
units
1500
1000
500
-500
-1000
-1500
10
11
12
13
14
15
16
min
Figure 6a. First (top) and sixth (inverted) injections of 10% fish oil spiked with Aroclor 1260. Seven Deans switch cuts were made from this DB-XLB column in order to isolate PCBs 28, 52, 101, 118, 138, 153, and 180. The DB-XLB column was backflushed after each run, preventing build-up of fish oil residue. The comparison shows that there was no shift in retention times caused by fish oil accumulation.
101
units
180 153 138
6000
4000
118 28 52
2000
-2000
118
-4000
-6000
-8000 9 10 11 12 13 14 15 min
Figure 6b. Chromatogram of the seven PCB congeners and interferences that were cut from the DB-XLB column to the DB-200. The first chromatogram (top) and sixth (inverted) are identical, providing further proof of retention time stability. Any retention time shift on the primary column would severely alter the appearance of the secondary chromatogram.
Conclusions
This paper demonstrates that it is possible to analyze PCBs in fish oil without performing laborious sample cleanup prior to GC injection. A Deans switch was used to cut seven target PCBs (28, 52, 101, 118, 138, 153, and 180) from a DB-XLB column for further separation on a DB-200 column. This produced nearly baseline separation of the target PCBs. Only congener 118 was not well separated from co-eluting PCBs. Further refinement of the oven temperature program would be needed to isolate this congener. It has been estimated that about two-thirds of the fish oil remained on the primary GC column at the end of the run. By setting the Deans switch to the backflush mode for just 2.4 minutes at the end of each run, this material was swept backwards through the column and out the split vent. There was no evidence for retention time shifts or carryover from run to run.

The information contained in this publication is intended for research use only and is not to be followed as a diagnostic procedure. Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2007 Printed in the USA January 5, 2007 5989-6095EN
Improving Productivity and Extending Column Life with Backflush
Application Brief
Chin-Kai Meng
All Industries
A previous application note [1] has shown that multiple GC signals and MS signals can be acquired from a single sample injection. When a 3-way splitter is connected to the end of a column, column effluent can be directed proportionally to two GC detectors as well as the MSD. This multi-signal configuration provides full-scan data for library searching, SIM data for quantitation, and element selective detector data for excellent selectivity and sensitivity from complex matrices. The system used in this study consists of a 7683ALS, a 7890A GC with split/splitless inlet, 3-way splitter, ECD, dual flame photometric detector (DFPD), and a 5975C MSD. Figure 1 shows four chromatograms from a single injection of a milk extract. The synchronous SIM/scan feature of the 5975C MSD provides data useful for both screening (full scan data) and quantitation (SIM data). DFPD provides both P and S signals without the need to switch light filters. Noticeably in the full scan TIC in Figure 1, a significant number of matrix peaks were observed after 32 minutes. It is not uncommon to add a bake-out oven ramp to clean the column after analyzing complex samples. The bake-out period is used to quickly push the late eluters out of the column to be ready for the next injection. Therefore, it is common to use a higher oven temperature than required for the analysis and an extended bake-out period at the end of a normal
Full scan TIC
Highlights
Backflush a simple technique to remove high boilers from the column faster and at a lower column temperature to cut down analysis time and increase column lifetime. The milk extract example shows that a 7-minute 280 C backflush cleaned the column as well as a 33-minute 320 C bake-out. The cycle time was reduced by more than 30%. Using backflush, excess column bleed and heavy residues will not be introduced into the MSD, thus reducing ion source contamination.
SIM
ECD
DFPD(P)
10
15
20
25
30
35
40
Figure 1.
Four chromatograms collected simultaneously from a single injection of a milk extract.
over program to clean out the column, which adds to the cycle time and shortens the column lifetime. Adding the bake-out period to the milk extract analysis, additional matrix peaks were observed even up to 72 minutes, while target compounds already eluted before 42 minutes. This means that 30 minutes were lost in productivity for each injection. Backflush [2] is a simple technique to drastically decrease the cycle time by reversing the column flow to push the late eluters out of the inlet end of the column. Late eluters stay near the front of the column until the oven temperature is high enough to move them through the column. When the column flow is reversed before the late eluters start to move down the column, these late eluters will take less time and at a lower oven temperature to exit the inlet end of the column. There are many benefits in using backflush: Cycle time is reduced (no bake-out period, cooling down from a lower oven temperature) Column bleed is reduced (no high-temperature bake-out needed), resulting longer column life Ghost peaks are eliminated (no high boilers carryover into subsequent runs) Contamination that goes into the detector is minimized, which is especially valuable for the MSD (less ion source cleaning) of 320 C. A column effluent splitter or QuickSwap is required to do the backflush.
References
1. Chin-Kai Meng and Bruce Quimby, Identifying Pesticides with Full Scan, SIM, ECD, and FPD from a Single Injection, Agilent Application Note, 5989-3299EN, July 2005. 2. Matthew Klee, Simplified Backflush Using Agilent 6890 GC Post Run Command, Agilent Application Note, 5989-5111EN, June 2006.
Acknowledgement
Milk extract is courtesy of Dr. Steven Lehotay from USDA Agricultural Research Service in Wyndmoor, Pennsylvania, USA.
Figure 2 shows three total ion chromatograms from the Agilent 7890A GC/ 5975C MSD. The top chromatogram is a milk extract analysis with all the target compounds eluted before 42 minutes (over program goes to 280 C). However, an additional 33-minute bake-out period at 320 C was needed to move the high boilers out of the column. This bake-out period was almost as long as the required time to elute all target compounds. The middle chromatogram is the same milk extract analysis stopped at 42 minutes with a 7-minute backflush post-run at 280 C added to the analysis. The bottom chromatogram is a blank run after the backflushing was completed. The blank run shows that the column was very clean after backflushing. The example shows that a 7-minute backflush cleaned the column as well as a 33-minute bake-out. The milk extract example in Figure 2 illustrates the backflush technique in reducing cycle time and column bleed. The cycle time was reduced by more than 30% and the column was kept at 280 C, without going to the bake-out temperature

It took an additional 33 min and heating the column to 320 C to remove these high boilers
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2006 Printed in the USA December 26, 2006 5989-6018EN
Run stopped at 42 min and backflushed at 280 C for 7 mins Blank run after backflushing showing the column was clean
5 10 15 20 25 30 35 40 45 50 55 60 65 70 min
Figure 2.
Three total ion chromatograms comparing the results with and without backflush.
Using RTL and 3-Way Splitter to Identify Unknown in Strawberry Extract Application Brief
Chin-Kai Meng
Fruit and vegetable extracts are usually very complex to analyze. It is common to use the very selective GC detectors, for example NPD, ECD, and FPD, to look for trace pesticide residues in the extracts. Mass spectrometry is most often used to confirm the hits from GC detectors. A previous application note [1] describes a GC/MS system with a three-way splitter added to the end of the column. The column effluent could be split three ways to two GC detectors and the MSD. The splitter system is therefore capable of providing up to four signals (two GC signals, SIM, and full-scan chromatograms) from a single injection. The combination of element selective detectors, SIM/scan, and Deconvolution Reporting Software (DRS) makes a very powerful pesticide analysis system [2]. The trade-off is the decrease of analyte concentration in any detector due to the flow splitting at the end of the column. The system used for this study consists of an Agilent 7890A GC with split/splitless inlet, a three-way splitter, ECD, dual flame photometric detector (DFPD), and 5975C MSD. Figure 1 shows chromatograms from 2 separate injections (each injection provides two GC signals) of the same strawberry extract without any hardware or filter changes. All of the target compounds were found and confirmed by DRS, GC, and MS signals except the unknown peak at about 41 minutes. The peak shows responses from ECD, DFPD(S) and DFPD(P). However, no peak was observed in the MS full-scan signal. This makes it difficult to confirm the unknown peak using the full-scan TIC. Since the analysis was retention time locked, it is therefore possible to find potential matches by examining the RTL pesticide database (part number G1672AA). The unknown compound, containing electron-capturing atoms (for example, Cl or O), P, and S atoms, would have a target retention time inside the
Highlights
Splitter+an inert, easy-to-use capillary flow technology that splits column effluent to multiple detectors (for example, MSD, DFPD, and ECD). The splitter configuration provides a comprehensive screening and quantitative system. By combing RTL, element-selective detector chromatograms, and the RTL pesticide database, a trace level pesticide residue was identified without the full-scan mass spectrum.
TIC
ECD
DFPD (S)
DFPD (P)
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
Figure 1.
Unknown compound detected by GC signals not found in strawberry extract TIC.
Table 1.
Compound List Extracted from the RTLPest3.tab File CAS 117337196 191242 3383968 60044260 83794 Mol form C15H15CIFN3O3S2 C22H12 C16H20O6P2S3 C12H4Br6 C23H22O6 Mol wt 403.9 276.3 466.5 627.6 394.4 R.T. 39.10 39.13 40.74 40.93 41.70 Target Ion 403 276 466 308 192 Q1 56 277 125 468 191 Q2 405 138 93 148 394 Q3 232 275 109 154 177
Name Fluthiacet-methyl Benzo[g,h,i]perylene Temephos PBB 169 hexabrombiphenyl Rotenone
41 \ 0.5-minute window (that is, 40.5 to 41.5 min) in the database, if it is included in the database. Table 1 is a portion of the RTLPest3.tab file1 opened in Microsoft Excel. The compound temephos at locked retention time 40.74 min meets all the criteria for the unknown peak. To further confirm peak identity, extracted ion chromatograms (EICs) of the four major ions of temephos were plotted. Figure 2 shows EICs of target ion and three qualifiers (ions 466, 125, 93, and 109 from Table 1) of temephos. Although the ion intensities were weak, the noticeable presence of all four ions at 40.9 min helped to confirm that the unknown peak was temephos.
1. The RTLPest3.tab file is created in the C:\Database directory while executing the Tools\List Screen Database command (in MSD Enhanced Data Analysis software) and selecting the RTLPest3.scd from the C:\Database directory.
200 100 0
Ion 466
200 100 0
Ion 125
200 100 0
Ion 93
200 100 0
Ion 109
36
37
38
39
40
41
42
43
44
Figure 2.
EICs of target ion 466 (temephos) and three qualifier ions.
References
1. Chin-Kai Meng and Bruce Quimby, Identifying Pesticides with Full Scan, SIM, ECD, and FPD from a Single Injection, Application Note, 5989-3299, July 2005. 2. Mike Szelewski and Bruce Quimby, New Tools for Rapid Pesticide Analysis in High Matrix Samples, Application Note, 5989-1716, October 2004.
Acknowledgement
Strawberry extract is courtesy of Dr. Steven Lehotay from USDA Agricultural Research Service in Wyndmoor, Pennsylvania, USA.

Copyright 2006 Agilent Technologies All Rights Reserved. Reproduction, adaptation, or translation without prior written permission is prohibited, except as allowed under the copyright laws. Printed in the USA December 13, 2006 5989-6007EN
Improving the Effectiveness of Method Translation for Fast and High Resolution Separations Application
Author
Michael Woodman Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Abstract
The increased availability of sub-2-micron (STM) columns and increased demand for methods friendly to mass spectrometers has led to strong trend toward conversion of existing HPLC methods to smaller diameter and smaller particle size columns. While the conversion is a simple mathematical exercise requiring the scaling flow rates, gradient times and injection volumes, many users observe less than perfect results. Here we look closely at the problem and propose calculations that improve the speed and/or resolution in a more predictable and beneficial way.
Simplistically, a column of 250-mm length and containing 5-m particles can be replaced by a 150-mm length column packed with 3-m particles. If the ratio of length to particle size is equal, the two columns are considered to have equal resolving power. Solvent consumption is reduced by L1/L2, here about 1.6-fold reduction in solvent usage per analysis. If an equal mass of analyte can then be successfully injected, the sensitivity should also increase by 1.6-fold due to reduced dilution of the peak as it travels through a smaller column of equal efficiency. LC/MS (Liquid Chromatography/Mass Spectrometry) ionization sources, especially the electrospray ionization mode, have demonstrated greater sensitivity at lower flow rates than typically used in normal LC/UV (UltraViolet UV/VIS optical detection) methods, so it may also be advantageous to reduce the internal diameter of a column to allow timely analysis at lower flow rates. The relationship of flow rate between different column diameters is shown in Equation 1.
Flowcol. 1 Diam.column2 Diam.column1
2
Introduction
Methods developed on older columns packed with large 5- or 10-m particles are often good candidates for modernization by replacing these columns with smaller dimension columns packed with smaller particle sizes. The potential benefits include reduced analysis time and solvent consumption, improved sensitivity and greater compatibility with mass spectrometer ionization sources.
= Flowcol. 2
(eq. 1)
The combined effect of reduced length and diameter contributes to a reduction in solvent consumption and, again assuming the same analyte mass can be injected on the smaller column, a proportional increase in peak response. We normally scale the injection mass to the size of the column,
though, and a proportional injection volume would be calculated from the ratio of the void volumes of the two columns, multiplied by the injection volume on the original column.
Inj. vol.col. 1 Volumecolumn2 Volumecolumn1 = Inj. vol.col. 2 (eq. 2)
For isocratic separations, the above conditions will normally result in a successful conversion of the method with little or no change in overall resolution. If one wishes to improve the outcome of the method conversion, though, there are several other parameters that should be considered. The first of these parameters is the column efficiency relative to flow rate, or more correctly efficiency to linear velocity, as commonly defined by van Deemter [1] and others, and the second is the often overlooked effect of extracolumn dispersion on the observed or empirical efficiency of the column. Van Deemter observed and mathematically expressed the relationship of column efficiency to a variety of parameters, but we are most interested here in his observations that there is an optimum linear velocity for any given particle size, in a wellpacked HPLC column, and that the optimum linear velocity increases as the particle size decreases. Graphically, this is often represented in van Deemter plots as shown in Figure 1, a modified version of the original plot [2]. In Figure 1 we observe that the linear velocity at which 5-m materials are most efficient, under the conditions used by the authors, is about 1 mm/sec. For 3.5-m materials the optimum linear velocity is about 1.7 mm/sec and has a less distinct opti0.02
mum value, suggesting that 3.5-m materials would give a more consistent column efficiency over a wider flow range. For the 1.8-m materials, the minimum plate height, or maximum efficiency, is a broad range beginning at about 2 mm/sec and continuing past the range of the presented data. The practical application of this information is that a reduction in particle size, as discussed earlier, can often be further optimized by increasing the linear velocity which results in a further reduction in analysis time. This increase in elution speed will decrease absolute peak width and may require the user to increase data acquisition rates and reduce signal filtering parameters to ensure that the chromatographic separation is accurately recorded in the acquisition data file. The second important consideration is the often overlooked effect of extracolumn dispersion on the observed or empirical efficiency of the column. As column volume is reduced, peak elution volumes are proportionately reduced. If smaller particle sizes are also employed there is a further reduction in the expected peak volume. The liquid chromatograph, and particularly the areas where the analytes will traverse, is a collection of various connecting capillaries and fittings which will cause a measurable amount of bandspreading. From the injector to the detector flow cell, the cumulative dispersion that occurs degrades the column performance and results in observed efficiencies that can be far below the values that would be estimated by purely theoretical means. It is fairly typical to see a measured dispersion of 20 to 100 L in an HPLC system. This has a disproportionate effect on the smallest columns and smallest particle sizes, both of which are expected to yield the smallest
Plate height (mm)
0.015
0.01
0.005
5.0 m SB-C18 3.5 m SB-C18 1.8 m SB-C18
Lin. vel. 4.6 mm 3 mm 2.1 mm 1 mm
mm/sec mL/min mL/min mL/min mL/min
1 0.7 0.3 0.14 0.033
2 1.4 0.6 0.29 0.066
3 2.1 0.9 0.44 0.1
4 2.8 1.2 0.58 0.133
5 3.5 1.5 0.73 0.166
Figure 1.
van Deemter plot with various flow rates and particle sizes.
possible peak volumes. Care must be taken by the user to minimize the extracolumn volume and to reduce, where practical, the number of connecting fittings and the volume of injection valves and detector flow cells. For gradient elution separations, where the mobile phase composition increases through the initial part of the analysis until the analytes of interest have been eluted from the column, successful method conversion to smaller columns requires that the gradient slope be preserved. While many publications have referred to gradient slope in terms of % change per minute, it is more useful to express it as % change per column volume. In this way, the change in column volume during method conversion can be used to accurately render the new gradient condition. If we think of each line of a gradient table as a segment, we can express the gradient by the following equation:
Experimental Conditions
System Agilent 1200 Series Rapid Resolution LC consisting of: G1379B micro degasser G1312B binary pump SL G1367C autosampler SL, with thermostatic temperature control G1316B Thermostatted column compartment SL G1315C UV/VIS diode array detector SL, flow cell as indicated in individual chromatograms ChemStation 32-bit version B.02.01 Columns Agilent ZORBAX SB-C18, 4.6 mm 250 mm, 5 m Agilent ZORBAX SB-C18, 3.0 mm 150 mm, 3.5 m Mobile phase conditions Organic solvent: Aqueous solvent: Gradient Conditions Gradient slope: 7.8% or 2.3% per column volume, as indicated. See individual chromatograms for flow rate and time Acetonitrile 25 mm phosphoric acid in Milli-Q water
% Gradient slope =
(End% Start%) #Column volumes
(eq. 3)
Sample Standard mixture of chlorinated phenoxy acid herbicides, 100 g/mL in methanol
Note that the use of % change per column volume rather than % change per minute frees the user to control gradient slope by altering gradient time and/or gradient flow rate. A large value for gradient slope yields very fast gradients with minimal resolution, while lower gradient slopes produce higher resolution at the expense of increased solvent consumption and somewhat reduced sensitivity. Longer analysis time may also result unless the gradient slope is reduced by increasing the flow rate, within acceptable operating pressure ranges, rather than by increasing the gradient time. Resolution increases with shallow gradients because the effective capacity factor, k*, is increased. Much like in isocratic separations, where the capacity term is called k', a higher value directly increases resolution. The effect is quite dramatic up to a k value of about 5 to 10, after which little improvement is observed. In the subsequent examples, we will see the results associated with the calculations discussed above.
Results
The separation was initially performed on a standard 4.6 250 mm, 5-m ZORBAX SB-C18 column thermostatted to 25 C (Figure 2) using conditions referenced in US EPA Method 555. The method was then scaled in flow and time for exact translation to a 3.0 150 mm, 3.5-m column (Figure 3). Solvent consumption is reduced from 60 mL to 15.5 mL per analysis. The separation was then re-optimized for faster separation with the identical slope, 7.8%, by increasing the flow rate from 0.43 to 1.42 mL/min, and proportionately reducing the gradient time (Figure 4). Finally, increased resolution is demonstrated by keeping the original times used in Figure 3 with the increased flow rate (Figure 5). This yields a gradient with identical time but a reduced slope of 2.3%. The increased resolution of peaks 4 and 5 is readily apparent. The conditions in Figure 4, 7.8% slope at increased linear velocity on 3.0 150 mm, 3.5-m material, yield a separation with comparable resolution to the original 4.6 250 mm method, but with only a 12-minute total analysis time. This is excellent for
12.557
mAU 350 300
14.380
17.779
19.414
21.063
250 200 150 100 50 0 12.5 15
18.871
23.050
24.667
13.194
17.607
17.5
20
22.5
25
27.5
Conditions EPA Method 555 with ZORBAX SB-C18 columns and fast DAD detector ZORBAX SB-C18 4.6 mm 250 mm, 5 m Column temp: 25 C Gradient: 10% to 90% ACN vs. 25 mM H3PO4 Gradient slope: 7.8% ACN/column volume Analysis flow rate: 1 mL/min Group A Compounds Total analysis time: Detection: Figure 2. 60 min UV 230 nm, 10-mm 13-L flow cell, filter 2 seconds (default)
Gradient separation of herbicides on 4.6 250 mm 5-m ZORBAX SB-C18.
8.781
800 700 600 500 400 300 200
9.990
mAU
Conditions: EPA Method 555 with ZORBAX SB-C18 columns and fast DAD detector ZORBAX SB-C18 3.0 mm 150 mm, 3.5 m Column temp: 25 C Gradient: 25 mm H3PO4/ACN, 0% to 90% ACN in 18 minutes Gradient slope: 7.8% ACN/column volume Analysis flow rate: 0.43 mL/min Detection: UV 230 nm, 3-mm 2-L flow cell, filter 0.2 seconds Total analysis time: 36 min.
12.061 12.831 16.314
14.106
13.046
15.317
15.786
17.081
9.120
0 8 10 12 14 16 18 min
Figure 3.
Gradient separation of herbicides on 3.0 150 mm, 3.5-m ZORBAX SB-C18.
13.854
100
18.348
29.595 min
400
300
2.674
3.011
mAU
Conditions EPA Method 555 with ZORBAX SB-C18 columns and fast DAD detector ZORBAX SB-C18, 3.0 mm 150 mm, 3.5 m Column temp: 25 C Gradient: 25 mM H3PO4/ACN, 10% to 90% ACN in 5.4 min. Gradient slope: 7.8% ACN/column volume Analysis flow rate: 1.42 mL/min Detection: UV 230 nm, 3-mm 2-L flow cell, filter 0.2 seconds Total analysis time: 12 min.
3.620
3.850 3.919
4.240
200
100
2.780
4.611
4.5
0 2 2.5 3 3.5 4 5 5.5 min
Figure 4.
High speed gradient separation of herbicides on 3.0 150 mm, 3.5-m ZORBAX SB-C18.
3.964
mAU 400 350 300 250
Conditions EPA Method 555 with ZORBAX SB-C18 columns and fast DAD detector ZORBAX SB-C18, 3.0 mm 150 mm, 3.5 m Temp: 25 C Gradient: 25 mM H3PO4/ACN, 10% to 90% ACN in 18 min. Gradient slope: 2.3% ACN/column volume Analysis flow rate: 1.42 mL/min Detection: UV 230 nm, 3-mm 2-L flow cell, filter 0.2 seconds Total analysis time: 36 min.
6.793 7.583 11.291 10.257 11.465 8.905
150 100
0 4 6 8 10 12 min
Figure 5.
Reduced slope gradient separation of herbicides on 3.0 150 mm, 3.5-m ZORBAX SB-C18.
12.692
4.093 4.253
50
7.912
10.056
200
4.933
4.743
4.914
high throughput screening and quantitation of a large number of samples. Figure 5, with the gradient slope reduced to 2.3%, results in a high-resolution separation with a calculated R value of 3.3 vs. the standard 3.0 150 mm separation value of 1.9, for the critical pair seen in Figure 5 at 7.5 to 8 minutes. In Table 1 the column has been replaced with a low dead volume connecting union in a system fitted with 0.12-mm id capillary tubing at all points of sample contact. A 1-L injection of dilute actone
Table 1. Volumetric Measurements of Various Flow Cells Elution volume (L) 11 14 Half height width (L) 5 6 5 Sigma width (L) 12 18
Conclusion
Careful analysis of the existing gradient conditions, coupled with an awareness of the need to accurately calculate new flow and gradient conditions can lead to an easy and reliable conversion of existing methods to new faster or higher resolution conditions. In addition, awareness of extracolumn dispersion, especially with small and high resolution columns, will ensure good column efficiency which is critical to a successful translation of the method.
References
1. J. J. van Deemter, F. J. Zuiderweg, A. Klinkenberg, Chemical Engineering Science 1956, 5, 271289 2. The Influence of Sub-Two Micron Particles on HPLC Performance, Agilent Technologies, application note 5989-9251EN, May 2003
Flow cell New SL 2 L 3 mm Micro 6 mm 1.7 L (n = 2) Semi-micro 6 mm 5 L (n = 2) Standard 10 mm 13 L New SL 10 mm 13 L
13
6.5
18.5
26 27
11 11
26 25

is made to determine the bandspreading contribution of the system, with various flow cells. Multiple flow cells were tested, and the average result reported, where possible. The elution volume summarizes the total volume of all tubing in the system. While the absolute volume from the 2-L to the 13-L flow cells is 11 L, we observe an increase of 15 to 16 L because of the larger diameter inlet tubing integral to the larger volume flow cells.
Screening for Hazardous Chemicals in Homeland Security and Environmental Samples Using a GC/MS/ECD/FPD with a 731 Compound DRS Database Application Note
Homeland Security, Environmental
Authors
Bruce Quimby Mike Szelewski Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 UASA
interpretation of the MS data, especially in samples with high matrix contamination. The combination of selective GC detectors, SIM/Scan, and deconvolution makes a very powerful hazardous chemical analysis system that shows significant progress toward the above goals.
Introduction
In recent years, there has been increasing concern over the release of hazardous chemicals through either accidental or intentional acts. Both the homeland security and environmental communities recognize the need for preparing analytical laboratories that can respond quickly to such incidents. The terms toxic industrial chemicals/toxic industrial materials (TIC/TIM) are used in homeland security to refer to hazardous chemicals, while the environmental community uses different terminology like hazardous materials. In either case, the challenge is to develop laboratory methods with the capability of identifying any hazardous chemical(s) involved in an incident and to be able to measure its concentration in collected samples. There are several significant challenges to face when developing methods for this analysis. The methods must able to: Rapidly and accurately identify the specific toxic agents involved Measure concentration correctly at high levels of agent at the epicenter (high dynamic range) Measure concentration correctly at low levels of agent at perimeters and during decontamination (low detection limits)
Abstract
Response to homeland security or environmental incidents involving hazardous chemicals requires first, the rapid and accurate identification of the chemical agent(s) involved and second, the quantitative measurement of that agent in large numbers of samples to aid in managing the response. Given the unknown nature of the analytes and the complexity of matrices that could be encountered, developing analytical methods for this analysis is challenging. The approach described in this work uses a gas chromatography/mass spectrometry (GC/MS) system with a micro-fluidic splitter added to the end of the column. The splitter divides the column effluent between the MS and either a dual-wavelength flame photometric detector (DFPD) or a micro electron-capture detector (ECD) and a single-wavelength FPD. This approach allows the simultaneous collection of MS and two channels of selective GC detector data from a single injection. This multisignal configuration provides: full-scan MS data for library searching, selective ion monitoring (SIM) data for trace analysis, ECD and FPD data for excellent selectivity and sensitivity in complex matrices. The systems use retention time locking (RTL) to produce retention times (RTs) that precisely match those in a 731 compound database of hazardous chemicals. Deconvolution Reporting Software (DRS) is used to provide fast and accurate
Be highly selective over matrix interferences (wood smoke, fuels, burning tires, etc.) to minimize both false positives and false negatives Indentify as many toxic agents as possible Handle large numbers of samples It is clear that there is no single analytical technique that can be used for detecting all possible hazardous chemicals. However, one technique that is widely applicable for the identification and measurement of broad classes of hazardous chemicals is GC/MS. GC/MS is widely used in laboratories worldwide for the analysis of thousands of different chemicals. GC/MS methods are typically developed to analyze between 10 and 100 individual compounds. A target compound is deemed to be present if the target ion and two or three qualifier ions, with specific abundance ratios, fall within a defined RT window. The identity of the target may be further confirmed by comparison of the scan at the apex of the peak with a library reference spectrum. Matrix interferences are usually minimized by optimizing a combination of the sample preparation, GC, and MS parameters. Since most methods only deal with at most a few matrix types, the ions chosen for identification purposes can be selected such that they are minimized in the matrix. With the limited number of targets addressed by the method, recalibration of response factors, RTs, and qualifier ion abundance ratios can be accomplished with the injection of a few calibration mixtures. General screening methods for very large numbers of targets in widely varying and complex matrices offer a new set of challenges for the method developer. When screening for hundreds of targets, several factors must be addressed: Use of sample preparation to reduce matrix interferences is now significantly limited because rigorous cleanup steps may unintentionally remove targets. This reduced level of cleanup can result in significantly higher levels of matrix interferences to contend with. Recalibration of response factors, RTs, and qualifier abundance ratios is difficult or impossible because of the large number of targets. The methods may be deployed in laboratories without access to standards for all of the targets.
The time required for data review of hundreds of targets in complex matrices can become unmanageably large. Even with a very large database of targets, it is possible that hazardous chemicals not in the target list could be present in a sample. Recently, several techniques have become available to help address the above set of challenges. RTL produces RTs that precisely match from instrument-to-instrument and to those in a database [1]. This eliminates the need for recalibration of the individual RTs and timed events. The introduction of reliable and inert microfluidic splitters allows for the simultaneous collection of mass spectral data and, for example, phosphorus, sulfur, and/or electron capture data [2]. The selective detector chromatograms can highlight suspect compounds even if they are not in the MS target list. They can also offer an alternative means for quantitation of target analytes. The introduction of the synchronous SIM/Scan feature allows for the simultaneous acquisition of both full scan and SIM data from the same injection [2, 3]. The scan data can be used for screening the full list of targets in the database while the SIM data looks for a high priority subset of compounds down to very low levels. One of the most significant tools developed for dealing with complex matrices is Agilents Deconvolution Reporting Software (DRS) [4]. It uses advanced computational techniques to extract the spectra of targets from those of overlapped interference peaks. It then compares the extracted spectrum with a library to determine if the target is present. Any hits are confirmed by searching against the main NIST MS reference library. This process is automated and provides significant time savings in data interpretation. Since it deals with the entire spectrum instead of just four ions, DRS can often correctly identify a target in the presence of interferences where the typical approach would fail. The use of DRS substantially reduces the number of both false positives and false negatives. This application note describes the combination of the above techniques with a database of 731 hazardous chemicals, the Agilent Hazardous Chemical DBL (HCD), to be used for screening purposes. The compounds were chosen because of their significance in environmental or food safety analysis. The reasoning is that if the materials are manufactured in significant quantities and are toxic, they would be likely to appear in an
environmental method. The pesticides are included because many exhibit toxicity. The list is comprised of: Chlorinated Dioxins and Furans: EPA 8280A, 10 compounds Polychlorinated biphenyls: EPA 8082, 19 compounds Volatiles: EPA 502/524, 60 compounds Semivolatiles: EPA 8270C Appendix IX, 140 compounds Pesticides: Agilent RTL Pesticide Database (adapted), 567 compounds Total: 796 compounds, with 65 compounds in two groups, or 731 individual compounds The names of all the compounds in the database are listed in Appendix A at the end of this note. The above list by no means contains all of the hazardous chemicals that could be encountered. However, it does screen for a large number of known hazards and with the addition of selective detection can highlight other nontarget compounds that may be of interest.
and where the fuel components are not of interest. In the examples shown below the database with hydrocarbons removed was used, since fuels were used as prototype matrices.
System Configuration
The system configurations used are shown in Figure 1A and 1B.
A
Auto-sampler Phosphorus FPD
3-Way effluent splitter with makeup
AUX EPC 3.8 psig
ECD
Column 6890N GC
5975 Inert MSD
The chromatographic conditions chosen for development of the database are general in nature and are compatible with the analysis of other types of compounds beyond those in the table. For example, laboratories with access to calibration standards for chemical warfare agents (CWA) can add CWA data to the tables and screen for them as well. The RTs for compounds in the database were collected with the column outlet pressure at 3.8 psig using a microfluidic splitter. This was done to assure that the RTs observed during sample analysis would closely match those in the database when a microfluidic splitter or QuickSwap is used. The chromatographic conditions for the database were chosen to be compatible with the method translation technique. Constant pressure mode was used in the GC inlet so that method translation can be used to precisely time scale the methods for faster operation [5]. Provided with the Agilent Hazardous Chemicals DBL are the files to run the analysis precisely threefold (3X) and sevenfold (7X) faster than the primary database (1X). Also, each of the three-speed variations of the database are provided in two forms: one with the entire set of 731 compounds and one with the 36 aromatic hydrocarbons removed. The latter is provided for use with samples known to contain fuels
Auto-sampler Dual Flame Photometric Detector Sulfur Phosphorus 3-Way effluent splitter with makeup
AUX EPC 3.8 psig
Column 6890N GC
5975 Inert MSD + Performance electronics
Figure 1.
System configurations. A). GC/MS/ECD/FPD system used for 1X and 3X screening analyses. B). GC/MS/DFPD system used for 7X screening analyses.
Key components are: Fast Oven The primary 1X method only requires the 120V oven. With the 6890N 240V oven (option 002), the screening analysis method can be run precisely three times faster (14.33 min) using a 15-m HP-5MS column. If the 240V GC is further equipped with SP1 2310-0236 (puts MSD interface in back of oven under rear injection port) and
3
using the G2646-60500 oven-insert accessory, the speed can be increased to seven times faster (6.14 min) with a 5-m HP-5MS column. Note that use of the oven insert prevents use of the front inlet and detector positions. Only one detector is available for splitting. The DFPD is a good choice for this configuration, as it uses only one detector position but generates two signals. ECD The 6890N Option 231 is a ECD. The signal from the electron capture detector (ECD) is collected, stored, and processed by the MS ChemStation simultaneously with the MS data. ECDs are selective in nature and exhibit very sensitive response to halogenated compounds, with detection limits below 1 pg for polyhalogenates. They also respond to several other functional groups like nitro compounds. They do, however, also respond to some fairly low-priority compounds, like phthalate esters. The ECD data can be used in several ways. Nontarget halogenated or nitro compounds are highlighted. The presence of an electrophore at the RT of an identified compound can be used to support confirmation of identity. The response on the ECD can be used for quantitative analysis, but only after calibration with a standard, as the response factors are compound dependent and can vary significantly with compound class. Single FPD The 6890N Option 240 is a single FPD. It is used to selectively detect either sulfur or phosphorus. The detector is usually run in the phosphorus mode to highlight such compounds as organophosphorus pesticides and nerve agents. In the phosphorus mode, the detector is highly selective (>106) with a very low (~0.050 pg) detection limits for phosphorus. The ability of the FPD to uncover nontarget organophosphorus compounds like new pesticides or designer nerve agents is especially helpful. The presence of phosphorus at the RT of an identified compound can be used to support confirmation of identity. Because the response per unit weight of phosphorus is relatively consistent from compound to compound, the FPD can be used for semi-quantitative analysis in situations where no calibration standard is available for an identified analyte. Dual FPD The 6890N Option 241 is a DFPD with two optical detection channels that measures sulfur and phosphorus simultaneously. The DFPD sulfur response is also selective (>104) and sensitive (detection limits <10 pg) , although not as much as phosphorus.
4
The sulfur signal is also quadratic with respect to the amount of sulfur injected. It is often used to detect sulfur-mustard agents and for confirmation of sulfur-containing pesticides. The response per unit weight of sulfur is relatively consistent from compound to compound, but varies more than that of the phosphorus signal. Microfluidic Splitter The 6890N Option 890 (3-way splitter) or Option 889 (2-way splitter) uses diffusion-bonded plate technology combined with metal column ferrules to make an inert, easy-to-use, leak free, hightemperature column-effluent splitter. The splitter uses Auxiliary EPC for constant pressure makeup (6890N Option 301). The Auxiliary EPC makeup can be pressure programmed at the end of the run to higher pressure, while at the same time the inlet pressure is lowered to near ambient. This causes the flow in the column to reverse direction, backflushing heavy materials out the split vent of the inlet. Backflushing can greatly reduce analysis times for samples that contain high-boiling matrix components [6]. The Aux EPC also allows column changing and maintainance without venting the MSD. When the column fitting is removed from the splitter, helium from the makeup supply purges the fitting, preventing air from entering the MSD. If the column is attached to the splitter but removed from the inlet, helium flows backwards through the column and out the inlet end. Inlet maintainance or column headtrimming can be done without cooling and venting the MSD to prevent sucking air into a hot source. MSD System The 5975 inert MSD with performance turbo (G3243A) or 5973N inert MSD with performance electronics and performance turbo (G2579A), EI (electron impact ionization mode) MSD is used. These configurations provide faster full scan rates while maintaining sensitivity. The scan rates are compatible with the narrower peaks generated by fast chromatography. The performance turbo pump is required to handle the higher flows associated with the screening method. Synchronous SIM/Scan The D.02.00 (or higher) revision of the Agilent MSD ChemStation is used because it supplies the synchronous SIM/Scan feature. SIM/Scan operates by collecting SIM data every other cycle and scan data on alternate cycles throughout the entire chromatogram. The signal-to-noise performance of the collected SIM and scan data is virtually identical to that obtained with SIM-only and scan-only
methods. As with conventional SIM methods, not all 731 targets can be monitored in a single run due to the required time separation between SIM groups. In general, the acquisition of SIM data is set up to collect high-priority targets at very low levels. Examples would be the chlorinated dioxins and CWAs. DRS Software (G1716AA) Spectral deconvolution of the MS data enables identification of analytes in the presence of overlapped matrix peaks [4]. This significantly reduces chromatographic resolution requirements, which allows detection of targets in higher levels of matrix or can be used with fast chromatography to shorten analysis times. DRS uses the AMDIS deconvolution program from NIST, originally developed for trace chemical-weapons detection in complex samples. DRS presents the analyst with three distinct levels of compound identification: ChemStation, based on RT and four-ion agreement AMDIS, based on cleaned spectra full-ion matching and locked RT NIST05 search using a 163000 compound library Hazardous Chemical DBL (G1671AA) This supplies the mass spectral library, method, and DRS files for the 731 compound-screening method.
volatility from gases to large polynuclear aromatic hydrocarbons (PAHs). Splitless injections are usually incompatible with the lowest boiling volatiles due to problems with the solvent. For low matrix samples where semivolatiles are of interest, splitless injections can be used. For ambient headspace analysis [7], the conditions are listed separately at the bottom of Table 1. The liner used for ambient headspace was 1-mm id straight through (no glass wool) and Siltek coated (Restek, part number 20973-214.5). The auto injector parameters are critical in ambient headspace and are listed in Table 1. The volatiles samples run by ambient headspace were prepared as described in Reference 7. While the targets in the table cover a very broad range of boiling points, it is usually not practical to screen for all of them in one run. This is because an analysis for semivolatile compounds would be done with a solvent that would occlude the lowest boiling volatiles in the table. Conversely, a method for injecting the lowest boiling compounds would usually not be suitable for the highest boiling. The MSD solvent delays listed in Table 1 are based on isooctane as the solvent in a semivolatiles analysis. If a lower boiling solvent is used, it may be possible to reduce these delays accordingly. Some of the target compounds were found to have sufficiently high boiling points to require higher inlet and detector temperatures. These were the higher molecular weight PAHs, the polychlorinated dioxins, and the polychlorinated furans. For these compounds the inlet temperature, MS source, and transfer line were also raised to 300 C. Without this increase in temperature, the compounds would exhibit tailing and in some cases reduction in signal. The trade-off with temperature is that the performance of some thermally labile compounds is degraded at the higher temperatures. The MSD data acquisition sampling rates listed in Table 1 are for scan mode only. For volatiles analysis, the scan rate is increased one step. It is also increased one step when SIM/Scan is used. In SIM/Scan mode the SIM dwell time was set to 40 milliseconds for each ion monitored. The microfluidic splitter parameters are chosen to provide the desired flow ratio between detectors while meeting the flow requirements of the detectors used. A primary consideration is to make sure that the flow to the MSD does not exceed ~4 mL/min while collecting analyte data. It was also desired to split the effluent equally between the DFPD and MSD in the 2-way split configuration. In the 3-way configuration, the split to the ECD was reduced
5
Instrument Operating Parameters

The instrument operating parameters used (unless noted otherwise) are listed in Table 1. These are starting conditions and may have to be optimized. The split/splitless injection port was used for all work described here. It was chosen for its flexibility, allowing splitless injections for clean samples and split injections for dirty or high-concentration samples. It is also compatible with column backflushing. For all cases (except ambient headspace), the inlet liner used was the 4-mm id Siltek Cyclosplitter (Restek, part number 20706-214.1). This inlet liner was found to be of low activity, as it does not contain glass wool. Proper mixing for split injections is done by the internal liner geometry. Except as noted, split injections with a split ratio of 10:1 were used. For high matrix samples, this roughly matches the amount of matrix injected with the column capacity. If excess amounts of matrix are injected, the RTs of targets can shift. Split injection is also the easiest and most reliable way of screening samples for analytes ranging in
Table 1.
Gas Chromatograph and Mass Spectrometer Conditions Original 1X Method 3X Method 7X Method
GC Agilent Technologies 6890N 7683 Autoinjector and Tray Inlet Mode Injection type Injection volume (uL) Inlet temp ( C) Pressure, nominal (psig) RT Locking compound RT Locking time (min) Split ratio Gas saver Gas type Oven Voltage (VAC) Initial oven temp (C) Initial oven hold (min) Ramp rate (C/min) Final temp (C) Final hold (min) Total run time (min) Equilibration time (min) Column Type Agilent part number Length (m) Diameter (mm) Film thickness (um) Outlet pressure (AUX EPC, psig) FPD or DFPD Type Temperature (C) Hydrogen flow (mL/min) Air flow (mL/min) Mode: Constant makeup flow Nitrogen makeup flow (mL/min) Data rate (Hz) ECD Temperature (C) Nitrogen makeup flow (mL/min) Mode: Constant makeup flow Data rate (Hz) AUX EPC Pressure Pressure (psig) Gas type EPC Split/Splitless Constant pressure Split 1.0 250 31.17 Tripropyl phosphate 12.874 10:1 Off Helium EPC Split/Splitless Constant pressure Split 1.0 250 23.96 Tripropyl phosphate 4.291 10:1 Off Helium EPC Split/Splitless Constant pressure Split 1.0 250 8.84 Tripropyl phosphate 1.839 10:1 Off Helium
120 or 240 40 2 10 300 15 43.00 0.5
240 40 0.667 30 300 5 14.33 0.5
240 (and pillow) 40 0.286 70 300 2.143 6.14 0.5
HP 5-MS inert 19091S-433i 30 0.25 0.25 3.8
HP 5-MS 19091S-431 15 0.25 0.25 3.8
HP 5-MS Custom 5 0.25 0.25 3.8
Dual, S and P 250 75 100 60 10
30 0 60 5
300 60 10
N/A N/A N/A
3.8 Helium
3.8 Helium
3.8 Helium
Table 1.
Gas Chromatograph and Mass Spectrometer Conditions (Continued)
MSD Agilent Technologies Tune file Mode Solvent delay (min) EM voltage Low mass (amu) High mass (amu) Threshold Sampling Scans/s Quad temp (C) Source temp (C) Transfer line temp (C) Splitter Type 6890N option number Flow ratio [Deactivated fused silica tubing] MSD restrictor length (m) MSD restrictor id (mm) FPD/DFPD restrictor length (m) FPD/DFPD restrictor id (mm) ECD restrictor length (m) ECD restrictor id (mm) Ambient Headspace Inlet Mode Injection type Inlet temp ( C) Pressure, nominal (psig) RT locking compound RT locking time (min) Split ratio Gas saver Gas type Autoinjector Sample washes Sample pumps Injection volume (L) Syringe size (L) PreInj Solvent A washes PreInj Solvent B washes PostInj Solvent A washes PostInj Solvent B washes Viscosity delay (s) Plunger speed Pre-injection dwell (min) Post-injection dwell (min) Sampling depth (mm) [critical!]
5973 inert with Performance Electronics Atune.U Scan 0.40 Atune voltage 35 565 0 0 9.46 150 230 280
3 way 890 1:1:0.1 MSD:FPD:ECD 1.44 0.18 0.53 0.18 0.51 0.10
3 way 890 1:1:0.1 MSD:FPD:ECD 1.44 0.18 0.53 0.18 0.51 0.10
2 way 889 1:1 MSD:DFPD 1.44 0.18 0.53 0.18 N/A N/A
EPC Split/Splitless Constant pressure Split 200 31.17 Tripropyl phosphate 12.874 1:1 Off Helium
0 3 50 100 0 0 1 3 5 Fast 0 0 20
to 1/10th that going to the MSD and FPD because of the extreme sensitivity of the detector. The lengths and diameters of the detector restrictors were calculated using the spreadsheet calculator included with the splitter. The peak recognition windows used in the Agilent ChemStation were set to 0.2 min and in AMDIS to 12 s. these values were found to be sufficiently wide enough to compensate for some RT drift yet narrow enough to minimize the number of false positives. The minimum match factors setting in AMDIS was set to 45. This value seemed to give the least number of false positives and false negatives.
relatively non-polar volatiles in water. It is convenient for labs that need to screen samples for volatiles but do not have a dedicated headspace sampler. The conversion from liquid sampling to ambient headspace simply requires changing the inlet liner and the autosampler syringe. Figure 2 shows the chromatograms from a run using the system in Figure 1A. A mixture of 14 halogenated volatiles was spiked into water at 2 ppm. Fifty microliters of the approximately 1 mL of headspace in the vial was injected. With the exception of peaks 3 and 4, which coelute, the compounds are well separated. The ECD chromatogram is inverted for comparison with the MS total ion chromatogram from the full-scan data. All of the volatiles respond on the ECD, although the response to compounds 1, 2, and 8 is significantly lower than for the rest of the compounds. In general with an ECD the response to a compound increases dramatically with the number of halogens in the molecule. Since none of the compounds contain phosphorus, there is no response on the FPD. Figure 3 shows the DRS report for the sample. For each compound identified, the RT, Chemical Abstracts number (CAS#), and compound name are listed. A line is generated in the report if a compound is found by either the Agilent ChemStation, AMDIS, or both.
Results
Volatiles To evaluate the HCD method for volatiles analysis, headspace injection was chosen. Headspace injections are usually done with an automated heated sampler specifically designed for the purpose. Ambient headspace [7] is a variant of the technique that uses a gastight syringe in the liquid autosampler and injects the headspace from a 2-mL vial. It is unheated, and is thus limited to compounds that are volatile at room temperature. Ambient headspace works well for the analysis of
3,4 7 2 1 5 6 8 10 9 11 12
TIC: 2ppmMIX 3_Only_simscan.D\DATA.MS
14 13
TIC
ECD
FPD
2 4 6 8 10 12
5) 6) 7) 8) 9)
10) 11) 12) 13) 14)
Figure 2. 8
Ambient headspace analysis of volatile organics in water, spiked at 2 ppm per component.
Figure 3.
DRS report for the analysis in Figure 2.
The report shows that a compound has been determined as present by the Agilent ChemStation if a value appears in the Agilent ChemStation Amount column. This means the identification criteria set in the DATA ANALYSIS section of the method have been met. Typically the criteria are that the target ion is present and all three qualifier ions are present in ratios that fall within the percent uncertainty values for that compound. The Agilent ChemStation Amount listed is a very rough approximation of the amount of the compound, in nanograms, reaching the MS. This is based on the response factor originally observed when the HCD table data was collected. Since valid quantitation requires recent recalibration of response factors on the specific instrument used for analysis, the numbers in this column should never be used to report concentrations of identified analytes. The error in these values can easiliy be a factor of 10 or higher. The purpose of the listed values is to give an approximate amount that can be used to guide standard preparation for quantitative calibration of the compound, if needed. The match value listed under the AMDIS column is the degree to which the extracted (deconvolved) spectrum of the peak at that RT matched the spectrum in the HCD AMDIS target library. The higher this number, the better the spectra agree. The
column R.T. Diff sec. lists the difference in seconds between the observed RT and that in the AMDIS target library. The lower this number, the better the RTs agree. The NIST column lists the reverse-match quality of the extracted spectrum compared with the NIST05 main library spectrum with the same CAS#. The entry Hit Num. is the number of the hit in the NIST search results that has the same CAS# as the identified compound. The higher the reverse-match value and the lower the hit number, the better the extracted spectrum matches with NIST05. The NIST column serves as a second opinion on the identity of the extracted spectrum. The analysis in Figure 2 is of course an easy one, but serves to demonstrate how the system works. All 14 spiked compounds were found by both the Agilent ChemStation and AMDIS. The certainty of identification is very high because: The target ion and three qualifier ions are present in appropriate ratios and at the appropriate time as determined by the Agilent ChemStation The deconvolved spectrum and the RT at which it appears closely matches the data in the AMDIS target library.
9
The extracted spectrum of the identified compound also matches the spectrum with the same CAS # in the NIST05 library. The compounds all have a significant response on the ECD, as expected from their halogen content. To challenge the system in a more realistic way, the effect of matrix and dilution of the analytes was studied. Additional samples were prepared that contained: the same 2-ppm mixture of analytes plus 100 ppm of pump gasoline; 100 ppb of analytes only; and 100 ppb of analytes plus 100 ppm of pump gasoline. Figure 4 shows the chromatograms from the 100 ppb of analytes with 100 ppm of gasoline. The complexity of the TIC chromatogram illustrates the severe matrix challenge presented by the thousand-fold excess of gasoline. In the ECD chromatogram, interference peaks are now apparent. However, with the exception of peaks 1, 2, 8, and 12, all of the analytes peaks are still visible above the matrix interferences.
Table 2 summarizes the results from the matrix and dilution experiments. In the sample that was 2 ppm of analytes with 100 ppm of gasoline, the Agilent ChemStation (column labeled Quant) found all but two of the compounds. Those two compounds had qualifier ions out of range due to interferences from the matrix. AMDIS successfully found all 14 compounds. Also, with the exception of compound 8, all of the analytes were clearly visible above the matrix responses on the ECD chromatogram. In the sample that contained 100 ppb of analytes but without gasoline, quant found 7 of the 14 analytes. Using full-scan data, the signal to noise ratio for most of the analytes at the 100-ppb level is very low. This results in difficulties with finding the qualifier ions in ratios that fall within the specified uncertainty range in the quant calibration table. AMDIS found 11 of the 14 compounds. Peak 3 was not found due to a severe overlap with the coeluting peak number 4. Peaks 9 and 13 were missed by AMDIS because the signal to noise ratio was too low.
TIC
1 2 5 6
8 7 11
12
ECD
9 10 13 14
3,4
FPD
2 4 6 8 10 12
5) 6) 7) 8) 9)
10) 11) 12) 13) 14)
Figure 4.
Ambient headspace analysis of volatile organics in water. Analytes at 100 ppb plus pump gasoline at 100 ppm.
10
Table 2.
Effect of Matrix and Concentration on DRS Results 2 ppm STD only 2 ppm STD with 100 ppm gasoline Quant AMDIS (ng) (match) 2.47 7.34 5.59 7.71 4.81 3.05 3.50 5.32 2.41 1.85 2.40 3.54 12 93 98 64 97 98 84 96 97 95 99 98 98 90 89 14 0.65 7 0.07 0.12 0.37 0.21 0.40 0.21 100 ppb STD only Quant (ng) AMDIS (match) 73 89 Overlap 90 88 53 72 66 S/N 89 48 79 S/N 75 10 0.36 7 0.31 0.19 0.14 0.22 0.30 0.23 100 ppb STD with 100 ppm gasoline Quant AMDIS (ng) (match) 65 85 Overlap 82 74 Overlap Overlap 46 66 88 53 75 59 52 11
RT (min) 1.491 1.536 1.793 1.863 2.317 2.658 2.735 2.938 3.250 4.003 5.151 5.283 8.208
Compound 1,2-Dichloroethane 1,1-Dichloropropylene 1,2-Dichloropropane Trichloroethylene cis-1,3-Dichloropropylene trans-1,3-Dichloropropylene 1,1,2-Trichloroethane 1,3-Dichloropropane 1,2-Dibromoethane 1,1,1,2-Tetrachloroethane 1,1,2,2-Tetrachloroethane 1,2,3-Trichloropropane 3-Chloro-1,2-dibromopropane Total Found
Peak Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Quant (ng) 2.27 7.60 4.92 7.58 4.39 3.30 2.82 3.39 2.60 5.15 2.38 1.89 1.62 16.46 14
AMDIS (match) 97 100 95 99 98 97 99 98 91 100 99 98 93 94 14
10.435 Hexachlorobutadiene
With 100 ppm of gasoline added to the 100-ppb sample, quant again found 7 of the 14 compounds and AMDIS again found 11 of the 14. Curiously, in both cases some of the compounds missed in the absence of matrix were now found. It is possible that the presence of matrix enhances the concentration of some of the analytes in the headspace. The compounds missed in quant were again the result of low signal to noise and/or interference. In AMDIS the three missed peaks were due to severe interferences from the gasoline. As indicated above, the ECD response from 10 of the 14 compounds was still visible above the peaks due to interferences. SIM/Scan The quant data in Table 2 was generated using full scan mode. Peak 13 was missed in quant due to low signal to noise ratio. SIM/Scan mode can be
used to collect SIM data simultaneously with the scan data. The 100 ppb plus 100-ppm gasoline sample was run in SIM/Scan mode with SIM groups for each of the 14 analytes. Figure 5 compares the target and qualifier extracted ion chromatograms in both modes with the ECD response for peak 13. The signal-to-noise (peak to peak) for the target ion increases from 34 in full scan mode to 433 in SIM mode. The peaks lost in quant due to low signal-to-noise were all recovered in SIM mode. This example demonstrates the power of SIM/Scan when looking for high-priority targets at low levels. If necessary, the ECD could also be used for quantitation, as it has a high signal to noise ratio and is free from interference.
11
Ion 157 Scan Ion 75 Scan Ion 155 Scan
s/n (pk-pk) = 34
Ion 39 Scan s/n (pk-pk) = 433
Ion 157 SIM Ion 75 SIM Ion 155 SIM
Ion 39 SIM
ECD
s/n (pk-pk) = 122
8.0
8.1
8.2
8.3
8.4
Figure 5.
Target and qualifier extracted ion chromatograms for peak 13 (3-Chloro-1,2-dibromopropane) in Figure 4. SIM, scan, and ECD data collected simultaneously.
AMDIS Figure 6 illustrates the ability of AMDIS to clean the interference ions from the spectrum of an analyte. The raw spectrum at the top of Figure 6 was taken at the apex of peak 13 in the 100 ppb plus 100-ppm gasoline sample. When searched against the NIST05 library using the NIST search program, the actual compound (3-Chloro-1,2-dibromopropane) was the 70th hit in the search results. Using manual subtraction of nearby spectra in the Agilent ChemStation data analysis program improved the quality of the spectrum so that it was now the second hit when searched in NIST. This is a tedious process, however, when dealing with a large number of analytes. The spectrum as deconvolved by AMDIS is shown in Figure 6 above the
NIST05 library spectrum. When this spectrum is searched, it is the first hit in the results. The automated deconvolution provided by AMDIS saves an enormous amount of time in the data review process. Fast Methods When a retention time locked database is constructed, the RTs are (or at least should be) collected under the highest resolution conditions expected for the application. If the database is collected under constant pressure mode, method translation can then be used to adjust the speed of the method to meet the needs of different situations.
12
100
155 44 75 55 81 97 117 132 188
50
Raw spectrum
211 227 246 261 280 300
0 157
100 75 38 0 30 60 90 120 58 95 115 132
50
Manual subtraction
188 150 157 180 211 210 230 246 240 261 270 282 300 300
100 75 50 39 49 0 49 50 100 30 50 39 75 70 90 110 130 61 85 93 93 105 136 119 129
Deconvolved spectrum
188 187 199
NIST 05 library
157 150 170 190
Figure 6.
Comparison of raw, manually subtracted, AMDIS deconvoluted, and NIST05 reference spectra for peak 13 (3-Chloro-1,2-dibromopropane) in Figure 4.
The 3X method uses RTs in its database that are simply the RTs from the 1X method divided by exactly 3. The 7X method likewise uses RTs that are 1/7 of those in the original database. The quality of RTs matching between the two new faster methods and the new divided databases is demonstrated in Figure 7. Three different mixtures containing 13 chlorinated hydrocarbons and 36 pesticides were run with the two methods. The RTs were compared to those in the two new databases. The graph at the top of Figure 7 plots the database RT on the x-axis versus the difference of the measured RT from the database on the y axis. If the RT matching were perfect, the plot would be a straight horizontal line at zero height on the y axis. The maximum deviation from the table values for the 3X method was 0.047 min. The plot
indicates that a peak recognition window of 0.1 min should be sufficient. The maximum deviation in the 7X plot at the bottom of Figure 8 is +0.032 min indicating that the same peak recognition window could be used here as well. In general the RTs in scaled methods agree very well with the predicted RTs. The conditions for the two higher-speed methods were chosen to increase speed while maintaining the same column capacity. The capacity is important for both the dynamic range of quantitative measurements and for minimizing analyte RT shifts in samples with high levels of matrix. In gas chromatography, the well-known triangle of speed, resolution, and capacity dictates that if the capacity is to be maintained and the speed is to be increased, then the resolution will decrease.
13
0.100
0.000
_0.050
_0.100 0.0 2.0 4.0 6.0 Table RT (min) 8.0 10.0 12.0
0.100
0.000 _0.050 _0.100 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Table RT (min)
Figure 7.
Difference between scaled HCD table and experimental retention times for 50 compound test set. Y axis is table value minus experimental, X axis is table RT. Top plot is 3X, bottom is 7X.
Figure 8 shows three sets of chromatograms using the HCD database at three different speeds. The sample consists of nine organophosphorus pesticides (identified in the caption to Figure 8) at 50 ppm and a matrix consisting of an equal volume mixture of gasoline, kerosene, and diesel fuel spiked at 50,000 ppm total mixture. The 1X and 3X data were collected on the three-way splitter instrument and the 7X was collected on the DFPD instrument. All nine compounds also contained sulfur as can be seen in the DFPD sulfur chromatogram at the bottom of Figure 8. Note that the sulfur tails somewhat compared to the phosphorus.
14
TIC
1 2 3 4 5 6 7 8 9
1X
P (FPD) ECD
12
16
20
24
28
TIC
3X
P (FPD) ECD
1 2 3 4 5 6 7 8 9
TIC: 50_OP_50K_GKD.D\DATA.MS
TIC
7X
P (FPD) S (FPD)
1 2 3 4
Peak identities
1) 2) 3) 4)
O,O,O-triethyl phosphorothioate Thionazin Sulfotepp Phorate
5) 6) 7) 8) 9)
Dimethoate Disulfoton Methyl parathion Parathion Famphur
Figure 8.
Comparison of 1X, 3X, and 7X chromatograms. 1X and 3X were run on GC/MS/ECD/FPD system, 7X on GC/MS/DFPD.
15
Figure 9 expands the RT region of the phosphorous chromatogram containing peaks 3, 4, and 5 from Figure 8. The decrease in resolution with increasing speed is clearly evident. If only the standard target and three qualifier ion approach is used, the loss in resolution causes a significant problems. With the 1X method, all nine of the analytes are identified and eight false positives are reported. With the 3X method, all analytes are again found but now with 25 false positives. With the significantly decreased resolution of the 7X method, only seven of the nine analytes are identified and 48 false positives are reported.
3
The situation is much different when using the approach described here. Even in the worst situation, the 7X method, AMDIS finds all nine analytes with high-quality matches and only three false positives. The DRS report for the 7X analysis is shown in Table 3. To simplify the table, the 48 false positives that only appear in the quant column are not shown. The analyte compounds are shown in bold. All show close RT and high-quality spectral matches to both the AMDIS target library and to the NIST05 library.
1X
16.50
17.75
3X
5.50
5.916
7X
2.365
2.544
Peak identities
3) Sulfotepp 4) Phorate 5) Dimethoate Figure 9. Comparison of FPD phosphorus chromatograms from 1X, 3X, and 7X runs in Figure 8.
16
Table 3.
DRS Report for 7x Analysis of 50 ppm Pesticides In 50,000 ppm Gasoline/Kerosine/Diesel Matrix Agilent ChemStation amount (ng) 13.92 AMDIS match 71 69 46 64 55 91 88 90 84 92 92 91 93 RT Diff (sec.) 9.5 0.7 7.6 0.6 2.3 0.5 0.5 0.6 0.7 0.6 0.6 0.7 0.8 NIST reverse match 74 71 74 80 85 85 83 85 85 88 82 85 85 Hit number 50 1 1 3 1 1 1 1 1 1 1 1 1
RT 0.973 1.380 1.520 1.520 2.113 2.138 2.138 2.275 2.417 2.427 2.485 2.619 2.748 2.901 3.360
Cas # 98862 126681 94597 52417502 132649 90437 2131411 297972 3689245 298022 60515 298044 298000 56382 52857
Compound name Acetophenone O,O,O-triethyl phosphorothioate Safrole Benzeneacetaldehyde, ,2,5-trimethylDibenzofuran o-Phenylphenol Naphthalene, 1,4,5-trimethylThionazin Sulfotepp Phorate Dimethoate Disulfoton Methyl parathion Parathion (ethyl) Famphur (48 quant-only hits not shown)
0.35
89.2 23.31 27.34 22.7 25.12
The peak at 0.973 minutes is a reasonable spectral match to acetophenone, but the large time difference and being the 50th hit in the NIST search results suggests that this is not the compound. The peak at 1.520 min is a poor spectral match with a large time difference. The absence of a NIST reverse search and hit entry means that the listed compound was not in the top 100 hits in the NIST search. The next compound listed at 1.520 min is the top entry from the NIST search. It is quite clear that safrole is not present. The peak at 2.113 min, dibenzofuran, was not one of the analytes added to the sample. However, it probably is present in the diesel fuel matrix. Its presence is supported by both reasonably good spectral matches and close time matching with a database. The last extraneous peak at 2.138 min is also questionable. The time match is somewhat poor and the NIST reverse search suggests the identification is not correct.
All nine analytes are detected with the FPD on both the phosphorus and sulfur chromatograms. All analytes except peak 1 are detected selectively on the ECD as well. These results suggest that while the loss of resolution in going to 7X is unacceptable when using only conventional screening approaches, with the method discussed here, it is a viable option. By using the DRS report combined with the selective detector data, the number of false positives and false negatives are significantly reduced. For those situations where speed is a critical factor, for example in response to homeland security incidents, the fastest method may be the one of choice. For many laboratories, the 3X method would be an attractive choice. It has higher resolution than the 7X and higher speed than the 1X and still allows the use of two GC detectors in parallel with the MSD. It also only requires a 240V oven, not the repositioning of the MSD to the back position.
17
Conclusions
The systems described here offer several advantages when screening samples for the presence of hazardous chemicals. The advantages derive from a combination of techniques that result in both faster and more accurate screening results. Retention time locked target database of 731 hazardous chemicals for screening with MS Microfluidic splitter - using selective detection simultaneous with MS data for added confirmation, finding non-target suspect compounds, and alternate quantitation SIM/Scan - Acquire SIM data on high priority targets simultaneously with scan data. Saves time by eliminating need to run samples in both modes. DRS - automated deconvolution dramatically increases accuracy of target identification, even in the most challenging matrices. The reduction of data interpretation from hours to minutes is especially useful for response to hazardous chemical incidents. Fast chromatography using shorter columns, faster ovens, and backflushing to greatly reduce run times. This combination of techniques offers a viable solution to the hazardous chemicals challenge.
3. Chin-Kai Meng, Improving Productivity with Synchronous SIM/Scan, Agilent Technologies, publication 5989-3108EN www.agilent.com/chem 4. Philip Wylie, Michael Szelewski, Chin-Kai Meng, Christopher Sandy, Comprehensive Pesticide Screening by GC/MSD using Deconvolution Reporting Software, Agilent Technologies, publication 5989-1157EN www.agilent.com/chem 5. B. D. Quimby, L. M. Blumberg, M. S. Klee, and P. L. Wylie, Precise Time-Scaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking, Agilent Technologies, publication 5967-5820E www.agilent.com/chem 6. Michael J. Szelewski and Bruce Quimby, New Tools for Rapid Pesticide Analysis in High Matrix Samples, Agilent Technologies, publication 5989-1716EN www.agilent.com/chem 7. Michael J. Szelewski and Bruce D. Quimby, Ambient Headspace GC and GC-MSD Analysis of Nonpolar Volatiles in Water, Agilent Technologies, publication 5968-9455E www.agilent.com/chem

References
1. Retention time locking (RTL), Vince Giarrocco, Bruce Quimby, and Matthew Klee, Retention Time Locking: Concepts and Applications, Agilent Technologies, publication 5966-2469E www.agilent.com/chem 2. Chin Kai-Meng and Bruce Quimby, Identifying Pesticides with Full Scan, SIM, ECD, and FPD from a Single Injection, Agilent Technologies, publication 5989-3299EN www.agilent.com/chem
18
Appendix A

Volatiles: EPA 502/524, 60 compounds
1,1,1,2-Tetrachloroethane 1,1,1-Trichloroethane 1,1,2,2-Tetrachloroethane 1,1,2-Trichloroethane 1,1-Dichloroethane 1,1-Dichloroethylene 1,1-Dichloropropylene 1,2,3-trichlorobenzene 1,2,3-Trichloropropane 1,2,4-trichlorobenzene 1,2,4-trimethylbenzene 1,2-Dibromoethane 1,2-dichlorobenzene 1,2-Dichloroethane 1,2-Dichloropropane 1,3,5-trimethylbenzene 1,3-dichlorobenzene 1,3-Dichloropropane 1,4-dichlorobenzene 2,2-Dichloropropane 2-chlorotoluene 3-Chloro-1,2-dibromopropane 4-chlorotoluene Benzene Bromobenzene Bromochloromethane Bromodichloromethane Bromoform Bromomethane Carbon Tetrachloride Chlorobenzene Chlorodibromomethane Chloroethane Chloroform Chloromethane cis-1,2-Dichloroethylene cis-1,3-Dichloropropylene Dibromomethane Dichlorodifluoromethane Ethylbenzene Hexachlorobutadiene Isopropylbenzene Methylene Chloride m-xylene Naphthalene n-butylbenzene n-propylbenzene o-Xylene p-isopropyltoluene p-xylene Styrene tert-butylbenzene Tetrachloroethylene Toluene trans-1,2-Dichloroethylene trans-1,3-Dichloropropylene Trichloroethylene Trichlorofluoromethane Vinyl chloride 4,4'-DDD 4,4'-DDE 4,4'-DDT 4-aminobiphenyl 4-bromophenyl phenyl ether 4-chloro-3-methylphenol 4-chloroaniline 4-chlorophenyl phenyl ether 4-nitroaniline 4-nitrophenol 4-nitroquinoline-1-oxide 5-nitro-o-toluidine 7,12-dimethylbenz[a]anthracene a,a-dimethylphenethylamine Acenaphthene Acenaphthylene Acetone Acetophenone Aldrin Alpha-BHC (alpha-HCH) Aniline Anthracene Aramite (total) Benz[a]anthracene Benzene Benzo[a]pyrene Benzo[b]fluoranthene Benzo[ghi]perylene Benzo[k]fluoranthene Benzyl alcohol Beta-BHC (beta-HCH) Bis(2-chloroethoxy)methane Bis(2-chloroethyl) ether Bis(2-chloroisopropyl) ether Bis(2-ethylhexyl)phthalate Butyl benzyl phthalate Chlorobenzilate Chrysene Delta-BHC (delta-HCH) Diallate (total) Dibenz[a,h]anthracene Dibenzofuran Dieldrin Diethyl phthalate Dimethoate Dimethyl phthalate Di-n-butyl phthalate
Semivolatiles: EPA 8270C Appendix IX, 140 compounds

1,2,4,5-tetrachlorobenzene 1,2,4-trichlorobenzene 1,2-dichlorobenzene 1,3,5-trinitrobenzene 1,3-dichlorobenzene 1,4-dichlorobenzene 1,4-naphthoquinone 1-naphthylamine 2,3,4,6-tetrachlorophenol 2,4,5-trichlorophenol 2,4,6-trichlorophenol 2,4-dichlorophenol 2,4-dimethylphenol 2,4-dinitrophenol 2,4-dinitrotoluene 2,6-dichlorophenol 2,6-dinitrotoluene 2-acetylaminofluorene 2-chloronaphthalene 2-chlorophenol 2-methyl-4,6-dinitrophenol 2-methylnaphthalene 2-naphthylamine 2-nitroaniline 2-nitrophenol 2-picoline 3,3'-dichlorobenzidine 3,3'-dimethylbenzidine 3-methylcholanthrene 3-nitroaniline
19
Di-n-octyl phthalate Dinoseb Diphenylamine Disulfoton Endosulfan I Endosulfan II Endosulfan sulfate Endrin Endrin aldehyde Ethyl methanesulfonate Famphur Fluoranthene Fluorene Gamma-BHC (lindane) Heptachlor Heptachlor epoxide -isomer B Hexachlorobenzene Hexachlorobutadiene Hexachlorocyclopentadiene Hexachloroethane Hexachlorophene Hexachloropropene Indeno[1,2,3-cd]pyrene Isodrin Isophorone Isosafrole Kepone m-cresol (3-methylphenol) m-dinitrobenzene Methapyrilene Methoxychlor Methyl methanesulfonate Methyl parathion Naphthalene Nitrobenzene N-nitrosodiethylamine N-nitrosodimethylamine N-nitrosodi-n-butylamine N-nitrosodi-n-propylamine N-nitrosodiphenylamine N-nitrosomethylethylamine N-nitrosomorpholine (4-nitrosomorpholine) N-nitrosopiperidine (1-nitrosopiperidine) N-nitrosopyrrolidine (1-nitrosopyrrolidine) O,O,O-triethyl phosphorothioate o-cresol (2-methylphenol) o-toluidine p-(dimethylamino)azobenzene Parathion (ethyl) p-cresol (4-methylphenol) Pentachlorobenzene Pentachloroethane Pentachloronitrobenzene Pentachlorophenol Phenacetin 20
Phenanthrene Phenol Phorate p-phenylenediamine Pronamide Pyrene Pyridine Safrole Sulfotepp Thionazin
Chlorinated Dioxins and Furans: EPA 8282, 19 compounds

2,3,7,8-Tetrachlorodibenzo-p-dioxin 1,2,3,7,8-Pentachlorodibenzo-p-dioxin 1,2,3,4,7,8-Hexachlorodibenzo-p-dioxin 1,2,3,4,6,7,8-Heptachlorodibenzo-p-dioxin Octachlorodibenzo-p-dioxin 2,3,7,8-Tetrachlorodibenzofuran 1,2,3,7,8-Pentachlorodibenzofuran 1,2,3,4,7,8-Hexachlorodibenzofuran 1,2,3,4,6,7,8-Heptachlorodibenzofuran Octachlorodibenzofuran
Polychlorinatedbiphenyls: EPA 8082, 19 compounds

2-chlorobiphenyl 2,3-dichlorobiphenyl 2,2',5-trichlorobiphenyl 2,4',5-trichlorobiphenyl 2,2',5,5'-tetrachlorobiphenyl 2,2',3,5'-tetrachlorobiphenyl 2,3',4,4'-tetrachlorobiphenyl 2,2',4,5,5'-pentachlorobiphenyl 2,2',3,4,5'-pentachlorobiphenyl 2,3,3',4',6-pentachlorobiphenyl 2,2',3,5,5',6-hexachlorobiphenyl 2,2',4,4',5,5'-hexachlorobiphenyl 2,2',3,4,5,5'-hexachlorobiphenyl 2,2',3,4,4',5'-hexachlorobiphenyl 2,2',3,4',5,5',6-heptachlorobiphenyl 2,2',3,4,4',5',6-heptachlorobiphenyl 2,2',3,4,4',5,5'-heptachlorobiphenyl 2,2',3,3',4,4',5-heptachlorobiphenyl 2,2',3,3',4,4',5,5',6-nonachlorobiphenyl
Pesticides: Agilent RTL pesticide database (adapted), 567 compounds

1,2,4-Trichlorobenzene 1,2-Dibromo-3-chloropropane 17a-Ethynylestradiol 2-(1-naphthyl)acetamide 2-(2-Butoxyethoxy)ethyl thiocyanate 2-(Octylthio)ethanol
2,3,4,5-Tetrachlorophenol 2,3,4,6-Tetrachlorophenol 2,3,5,6-Tetrachlorophenol 2,3,5-Trichlorophenol 2,3,5-Trimethacarb 2,3,5-Trimethylphenyl methyl carbamate (Trimethacarb) 2,3,7,8-Tetrachlorodibenzofuran 2,3,7,8-Tetrachlorodibenzo-p-dioxin 2,4,5-T methyl ester 2,4,5-Trichlorophenol 2,4,6-Trichlorophenol 2,4-D methyl ester 2,4-D sec-butyl ester 2,4-DB methyl ester 2,4-Dichlorophenol 2,4-Dichlorophenyl benzenesulfonate 2,4-Dimethylaniline 2,6-Dichlorobenzonitrile 2,6-Dimethylaniline 2-[3-Chlorophenoxy]propionamide 2-Ethyl-1,3-hexanediol 2-Hydroxyestradiol 2-Methylphenol 2-Phenoxypropionic acid 3,4,5-Trimethacarb 3,4-Dichloroaniline 3,5-Dichloroaniline 3-Chloroaniline 3-Hydroxycarbofuran 4,4'-Dichlorobenzophenone 4,6-Dinitro-o-cresol (DNOC) 4-Chloroaniline 4-Methylphenol 5,7-Dihydroxy-4'-methoxyisoflavone 9,10-Anthraquinone Acephate Acetochlor Acifluorfen methyl ester Alachlor Aldrin Allidochlor Ametryn Amidithion Aminocarb Amitraz Ancymidol Anilazine Aniline Atraton Atrazine Azaconazole Azamethiphos Azinphos-ethyl Azinphos-methyl Aziprotryne Azobenzene Barban
Benalaxyl Benazolin-ethyl Bendiocarb Benfluralin Benfuresate Benodanil Bentazone Bentazone methyl derivative Benthiocarb Benzo(a)pyrene Benzophenone Benzoylprop ethyl b-Estradiol BHC alpha isomer BHC beta isomer BHC delta isomer Bifenox Bifenthrin Binapacryl Bioallethrin Bioallethrin S-cyclopentenyl isomer Bioresmethrin Biphenyl Bis(2-ethylhexyl)phthalate Bisphenol A Bitertanol I Bitertanol II Bromacil Bromobutide Bromocyclen Bromophos Bromophos-ethyl Bromopropylate Bromoxynil Bromoxynil octanoic acid ester Buprofezin Butachlor Butamifos Butoxycarboxim Butralin Butyl benzyl phthalate Butylate Butylated hydroxyanisole Captafol Captan Carbaryl Carbetamide Carbofuran Carbofuran-3-keto Carbophenothion Carbosulfan Carboxin Chinomethionat Chloramben methyl ester Chloranocryl Chlorbenside Chlorbromuron
Chlorbufam Chlordecone Chlordimeform Chlorfenethol Chlorfenprop-methyl Chlorfenson Chlorfenvinphos Chlorflurecol-methyl ester Chlormefos Chlornitrofen Chlorobenzilate Chloroneb Chloropropylate Chlorothalonil Chlorotoluron Chlorpropham Chlorpyrifos Chlorpyrifos Methyl Chlorthal-dimethyl Chlorthiamid Chlorthion Chlorthiophos Chlorthiophos sulfone Chlorthiophos sulfoxide Chlozolinate cis-Chlordane Clomazone Coumaphos Crimidine Crotoxyphos Crufomate Cyanazine Cyanofenphos Cyanophos Cycloate Cycluron Cyfluthrin I Cyfluthrin II Cyfluthrin III Cyfluthrin IV Cyhalothrin I (lambda) Cymoxanil Cypermethrin I Cypermethrin II Cypermethrin III Cypermethrin IV Cyprazine Cyprofuram Cyromazine d-(cis-trans)-Phenothrin-I d-(cis-trans)-Phenothrin-II Dazomet Decachlorobiphenyl Deltamethrin Demephion
Demeton-S Demeton-S-methylsulfon Desbromo-bromobutide Desmedipham Desmetryn Dialifos Di-allate I Di-allate II Diamyl phthalate Diazinon Dibrom (naled) Dicamba Dicamba methyl ester Dicapthon Dichlofenthion Dichlofluanid Dichlone Dichlormid Dichlorophen Dichlorprop Dichlorprop methyl ester Dichlorvos Diclobutrazol Diclofop methyl Dicloran Dicrotophos Dicyclohexyl phthalate Dicyclopentadiene Dieldrin Diethatyl ethyl Diethofencarb Diethyl dithiobis(thionoformate) (EXD) Diethyl phthalate Diethylene glycol Diethylstilbestrol Difenoconazol I Difenoconazol II Diflufenican Dimefox Dimethachlor Dimethametryn Dimethipin Dimethoate Dimethylphthalate Dimethylvinphos(z) Dimetilan Di-n-butylphthalate Diniconazole Dinitramine Dinobuton Dinocap I Dinocap II Dinocap III Dinocap IV
21
Dinoseb Dinoseb acetate Dinoseb methyl ether Dinoterb Dinoterb acetate Di-n-propyl phthalate Dioxacarb Dioxathion Dioxydemeton-S-methyl Diphacinone Diphenamid Diphenylamine Dipropetryn Disulfoton Ditalimfos Dithiopyr Diuron Dodemorph I Dodemorph II Drazoxolon Edifenphos Endosulfan (alpha isomer) Endosulfan (beta isomer) Endosulfan ether Endosulfan lactone Endosulfan sulfate Endrin Endrin aldehyde Endrin ketone EPN Epoxiconazole EPTC Erbon Esfenvalerate Esprocarb Etaconazole Ethalfluralin Ethiofencarb Ethiolate Ethion Ethofumesate Ethoprophos Ethoxyquin Ethylenethiourea Etridiazole Etrimfos Famphur Fenarimol Fenazaflor Fenbuconazole Fenchlorphos Fenfuram Fenitrothion Fenobucarb Fenoprop Fenoprop methyl ester Fenoxycarb 22
Fenpropathrin Fenson Fensulfothion Fenthion Fenthion sulfoxide Fenuron Fenvalerate I Fenvalerate II Fepropimorph Flamprop-isopropyl Flamprop-methyl Fluazifop-p-butyl Flubenzimine Fluchloralin Flucythrinate I Flucythrinate II Flumetralin Fluometuron Fluorodifen Fluotrimazole Flurenol-butyl ester Flurenol-methylester Fluridone Flurochloridone I Flurochloridone II Fluroxypyr-1-methylheptyl ester Flusilazole Flutolanil Flutriafol Fluvalinate-tau-I Fluvalinate-tau-II Folpet Fonofos Formothion Fuberidazole Furalaxyl Furathiocarb Furmecyclox Heptachlor Heptachlor epoxide Heptachlor exo-epoxide isomer B Heptenophos Hexabromobenzene Hexachlorobenzene Hexachlorophene Hexaconazole Hexazinone Hexestrol Imazalil Ioxynil Iprobenfos Iprodione Isazophos Isobenzan Isobornyl thiocyanoacetate Isocarbamide Isodrin
Isofenphos Isomethiozin Isoprocarb Isopropalin Isoprothiolane Isoproturon Isoxaben Isoxathion Jodfenphos Kinoprene Lenacil Leptophos Leptophos oxon Lindane Linuron Malathion Malathion-o-analog MCPA methyl ester MCPB methyl ester m-Cresol Mecarbam Mecoprop methyl ester Mefenacet Mefluidide Menazon Mephosfolan Mepronil Metalaxyl Metamitron Metasystox thiol Metazachlor Methacrifos Methamidophos Methfuroxam Methidathion Methiocarb Methiocarb sulfone Methiocarb sulfoxide Methomyl Methoprene I Methoprene II Methoprotryne Methoxychlor Methyl paraoxon Methyl parathion Methyl-1-naphthalene acetate Methyldymron Metobromuron Metolachlor Metolcarb Metribuzin Mevinphos Mirex Molinate Monalide Monocrotophos Monolinuron
Myclobutanil N,N-Diethyl-m-toluamide N-1-Naphthylacetamide Naphthalic anhydride Napropamide Nicotine Nitralin Nitrapyrin Nitrofen Nitrothal-isopropyl N-Methyl-N-1-naphthyl acetamide Norflurazon Nuarimol o,p'-DDD o,p'-DDE o,p'-DDT Octachlorostyrene o-Dichlorobenzene Omethoate o-Phenylphenol Oryzalin Oxabetrinil Oxadiazon Oxadixyl Oxamyl Oxycarboxin Oxychlordane Oxydemeton-methyl Oxyfluorfen p,p'-DDD p,p'-DDE p,p'-DDT Paclobutrazol Paraoxon Parathion p-Dichlorobenzene Pebulate Penconazole Pendimethalin Pentachloroaniline Pentachloroanisole Pentachlorobenzene Pentachloronitrobenzene Pentachlorophenol Pentanochlor Permethrin I Permethrin II Perthane Phenamiphos Phenkapton Phenoxyacetic acid Phenthoate Phorate Phosalone Phosfolan Phosmet Phosphamidon I
Phosphamidon II Phthalide Picloram methyl ester Pindone Piperalin Piperonyl butoxide Piperophos Pirimicarb Pirimiphos-ethyl Pirimiphos-methyl Plifenat p-Nitrotoluene Pretilachlor Probenazole Prochloraz Procymidone Profenofos Profluralin Promecarb Prometon Prometryn Propachlor Propamocarb Propanil Propargite Propazine Propetamphos Propham Propiconazole-I Propiconazole-II Propoxur Propyzamide Prothiofos Prothoate Pyracarbolid Pyrazon Pyrazophos Pyrazoxyfen Pyributicarb Pyridaben Pyridaphenthion Pyridate Pyridinitril Pyrifenox I Pyrifenox II Pyrimethanil Pyroquilon Quinalphos Quinoclamine Quizalofop-ethyl Resmethrin S,S,S-Tributylphosphorotrithioate Sebuthylazine Secbumeton Simazine Simetryn Sulfotep
Sulfur (S8) Sulprofos Swep Tamoxifen TCMTB Tebuconazole Tebutam Tecnazene Temephos Terbacil Terbucarb Terbufos Terbumeton Terbuthylazine Terbutryne Tetrachlorvinphos Tetradifon Tetraethylpyrophosphate (TEPP) Tetramethrin I Tetramethrin II Tetrapropyl thiodiphosphate Tetrasul Thenylchlor Thiabendazole Thiofanox Thiometon Thionazin Tiocarbazil I Tiocarbazil II Tolclofos-methyl Tolylfluanid trans-Chlordane Triadimefon Triadimenol Tri-allate Triamiphos Triazophos Tributyl phosphate Tributyl phosphorotrithioite Trichlorfon Trichloronate Triclopyr methyl ester Tricyclazole Tridiphane Trietazine Triflumizole Trifluralin Tryclopyrbutoxyethyl Tycor (SMY 1500) Uniconizole-P Vamidothion Vernolate Vinclozolin
23
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Agilent Technologies, Inc. 2006 Printed in the USA February 9, 2006 5989-4834EN
New Tools for Rapid Pesticide Analysis in High Matrix Samples Application
Food Analysis
Authors
Mike Szelewski and Bruce Quimby Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Tools like Method Translation [1] have made it straightforward to reduce analysis time by a known factor and maintain the exact relative elution order of the analytes. The use of appropriate shorter and smaller diameter columns can maintain the same resolution while achieving a much shorter analysis time. One application area where this approach has met difficulty, however, is the gas chromatography/ mass spectrometry (GC/MS) analysis of pesticides in complex matrices like food. This application requires that speed-up schemes maintain column capacity in order to handle the large matrix peaks and achieve low detection limits for analytes. Since chromatography is governed by the triangle of speed, resolution, and capacity, resolution must be sacrificed to increase speed at the same capacity. The problem is that chromatographic resolution is also needed to confirm the identity of any target analytes detected in the presence of interferences from the sample. In this note, the reduction in chromatographic resolution in faster analysis is more than adequately compensated for by use of spectral deconvolution [2] and simultaneous element-selective detection for the confirmation step. The system consists of a GC/MS with a dualwavelength flame photometric detector (DFPD) for the simultaneous collection of phosphorus, sulfur, and mass spectral data. The GC column effluent is split between the two detectors in the ratio of 2:1 in favor of the mass selective detector (MSD). The system is retentiontime locked to the Agilent pesticide library [1] scaled to threefold faster times, which contains the
Abstract
Recent developments in GC/MS hardware and software make it possible to analyze samples with high levels of matrix contamination much faster than ever before. New tools such as mass spectral deconvolution, reliable and inert effluent splitters, and column backflushing capabilities can be combined to produce large time savings. By accelerating the chromatographic run, post-run bakeout, and data interpretation steps, analysis times can be shortened by at least three-fold versus conventional methods. These tools are especially useful in analyses with high levels of matrix background, such as the inspection of the food supply for contaminants. In addition to monitoring for pesticide residues, the threat of terrorism has recently raised concerns over deliberate contamination of food with other toxic materials. This article describes a GC/MS system for the rapid screening of foodstuffs for chemical contaminants with a special emphasis on pesticides, organophosphorus, and organosulfur compounds.
Introduction
Techniques for decreasing the analysis time of gas chromatography (GC) methods have been developed in recent years.
retention times and spectra for 567 pesticides used worldwide. Samples are analyzed with MS in fullscan electron-impact ionization (EI) mode. The combination of precise retention times, elemental, and mass spectral data is used to screen for specific target compounds. The flame photometric detector (FPD) data also highlights any non-target, P- or S-containing compounds for identification by MS. The MS data is screened using the standard quantitation software based on retention time (RT), ion ratios, and spectral cross correlation. The MS data is also processed using spectral deconvolution software, which greatly reduces spectral interferences from the matrix. The deconvoluted spectra are then searched against a table of targets. Any hits are confirmed by searching against the main NIST library. This process is automated by the Agilent Deconvolution Reporting Software (DRS), also providing significant time savings in data interpretation.
The system described here uses column backflushing, a technique used to save large amounts of time with complex samples. Backflushing is done with the splitter hardware. This technique removes heavy residues from the column much faster and at lower temperatures than the conventional bakeout step at the end of the run. This reduces MS source contamination by preventing the higher levels of column bleed and heavy matrix components from entering the MSD. It also increases the column lifetime. The approach used thus reduces analysis time in three major ways: shortening the chromatographic run time; automating data interpretation; and reducing bakeout time. Other notable advantages are the ability to change columns and/or inlet liners without venting the MSD, and a reduced need for MS source cleaning. System Configuration The system configuration used is shown in Figure 1. Key components are:
Auto-sampler Dual Flame Photometric Detector

Sulfur Phosphorus
AUX EPC
Effluent Splitter with Makeup
0.814 m 0.18 mm id
Column
1.1 m 0.18 mm id
6890N GC
15 m 0.25 mm id 0.25 um HP-5MS
5973 Inert MSD
Figure 1.
System configuration.
Key Components Fast Oven With the 6890N 220V oven (option 002), the pesticide analysis method can be run precisely 3 times faster (14 min) using a 15 m HP-MS column. If the 220V GC is further equipped with SP1 2310-0236 (puts MSD interface in back of oven under rear injection port) and the G2646-60500 oven insert accessory (reduces oven volume twofold), the speed can be increased to 4.8 times faster (9 min). The cool-down time of the oven is also reduced. Dual FPD 6890N Option 241 is a single flame photometric detector with two optical detection channels, one for sulfur and one for phosphorus. The signals from the DFPD are collected, stored, and processed by the MS ChemStation simultaneously with the MS data. The FPD data can be used in several ways. Nontarget organophosphorus compounds like new pesticides or designer nerve agents are highlighted. The presence of an element at the retention time of an identified compound can be used to support confirmation of identity. The response on the FPD can be used for quantitative or semi-quantitative analysis, especially for situations where no calibration standard is available for an identified analyte. Microfluidic Splitter 6890N Option 889 uses diffusion bonded plate technology combined with metal column ferrules to make an inert, easy-to-use, leakfree, high-temperature column effluent splitter. The splitter uses Auxillary (Aux) electronic pneumatics control (EPC) for constant pressure makeup (6890N Option 301). The Aux EPC makeup can be pressure programmed at the end of the run to higher pressure, while at the same time the inlet pressure is lowered to near ambient. This causes the flow in the column to reverse direction, backflushing heavy materials out the split vent of the inlet. The Aux EPC also allows column changing and maintainance without venting the MSD. When the column fitting is removed from the splitter, helium from the makeup supply purges the fitting, preventing air from entering the MSD. If the column is attached to the splitter but removed from the inlet, helium flows backwards through the column and out the inlet end. Inlet maintainance or column headtrimming can be done without cooling and venting the MSD and air is not introduced into a hot source. MSD System The 5973N Inert with Performance Electronics and performance turbo (G2579A) EI MSD is used. This configuration provides faster full scan rates while maintaining sensitivity. The scan rates are compatible with the narrower peaks
generated by fast chromatography. The performance turbo pump is required to handle the higher flows associated with fast chromatography and backflushing. DRS Software (G1716AA) Spectral deconvolution of the MS data allows identification of analytes in the presence of overlapped matrix peaks. This significantly reduces chromatographic resolution requirements, allowing much shorter analysis times. DRS utilizes the AMDIS deconvolution program from NIST, originally developed for trace chemical weapons detection in complex samples. DRS presents the analyst with three distinct levels of compound identification: (1) ChemStation, based on retention time and four ion agreement; (2) AMDIS, based on cleaned spectra full ion matching and locked retention time; (3) NIST02 search using a >147,000 compound library. Instrument Operating Parameters The recommended instrument operating parameters are listed in Table 1. These are starting conditions and may have to be optimized. Split injection was used to match the amount of matrix to the column capacity. Citrus oils cause retention shifts if excess sample is injected. Splitless injection could be used for samples with significantly less matrix. The inlet liner was found to be of low activity, as it does not contain glass wool. Proper mixing for split injections is done by the internal liner geometry. The 6890 220V oven was needed for the ramps described in Tables 1 and 2. This oven program is necessary for the precise 3 speed increase of the RTLocked pesticide database. The 15-m HP-5ms column has the same phase ratio as the 30 m column traditionally used for the 1 method. This shorter column allows a flow rate for a 3 precisely scaled faster method. The outlet is listed as unspecified because the column connects to the splitter. The splitter pressure is operated at a constant 3.8 psig using an auxillary EPC module. The 5973 inert Performance Electronics data acquisition sampling rate was set to 1, which is faster than the typical setting of 2. Signal-to-noise is improved over previous systems at faster sampling rates. More data points allows for easier integration and better deconvolution to compensate for the loss in resolution using a shorter column. The microfluidic splitter parameters are chosen to provide the desired split ratio between detectors
3
while meeting the flow requirements of the detectors used. A primary consideration with the current system is to make sure that the flow to the MSD does not exceed ~4 mL/min while collecting analyte data. It was also desired to split the effluent 2:1 in favor of the MSD. These parameters were entered into the spreadsheet calculator (included with the splitter), which calculated the lengths and diameters of the detector restrictors
Table 1. GC Inlet Mode Inlet temp Pressure Split ratio Split flow Total flow Gas saver Gas type Inlet Liner Oven Oven ramp Initial Ramp 1 Ramp 2 Ramp 3 Ramp 4 Total run time Total run time Equilibration time Oven max temp Column Length Diameter Film thickness Mode Inlet Outlet Outlet pressure Back Detector (FPD) Temperature Hydrogen flow Oxidizer flow Oxidizer gas type Mode Makeup flow Makeup gas type Flame Lit offset Photo multiplier Gas Chromatograph and Mass Spectrometer Operating Parameters Agilent Technologies 6890 EPC Split/Splitless Split, 1.0 L injected 250 C 23.84 psi 10:1 44.1 mL/min 48.1 mL/min Off Helium Siltek Cyclosplitter, 4 mm id, Restek part number 20706-214.1 220V C/min 75 9 24 50 Next C 70 150 200 280 320 Hold min 0.67 0.00 0.00 3.33 (end of pesticide ramp) 50.0 (end of oil elution)
13.96 min to elute pesticides 64.76 min to elute heavy components from citrus oils 0.5 min 325 C Agilent Technologies HP-5MS, p/n 19091S-431 15.0 m 0.25 mm 0.25 m Constant Pressure = 23.84 psi Front Unspecified 3.8 psi (aux pressure to splitter) 250 C 75.0 mL/min 100.0 mL/min Air Constant makeup flow 60.0 mL/min Nitrogen On 5.00 On
Table 1.
Gas Chromatograph and Mass Spectrometer Operating Parameters (Continued) 5 Hz Back detector On 0.0 (Off) 0 Off 0 Signal 2 Data rate: Type: Save data: Zero: Range: Fast Peaks: Attenuation: 5 Hz Front detector On 0.0 (Off) 0 Off 0
Signal 1 Data rate Type Save data Zero Range Fast Peaks Attenuation AUX Pressure 5 Description Gas type Initial pressure Initial time MSD Tune file Mode Solvent delay EM voltage Low mass High mass Threshold Sampling Scans/sec Quad temp Source temp Transfer line temp Splitter Split ratio MSD restrictor DFPD restrictor
Helium 3.80 psi 0.00 min (this value will follow oven ramp) Agilent Technologies 5973 Inert Performance Electronics Atune.U Scan 1.00 min Atune voltage 45 amu 450 amu 0 1 6.68 150 C 230 C 280 C Agilent 6890N Option 889 2:1 MSD:DFPD 1.1 m 0.18 mm id deactivated fused silica tubing 0.81 m 0.18 mm id deactivated fused silica tubing
Backflush Instrument Operating Parameters Instrument operating conditions for backflushing are shown in Table 2. These are starting parameters and will have to be optimized depending on sample matrix. Conditions listed here are only those that are different from the Table 1 conditions. Instead of baking the column at 320 C for 50 min, the heavy matrix material is backflushed through the column inlet, out the split vent. This is accomplished in 5 min at 300 C, saving 45 min of run time, preserving column life, and shortening cool down time. The column inlet pressure is reduced to 1.1 psi by using the ramped pressure feature of the EPC. At the end of the backflush time, it is ramped back to initial conditions.
Simultaneous with ramping the inlet pressure down to 1.1 psig, the Aux EPC splitter pressure is ramped up to 23 psig. This increase in pressure at the column outlet, along with the decrease in inlet pressure, backflushes the column. The backflush pressure for the Aux EPC is set to limit the flow to the MSD to < 8 mL/min. This is calculated using the Agilent Flow Calculation software program, also supplied with the splitter kit. The calculator is used to find the pressure which gives the desired flow through the MSD splitter restrictor (1.1 m 0.18 mm id) at the backflushing temperature 300 C. The result was 23.6 psig which would produce a flow of 8 mL/min, so 23 psig was used. The backflush time is determined empirically and depends on the sample matrix. The process is to try a backflush run followed by a blank run with the conventional long-hold to see if the heavy materials
are completely removed. If not, the process is repeated with a longer backflush time. As a very rough guide, start with a backflush time of five void times for the backwards flow. Using the Agilent Flow Calculation software with the inlet pressure at 23 psig, the outlet pressure at 1.1 psig, and the temperature at 300 C, the hold-up time (void time) is 0.423 min. The rough guide says that the column should be backflushed for 2.12 min. This works for removing most heavies, but 5 min is required in this case to remove them all. At the end of the backflush time, the Aux EPC is ramped back to initial conditions. The MSD is time-programmed to collect data over the time range of 1 to 13.96 min. All of the pesticides elute during this time range.
Table 2.
Backflush Gas Chromatograph and Mass Spectrometer Operating Parameters. Use Table 1 Conditions Except for These Differences C/min 75 9 24 50 Next C 70 150 200 280 300 Hold min 0.67 0.00 0.00 3.33 (end of pesticide ramp) 5.40 (end of oil backflush)
Backflush Oven Conditions Oven ramp Initial Ramp 1 Ramp 2 Ramp 3 Ramp 4 Total run time Total run time
13.96 min to elute pesticides 19.76 min to elute heavy components from citrus oils
Backflush Column Conditions Mode Ramped Pressure Press ramp psi/min Initial Ramp 1 99 Ramp 2 99 Backflush AUX 5 Conditions Press ramp psi/min Initial Ramp 1 99 Ramp 2 99 Backflush MSD Conditions Timed MS detector entries Time (min) 13.96
Next psi 23.84 1.1 23.84 Next psi 3.8 23.0 3.8
Hold min 13.96 (end of pesticide ramp) 5.57 (backflush time) 0.00 Hold min 13.96 (end of pesticide ramp) 5.61 (backflush time) 0.00
State (MS on/off) Off
Results
A mixture of 25 pesticides was run at 1, 3, and 4.8 speeds. The Total Ion Chromatograms (TICs) are shown in Figure 2. There is some loss in resolution, but the loss is limited because the shorter columns are run at flows closer to the optimum. The resolution losses are mitigated by using the faster scanning capabilities of the performance electronics together with DRS.
1 30 m
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
3 15 m
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
11.00
12.00
13.00
4.8 10 m
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
5.50
6.00
6.50
7.00
7.50
8.00
8.50
Figure 2.
TICs of pesticide test mix at three different scaled speeds. All columns have the same phase ratio.
A neat lemon oil was analyzed using the 3 speed conditions. The TIC is shown in Figure 3 with the DFPD phosphorus (P) and sulfur (S) data channels. The ChemStation software can simultaneously acquire two signals of GC data with the MSD data. The pesticides elute within the 14 min window shown, but the matrix continues to elute for an additional 50 min at 320 C.
TIC
7.441 min
FPD (P)
7.257 min
FPD (S)
10
12
14
Figure 3.
MS and DFPD data from lemon oil analyzed with 3x method.
The P and S chromatograms indicate the possible presence of numerous pesticides. The largest P peak, 7.441 min, also contains S. A PBM reverse Library search identified the peak as Methidathion (C6H11N2O4PS3) with a match quality of 45. It was also identified using a target ion and three qualifier ions. The raw apex spectrum of the P peak at 7.257 min is shown in the top of Figure 4. No match was found for a P-containing compound in the top 20 library search results. It was also not identified by the ChemStation target and qualifier ion criteria due to out-of-range ratios.
69 81 93 55 131 109
Raw spectrum at 7.257 min

121 137 159 171 187
180
206
200
227 241
220 240
256 270 283 296

260 280 300
314
329
320
40
60
80
100
120
140
160
100
58 86 76 65
131 97 159
Deconvoluted spectrum
296
171 198 226 252
329
0
86 97
50 80 110
296
100
159 131
140 170
329
Library spectrum of Mecarbam

200 230 260 290 320
Figure 4.
Top - Raw mass spectrum of peak at 7.257 min in lemon oil. Bottom - Deconvoluted spectrum of 7.257 min peak compared to NIST02 library spectrum of Mecarbam.
The DRS report is shown in Figure 5. The peak at 7.257 min is clearly identified as Mecarbam by the DRS software. The deconvoluted spectrum has a match factor of 72, compared to both the Agilent Pesticide AMDIS database and the NIST02 library. Additionally, the DRS report shows the elution time to be only 0.5 s from expected. Further confirmation is the presence of P with S barely visible. The deconvoluted spectrum for the peak at 7.257 min is shown at the bottom of Figure 4, together with the NIST library spectrum of Mecarbam. The DRS report displays the amount for compounds found by the normal ChemStation quant process. The compounds must be properly
calibrated to have a meaningful value. In this lemon oil, the ChemStation found four compounds. The AMDIS portion of DRS found two of the same compounds and an additional five compounds missed by the ChemStation. The NIST02 library search confirmed all of the compounds found by AMDIS using the NIST02 >147,000 compound library. The DRS results show good match quality at the locked retention times for seven target compounds. No single software package can produce this same confidence level in compound identification. An experienced analyst would take 14 hours to process this complex sample manually with each of the three software packages used by DRS. DRS produced this report is less than two minutes.
Figure 5.
DRS Report for lemon oil.
10
Backflushing Citrus oils contain significant amounts of material that elute after the last pesticide. This requires a 150-min hold at 320 C to elute all of the heavy material with a 1 method. The total run time for the 1 method is therefore 195 min, as shown in Table 3.
Table 3. Method Run Time Comparison 30 m 1 min 42 195 n/a 15 m 3 min 14 65 20 10 m 4.8 min 8.75 40.6 12.5
The 3 method requires a 50-min hold at 320 C, as shown at the top in Figure 6, resulting in a 65-min run time. With backflushing, all of this heavy material is removed in 5 min at 300 C, as shown in the bottom of Figure 6. This is a 9-fold reduction in analysis time compared to the 1 method. 4.8x Method Using the 220V oven, SP1 2310-0236, and oven insert accessory, the method can be scaled to 4.8 faster, as shown in Table 3. There is a practical limit to the amount of matrix that can be tolerated with the reduced resolution using a 10-m column. However, for matrices less complex than a citrus oil, the 4.8 method can be successfully used to save even more time. The Performance Electronics allows running at faster scan speeds while maintaining signal/noise ratio. Sufficient points across a peak are maintained even with faster chromatography.
Column Speed Run time Pesticides No backflush matrix With backflush matrix
Normal bakeout: 65 min run
Backflush time range
Backflush: 19 min run
10
20
30
40
50
60
Figure 6.
Top - 3 lemon oil analysis with 50 min bakeout at 320 C. Bottom - 3 lemon oil analysis with 5 min backflush at 300 C.
11
Conclusions
New tools are available for the analysis of pesticides in complex matrices. These tools can be combined to significantly reduce analysis and data processing time. Fast oven - programming rates needed for faster runs Shorter column - 3 faster runs with sufficient resolution Microfluidic splitter - confidence in results using selective detection simultaneous with MS data Backflush - additional 3 reduction in run time with lower column temperatures for extended lifetime Performance Electronics - maintain signal/noise at faster sampling rates DRS - three levels of target analyte identification in less than two minutes Using the above tools, the run time for analysis of lemon oil was reduced from 195 minutes to 20 minutes (nine-fold). DRS reduced the data analysis from hours to minutes.
References
1. B. D. Quimby, L. M. Blumberg, M. S. Klee, and P. L. Wylie Precise Time-Scaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking Agilent Technologies publication 5967-5820E www.agilent.com/chem 2. P. Wylie, M. Szelewski, and C. K. Meng Comprehensive Pesticide Screening by GC/MSD using Deconvolution Reporting Software Agilent Technologies publication 5989-1157EN www.agilent.com/chem

Precise Time-Scaling of Gas Chromatographic Methods Using Method Translation and Retention Time Locking Application
Gas Chromatography May 1998
Authors
B. D. Quimby, L. M. Blumberg, M. S. Klee, and P. L. Wylie Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Key Words
Pesticides, GC, GC-AED, retention time locking, RTL, method translation, scalable RT libraries
the observed element content of the peak. The combination of time and element content narrows rapidly the possible compounds that could have produced the heteroatom response to a few pesticides. The element-selective detection is done with either gas chromatography-atomic emission detection (GC-AED), which can screen for all the individual elements found in pesticides, or with a combination of other selective detectors like the electron capture detector (ECD), the nitrogen-phosphorus detector (NPD), the flame photometric detector (FPD), or the electrolytic conductivity detector (ELCD). The GC-AED technique can also be used to calculate element ratios and to quantitate unknown peaks that are detected because of its equimolar element response factors. The measured element ratios can be used to further distinguish between possible identities of detected heteroatomic compounds, often resulting in a single entry as the likely identity of a given peak. With compound-independent calibration, the amount of the unknown can be calculated using element response factors generated with a different standard compound.
Introduction
Interest in the analysis of pesticide residues has been increasing recently, in part due to the discovery that some of these compounds act as endocrine disrupters. Agilent Technologies has responded to the need for rapid, accurate, and comprehensive screening analysis for pesticides by developing a method to screen for 567 pesticides and suspected endocrine disrupters. The method uses element- selective detection and a retention time locked library of retention times to find and identify pesticides in a sample.1 In the method, sample extracts are run with element-selective detection using a prescribed set of chromatographic conditions and with the column retention time locked to the retention times in a table. If any peaks containing heteroatoms are observed, the section of the table corresponding to a small time window around the observed peak is searched. The time search results are further sorted using
Abstract
Complete development of a gas chromatographic method often involves a significant amount of effort. Once a method is completed, retention time locking (RTL) can be used to implement the method and to obtain the same retention times on multiple systems. This application note describes how to use method translation combined with RTL to implement precise time-scaled versions of a method on multiple instrument types. This allows the original method to be re-used with minimal effort, while optimizing the method for a given sample type or instrument setup. In this way, the utility of the original method is extended greatly, increasing the payback on the investment in its development and optimizing its use for specific analyses. In this note, the Agilent RTL Pesticide Library method is used as an example. The steps involved in precise time-scaling of the method to different speeds, detectors, and columns are presented.
Once the element-selective screen is completed, samples that contain any suspect compounds are run on a GC with mass spectral detection (GC-MS) system that is retention time locked to the pesticide method, thus having the same retention times as the element-selective detectors. Using the possible identities generated from the element screen, the GC-MS data is evaluated to decide which (if any) of the possible identities for suspect peaks is correct. The confirmation process is simplified greatly because the element screen usually yields only a few possibilities and because the retention time in the GC-MS run is accurately known. In practice, extracted ion chromatograms for characteristic ions of each possible compound are used to determine the identity of suspect compounds. This screening method minimizes false negatives, even in dirty samples, by using element-selectivity and time in the initial screen. With elementselective detection, all compounds containing chlorine, phosphorus, nitrogen, etc. are detected. Even if a detected heteroatomic compound is not in the table, its presence is known, and it can be marked for further GC-MS evaluation. By using GCMS for confirmation, false positives are also minimized. The RTL Pesticide Library method is a good example of a method in which a substantial investment of time and material has been made. As with many methods intended for use in multiple laboratories, it would be desirable to be able to scale the method for use in different situations of sample type and instrument setup. Because the method relies on the measured retention times of 567 compounds, it would be impractical to remeasure all the retention times
whenever the method is modified, for example, to increase its speed. Method translation24 is a calculation technique developed at Agilent Technologies that allows a capillary column GC method to be translated to different chromatographic conditions. The technique calculates the required changes in inlet pressure and oven temperature ramp rates and hold times required to maintain peak elution order identical to that of a reference method. In this way, the speed of an analysis can be scaled predictably to accommodate the needs of a specific sample or instrument type. The inlet pressure calculated for the new version of a method by the method translation software is based on the assumed or nominal dimensions of the column. As such, the calculated inlet pressure will provide a close, but not exact, match to the desired scaled retention times. To match precisely the retention times of the scaled method to the desired scale factor, the new method must be retention time locked. Retention time locking3 (RTL) is a technique developed by Agilent Technologies whereby the inlet pressure required to match retention times precisely is calculated from a calibration curve of inlet pressure versus retention time. Using method translation followed by RTL allows a method to be scaled by a precisely known factor. Once the chromatography has been scaled, a retention time table, such as the RTL Pesticide Library, can then be scaled by the same factor, resulting in a new library whose retention times match those of the scaled method precisely. The steps required to scale the method are: 1. Determine the desired scale factor for the new method.
2. Use the method translation software4 to calculate the inlet pressure and oven temperature adjustments to obtain the desired scaling of the method. The scale factor is the speed gain value reported in the method translation software. Make sure that the new method parameters are consistent with the hardware capabilities of where the new method will be used. 3. Perform the RTL calibration runs for the new method. Alternatively, the method translation software can be used to calculate the RTL calibration points for the new method using those from the original method. 4. Retention time lock the new method using the locking reference standard from the original method. The new method should be locked to the original reference standard retention time divided by the scale factor. 5. Export the retention time table as a text file using the EXPORT function in the RTL SEARCH menu of the RTL ChemStation software. 6. Divide the retention times in the table by the scale factor in a spreadsheet program like Microsoft Excel. 7. Re-import the new, scaled table. 8. Run a representative test mixture to validate the scaled method. Several examples of scaling the HP RTL Pesticide Library are presented below.
Experimental
All data were collected on Agilent 6890 Series GC systems. All systems were equipped with: Electronic pneumatics control (EPC)
Split/splitless inlet Automatic liquid sampler

Locking GC-MS with Other GC Detectors
When using selective GC detectors in conjunction with GC-MS, one problem that is encountered is knowing the relationship between retention times on the selective detector and that of the GC-MS. In GC-MS, the outlet pressure of the column is vacuum, while with most other GC detectors, the outlet pressure of the column is at or near atmospheric pressure. This difference in outlet pressures results in large differences in retention time between GC with MS detection and GC with other detectors. Comparison of GC-FID, a general detector, with GC-MS is reasonably straightforward, because the total ion chromatogram (TIC) of the GC-MS system has similar response to the FID. Retention times on the GC-MS system corresponding to those on the GC-FID can be determined by looking for similar patterns of response. With selective detectors, this is much more difficult because the response patterns from selective detectors usually do not resemble the TIC. For this reason, matching the retention times of selective detectors precisely with the GC-MS system simplifies data analysis greatly. In this first example of scaling the RTL Pesticide Library, the method will be scaled from the GC-AED method to the GC-MS method. In this case, the desired scale factor is exactly 1, that is, the GC-MS retention times are desired to be exactly the same as those of the GC-AED. The first step is to use the method translation software to determine the GC conditions to use for GC-MS.
The GC-AED system also included an Agilent G2350A atomic emission detector with GC-AED ChemStation software (rev B.00.00) for Microsoft Windows NT. The GC-micro-ECD system was controlled by Agilent GC ChemStation software (rev A.05.04). Both the GC-AED and the GC-micro-ECD ChemStations contained RTL software for GC ChemStation (G2080AA) and the Retention Time Locking Pesticide Library for GC ChemStation (G2081AA). The GC-MS system (G1723A) used consisted of an 6890 Series GC equipped with an Agilent 5973 mass selective detector (MSD). The process for retention time locking the GC-MS system is described in reference 2. All systems except the micro-ECD instrument used 30 m 0.25 mm id 0.25 m HP-5MS columns (part no. 19091S-433). The Agilent micro-ECD instrument used 10 m 0.1 mm id 0.1 m HP-5 column (part no. 19091J-141). RTL measurements were made with a solution of dichlorvos, methyl chlorpyrifos, and mirex, each at 10-ppm concentration in acetone. All injections were 1- L splitless, except for the micro-ECD experiments, which were 1- L split 100:1. In all methods, inlets were operated at 250 C and detectors at 300 C. Method translation requires inlets to be run in constant pressure mode to obtain precise scaling of retention times. Thus, all methods discussed in the note were run in this mode.
Figure 1 shows the method translation software. The original method conditions for the GC-AED pesticide method are entered in the column labeled Original Method. The column dimensions, carrier gas type, inlet pressure, outlet pressure, ambient pressure, and oven temperature program are entered here. Note that the inlet pressure is in psi (gauge), while the outlet pressure and ambient pressure are psi (absolute). The original method here is being used on a GC-AED system, so the outlet pressure is entered as atmospheric pressure plus 1.5 psi, the operating pressure of the GC-AED. The Criterion parameter is set to None, which allows the user to select a specific value of speed gain by adjusting the value of hold-up time for the translated method (see figure 1). In the column labeled Translated Method, the parameters of column dimensions, carrier gas type, outlet pressure, and ambient pressure for the GC-MS method are entered. Note that the inlet pressure and oven program are not entered; they are calculated by the program. To set the speed gain to a desired value, take the calculated value of hold-up time in the first column (0.996060 minute) and divide it by the scale factor. Because in this case the desired scale factor (speed gain) is 1, the same hold-up time for both the GC-AED and the GC-MS methods is required. Clicking the radio button next to the hold-up time in the Translated Method column will do this automatically. The method translation indicates that to obtain the same retention times on the GC-MS system as on the GC-AED, use all the same method parameters
except inlet pressure. Instead of using 27.6 psi as is used on the GC-AED, method translation calculates that 17.93 psi on the GC-MS system will result in matching retention times. As mentioned above, this inlet pressure is calculated on the assumed dimensions of the column in the GC-MS system. To get the retention times to match precisely, RTL3 is used. To retention time lock the GC-MS method to the GC-AED method in this example, it is necessary to construct an RTL calibration file for the GC-MS system. Construction of this file only needs to be done once. All subsequent users of the GC-MS method will then be able to use this calibration file for a similarly configured GC-MS instrument. The RTL calibration file is constructed by running five calibration runs of the target compound, in this case methyl chlorpyrifos, at five different inlet pressures. The runs are made at conditions identical to the nominal method except that four of the runs are made at different pressures. The pressures used are typically: Target pressure 20% Target pressure 10% Target pressure (nominal method pressure) Target pressure + 10% Target pressure + 20%
Figure 1. Method translation software showing scaling HP RTL Pesticide Method from GCAED conditions to GC-MS with a scale factor of 1. the nominal method pressure, and the retention time is observed. The pressure and resulting retention time are then entered into the (Re)Lock New Column menu item of the RTL software to calculate the correct pressure for obtaining locked retention times. Normally, the RTL calibration for a new method is determined by actually making the five calibration runs. In the current example, methyl chlorpyrifos would be run at: 17.93 psi 20% = 14.34 psi 17.93 psi 10% = 16.14 psi 17.93 psi (nominal method pressure) 17.93 psi + 10% = 19.72 psi 17.93 psi + 20% = 21.56 psi calibration data, method translation can be used to calculate the new RTL calibration points. This is useful when you want to try a scaled method rapidly and save the time required in making the five runs. (Note: For methods that will be used extensively, the five-runs approach may provide a somewhat better calibration. It is recommended that for these methods, the standard calibration be performed.) To calculate the five RTL calibration pairs of pressure and retention time for the GC-MS method from those of the GC-AED method: Take the inlet pressure used for each original GC-AED RTL calibration run, and enter it into the method translation software for the inlet pressure of the original method. Make sure the hold-up times are locked, giving a speed gain of 1.
The retention time of the target compound is determined for each run. The resulting set of five pressures and corresponding retention times is then entered in the RTL calibration dialog box for the method and saved with the method. To lock the method on the GC-MS setup, the target compound is run at
However, because the new GC-MS method is scaled from an existing GC-AED method that already has RTL
The inlet pressure calculated in the Translated Method column will now change to a new value, corresponding to the pressure that would be obtained if the calibration run were made on a GC-MS system. This pressure is used with the retention time obtained for the corresponding GC-AED calibration run as a calibration point for the GC-MS method.
When all five points have been calculated in this way, they are entered into the RTL calibration dialog box for the GC-MS method and saved with the method. Table 1 lists the original RTL calibration pressures and times with the calculated pressures and times for the GC-MS method. To test the accuracy of using a predicted RTL calibration file for GC-MS, a real calibration set was measured on the GC-MS system. The data is shown in the first two columns of table 2. (Note: The calibration points are spaced ~ 5% apart in pressure instead of the typical 10%.) A GC-MS RTL calibration file was constructed with these measured points. For each point, the locking pressure required to lock the method was calculated and is shown in column 3 of table 2. The locking pressure is the pressure determined by the RTL software that would make methyl chlorpyifos have a retention time of 16.596 minutes. This is determined by entering the pressure and retention time for each point into the (Re)Lock New Column menu item of the RTL software. If the calibration is done correctly, the locking pressures determined from each point should be very similar, as they are in column 3 of table 2.
Column 4 of table 2 shows the locking pressures for the same set of runs but determined using the GC-MS RTL calibration points calculated using method translation. The calculated data provide locking pressures that agree well with those based on measured data. The range in locking pressures pressure is only from 17.72 to 17.75 psi. This range of 0.03 psi corresponds to only about a 0.006-minute range in the retention time of methyl chlorpyrifos. Figure 2 shows the locked chromatograms from a threecomponent mixture run on GC-AED and GC-MS systems. As can be seen, the retention times are well matched between the two methods.
The RTL Pesticide Library contains the retention times of the 567 pesticides measured with GC-FID. The values measured with the FID would be the same observed with any detector that is operated at or near atmospheric pressure. Because retention time matching is critical in this application, the retention times for all the compounds in the table were also measured on the GC-MS system after scaling as described here. Figure 3 is a plot of the difference between the retention times measured on the GC-FID and the GC-MS systems. The plot shows the retention times match well within 0.1 minute out to 30 minutes. A few compounds at the end deviate outside this window, with one compound 0.2-minute different. The
Table 1.
RTL Calibration Points from Original GC-AED Method and Calculated Points for GC-MS
GC-MS RTL Calibration Calculated Pressure (psi) 24.27 21.18 17.934 14.654 11.449 Calculated Ret Time (min) 15.346 15.919 16.578 17.338 18.242
GC-AED RTL Calibration Pressure (psi) 33.1 30.4 27.6 24.8 22.1 Ret Time (min) 15.346 15.919 16.578 17.338 18.242
Table 2.
Comparison of Locking Pressures Calculated Using Measured and Predicted GC-MS RTL Calibration Data
Locking Pressures Using Measured RTL Cal Points Pressure (psi) 17.73 17.72 17.72 17.74 17.72 Using Calculated RTL Cal Points Pressure (psi) 17.75 17.73 17.72 17.74 17.74
GC-MS Locking Runs Measured GC-MS RTL Cal Points Pressure ( psi) 20 19 18 17 16 Ret Time (min) 16.127 16.326 16.536 16.760 16.988
deviation is clearly largest in the isothermal hold region, which starts at 31.87 minutes. This effect is seen with GC-MS, but not with scaling to other atmospheric pressure detectors. While the cause is not yet clearly understood, it appears related to the vacuum outlet pressure of the GC-MS column. Although this level of matching is very good, the table includes both the GC-FID and GC-MS retention times so that smaller time windows can be used in searching unknowns.
Gaining Speed in the Same Instrument Setup

In the analysis of pesticide residues in food, there are usually only a few compounds encountered in any one sample. Because the screening method uses selective detectors, it makes sense to consider trading speed for chromatographic resolution. Selective detectors respond to only those compounds containing a specific heteroatom(s), and the chromatography only needs to resolve those compounds from each other, not from every other compound in
1
the matrix. This approach can save a significant amount of analysis time. In this example of scaling the RTL Pesticide Library, the method will be increased in speed at the expense of chromatographic resolution. The first consideration is by what factor to increase the speed. The method translation software is useful for determining this. A candidate speed gain, in this example threefold, is entered into the method translation software. The resulting inlet pressure and oven temperature ramp rates are then inspected to see if the instrument on
Locking GC-AED with Other GC Detectors

When the method translation step is done to scale the GC-AED method to other atmospheric pressure detectors, the only different parameter to enter is the outlet pressure. The outlet pressure for the GC-AED method is 16.2 psi and that for the others is 14.696 psi. The method translation calculates that the nominal GC-AED inlet pressure of 27.6 psi would be changed to 26.29 psi for the other atmospheric detectors. This difference (<5%) is so small that it can be neglected, because corrections in this range are compensated easily by the retention time locking step. Thus, the method conditions and RTL calibration points used with GC-AED are interchangeable with FID, NPD, ECD, FPD, and other atmospheric detector methods. Note that this would not always be the case. If for example, a method is being scaled that uses a very low inlet pressure, the 1.5-psi difference in outlet pressure could become significant. It is best to check the method with method translation and see if the inlet pressure will change by >10%. If it does, it would be advisable to collect (or translate) a new RTL calibration centered around the translated nominal inlet pressure.
GC-MS
2
10
15
20
25
30
35
GC-AED
10
15
20
25
30
35
40 min
Figure 2. GC-AED chlorine and GC-MS TIC chromatograms of three-component locking mixture. Peak identifications: 1. dichlorvos, 2. methyl chlorpyrifos, 3. mirex.
0.300 0.200 Difference (mm) 0.100 0.000 0 -0.100 -0.200 -0.300 Retention Time (min) 5 10 15 20 25 30 35 40 45
Figure 3. Difference plot of GC-MS and GC-FID retention times in RTL Pesticide Library.
which the new method will be run is compatible with those parameters. Figure 4 shows the method translation software with the data entered for a speed gain of 3. Note that columns for Original Method and Translated Method are set up as in the previous example with two exceptions. Because the scaling is from GC-AED to GC-AED, the outlet pressure in both columns is entered as 16.2 psi. The second and most significant difference is the holdup time. The desired speed gain is 3. To set the speed gain, the calculated value of hold-up time in the first column (0.996060 minute) is divided by exactly 3. This value (0.33202 minute) is entered for the hold-up time in the second column. This will force the speed gain to exactly 3. The inlet pressure and oven temperature ramp for the new threefold speed method are now calculated. The calculated inlet pressure is 87.862 psi, which is compatible with the EPC module on the current system (maximum 100 psi). Note that the helium source supplying the GC must be capable of reaching 100 psi of helium. An optional 150-psi EPC module is available for the HP 6890 GC to provide additional inlet pressure, if necessary. The oven temperature program calculated for the new method has the first ramp listed as 75 C/min. This ramp rate is compatible with the 240-V oven option on the current instrument but would not work with a 120-V oven, which is limited to about 50 C/min in this temperature range. With a 120-V oven, the speed gain would be limited to about 2. The next step is to calculate the RTL calibration points from the original
GC-AED method. This is done by the same process as shown in the GC-MS scaling above. In this case, when one of the original method RTL calibration pressures is entered, the resulting holdup time must be divided by 3 and entered for the holdup time in the Translated Method column. This will force the speed gain back to 3. The resulting inlet pressure is then paired with the retention time of the corresponding original GC-AED calibration run, but divided by 3 as a calibration point for the new method. Table 3 shows the RTL calibration points from the original GC-AED method and calculated points for the threefold speed gain (3 ) method. When the calibration data is entered into the RTL calibration dialog box, the target time for methyl chlorpyrifos is entered as 5.532 minutes, which is 16.596 minutes divided by 3.
Table 4 compares the locking pressures determined with measured and with calculated RTL calibration points. As in the above GC-MS example, the range of the locking pressures from the calculated data is only 0.11 psi (87.88 to 87.99), which corresponds to ~ 0.003 minute. Figure 5 compares the chromatograms of the RTL locking mixture from both the original and the 3 scaled methods. Note that while the chromatographic resolution is reduced, the speed is increased by a factor of 3. Figure 6 shows a plot of the difference between the RTL Pesticide Library retention times, divided by 3, and those of the 3 method. The data were taken with a 36-component subset of the library. The plot shows the retention times match well within 0.05 minute for all compounds, even
Figure 4. Method translation software showing scaling RTL Pesticide method scaled to threefold faster method.
those in the 3.3-minute hold time at the end of the run.
Table 3.
RTL Calibration Points from Original GC-AED Method and Calculated Points for Threefold Speed Gain (3 ) Method
3x GC-AED RTL Calibration Calculated Pressure (psi) 106.21 97.23 87.86 78.44 69.31 Calculated Ret Time (min) 5.115 5.306 5.526 5.779 6.081
GC-AED RTL Calibration
Gaining Speed with a Small-Bore Column

In the previous example, speed was gained at the expense of resolution. In this example, speed will be gained while maintaining most of the resolution but sacrificing capacity. This is done by scaling the original method to a 0.1-mm id column. In scaling to columns of a different diameter, there are two important considerations that must be obeyed to obtain precise matching to a library or reference method. The first is that the stationary phase composition must be the same as that used in the original method. The second is that the phase ratio of the column being scaled to must be the same as that of the reference method. Columns of the same phase ratio have the same ratio of inner diameter to film thickness. Because the reference method was developed on a column with 0.25 mm id 0.25 m film thickness, scaling to a 0.1-mm id column will require a 0.1- m film thickness. A 10-m column of these dimensions was chosen for this example. The micro-ECD for the 6890 GC is extremely sensitive, with detection limits in the low femtogram range for polyhalogenated pesticides. These detection limits are so low that it is reasonable to consider using split mode for a rapid screening method. Using split mode with a split ratio of 100 still gives a detection limits in the range of a few picograms. The split is also more compatible with the relatively low capacity of the column.
Pressure (psi) 33.1 30.4 27.6 24.8 22.1
Ret Time (min) 15.346 15.919 16.578 17.338 18.242
Table 4.
Comparison of Locking Pressures Calculated Using Measured and Predicted 3 GC-AED RTL Calibration Data
Locking Pressures Using Measured RTL Cal Points Pressure (psi) 87.99 87.94 87.99 87.99 87.97 Using Calculated RTL Cal Points Pressure (psi) 87.99 87.95 87.99 87.96 87.88
3x GC-AED Locking Runs Measured 3x GC-AED RTL Cal Points Pressure ( psi) 97 92 87 82 77 Ret Time (min) 5.319 5.433 5.557 5.689 5.832
3 1
GC-AED (1x)
2
10
15
20
25
30
35
40 min
GC-AED (3x)
10
12
min
Figure 5. Chlorine chromatograms from original and 3x GC-AED methods of three-component locking mixture. Peak identifications: 1. dichlorvos, 2. methyl chlorpyrifos, 3. mirex.
Difference (min)
Figure 7 shows the method translation from the GC-AED method to the 0.1-mm id column with a scale factor of 3. A speed gain of 3 was again chosen based on oven and inlet limitations as described above. The same scaling process as used above is followed. The RTL calibration points for the new 3 0.1-mm micro-ECD method were both calculated with method translation and measured. Table 5 shows the calculated values. When the locking pressures from the measured and calculated values were examined, the calculated values provided much poorer predictions of locking pressure than expected. The pressure required to actually lock the column was confirmed to be 65.95 psi, as predicted by the measured RTL calibration data. Method translation had predicted the inlet pressure would be 58.514 psi for an assumed 10-m column length. Because the actual locking pressure was noticeably higher, this suggests that the actual column length was longer and/or the column diameter was smaller and/or the film thickness larger than the assumed values. As an experiment, it was assumed that the problem was in the assumed length of the column used in calculating the RTL calibration points. The column length entry for the 0.1-mm column was iteratively adjusted until the calculated inlet pressure matched the actual locking pressure, 65.95 psi. This resulted in a calculated column length of 10.5622 m. A new set of calculated RTL calibration points were calculated using 10.5622 m as the length of the 0.1-mm column. The results are shown in table 6.
0.2 0.15 0.1 0.05 0 0 -0.05 -0.1 -0.15 -0.2 Retention Time (min) 2 4 6 8 10 12 14
Figure 6. Difference plot of RTL Pesticide Library (GC-FID) retention times divided by 3 minus 3 GC-AED retention times for 36-compound subset of the library.
Figure 7. Method translation software showing scaling RTL pesticide method scaled to a threefold faster method on a 10-m 0.1-mm id column.
Table 7 shows a comparison of locking pressures calculated using measured and predicted 3 0.1-mm id micro-ECD calibration data. The range of locking pressures from the measured data (66.03 to 65.93) only corresponds to a spread in retention times of about 0.004 minute. However, with the data calculated based on a 10-m assumed length, the spread (66.38 to 63.18) is much larger and would correspond to a time range of 0.14 minute. The locking pressures calculated using the 10.5622 value are much more consistent with the measured values. The range in retention times would be ~ 0.03 minute if all the calculated points are used, and if the first value in column 5 is ignored, the range drops to ~ 0.005 minute. The fact that the agreement in locking pressures is much improved by using 10.56 m instead of 10 m suggests that length is probably the largest contributor to the discrepancy. These results should reinforce the recommendation that if a method is to be used extensively, it is prudent to obtain measured RTL calibration data. It should be noted, however, that even with the RTL calibration from the 10-m assumed length, the worst consequence would be that the RT locking step would need to be repeated an extra time to get a more precise match. Figure 8 compares the chromatograms of the RTL locking mixture from both the original and the 3 0.1-mm id micro-ECD methods.
Table 5.
RTL Calibration Points from Original GC-AED Method and Calculated Points for 3 0.1-mm id Micro-ECD Method Assuming 10-m Column Length
3x Micro-ECD RTL Calibration Calculated Pressure (psi) 71.03 64.90 58.51 52.11 45.91 Calculated Ret Time (min) 5.115 5.306 5.526 5.779 6.081
Table 6.
RTL Calibration Points from Original GC-AED Method and Calculated Points for 3 0.1-mm id Micro-ECD Method Assuming 10.5622-m Column Length
3x Micro-ECD RTL Calibration Calculated Pressure (psi) 80.03 73.13 65.95 58.74 51.75 Calculated Ret Time (min) 5.115 5.306 5.526 5.779 6.081
Table 7.
Comparison of Locking Pressures Calculated Using Measured and Predicted 3 0.1-mm id Micro-ECD Calibration Data
Locking Pressures Using Measured Using 10-m Calculated Using 10.56-m Calculated RTL Cal Points RTL Cal Points RTL Cal Points Pressure (psi) 65.95 66.03 65.95 65.93 66.00 Pressure (psi) 66.38 65.77 65.12 64.36 63.18 Pressure (psi) 65.30 65.85 65.96 65.95 65.90
3x Micro-ECD Locking Runs Measured 3x Micro-ECD RTL Cal Points Pressure (psi) 48.81 52.66 58.51 64.36 70.22 Ret Time (min) 6.323 6.041 5.797 5.585 5.396
10
Note that while the most of the chromatographic resolution is preserved, the speed is increased by a factor of 3. After being locked, the three peaks in the 3 micro-ECD method had retention times of 1.924, 5.533, and 9.963 minutes, respectively. These values are very close to the RTL Pesticide Library retention times for the three compounds divided by 3: 1.932, 5.532, and 9.949. The fact that the largest difference between the scaled table and the 3 micro-ECD method is only 0.014 minute again demonstrates the precision of retention time matching achievable with the scaling technique described here.
References
1. P. L. Wylie and B. D. Quimby, A Method Used to Screen for 567 Pesticides and Suspected Endocrine Disrupters, Hewlett-Packard Company, Application Note 228-402, Publication 5967-5860E, April 1998. 2. M. Klee and V. Giarrocco, Predictable Translation of Capillary GC Methods for Fast GC, Hewlett-Packard Company, Application Note 228-373, Publication 5965-7673E, March 1997. 3. V. Giarrocco, B. D. Quimby, and M. S. Klee,Retention Time Locking: Concepts and Applications, Hewlett-Packard Company, Application Note 228-392, Publication 5966-2469E, December 1997. 4. Capillary Column Method Translator, user contributed software, free download from: www.hp.com/go/mts.
Conclusions
Using method translation combined with retention time locking provides a means of extending the usefulness of existing capillary GC methods. The ability to precisely scale a method to meet the needs of different samples and instrument types greatly reduces the effort required to re-use methods, thus saving time and money.
3 1
GC-AED (1x)
2
10
15
20
25
30
35
40 min
GC-micro-ECD (3x)
10
12
min
Figure 8. Chlorine chromatogram from 1 GC-AED method (top) and 3 micro-ECD method (bottom) of three-component locking mixture. Peak identifications: 1. dichlorvos, 2. methyl chlorpyrifos, 3. mirex.
11
Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. Information, descriptions, and specifications in this publication are subject to change without notice. Microsoft and Windows NT are U.S. registered trademarks. Copyright 2000 Agilent Technologies, Inc. Printed in the USA 3/2000 5967-5820E
Retention Time Locking: Concepts and Applications
Application
Gas Chromatography December 1997
Authors
Vince Giarrocco Bruce Quimby Matthew Klee Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Key Words
Retention time locking, method validation, styrene analysis, ASTM D 5135, capillary gas chromatography, laboratory productivity
Introduction
Retention time is the fundamental qualitative measurement of chromatography. Most peak identification is performed by comparing the retention time of the unknown peak with that of a standard. It is much easier to identify peaks and validate methods if there is no variation in the retention time of each analyte. However, shifts in retention time occur frequently. Routine maintenance procedures such as column trimming alter retention times. In a multi-instrument laboratory running duplicate methods, the retention times for each instrument will differ
Abstract
The concepts and applications of retention time locking (RTL) are described. RTL simplifies the process of transferring methods from chromatographic instrument to chromatographic instrument, column to column, and detector to detector. The analysis of impurities in styrene according to ASTM D 5135 is used to demonstrate the efficacy of the approach. Using RTL, the retention times matched within an average of 0.16% (0.020.03 minute) in constant pressure modes.
from each other, even when run under nominally identical conditions. These differences in retention times mean that each instrument must have a separate calibration and integration event table, making it time-consuming to transfer methods from one instrument to another. Differences in retention time also complicate comparison of data between instruments and over time. Retention time locking (RTL) is the ability to very closely match chromatographic retention times in any Agilent 6890 gas chromatograph (GC) system to those in another 6890 GC system with the same nominal column. There are several subtle effects that combine to cause retention time differences between similarly configured GC systems. Columns of the same part number can vary slightly in length, diameter, and film thickness.
GC pneumatics can have small variations in the actual inlet pressure applied at a given setpoint. The actual temperature of the GC oven also has minute but real deviations from the indicated value. The sum of these and other effects result in the observed retention time differences between similarly configured GC systems. The pneumatics and oven temperature control of the 6890 GC have advanced the state of the art in GC hardware accuracy and precision. Agilents advances in fused silica capillary column technology have resulted in highly reproducible column-to-column retention characteristics. With these advances, retention time precision for a given peak in a single GC setup is usually better than 0.01 minute. However, even with these advances in columns and instrument hardware, the sum of the effects mentioned above can cause retention time differences between identically configured GC systems of as much as 0.4 minute. It would be impractical to control all of the instrument and column variables to a degree where retention time differences between similarly configured GC systems are removed. There is, however, a means of greatly reducing these differences. By making an adjustment in the inlet pressure, the retention times on a given GC setup can be closely matched to those of a similarly configured GC system. RTL is based on this principle. The process of RTL is to determine what adjustment in inlet pressure is necessary to achieve the desired match in retention times. Agilent RTL software (G2080AA), which integrates into the Agilent GC ChemStation (version A.05.02 or later), provides the tool required to determine the correct inlet pressure quickly and simply.
There are several advantages gained by using RTL in the laboratory. Peak identification becomes easier and more reliable. It is easier to compare data both between instruments and over time. Comparison of data when using different detectors for analyte identification is simplified. Transferring methods from instrument to instrument or laboratory to laboratory is easier because calibration time windows normally will not require readjustment. Validation of system performance is easier. With locked GC methods, the development and use of retention time data bases for unknown identification is much more straightforward. To maintain a locked method, RTL should be performed whenever: The column is changed or trimmed The method is installed on a new instrument A detector of different outlet pressure is used System performance is validated Troubleshooting chromatographic problems
Systems where the predicted locking pressure falls outside the range of the current calibration
A specific solute (usually one found in the normal method calibration standard) must be chosen and then used for both developing the locking calibration and locking all future systems. The solute, or target peak, should be easily identifiable, symmetrical, and should elute in the most critical part of the chromatogram. Solutes that are very polar or subject to degradation should be avoided. Once the target solute has been chosen and all other chromatographic parameters of the method have been determined, five calibration runs are performed. The runs are made at conditions identical to the nominal method except that four of the runs are made at different pressures. The pressures used are typically: Target pressure 20% Target pressure 10% Target pressure (nominal method pressure) Target pressure + 10% Target pressure + 20%
To lock a given method for the firsttime or for the reasons below, one must first develop a retention time versus pressure (RT vs. P) calibration. Even when using columns with the same part number (same id, stationary phase type, phase ratio, and same nominal length), separate/different locking calibration curves are needed when using: Systems with different column outlet pressures (FID/atmospheric, MSD/vacuum, AED/ elevated) Columns differing from the nominal length by more than 15% (e.g., due to trimming)
The retention time of the target compound is determined for each run. The resulting five pairs of inlet pressures and corresponding retention times are entered into the ChemStation software to generate an RTL calibration file. Figure 1 shows the dialog box used to enter the calibration data. After the data is entered, a plot is displayed, as shown in figure 2. The maximum departure of the fitted curve from the data is given for both time and pressure. If the fit is acceptable, the retention time versus pressure calibration is stored and becomes part of the GC
method. This calibration need only be generated once. Subsequent users of the method can use this calibration when running the method on a similar instrument setup, regardless of location. To relock a system or lock a new one: 1. Set up the method conditions and run a standard containing the target compound. 2. Enter the actual retention time of the target compound into the (Re)Lock current method dialog box (see figure 3). 3. Update the 6890 method with the new calculated pressure, and save the method. 4. Validate the retention time lock by injecting the standard at the new pressure, and compare the retention time obtained to the desired retention time. 5. Repeat steps 2 to 4, if necessary. Figure 1. Dialog box used for entering retention time locking calibration data
A Note on Constant Flow versus Constant Pressure Modes of EPC Operation

Many GC chromatographers prefer to use the constant flow mode of EPC operation. In this mode, inlet pressure increases automatically to maintain constant outlet flow rate as the oven temperature increases during the run. Constant flow mode reduces run time and ensures that flow-sensitive detectors see a constant column effluent flow. The constant pressure mode of EPC operation is also popular. In this mode, the pressure remains constant during the run (outlet flow will decrease as temperature increases). For those wishing to reduce run time in constant pressure mode, a higher pressure can be chosen. For
Figure 2. Plot of calibration data as displayed by RTL software
Figure 3. Dialog box used to calculate locking pressure and update the 6890 method
flow-sensitive detectors, one can set constant column flow + makeup via the 6890 keyboard or ChemStation. In this mode, the makeup flow is increased as the column flow decreases to keep the sum of the two constant. The underlying theory of RTL predicts that constant pressure mode of EPC provides the closest matching of retention times. If one desires to compare data from systems with very different configurations, such as GC/FID to GC/MSD, it is best to use constant pressure mode. As can be seen from the styrene analysis data herein, retention time matching between systems of the same configuration (GC/FID, in this case) is still quite good in the constant flow mode. This application note shows the use of RTL to lock retention times between multiple chromatographic instruments, columns, and detector types and demonstrates RTL in both constant flow and constant pressure modes.
Temperature program: 80 C (9 min), 5 C/min to 150 C
The inlet pressures/flows used are indicated with each chromatogram. A third 6890 Series GC was also used. This system was equipped with an Agilent 5973 mass selective detector (MSD) and was used for peak identification. The GC-MSD chromatographic parameters used were the same as the GC systems noted above except for the inlet pressures as indicated.
The sample was then run at four other pressures to collect the five data pairs for RTL calibration. Because this method was run in constant flow mode, the pressures entered into the RTL software were the initial pressures. The =-methylstyrene peak (peak 10) was chosen as the target compound. The calibration data are shown in figure 1. The method conditions and RTL calibration were then moved to GC system 2, a different GC and column. The sample was run at the original method inlet pressure of 18.2 psi. The chromatogram obtained using this scouting run is overlaid on the original chromatogram in figure 5. The retention times shifted about 0.3 minute on the second GC. This is a typical result obtained when trying to replicate an analysis on a second instrument or with a second column. The retention time of =-methylstyrene was entered into the RTL software

GC-FID to GC-FID Locking
Figure 4 shows the original chromatogram (GC system 1) obtained from running a styrene sample under the conditions specified in ASTM D 5135.1 Many of the typical impurities found in styrene are found here. The phenylacetylene peak represents about 60 ppm. The peaks are identified in table 1.
pA 28 26 24 2
Experimental
Two 6890 Series GC systems were used. Each system was equipped with: Electronic pneumatics control (EPC) Split/splitless inlet (250 C, He carrier gas, split 80:1) Automatic liquid sampler
10
6 22 20 18 16 5 11 7 14 12 12 1 8 13
GC ChemStation (version A.05.02) Flame ionization detector (FID) 60 m 0.32 mm, 0.5 mm HP-INNOWax column (part no. 19091N-216)
2.5
7.5
10
12.5
15
17.5
20
22.5
min
Figure 4. Styrene sample run on GC system 1 at 18.2 psi initial pressure, constant flow mode
dialog box on GC system 2, as shown in figure 3. The RTL software indicated the initial pressure should be modified from 18.2 psi to 18.96 psi. The new initial pressure was entered into the method and saved. Figure 6 compares the chromatograms obtained from the original run and after locking retention times using the =-methylstyrene. Table 2 compares the retention times before and after using this approach. The retention times are now closely matched.
Table 1.
Peak # 1 2 3 4 5 6 7
Peak Identities for Figure 4

Name Nonaromatics Ethylbenzene p-Xylene m-Xylene i-Propylbenzene o-Xylene n-Propylbenzene Peak # 8 9 10 11 12 13 Name p/m-Ethyltoluene Styrene =-Methylstyrene Phenylacetylene >-Methylstyrene Benzaldehyde
pA 27.5 25.6
Ethylbenzene
=-Methylstyrene
10.318 min 22.5 20 10.658 min
17.778 min 18.099 min
GC-FID to GC-MSD Locking

17.5
A second experiment was conducted to lock the original method from GC system 1 to the GC-MSD. This is useful for identification of unknown impurities that show up in the FID chromatogram. For example, there is a shoulder evident on the front side of the phenylacetylene peak in figure 4. It would simplify locating the impurity in the GC-MSD data if the retention times closely matched that of the GC-FID. Because constant pressure mode is preferred when comparing data from FID and MSD systems, constant pressure mode was chosen, and the styrene sample was re-run on GC system 1 at 18.2 psi for reference. The next step was to determine the chromatographic conditions to be used on the GC-MSD. The Agilent method translation software tool was used to calculate the conditions necessary to have the peaks elute in the identical order on the two systems.2,3 Because the retention times need to match, the dead time and temperature program used for running the GC-MSD must be the same as the GC
15 12.5 10
Original (GC system 1, column 1)
Scouting (GC system 2, column 2) 5 7.5 10 12.5 15 17.5 20 22.5 min
Figure 5. Comparison of original chromatogram on GC system 1 with GC system 2 before retention time locking
pA 27.5 25.6 22.5 20 17.5 15 12.5 10
Ethylbenzene 10.318 min vs. 10.298 min
=-Methylstyrene
17.778 min vs. 17.776 min
Original (GC system 1, column 1)
Locked (GC system 2, column 2) 5 7.5 10 12.5 15 17.5 20 22.5 min
Figure 6. Comparison of original chromatogram on GC system 1 with GC System 2 after retention time locking
method. The pressure used, however, will be different due to the difference in column outlet pressure. The GC-MSD inlet pressure is calculated using the none mode of the method translation software (figure 7). In this mode, the holdup time between the two columns was forced to be identical to the GC-FID. This gives a speed gain of 1. The pressure calculated for use on the GC-MSD was 8.44 psi. Note that this calculated pressure is only the nominal pressure required to get similar retention times, not the exact locking pressure. A different RTL calibration is required for GC-MSD because the outlet pressure is vacuum, and that of the FID is atmospheric pressure. Five runs were made on the GC-MSD system bracketing the 8.44 psi nominal method pressure. Because the GC-MSD used in this study was not equipped with RTL software, a dummy method was created in GC system 1 and the GC-MSD RTL calibration data was entered into it. A scouting run of the Styrene sample was made on the GC-MSD, and the =-methylstyrene retention time was used for locking. The locking inlet pressure calculated with the dummy method was 7.9 psi and was entered into the GC-MSD. Figure 8 shows the resulting matched chromatograms from the GC-FID and GC-MSD. As seen in table 3, the retention times are now closely matched within 0.02 minute. Figure 9 shows the MSD first choice of library search result of the impurity that created the shoulder on the front side of the Phenylacetylene peak. RTL ensured that this shoulder remained separated on the MSD system and eluted at the same time
Table 2.
GC-FID Retention Times Before and After Locking for Styrene Impurities (Constant Flow Conditions). Chromatograms Shown in Figures 4, 5, and 6.
Original Run GC 1/Column 1 18.2 psi 10.318 10.616 10.858 11.985 12.533 13..360 17.778 18.806 20.248 24.097 GC2GC1 Before RTL 0.340 0.333 0.337 0.359 0.345 0.364 0.321 0.275 0.310 0.279 0.326 Scouting Run GC 2/Column 2 18.2 psi 10.658 10.949 11.195 12.344 12.878 13.724 18.099 19.081 20.558 24.376 GC2GC1 After RTL 0.020 0.026 0.022 +0.005 0.012 0.016 0.002 0.040 0.006 0.069 0.028 Locking Run GC 2/Column 2 19.0 psi 10.298 10.590 10.836 11.990 12.521 13.376 17.776 18.766 20.242 24.028
Component Ethylbenzene p-Xylene m-Xylene i-Propylbenzene o-Xylene n-Propylbenzene =-Methylstyrene* Phenylacetylene >-Methylstyrene Benzaldehyde Average , * Used in locking calculation
Figure 7. Method translation software provides scaled conditions for GC systems with different configurations
for easy comparison to the FID results.
Conclusions
Retention time locking facilitates replicating results from instrument to
instrument, from column to column, and from detector to detector by locking retention times. The retention times of a styrene sample analyzed according to ASTM D 5135 matched to within 0.06 minute after locking.
References
1. ASTM D 5135-95, Analyses of Styrene by Capillary Gas Chromatography, Annual Book of Standards, Volume 06.04, ASTM, 100 Bar Harbor Drive, West Conshohocken, PA 19428 USA. 2. M. Klee and V. Giarrocco, Predictable Translation of Capillary GC Methods for Fast GC Agilent Technologies, Inc., Application Note 228-373, Publication 5965-7673E, March 1997. 3. GC Pressure/Flow Calculator for Windows, Version 2.0 and Method Translation Tool Version 2.0. Available at http://www. chem.agilent.com/servsup/ usersoft/main.html.
GC-FID
GC-MSD, TIC
1.0
3.0
5.0
7.0
9.0
11.0 13.0
15.0
17.0
19.0
21.0
23.0
25.0 min
Figure 8. Comparison of chromatogram on GC system 1 with GC-MSD system after retention time locking, Constant Pressure Mode Table 3. GC-FID vs. GC-MSD, Method Translated then LockedRetention Times (Constant Pressure Conditions)
GC-FID Original 18.2 psi 10.315 10.620 10.869 12.038 12.613 13.492 18.276 19.406 21.008 25.475 GC-MSD 7.9 psi 10.338 10.642 10.890 12.053 12.630 13.508 18.267 19.389 20.987 25.415 Average
* Used in locking calculation
Component Ethylbenzene p-Xylene m-Xylene i-Propylbenzene o-Xylene n-Propylbenzene a-Methylstyrene* Phenylacetylene b-Methylstyrene Benzaldehyde
RT Difference min 0.023 0.022 0.021 0.015 0.017 0.016 0.009 0.017 0.011 0.060 0.021
Phenylacetylene
1-Ethenyl-3-methyl-benzene
17.6
18.0
18.4
18.8
19.2
19.6
20.0
20.4
20.8
min
Figure 9. GC-MSD identification of impurity in shoulder of phenylacetylene peak
Large Volume Injection for Gas Chromatography Using a PTV Inlet
Application
Gas Chromatography March 1997
Authors
Bill Wilson, Philip L. Wylie, and Matthew S. Klee Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808-1610 USA
Introduction
The demand for lower detection limits is important in environmental, pharmaceutical, food analysis, and other gas chromatography (GC) applications. This demand has driven instrument manufacturers to provide more sensitive instrumentation and procedures, which has prompted regulators to reduce allowable limits, and so on in a never-ending cycle. Improvements in sample handling, sample injection techniques, and detectors have all contributed to the ability to measure compounds at decreasing levels. Concentrating samples is an established approach for increasing method sensitivity. For many environmental analysis, this involves extraction followed by solvent evaporation, which generates large volumes of waste solvent and increases sample preparation time significantly. Recent advances, such as supercritical fluid extration (SFE), solid-phase extraction (SPE), solid-phase microextraction (SPME), and pressurized fluid extraction, are making inroads on liquid/liquid extractions, but these still involve additional sample preparation time.
Abstract
Reduced sample detection limits is a continuing goal in gas chromatography. Large sample injection volume is one possible approach. New inlets and injection techniques supporting large volume injection (LVI) have been developed in recent years. This paper discusses the uses and limitations of LVI, describes LVI with a programmable temperature vaporizer (PTV) inlet, and reviews the results of LVI analysis of pesticides and straight-chain hydrocarbons.
Detection limits can be reduced by lowering system background and interferences. This can be done through sample cleanup, such as florisil column chromatography, SPE, and SPME, and by using selective detectors that do not respond to the background. How-ever, the cleanup requires time, and the need for reduced limits has surpassed the sensitivity of even the best detectors. Large volume injection (LVI) is another approach to lower detection limits. The typical injection volume for capillary column analysis is 0.5 to 2 L. Agilent 6890 Series and 5890 gas chromatographs (GCs) allow approximately two times the normal injection volume (up to 5 L, depending on the solvent) using "pulsed" splitless injection. Injecting still larger volumes with standard techniques can lead to contamination of the system, irreproducible results, and loss of sample. With the new LVI technique, good chromatography can be obtained with injection volumes of 5 to 500 L or more. Table 1 summarizes several common ways to lower detection limits.
Key Words
Large volume injection, LVI, programmed temperature vaporizer, PTV, gas chromatography, GC, pesticide analysis, hydrocarbon analysis
Table 1. Approaches to Lowering Detection Limits

Concentrate the Sample Extraction followed by evaporation of solvent Solid-phase extraction (SPE) Solid-phase micro extraction (SPME) Headspace sampling Purge and trap (P&T) sampling Large volume injection (LVI) Transfer More Sample into the Column Cool on-column (COC) injection Splitless injection Large volume injection (LVI) Use More Sensitive or Selective Detectors Electron capture detector (ECD) Electrolytic conductivity detector (ELCD) Atomic emission detector (AED) Selected-ion monitoring mass spectrometry (SIM-MS) Nitrogen-phosphorus detector (NPD) Flame photometric detector (FPD) Decrease System Noise Selective detectors Sample cleanup to reduce interferences Use headspace or purge and trap sampling Decrease column bleed
In LVI, a large volume of sample is injected. The bulk of the solvent is evaporated before transfer of the sample to the analytical column and the start of the analytical sepa-ration. There are two primary techniques to eliminate solvent: PTV and cool oncolumn injection with solvent vapor exit (COC-SVE). COC-SVE is most appropriate for clean samples with volatile (early-eluting) compo-nents such as extracts of drinking water. The COC-SVE technique is discussed elsewhere.1
The Agilent 6890 Series GC uses a standard automatic sampler with syringe sizes up to 50 L. A 50- L syringe can inject up to 25 L. Multiple injections can be used with the PTV inlet when even larger volumes are required. With the 6890 GC system, delay between injections can be controlled as well as the number of injections or the total injection volume. Injection parameters are set through the Agilent ChemStation. A problem with multiple injections is the increased number of punctures of the GC inlet and vial septa. This reduces septum life and increases the possibility of contamination of sample and inlet. Using a "septumless head" for the inlet can eliminate the inlet problems. Figure 2A shows septum cap extract that contaminated the sample after the vial was pierced 40 times during several multiple-injection experiments. 100% Teflon septa minimize sample contamina-tion such as this, but once punctured, Teflon septa do not reseal. The PTV inlet can be considered a temperature-programmable split/ splitless inlet with the same basic configuration. While it can be used hot for split and splitless applications, this is not recommended because the volume of vaporized solvent may exceed the low internal volume of the PTV inlet. PTV is ideal for cold split or splitless applications,
avoiding most of the problems associated with hot inlets such as sample discrimination, liner overload, and sample decomposition. For large volume injections, the PTV is used in a "solvent vent" or "solvent elimination" mode. Sample is introduced into the inlet with the inlet temperature near the boiling point of the solvent and with a relatively high split ratio. The solvent (and low-boiling solutes) is vented while the higher boiling solutes (more than about 100C above the solvent boiling point) remain and are concentrated in the inlet. After a preset time, the split vent is closed and the inlet temperature increased to transfer the solutes and any residual solvent to the column for separation. Because the sample is evaporated from the inlet, nonvolatile sample components and degradation products remain behind in the inlet, minimizing column contamination. There is evidence that inlet contamination in PTVs influences subsequent injections less than in hot inlets. If contamination becomes an issue, the inlet liner is easily changed. Thus, PTV is a better choice for dirty samples than cool-on-column and split/ splitless inlet.
PTV
LVI with PTV is ideal for trace analysis of later eluting solutes (boiling points approximately 100C higher than the solvent) and for dirty samples. Typical injection volumes for solvent elimination PTV are 25 to 100 L. Injection volumes up to 1 mL have been demonstrated.2 The large sample volumes are injected by manual injection, by multiple sample injections from a standard automatic sampler, or by LVI with a variablespeed injector. An automatic injector is recommended for maximum reproducibility. A standard automatic sampler making repeat injections is more cost effective than purchasing a variable-speed sampler and requires less solvent for syringe cleaning.
For a PTV inlet to work well, inlet temperature must be programmed independently from the column oven. The 6890 provides this function. For example, the inlet can be heated with the split flow off to transfer the sample to the column before the oven temperature program begins. After sample transfer, the inlet can be heated further to bake off contaminants with a high split flow to minimize inlet contamination. LVI by PTV is not a good choice when the target compounds include highly volatile species because these lowboiling compounds are vented along with the solvent. The lowest boiling target compound should boil at least 100C above the solvent to have a reasonable chance of success. Table 2 lists the advantages and disadvantages of LVI by solvent elimination PTV.
Table 2. Advantages and Disadvantages of LVI by Solvent Elimination PTV

Advantages Most flexible LVI technique Good for late-eluting compounds such as pesticides, PAHs, etc Inlet protects column so one can use dirty samples Disadvantages Loss of volatile sample components Possibility of sample decomposition(although less than with split/splitless) More difficult to use than conventional inlets like split/splitless
Table 3. Software, Hardware, and Firmware Versions that Support LVI-PTV

Item ChemStation software G1513A injector G1512A ALS 6890 GC Software/Firmware A.04.02 or higher A.09.10 A.01.08 A.02.01 or higher
Experimental
A 6890 GC with electronic pneumatics control (EPC) was used. A G1916A automatic liquid sampler (ALS) with a G1513A controller performed sample injection. An Agilent ChemStation (version A.04.02) controlled the instruments and acquired and processed data. Table 3 lists the hardware and software revisions that support multiple injections, postdwell time, and the slow plunger mode required for LVI-PTV. Experimental conditions for the GC methods are given with the chromatograms.

Multiple Injections
Multiple injections are a straight-forward and reliable way to introduce large sample volumes into the inlet. Figure 1 shows the linearity obtained from two sets of multiple injections using a standard G1513 ALS. Depending on solvent type and injection volume, liquid sample may run down the liner and enter the column. If this occurs, the column may overload with solvent causing catastrophic peak splitting and possible damage to the stationary phase. To minimize this possibility, a packed liner should be used with multiple injections of more than 5 L. In addition to preventing solvent from flowing into the column, glass wool or other packing provides a surface to retain a film of solvent which, in turn, helps retain early-eluting compounds. Figure 2A shows loss of analytes eluting before C18. Figure 2B, with glass wool packing in the liner, shows complete recovery down to C14.
Area Count 12,000
Packed Liner
10,000 Response Factor RSD C23: 2.1% C22: 2.6% 6,000
8,000
4,000
2,000
0 0
50
100
150
200
250
Total Injection Volume, L
Figure 1. Linearity of multiple injections

pA 2500
Inlet = 40 C at injection Open baffle liner 50 mL/min purge for 2.5 min 30 m x 0.32 mm x 0.25 m HP-5
C18
2000
1500 C16 Septum extract 1000 C14 (vial pierced 40 times)
500 C12 C10 0 0 pA 2.5 5 7.5 10 12.5 Minutes 15 17.5 20 22.5
B
6000
C14 C16 C18
Glass wool packed liner 10 x 25 L injections
5000 4000 3000 2000 C12 1000 C10 0 0 2.5 5 7.5 10 12.5 Minutes
15
17.5
20
22.5
Figure 2. Comparison of packed and open baffle liners using PTV sample: C10-C44 in Hexane
PTV
Figure 3 is a chromatogram of an LVI of pesticides using a PTV inlet. By injecting 25 L divided into five injections, good response is obtained from a 0.01-ppm mixture COC-SVE is usually the preferred technique for samples with low-boiling components, because volatile compounds evaporate with the solvent (as shown in Figure 2A) with PTV. However, for volatile samples that are too dirty for COC-SVE sampling, the PTV inlet can be cryocooled below ambient temperature resulting in much better recovery of the low boilers. Figure 4 compares the recovery of C10 using PTV withthe inlet temperature at 40C and -10C during the injection step. There is 100% recovery of C10 with cryocooling. Much less solvent was eliminated at the lower temperature, which helped retain the early-eluting compounds. It is important to determine carefully the best injection parameters for LVI methods. Figure 5 shows the improvement in peak shape obtained for a sample containing pesticides when the initial PTV temperature, vent flow, and injection delay are optimized. Inlet temperature and vent flow influence the speed and extent of solvent removal. The chromatogram in figure 5B indicates that insufficient solvent was vented, thereby increasing the chance of carry-over,degrading peak shape, andincreasing ghost peaks.
1. Methamidophos 2. Acephate 3. Dimethoate 4. Diazinon
5 X 5 L Injections 5. Chlorothalonil 6. Chlorpyrifos-Methyl 7. Chlorpyrifos 8. Thiabendazole 9. Imazalil 10. Ethion 11. Phosmet 12. Azinphos-methyl
4 5 6 1 2 3
8 9
10
11 12
0 5 6 7 8 Minutes 9 10 11 12
Figure 3. LVI with solvent elimination PTV Pesticides: 0.01 ppm; PTV conditions: vent flow, 300 mL/min;vent pressure: 0 until 1 min; purge flow: 50 mL/min at 3.50 min; gassaver on at 4.70 min; PTV initial temperature: 20 C; PTV initial time:1.1 min; PTV rate: 700 C/min; PTV final temperature: 300 C; injection delay: 0.00 min; column: 30 m x 0.25 mm x 0.25 m HP-5MS
pA 3000 2000 C20 C16 1000 0 Hexane Solvent C14 C10 C12
PTV = 40 C
C23 100 % Recovery
0
pA 3000 2000 1000 0
6 C10
8 C14
10
12
C23 C16
PTV = 10 C
C12 C20
Conclusion
GC using LVI is useful for lowering detection limits. It has broad applicability for applications that require more sensitivity than can be obtained with standard injection volumes. PTV inlets are appropriate for LVI of lateeluting samples, for dirty samples, and for any cold split/splitless injection. Total automation of temperature, pressure, and flow simplifies
6 Minutes
10
12
Figure 4. PTV with cryocooled inlet
PTV use. LVI with solvent elimination PTV requires careful method development for maximum accuracy and reproducibility.
UsingCOC-SVE," Agilent Technologies, Application Note 228-377, Publication Number (23)5965-7923E, March 1997. 2. J. Staniewski and J. Rijks, HRC, 16(1993)182.
References
1. B. Wilson, et al., "Large Volume Injection for Gas Chromatography
Abundance 9e+06
9e+05 Initial PTV Temp = 20 C Vent Flow = 300 mL/min
7e+06
5e+05
5e+06
2e+05 7.60 8.00 8.40 8.70
3e+06
A
500000 0 5 1.2e+07 6e+07 9e+06 Initial PTV Temp = 35 C Vent Flow = 100 mL/min 6 7 8 9 10 11 12
4e+07
5e+06 2e+06
B
2e+07 8.20
8.60
9.00
5000000 0 5 6 7 Minutes 8 9 10 11 12
Figure 5. Effect of PTV setpoints on peak shape sample: pesticides at 1.0 ppm (peaks identified in figure 3); column: 30 m x 0.25 mm x 0.25 mm HP-5MS
Sample Preparation
Application Notes
Analysis of Pesticide Residues in Apple Using Agilent SampliQ QuEChERS AOAC Kits by GC/MS
Application Note
Food Safety
Authors
Limian Zhao, David Schultz, and Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19809-1610 USA
Abstract
This application note describes the use of a quick, easy, cheap, effective, rugged, and safe (QuEChERS) AOAC sample preparation approach for extraction and cleanup of 17 GC-amenable pesticide residues from multiple classes, in apple. The method employed involves initial extraction in a buffered aqueous/acetonitrile system, an extraction/ partitioning step after the addition of salt, and a cleanup step utilizing dispersive solid phase extraction (dispersive SPE). The two different dispersive SPE clean-up approaches used either a 1 mL or 8 mL sample volume and were evaluated in parallel after sample extraction. The target pesticides in the apple extracts were then analyzed by gas chromatography/mass spectrometry (GC/MS) operating in selective ion monitoring (SIM) mode. The method was validated in terms of recovery and reproducibility. The limit of quantitation (LOQ) for most pesticides is 10 ng/g; however, the pesticide Folpet has an LOQ of 50 ng/g in apple. This application employing SampliQ QuEChERS kits produced results well below the maximum residue limits (MRLs) for all the pesticides screened. The spiked levels for the recovery experiments were 10, 50, and 200 ng/g. Recoveries ranged between 70 and 136% (92.5% on average), with RSD below 15% (5.0% on average).
Introduction
The QuEChERS method for pesticide analysis was first introduced by USDA scientists in 2003. [1] The method was modified to address problematic pesticides by including a buffered extraction system [2]. After a full validation for more than 200 pesticides, this improved method was formalized and adopted as AOAC Official Method 2007.01. [3] In summary, the method uses a single-step buffered acetonitrile (1% HAc) extraction while simultaneously salting out water from the sample using anhydrous magnesium sulfate (MgSO4) to induce liquid-liquid partitioning. For cleanup, a dispersive solid phase extraction (dispersive SPE) step is employed using a combination of primary secondary amine (PSA) to remove fatty acids as well as other components, and anhydrous MgSO4 to reduce the remaining water in the extract. After mixing and centrifugation, the upper layer is ready for analysis. Gas Chromatography/Mass Spectrometry (GC/MS) has been widely used in pesticide analysis for many years, because many pesticides are volatile or semi-volatile they are GCamenable. Previously, we evaluated the performance of a SampliQ AOAC buffered extraction kit and SampliQ AOAC dispersive SPE kits for the analysis of polar pesticides in apple using LC/MS/MS for detection and quantification. [4] In this study, the performance of the SampliQ AOAC Buffered Extraction kit (PN 5982-5755) and SampliQ AOAC dispersiveSPE kits for General Fruits and Vegetables (p/n 5982-5022 and 5982-5058) were evaluated for the extraction of volatile and semi-volatile pesticides. Analysis was performed by GC/MS. Seventeen GC-amenable pesticides were selected which represent multiple classes, including non-polar organochlorine pesticides (OCs), certain organophosphorus pesticides (OPs) and organonitrogen pesticides (ONs). The MRLs of these pesticides are a function of both the pesticide class and food matrix and have been set at 10 ng/g or higher. Table 1 shows the chemical and regulatory information for these pesticides in apple.
Honeywell (Muskegon, MI, USA), and acetic acid (HAc) was from Sigma-Aldrich (St Louis, MO, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard (triphenyl phosphate, TPP) were purchased from Sigma-Aldrich (St Louis, MO, USA), Chem Service (West Chester, PA, USA), or Ultra Scientific (North Kingstown, RI, USA).
Solutions and Standards

A 1% acetic acid solution in ACN was prepared by adding 10 mL of acetic acid to 1 L of ACN. Standard and internal standard (IS) stock solutions (2 mg/mL of 11 pesticides) were made in MeOH, respectively, and stored at 20 C. A commercially available mix of 6 pesticides, at 20 g/mL in hexane was used directly. Three QC spiking solutions of 11 pesticides at 1.5, 7.5 and 30 g/mL were made fresh daily in 1:1 ACN/H2O containing 0.1% FA, while the 20 g/mL of 6 pesticides mix was directly used for QC spike. A 2.5 g/mL standard solution of 17 pesticides in ACN containing 0.1% FA was used to prepare the calibration curves in the matrix blank extract by appropriate dilution. A 15 g/mL of TPP spiking solution in 1:1 ACN/H2O containing 0.1% FA was used as the internal spiking standard (IS).
Equipment and Material

Agilent Gas Chromatograph (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent 5975C Series GC/MSD (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent SampliQ QuEChERS AOAC Extraction kits, p/n 5982-5755 (Agilent Technologies Inc., Wilmington, DE, USA). Agilent SampliQ QuEChERS AOAC dispersive SPE kits for General Fruits and Vegetables, p/n 5982-5022 and 59825058 (Agilent Technologies Inc., Wilmington, DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, So Painfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA)
Experimental
All reagents and solvents were HPLC or analytical grade. Acetonitrile (ACN), and methanol (MeOH) were from
Table 1. Name
Pesticides Chemical and Regulatory Information [58] Category Phenol Log P 3.18 pKa 9.4 Structure MRLs in apple (ng/g)* 20
s-Phenylphenol
OH
Carbaryl
Carbamate
2.36
10.4
50
NH O O
Dichlofluanid
Sulphamide
3.7
NA
5000
S N S N S
Dichlorvos Organophosphate 1.9 NA
Cl S F
10
Cl
O Cl Cl
Diazinon Organophosphate 3.69 2.6
P O
O
100
N O
S P O O
10
Chlorothalonil
Chloronitrile
2.94
NA
N Cl Cl Cl Cl
N
500
O
Dichlorobenzo phenone
Organochlorine
4.44
NA
Cl C Cl
(Continued)
Table 1. Name Folpet
Pesticides Chemical and Regulatory Information [58] Class Phthalimide Log P 3.02 pKa NA Structure MRLs in apple (ng/g)*
O N S Cl O Cl
Cl Cl Cl
Cl Cl Cl Cl
Cl O O S O Cl Cl Cl Cl Cl Cl O Cl Cl Cl Cl
3000
Cl
20
Chlordane
Cyclodiene organochlorine
2.78
NA
Cl
50
Endosulfan
Organochlorine
3.13
NA
Dieldrin
Chlorinated hydrocarbon
3.7
NA
10
Cl
50
DDE
Organochlorine
6.55
NA
Cl
Cl
Cl
Ethion Organophosphate 5.07 NA
O S P O S S S P O O
Cl
300
(Continued)
Table 1. Name
Pesticides Chemical and Regulatory Information [58] Class Organochlorine Log P 3.13 pKa NA Structure MRLs in apple (ng/g)* 50
Endosulfan sulfate
Cl Cl Cl Cl
Cl Cl O O S O O Cl Cl
10
Endrin ketone
Organchlorine
4.99
NA
Cl
Cl Cl Cl O
50
Permethrins
Pyrethroid
6.1
NA
Cl Cl O
Coumaphos Organothio phosphate 3.86 NA 100
O Cl
O P O S
*The MRLs numbers list in the table are for apple or lowest level in other fruit and vegetables. They could be higher in different commodities.
Instrument Condition
An Agilent GC/MS method for pesticides analysis was used for this study. [9]
GC conditions Auto-sampler: Inlet: Column: Agilent 7683 automatic liquid sampler Splitless Agilent 30 m 0.25 mm 0.25 m HP-5MS Ultra Inert (p/n 19091S-433UI) Helium in the constant pressure mode Chlorpyrifos-methyl locked to 16.596 min (nominal column head pressure = 22.0 psi) 70 C (2 min), 25 C/min to 150 C (0 min), 3 C /min to 200 C (0 min), 8 C/min to 280 C (11.5 min) 1.0 L Atune.u SIM (refer to Table 2 for settings in detail) 230 C, 150 C and 280 C respectively, 3.00 min Autotune voltage
Sample preparation
Sample comminution Organically grown, pesticide-free apples were purchased from a local grocery store. Approximately three pounds of apples were chopped into small, bean sized cubes. Skin was included, but the seeds were discarded. The chopped apple cubes were then placed into a clean plastic bag and frozen at 20 C overnight. The bag was massaged occasionally to make sure the cubes remained separate. The following day, only the required amount of frozen apple cubes was removed and thoroughly blended. Dry ice was added while comminuting, when possible. Samples were comminuted thoroughly to get the best sample homogeneity, ensuring there were no pieces of apple visible in the final sample. Extraction/Partitioning A 15 g ( 0.1g) amount of previously homogenized sample was placed into a 50 mL centrifuge tube (from the SampliQ QuEChERS extraction kit). QC samples were fortified with 100 L of appropriate QC spiking solution (11 pesticides) and 7.5, 37.5, and 150 L of 20 g/mL stock solution (6 pesticides mixture), respectively, yielding QC samples with concentra-
Carrier gas: Retention time locking: Oven temperature program: Injection volume: MS conditions Tune file: Mode: Source, quad, transfer line temperature: Solvent delay: Multiplier voltage: Table 2. Analyte (1) Dichlorvos (2) -Phenylphenol (3) Diazinon (4) Chlorothalonil (5) Carbaryl (6) Dichlofluanid (7) Dichlorobenzophenone (8) Folpet (9) -Chlordane (10) Endosulfan (11) Dieldrin (12) DDE (13) Ethion (14) Endosulfan sulfate TPP (IS) (15) Endrin ketone (16) Permethrin (17) Coumaphos
Instrument Acquisition Data Used for the Analysis of 17 Pesticides by GC/MS. SIM 184.9 170.1, 169.1 137.1, 179.1 265.9, 263.9 144 123, 167.1 139, 249.9 259.9, 261.9 372.9, 374.9 240.8, 238.8 262.8 245.9, 317.9 230.9 273.8 325.1, 326.1 316.9 183.1 362.0 Collection window (min) 3.0 6.5 6.5 9.5 13.5 14.65 14.65 16.0 16.0 17.5 17.5 18.8 18.8 20.5 21.35 21.8 21.8 22.3 22.3 23.2 23.2 25.0 23.2 25.0 25.0 26.4 26.4 27.2 27.2 28.0 28.0 28.5 30.0 32.5 30.0 32.5 RT (min) 5.8 8.8 14.5 14.8 16.8 18.4 19.2 21.6 22.0 22.6 23.9 24.0 26.0 26.8 27.7 28.2 31.4, 31.6 31.7
tions of 10, 50 and 200 ng/g. A 100 L amount of internal standard spiking solution (15 g/mL of TPP) was added to all samples except the control blank to yield a 100 ng/g concentration in each sample. Tubes were capped and vortexed for 1 min. A 15 mL amount of 1% HAc in ACN was added to each tube using the dispenser. An Agilent SampliQ QuEChERS extraction salt packet from the kit (PN 5982-5755) containing 6 g of anhydrous MgSO4, and 1.5 g of anhydrous NaOAc was added directly to the tubes. The salt bag was massaged carefully to break up any salt clumps before pouring. The tubes were examined to ensure that no powder was left in the threads or rims of the tubes. Sample tubes were sealed tightly and shaken vigorously for 1 min by hand to ensure that the solvent interacted with the entire sample and crystalline agglomerates were dispersed. Sample tubes were centrifuged at 4000 rpm for 5 min. Dispersive SPE Cleanup A 1 mL aliquot of the upper ACN layer was transferred to an Agilent SampliQ QuEChERS dispersive SPE 2 mL tube (p/n 5982-5022). An 8 mL aliquot was transferred to an Agilent SampliQ QuEChERS dispersive SPE 15 mL tube (p/n 5982-5058). The 2 mL tube contained 50 mg of PSA and 150 mg of anhydrous MgSO4; while the 15 mL tube contained 400 mg of PSA and 1200 mg of anhydrous MgSO4. The tubes were tightly capped and vortexed for 1 min. The 2 mL tubes were centrifuged with a micro-centrifuge at 13,000 rpm for 2 min, and 15 mL tubes in a standard centrifuge at 4000 rpm for 5 min. An aliquot from the extract, 500 L was transferred into an autosampler vial, and analyzed by GC/MS. Figure 1 shows the flow chart for the QuEChERS AOAC sample extraction procedure.
Weigh 15 g comminuted sample (0.01 g) in 50 mL centrifuge tube
Add 100 L of IS (TPP) solution, and QC spike solution if necessary, vortex 1 min
Add 15 mL of ACN containing 1% HAc
Add SampliQ AOAC QuEChERS Extraction salt packet
Cap and shake vigorously for 1 min
Centrifuge at 4000 rpm 5 min
Transfer 1 mL of upper ACN layer to SampliQ AOAC dispersive SPE 2 mL tube, or 8 mL to SampliQ AOAC dispersive SPE 15 mL tube
Vortex 1 min, centrifuge at 13000 rpm for 2 min for 2 mL tubes or at 4000 rpm for 5 min for 15 mL tubes
Transfer 500 L extract to autosampler vial
Analyze by GC/MS Figure 1. Flow chart of the Agilent SampliQ QuEChERS AOAC extraction procedure.

Using the SampliQ QuEChERS kits, the entire procedure is faster, easier, offers time and labor savings, while ensuring consistency. An analyst can process 4050 samples in just a few hours. The addition of a food sample with a high content of water directly to the salts creates an exothermic reaction, which can affect analyte recoveries, especially for volatile pesticides. Agilent's SampliQ extraction salts are uniquely prepared in an anhydrous package. The unique SampliQ anhydrous salts packet allows addition after adding organic solvent to the sample, as specified in the original QuEChERs method. In our previous study, the new design of SampliQ QuEChERS AOAC kits demonstrated excellent recovery and precision for a broad variety of semi-polar to polar pesticides using
7
LC/MS/MS. [4] There are many semi-volatile and volatile pesticides, so the use of GC/MS is applicable for the performance evaluation of the AOAC kits for the analysis of these groups of pesticides. The selectivity of GC/MS (SIM mode) is not as effective as that of LC/MS/MS (MRM mode). Furthermore, the final QuEChERS samples still contained food matrix impurities, which can be observed in the GC/MS chromatogram of blank apple extract. Therefore, it is important to carefully choose the monitored ions of each compound when setting up the SIM method. In general, the most abundant ions were selected in order to achieve the best sensitivity; however in a few instances the sensitivity was compromised to obtain better selectivity by using more unique but less abundant ions. As shown in Figure 2a, there are interference peaks apparent in the blank chromatogram; fortunately most pesticides are free of co-eluting interferences. There was an interference eluting at a retention time very close to that of s-phenylphenol, and can not be differentiated for quantitation.
The response of this interferent within the blank was integrated to be less than 20% response of s-phenylphenol peak at the LOQ (10 ng/g) sample. Therefore, the selectivity was considered acceptable for this compound. Figure 2 (a, b) shows the chromatograms of a blank apple extract and 50 ng/g fortified apple extract.
Linearity and Limit of Quantification (LOQ)

The linear calibration range for all of the pesticides was 0400 ng/g; excluding Folpet at 50400 ng/g due to poor sensitivity. Two different dispersive SPE volumes (1 mL and 8 mL) were used for evaluation and comparison; therefore, two calibration curves were generated from matrix blanks prepared
from each size. Each calibration curve was made at levels of 10, 20, 50, 100, 250, and 400 ng/g. The TPP was the internal standard (IS) at 100 ng/g in all cases. The calibration curves were generated by plotting the relative responses of analytes (peak area of analyte/peak area of IS) to the relative concentration of analytes (concentration of analyte/concentration of IS). Table 1 shows that the 10 ng/g quantification limits LOQ (10 ppb) and 50 ng/g LOQ for Folpet (50 ppb) established for pesticides are substantially lower than many MRLs for the pesticides in fruit and vegetables. The regression fit used for the calibration curves was the average response factor. Table 3 shows the linear term and RF relative standard deviation (%) for both 1 mL and 8 mL dispersive SPE.
2000 1800 1600 1400 Abundance 1200 1000 800 600 400 200 0
GC/MS Chromatogram of Apple Extracts, Blank Relative to Fortified Sample, 50 ng/g after Agilents SampliQ QuEChERS extraction and dispersive SPE, for General Fruits and Vegetables
IS
5.00 2000 1800 1600 1400 Abundance 1200 1000 800 600 400 200 0 5.00
10.00
15.00
20.00 Time (min)
25.00
30.00
35.00
40.00
IS 2 12 9 8 10 11
20.00 Time (min)
1
10.00
13
25.00
16 14 15
30.00
17
35.00 40.00
15.00
Figure 2.
GC/MS chromatogram of apple extract. (A) apple extract blank; (B) 50-ng/g fortified apple extract. Peak Identification: 1. Dichlorvos, 2. s-Phenylphenol, 3. Diazinon, 4. Chlorothalonil, 5. Carbaryl, 6. Dichlofluanid, 7. Dichlorobenzophenone, 8. Folpet, 9. -Chlordane, 10. Endosulfan, 11. Dieldrin, 12. DDE, 13. Ethion, 14. Endosulfan sulfate, 15. Endrin ketone, 16, Permethrin, 17. Coumaphos. IS. Triphenyl phosphate (TPP).
Table 3. Analytes Dichlorvos
Linearity of 17 Pesticides in Apple Extract 1 mL dispersive SPE Linear Term RF Rel Std Dev (%) 3.47e-001 1.37e-000 7.04e-001 6.84e-001 8.07e-001 1.04e-000 4.55e-001 3.88e-002 3.23e-001 8.56e-002 2.71e-001 1.43e-000 5.87e-001 2.72e-001 2.75e-001 9.71e-001 2.70e-001 11.4 10.7 10.9 13.7 14.1 12.8 11.4 19.5 10.4 15.2 6.2 8.4 19.7 9.6 10.1 9.4 15.6 8 mL dispersive SPE Linear Term RF Rel Std Dev (%) 3.87e-001 1.50e-000 7.39e-001 8.02e-001 1.01e-000 1.08e-000 4.60e-001 4.52e-002 3.31e-001 8.26e-002 2.59e-001 1.39e-000 5.63e-001 2.74e-001 2.75e-001 9.29e-001 2.72e-001 4.6 11.4 6.5 8.9 10.8 8.6 8.2 20.1 9.2 8.8 5.9 7.5 17.0 9.5 7.8 8.0 15.7
-Phenylphenol Diazinon Chlorothalonil Carbaryl Dichlofluanid Dichlorobenzophenone Folpet -Chlordane Endosulfan Dieldrin DDE Ethion Endosulfan sulfate Endrin ketone Permethrin Coumaphos
Recovery and Reproducibility

The recovery and reproducibility were evaluated by spiking pesticides standards in comminuted apple sample at levels of 10, 50 and 200 ng/g. These QC samples were quantitated against the matrix-spiked calibration curve. The analysis was performed in replicates of six (n=6) at each level. The recovery and reproducibility (shown as % RSD) data for 1 mL and 8 mL volume dispersive SPE are shown in Table 4 and Table 5, respectively. It can be seen from the results that all of the pesticides give excellent recoveries (average of 90.8% for 1 mL and 94.2% for 8 mL) and precision (average of 5.7% RSD for 1 mL and 4.3% RSD for 8 mL). As mentioned above, an
interferent was eluted very closely with s-phenylphenol. The selectivity was still acceptable because the interferent contributed less than 20% of LOQ; however, the contribution of the interference peak resulted in the higher recovery of this compound at low levels. Folpet is a notoriously unstable pesticide, and the main problems dealing with degradation and instability come from the N-trihalomethylthio functional group. [3, 10] Folpet was quantified, but the LOQ was found to be 50 ng/g due to poor sensitivity, however, recovery and reproducibility at 50 ng/g and above were acceptable (average recovery 85.5%, average reproducibility 10%).
Table 4.
Recovery and Repeatability of Pesticides in Fortified Apple With Agilent SampliQ 2 mL Dispersive SPE Tube (p/n 5982-5022); Recovery 90.8%, RSD 5.7% (avg) 10 ng/g fortified QC Recovery RSD (n=6) 86.8 113.4 98.6 86.1 96.1 90.0 97.8 79.6 69.8 90.6 84.0 90.9 79.8 85.2 87.9 87.8 7.0 6.3 2.3 10.0 9.0 7.0 7.6 4.4 9.2 10.9 4.8 1.8 1.9 12.0 2.8 5.1 50 ng/g fortified QC Recovery RSD (n=6) 83.9 96.3 87.3 84.4 93.8 84.6 95.0 74.4 88.9 91.2 86.6 89.4 103.5 80.4 80.7 93.8 89.7 11.6 6.5 2.8 5.3 8.3 2.9 6.2 9.1 4.3 5.3 3.2 3.8 1.4 4.6 3.6 2.0 3.0 200 ng/g fortified QC Recovery RSD (n=6) 81.5 100.5 90.4 93.2 99.1 94.6 102.2 95.7 95.3 96.2 92.8 95.4 116.5 86.8 91.8 94.0 90.0 5.5 3.6 4.9 7.6 8.2 5.0 4.3 11.0 4.4 5.2 4.8 4.5 5.0 5.6 4.5 4.4 6.4
Analytes Dichlorvos s-Phenylphenol Diazinon Chlorothalonil Carbaryl Dichlofluanid Dichlorobenzo phenone Folpet g-Chlordane Endosulfan Dieldrin DDE Ethion Endosulfan sulfate Endrin ketone Permethrin Coumaphos
10
Table 5.
Recovery and Repeatability of Pesticides in Fortified Apple With Agilent SampliQ 15 mL Dispersive SPE Tube (p/n 5982-5058); Recovery 94.2%, RSD 4.3% (avg) 10 ng/g fortified QC Recovery RSD (n=6) 103.4 125.8 96.0 96.5 97.7 91.7 98.8 80.9 80.3 81.2 86.1 106.5 91.6 76.2 97.9 82.3 4.2 8.7 4.5 3.0 3.9 5.1 9.3 3.5 7.3 3.4 1.8 3.6 4.6 3.3 1.6 6.7 50 ng/g fortified QC Recovery RSD (n=6) 85.6 99.2 82.3 82.8 91.4 83.7 96.2 88.4 87.5 84.1 93.1 92.4 122.2 87.7 82.4 104.7 86.5 8.1 4.4 2.1 5.2 4.4 1.0 4.7 4.0 3.3 3.6 2.0 3.4 2.0 4.0 3.9 1.1 2.5 200 ng/g fortified QC Recovery RSD (n=6) 97.2 105.4 88.4 97.7 101.9 93.7 105.3 72.5 94.8 98.6 98.7 98.9 136.3 93.0 91.8 106.6 89.3 7.2 5.0 6.3 4.5 5.0 5.1 4.3 6.0 5.0 3.0 3.9 3.9 4.2 4.1 4.1 4.2 5.1
Analytes Dichlorvos -Phenylphenol Diazinon Chlorothalonil Carbaryl Dichlofluanid Dichlorobenzo phenone Folpet -Chlordane Endosulfan Dieldrin DDE Ethion Endosulfan sulfate Endrin ketone Permethrin Coumaphos
Figure 3 shows the recovery and precision results for 1 mL dispersive SPE and 8 mL dispersive SPE. The two different dispersive SPE clean-ups were performed by transferring 1 mL or 8 mL of ACN extract from the same sample following the extraction step. In order to simplify the comparison, the average recovery and precision of three fortification concentrations were used for all pesticides. The results of each dispersive SPE clean-up appeared to be independent of volume used. Both approaches provided efficient and similar sample clean-up, and thus generated relatively equivalent results.
11
Exceptional Recoveries and Precision for 1 and 8 mL Volumes for Agilent SampliQ Dispersive SPE, 2 and 15 mL kits
120
1 mL
8 mL
100
80 Recovery (%)
60
40
20
0 Dichlorvos Carbaryl Dichlorfluanid Chlorothalonil -Phenylphenol Dichlorobenzophenone Endosulfan sulfate Endrin ketone -Chlordane Coumaphos Endosulfan Permetrin DDE Diazinon Dieldrin Ethion Forpet
Figure 3.
Recoveries and precision for 1 and 8 mL sample volumes employing Agilent SampliQ Dispersive SPE, 2 and 15 mL kits, respectively.
Conclusions
Agilent SampliQ QuEChERS AOAC method for General Fruits and Vegetables: Extraction and Dispersive SPE kits provided a simple, fast and effective method for the purification and enrichment of representative volatile to semi-volatile pesticides in apple. The recovery and reproducibility, based on matrix spiked standards, were acceptable for multiclass, multi-residue pesticide determination in apple. The impurities and matrix effects from apple did not interfere with the quantitation of target compounds. The LOQs of the pesticides were lower than regulated MRLs in apple. As the selected pesticides
represented a broad variety of different classes and properties, the Agilent SampliQ QuEChERS AOAC Extraction and Dispersive SPE kits for General Fruits and Vegetables is an excellent choice for other pesticides in similar food matricies.
Acknowledgements
The authors wish to thank Dr. Meng, Chin K. for his help with the GC/MS instrument, and Dr. Ball, Carol H. for many valuable suggestions.
12
References
1. M. Anastassiades, S. J. Lehotay, Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and "Dispersive Solid-Phase Extraction" for the Determination of Pesticide Residues in Produce, J. AOAC Int., 2003, 86, 412- 431. S. J. Lehotay, et al, Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005, 88, 615-629. S. J. Lehotay, et. al, Determination of Pesticide Residues in Foods by Acetonitrile Extraction and Partitioning with Magnesium Sulfate, Collaborative Study, J. AOAC Int., 2007, 90, 485-520. L. Zhao, D. Schultz, Evaluation of the QuEChERS AOAC Sample Preparation Kit for the Analysis of Pesticide Residues in Apples with LC/MS/MS Detection, Agilent Technologies publication 5990-3937EN. http://sitem.herts.ac.uk/aeru/footprint/en/index.htm http://www.m5.ws001.squarestart.ne.jp/foundation/ search.html http://www.mrldatabase.com/?selectvetdrug=0 http://www.ams.usda.gov/AMSv1.0/getfile?dDoc Name=PDP1995Summary P. L. Wylie, Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library, Agilent Technologies publication 5989-5076EN.
2.
3.
4.
5. 6. 7. 8. 9.
10. B. Mercedes et.al. Residue determination of captan and folpet in vegetable samples by gas chromatography/ negative chemical ionization-mass spectrometry. J. AOAC Int., 2006, 89, 1080-1087.

13
Agilent Technologies, Inc., 2010 Printed in the USA April 20, 2010 5990-4068EN
Screening of Pesticide Residues in Water by Sequential Stir Bar Sorptive Extraction-Thermal Desorption with GC/MS
Application Note
Food
Authors
Nobuo Ochiai and Kikuo Sasamoto GERSTEL K.K. 2-13-18 Nakane Meguro-ku, Tokyo, 152-0031 Japan
Abstract
The performance of sequential stir bar sorptive extraction (sequential SBSE) in combination with thermal desorption (TD)-GC/MS for screening of pesticide residues in water was described. Compared to conventional SBSE, sequential SBSE provides more uniform enrichment over the entire polarity and volatility range for organic pollutants at ultra trace levels in water. Sequential SBSE consists of an SBSE performed first on a 5-mL sample without modifier using one stir bar, then on the same sample after addition of 30% NaCl using a second stir bar. After extraction the two stir bars are placed in a single glass desorption liner and are thermally desorbed simultaneously. The presence of pesticide is elucidated with a retention time locked GC/MS method (RTL-GC/MS). Screening of pesticides at the ng/L level in river water samples was successfully carried out with sequential SBSE-TD-RTL-GC/MS operated in the scan mode.
Introduction
The determination of pesticide residues in environmental samples, such as water, soil, and agricultural products, has been a major subject for many years. This is because of their potential risk for human health, persistence and tendency to bio-accumulate. For water samples, analytical methods usually include extraction and enrichment steps for determining pesticide residues at very low levels (<sub-g/L). Therefore, miniaturized methods such as solid phase micro-extraction (SPME) and stir bar sorptive extraction (SBSE) were introduced because they are simple, solvent-free techniques that allow extraction and concentration in a single step. [1,2]. These sorptive extraction methods have been successfully applied to the determination of organic compounds in various sample matrices, such as water, soil, food, and biological fluid [3-5]. These methods also provide enhanced sensitivity because the extracted fraction (on a fiber or on a stir bar) can be introduced quantitatively into a GC system by thermal desorption. In addition, the enrichment factor for SBSE is higher than that of SPME because of the 50-250 times larger volume of extraction phase on the stir bar. Several authors indicated that the SBSE method allows high recovery and an extremely low limit of detection (LOD) at the sub-ng/L level, particularly for some solutes having hydrophobic characteristics [4, 5]. In 2008, Ochiai et al, developed an SBSE procedure termed sequential SBSE for exhaustive enrichment of organic pollutants in aqueous samples. This procedure is performed sequentially for one aliquot under two extraction conditions using two stir bars [6]. The technique counters the difficulty in recovering both hydrophobic and hydrophilic compounds. This new approach produced a remarkable improvement in recoveries for a set of 80 pesticides (log Ko/w: 1.70-8.35) in water, 82-113% for the majority and less than 80% for only five hydrophilic compounds. This application note describes the performance of sequential SBSE-thermal desorption (TD) in combination with retention time locked (RTL)-GC/MS for the screening of pesticide residues in water.
Sequential SBSE-TD-RTL-GC/MS
For the first SBSE, five milliliters of water sample were transferred to 10 mL headspace vials. A stir bar (Gerstel Twister; 24 L of PDMS) was added and the vial was sealed with a screw cap. SBSE of several samples was performed simultaneously at room temperature (24 C) for 60 min while stirring at 1500 rpm. After the first extraction, the stir bar was removed with forceps, dipped briefly in Milli-Q water, dried with a lint-free tissue, and placed in a glass thermal desorption liner. The glass liner was temporarily placed and stored in a sealed sample tray of the MPS 2. For the second extraction, 30 % NaCl was dissolved in the sample. Then, a second stir bar was added and the vial was capped again. The second extraction was performed under the same conditions as the first extraction. After the second extraction, the stir bar was removed with forceps, dipped briefly in Milli-Q water, dried with a lint-free tissue, and placed in the glass liner which contained the first SBSE stir bar. Finally, the glass liner was placed in the thermal desorption unit. No further sample preparation was necessary. Figure 1 shows the experimental setup of sequential SBSE. Reconditioning of stir bars was done after use by soaking in Milli-Q purified water and a mixture of methylene chloridemethanol (1:1) for 24 h each. Stir bars were then removed from the solvent and dried on a clean surface at room temperature for 1 h. Finally, the stir bars were thermally conditioned for 30 min at 300 C in a flow of helium. Typically, 30 extractions could be performed with the same stir bar.
1 st Extraction
2 nd Extraction
30% NaCl
Experimental
Instrumentation
Analyses were performed with a GERSTEL TDU thermal-desorption unit equipped with a MPS 2 auto-sampler (Gerstel) and a CIS 4 programmed temperature vaporization (PTV) inlet (Gerstel) installed on an Agilent 7890 GC with an Agilent 5975 Series GC/MSD triple-axis detector (TAD).
SBSE 1h@1500 rpm
SBSE 1h@1500 rpm
Solutes: log KO/W > 4
Solutes: log KO/W < 4
Figure 1.
Experimental setup of sequential SBSE
The two stir bars were thermally desorbed by the thermal desorption unit (TDU) with 50 mL/min desorption flow. Desorbed compounds were cryo-focused on quartz wool packed liner in the PTV inlet for subsequent GC/MS analysis. The analytical conditions are summarized in Table 1.
Table 1 Stir bar Sequential SBSE Analytical conditions GERSTEL Twister; 24 L PDMS 5 mL sample volume 60 min extraction for each SBSE Non-modifier for 1st SBSE 30 % NaCl for 2nd SBSE 1500 rpm stirring speed Splitless 50 mL/min desorption flow 40 C (0.2 min); 720 C/min; 280 C (5 min) Quartz wool packed liner Splitless 100 C (0.5 min); 12 C/s; 280 C (hold) 30 m 0.25 mm id 0.25 m Agilent HP-5ms Chlorpyrifos methyl locked to 15.59 min. 70 C (2 min); 25 C/min; 150 C; 3 C/min; 200 C; 8 C/min; 300 C (2 min) Scan mode Scan: m/z 58-510; 3.2 scans/s

Figure 2 shows a comparison of the recoveries of SBSE without modifier, SBSE with 30% NaCl, and sequential SBSE for representative pesticides with various log Ko/w values in natural water spiked at 500 ng/L. The recoveries for solutes with log Ko/w of less than 4.0 dramatically increased with salt addition, for example, for pirimicarb (carbamate; log Ko/w: 1.70), fenobucarb (carbamate; log Ko/w: 2.79), and pacrobutrazol (other; log Ko/w: 3.36), the recovery increased from 15% to 74%, 41% to 90%, and 31% to 95%, respectively. However, recovery for solutes with log Ko/w of more than 4.0 drastically decreased, for example, for terbufos (organophosphorus; log Ko/w: 4.24), pyridaben (other; log Ko/w: 5.47) and permethrin 1.2 (pyrethroid; log Ko/w: 7.43), the recovery decreased from 89% to 68%, 100% to 57%, and 101% to 54%, respectively. In contrast with conventional SBSE with or without salt addition, the sequential approach could eliminate the negative effect of the salt for solutes with log Ko/w of more than 4.0, while maintaining increased recovery for hydrophilic solutes with salt addition, resulting in high recovery. Therefore, using sequential SBSE, results in higher sensitivity for a wide range of solutes with different polarities. Figure 3 shows the mass chromatogram (m/z 304) of diazinon (log Ko/w: 3.86) in 5 mL of natural water spiked at 20 ng/L. Excellent sensitivity (peak-to-peak S/N of 40) and a well defined mass spectrum were obtained for ng/L level samples even with the scan mode.
TDU
PTV
Column RTL GC Oven MSD
120 100 Recovery (%) 80 60 40 20 0
SBSE SBSE with 30 % NaCl Seq. SBSE 900 800 700 600 500 400 300 200 100 0 12.00 13.00 14.00 15.00 Retention time (min) 16.00 17.00 Abundance
Figure 2.
Comparison of the recovery of SBSE without modifier, SBSE with 30 % NaCl, sequential SBSE for representative pesticides with various log Ko/w in natural water spiked at 500 ng/L.
Figure 3.
Mass chromatogram (m/z 304) of diazinon (log Ko/w: 3.86) in 5-mL natural water spiked at 20 ng/L.
Figure 4 shows a Wiley library search result, which obtained a match factor of 94. RTL GC/MS method can eliminate many false positives and give more confidence in compound identification with not only mass spectral information but also locked retention times. Recently, Agilent updated the RTL pesticide library with 926 pesticides, endocrine disruptors, and related compounds [7]. For the present paper, the presence of pesticides in river water sample is elucidated automatically via the RTL screener in combination with the RTL library for 926 compounds.
(a)
(b)
Figure 4.
Wiley library search result of diazinon (log Ko/w: 3.86) in 5-mL natural water spiked at 20 ng/L. (a) Measured mass spectrum (b) Wiley library spectrum.
Figures 5 and 6 show the screener software windows for the positive detection and identification of symetryn (log Ko/w: 2.90) and pyributicarb (log Ko/w: 5.34) in Ayase river water sample. The ratio of four qualifier ions are measured and compared with those listed in the database. The software window also shows the deviation of measured retention time with the
RTL value. After identification of these pesticides by the RTL screener, quantification was performed in six replicate analyses with standard addition calibration method. The determined concentrations were 240 ng/L (RSD: 2.9 %, n = 6) for symetryn and 25 ng/L (RSD: 8.9 %, n = 6) for piributicarb, respectively.
(a)
(b)
(c)
Figure 5.
Results screener window of the positive identification of symetryn (log Ko/w: 2.90) in Ayase river water (a) Mass chromatograms (m/z 213, 170, 155 and 198) (b) Measured mass spectrum (c) Expected and measured relative ion abundance ratio and deviation of the RTL value.
(a)
(b)
(c)
Figure 6.
Results screener window of the positive identification of pyributicarb (log Ko/w: 5.34) in Ayase river water (a) Mass chromatograms (m/z 165, 108, 181 and 93) (b) Measured mass spectrum (c) Expected and measured relative ion abundance ratio and deviation of the RTL value.
Conclusion
The combination of sequential SBSE, thermal desorption and RTL GC/MS operated in the scan mode can provide a very powerful system to screen wide range of pesticide residues at ultra trace levels in water.
References
1. 2. 3. 4. 5. 6. 7. C. L. Arthur, J. Pawliszyn, Anal. Chem, 1990, 62: 2145. E. Baltussen, P. Sandra, F. David, C. A. Cramers, J. Microcol. Sep, 1999, 11: 737. H. Lord, J. Pawliszyn, J. Chromatogr. A, 2000, 885: 153. F. David, P. Sandra, J. Chromatogr. A, 2007, 1152: 54. F. M. lancas, M. E. C. Queiroz, P. Grossi, I. R. B. Olivares, J. Sep. Sci, 2009, 32: 813. N. Ochiai, K. Sasamoto, H. Kanda, E. Pfannkoch, J. Chromatogr. A, 2008, 1200: 72. P. Wiley, Screening for 926 pesticides and endocrin disruptors by GC/MS with deconvolution reporting software and a new pesticide library, Agilent Technologies, publication 5989-5076EN, www.agilent.com/chem
Agilent Technologies, Inc., 2010 Printed in the USA January 19, 2010 5990-5217EN
Analysis of Pesticide Residues in Apple by GC/MS using Agilent SampliQ QuEChERS Kits for Preinjection Cleanup
Application Note
Food Safety
Authors
Karyn Usher West Chester University West Chester, PA 19383 USA Ronald E. Majors Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
The QuEChERS method, which stands for Quick, Easy, Cheap, Effective, Rugged and Safe was developed in 2003 by scientists at the USDA. This method was created to easily clean up and prepare food samples for multi-class, multi-residue pesticide analysis. This application note describes the use of the original, non-buffered QuEChERS method to prepare apple samples for residue analysis by gas chromatography/mass spectrometry (GC/MS). Fifteen pesticides of different classes were studied. The experiments were done using Agilent QuEChERS extraction kits for 10-g samples and dispersive kits for 1-mL sample volumes. The analysis was done by GC/MS using selective ion monitoring (SIM) mode. The limit of quantitation for all the pesticides studied was 10 ng/g in apple using this method. At 200 ng/g, the recoveries ranged from 89% to 102%, and at 10 ng/g, the recoveries ranged from 72% to103%. The relative standard deviations associated with these recoveries were less than 11% in all cases.
Introduction
Pesticides, which include herbicides, fungicides and other pest-control chemicals, have long been a part of agriculture. While their use can be widespread and beneficial, pesticides can also be harmful to both humans and animals. Because of this, pesticide handling is monitored by several agencies, including the United States Environmental Protection Agency (EPA) [1]. Of concern to the general population is the maximum residue levels (MRL) for pesticide in food items. In 2003, scientists at the USDA developed a method for the quick and easy cleanup of food samples for pesticide analysis.[2] This method was given the acronym QuEChERS, which stands for Quick, Easy, Cheap, Effective, Rugged and Safe. The method has since been modified for other analyses. The method for this analysis incorporates a simple acetonitrile/ water extraction facilitated by the addition of MgSO4, which salts out water from the sample and includes a liquid/liquid extraction with these two solvents. The extraction step is followed by a dispersive solid phase extraction that combines
Table 1. Analyte Dichlorvos Pesticide Chemical and Regulatory Information. Structure
both a primary secondary amine (PSA) and anhydrous MgSO4 to remove fatty acids and reduce the remaining water in the extract respectively. See the Agilent SampliQ QuEChERs Kit brochure (publication 5990-3562EN) or www.agilent.com/ chem/quechers for more information about QuEChERS and suggestions for analyses of different fruits and vegetables.
Experimental
Water (EMD Chemicals, Gibbstown, NJ) and acetonitrile (Burdick and Jackson, Muskegan, MI) were HPLC grade. The pesticides were all analytical grade. Dichlorvos (98.9%), and diazinone (99.5%) were purchased from Ultra Scientific (Kingstown, RI). Coumaphos was purchased from Honeywell Riedel De Haen (Seelze, Germany). All other pesticides were purchased from Chem Service (West Chester, PA). See Table 1 for the chemical and regulatory information for the pesticides used in this study. [3-5] The internal standard, triphenyl phosphate (TPP) was purchased from Sigma-Aldrich (St. Louis, MO).
Category Organophosphate
Log P 1.9
pKa NA
MRLs in apple 10
-phenylphenol
Phenol
3.18
9.4
20
Lindane
Organochlorine
3.69
NA
10
Diazinone
Organophosphate
3.69
2.6
100
Chlorpyrifosmethyl
Organophosphate
4.00
NA
500
(Continued)
Table 1. Analyte Chlorpyrifos
Pesticide Chemical and Regulatory Information. Structure Category Organophosphate Log P 4.55 pKa NA MRLs in apple 100
dichlorobenzophenone
Organochlorine
4.44
NA
500
Heptachlor-epoxide
Organochlorine
5.83
NA
10
Chlordane
Cyclodiene organochlorine 2.78
NA
20
Dieldrin
3.7
NA
10
DDE
organochlorine
6.55
NA
50
Endosulfan sulfate
Organochlorine
3.13
NA
50
Permethrins
Pyrethroid
6.1
NA
50
Coumaphos
Organophosphate
3.86
NA
100
Solutions
Individual stock solutions of the pesticides (2 mg/mL) were prepared in methanol and stored at -20 C. From these, the High-QC solution which was 20 g/mL for all the pesticides was prepared in 50:50 acetonitrile/water. The internal standard was 20 g/mL triphenyl phosphate in acetonitrile. The High-QC solution was used to prepare all other spiking solutions. A mid range spiking solution (Mid-QC) with a concentration of 5 g/mL was prepared in 50:50 acetonitrile/ water. A low range spiking solution (Low-QC) with a concentration of 1 g/mL was prepared in 50:50 acetonitrile/water.
added. The sample was shaken vigorously for 1 min, then 4 g of MgSO4 and 1 g of NaCl from the extraction salt packet (p/n 5982-5550) were added. The sample was vortexed for 1 min. A 100-L amount of internal standard solution was added, then the sample was centrifuged (Eppendorf 5810R 15 amp, Westbury, NY) for 5 min at 5000 rpm. See Figure 1.
10g homogenized apple in 50 mL centrifuge tube
Spike sample
Calibration Curve
A 2.5 g/mL standard working solution was prepared using the High-QC solution. A six-point calibration curve (10, 20, 50, 100, 250 and 400 ng/mL) was created by adding the appropriate volume of this 2.5 g/mL solution to the matrix blank extract. Internal standard solution was added to have a final concentration of 100 ng/mL.
Vortex 1 min
Add 10 mL ACN
Shake vigorously for 1 min
Sample Preparation
Certified organic, pesticide-free red delicious apples were purchased at a local grocery. Approximately 3 pounds of apples were diced into approximately 1-cm cubes. The seeds were discarded, but the skin was included. The cubes were placed in a plastic bag and frozen at -20 C overnight. For the first 5 hours in the freezer, the samples were massaged to prevent them from freezing together. When ready to perform the extraction, the amount of sample required was removed from the freezer. A coffee grinder (Mr. Coffee 2.3-oz coffee grinder, Shelton, CT) was used to comminute the sample. If necessary, dry ice may be added to aid this. The sample was checked to ensure that there were no large pieces or lumps remaining prior to extraction. [6].
4g MgSO4 and 1g NaCl (p/n 5982-5550)
Vortex 1 min
Add internal standard (TPP)
Centrifuge 5 min 5000 rpm
Figure 1.
Extraction using SampliQ QuEChERS kit.
QuEChERS Cleanup
Step 1, Extraction
The SampliQ Original QuEChERS Method (non-buffered) Extraction Kit, for use with 10g samples (p/n 5982-5550) was used for the extraction step. A 10-g (0.05g) amount of the homogenized apple sample was placed in a 50-mL centrifuge tube. 100 L of the appropriate spiking solution was added. The sample was vortexed (VWR vortex mixer model K-550-G, West Chester, PA) for 1 min, then 10 mL of acetonitrile were
Step 2 Dispersive SPE cleanup

The SampliQ QuEChERS Dispersive Kit for General Fruits and Vegetables was used for dispersive SPE cleanup (p/n 59825022). This kit removes polar organic acids, some sugars and lipids. One milliliter of the resultant solution was transferred to a 2-mL centrifuge tube containing 50 mg of PSA and 1 50 mg of MgSO4. This was vortexed for 30 sec, then centrifuged for 5 min (VWR micro-centrifuge model 235 B, West Chester, PA). A 0.5-mL amount of the resulting extract was transferred to a sample vial to be analyzed by GC/MS. See Figure 2.
1 mL upper layer in 2 mL centrifuge tube with 50 mg PSA and 150 mg MgSO4 (p/n 5982-5022)
Table 3.
Selective Ion Monitoring (SIM) mode conditions. RT (min) 6.9 10.4 15.7 16.6 18.9 21.5 23.0 24.0 24.6 25.4 25.5 27.9 28.6 32.1 32.2 SIM target 185.00 170.10 180.90 137.10 285.90 196.90 139.00 352.90 372.90 372.90 79.10 246.00 271.80 325.1 183.10 96.90 SIM qualifier 109.50 169.10 182.90 179.10 287.90 354.90 374.90 374.90 317.90 273.80 326.1 109.00 Collection window (min) 4.09.0 9.014.0 14.0-16.0 16.018.0 18.021.0 21.022.0 21.022.0 22.023.6 23.624.3 24.325.0 25.027.0 25.027.0 27.028.0 28.029.5 29.538.0 29.538.0
Peak Vortex for 30 sec
Analyte
1 Dichlorvos 2 -phenylphenol 3 Lindane Centrifuge for 5 min 4 Diazinone 5 Methyl-chlorpyrifos 6 Chlorpyrifos 0.5 mL into sample vial 8 Heptachlor-epoxide Inject 1.5 uL into GC/MS 9 -chlordane 10 -chlordane 11 Dieldrin 12 DDE 13 Endosulfan Sulfate I.S. TPP 14 Permethrin 15 Coumaphos
7 Dichlorobenzophenone 21.5
Figure 2.
Dispersive SPE using SampliQ QuEChERS kit.
Instrument Conditions
Samples were analyzed using an Agilent 7890A GC system with an Agilent 5975C Series GC/MSD (Agilent Technologies Inc., Santa Clara, CA). An Agilent GC/MS method for pesticide analysis was used with some minor modifications. (7) See Tables 2 and 3 for instrument conditions.
Table 2. GCMS instrument conditions

As shown in Figure 3b, the apple matrix blank sample had only a few peaks spread across the experimental collection times for the pesticides using the chosen GC/MS method. In the spiked sample (3a), all compounds except coumaphos were free of interferences and gave good linearity as shown in Table 4. The peak corresponding to coumaphos was difficult to integrate in some samples due to an irregular baseline, which is a possible reason for poor linearity. The QuEChERS method of sample preparation was proven to be quick, easy and effective for this type of analysis. When using the QuEChERS method, samples may still have some impurities that can show up in the chromatograms. In order to achieve the best sensitivity for the analytes of interest, SIM mode was used. Sensitivity for the pesticides was greatly increased by selecting ions corresponding to the analytes of interest to be monitored during different segments of the experiment. In most cases, the highest abundance ion for each analyte was chosen to give the best sensitivity. However, in some cases where selectivity was compromised by this choice, another less abundant ion was used for quantitation. For most of the analytes, a second qualifier ion was also used. The selected ions for each compound and the time segments during which they were monitored are given in Table 3.
GC conditions Injection source Inlet Column Carrier Gas Oven Temperature Program Manual Splitless Agilent J&W HP-5ms Ultra Inert, 30 m 0.250 mm, 0.25 m film (PN: 190915-433UI) Helium in constant flow mode 70 C (2 min) 25 C/min to 150 C (0 min) 3 C/min to 200 C (0 min) 8 C/min to 280 C (7 min) 1 L
Injection volume MS Conditions Tune File Mode: Source temperature Quad temperature Transfer line temperature Solvent Delay Multiplier Voltage
Atune.u SIM 230 C 150 C 280 C 4.00 minutes Autotune voltage
Figure 3. Table 4. Pesticide Dichlorvos
GC/MS chromatograms of apple extract. Peak identifications in Table 3. Regression data for pesticides Regression Equation y = 0.1243x - 0.01141 y = 0.6885x - 0.03763 y = 0.1719x - 0.02280 y = 0.1811x - 0.02608 y = 0.3242x - 0.05026 y = 0.1459x - 0.02455 y = 0.1995x - 0.02828 y = 0.07058x - 0.005587 y = 0.05601x - 0.001840 y = 0.2091x - 0.02544 y = 0.4609x - 0.05950 y = 0.1262x - 0.01675 y = 0.1327x + 0.03232 y = 0.06985x + 0.01864 R2 0.9965 0.9965 0.9967 0.9945 0.9943 0.9916 0.9937 0.9906 0.9917 0.9927 0.9923 0.9901 0.9897 0.9889 0.9543 Dichlorvos -phenylphenol Lindane Diazinone Methyl-chlorpyrifos Chlorpyrifos Heptachlor-epoxide g-chlordane a-chlordane Dieldrin DDE Endosulfan Sulfate Permethrin Coumaphos Table 5. Pesticide Recovery and reproducibility of pesticides in apple using the original QuEChERS method (n=4). High-QC 200ng/g Recovery RSD 99.4 89.5 92.6 102.1 98.5 100.2 95.4 95.9 93.5 99.9 92.7 99.5 97.6 96.6 2.8 6.3 4.2 4.4 3.1 1.2 0.6 3.9 2.0 2.6 1.8 1.9 2.3 2.1 3.0 Mid-QC 50ng/g Recovery RSD 96.7 79.6 88.5 98.8 90.0 95.6 89.1 85.6 90.0 85.8 93.6 87.1 90.8 93.0 79.6 10.8 6.8 9.7 5.5 4.3 4.0 6.4 5.4 6.8 6.9 5.3 5.7 2.8 3.4 3.5 Low-QC 10ng/g Recovery RSD 102.8 92.0 97.9 90.5 88.7 93.5 90.3 87.0 92.3 95.5 99.4 94.5 97.8 100.7 72.5 5.0 6.1 2.0 9.1 7.1 6.5 5.0 3.2 3.5 4.7 4.2 4.2 2.3 4.8 4.5
-phenylphenol Lindane Diazinone Methyl-chlorpyrifos Chlorpyrifos Heptachlor-epoxide -chlordane -chlordane Dieldrin DDE Endosulfan Sulfate Permethrin Coumaphos
Dichlorobenzophenone y = 0.1573x - 0.01840
Dichlorobenzophenone 99.4
Table 5 shows the recovery and reproducibility for each pesticide in the apple matrix spiked at three different concentrations (200 ng/g, 50 ng/g and 10 ng/g).
6
Conclusions
The results show that Agilent SampliQ QuEChERS kits offer an effective method of purification for pesticides in an apple matrix. The impurities remaining after the extraction and dispersive steps were minimal. When used in conjunction with the power of GC/MS in the SIM mode, this method of sample preparation offers a quick, easy and complete solution to quantitate pesticides in fruit matrices.
Acknowledgements
The authors would like to acknowledge Limian Zhao and David Schultz of Agilent Technologies for help with the GC/MS software and in acquiring the pesticide solutions. Also, thank you to Bill Davis and Abby Folk of Agilent Technologies for help with the GC/MS hardware configuration.
References
1. 2. http://www.epa.gov/opp00001/about/#balance M. Anastassiades, S. J. Lehotay, D. Stajnbaher, F.J. Schenck, Journal of AOAC International (JAOAC) 86, p. 412-431, 2003. http://www.ams.usda.gov/AMSv1.0/getfile?d DocName=PDP1995Summary http://www.mrldatabase.com/ http://ecfr.gpoaccess.gov/cgi/t/text/textidx?c=ecfr&sid=1c8cd959ef0d373fb7620f42c8445cca&tpl =/ecfrbrowse/Title40/40cfr180_main_02.tpl L. Zhao, D. Schultz, and J. Stevens, Analysis of Pesticide Residues in Apple using Agilent SampliQ QuEChERS European Standard EN Kits by LC/MS/MS Detection,Agilent Technologies, Santa Clara, CA, Publication 5990-3938EN (2009) P. Wylie, Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library, Agilent Technologies, Santa Clara, CA, Publication 5989-5076EN (2006)
3. 4. 5.
6.
7.
Analysis of Pesticide Residues in Apples using Agilent SampliQ QuEChERS AOAC Kit by LC/MS/MS Detection
Application Note
Food Safety
Authors
Limian Zhao, David Schultz, and Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19809 USA
Abstract
This application note describes the use of a quick, easy, cheap, effective, rugged, and safe (QuEChERS), Association of Analytical Communities (AOAC) Official Method 2007.01; sample preparation approach for extraction and cleanup of 16 pesticide residues in apple. The 16 pesticides chosen represent various classes of interest. The method employed involves initial extraction in a buffered aqueous/acetonitrile system, an extraction/partitioning step after the addition of salt, and then a cleanup step utilizing dispersive solid phase extraction (dispersive SPE). The two different dispersive SPE clean-up approaches (1 mL and 8 mL) were evaluated simultaneously after sample extraction. The target pesticides in the apple extracts were then determined by liquid chromatography coupled to an electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) operating in positive ion multiple reaction monitoring (MRM) mode. The method was validated in terms of recovery and reproducibility. The 5 ng/g limit of quantitation (LOQ) for pesticides in apple shown in this application was well below the maximum residue limits (MRLs). The spiking levels for the recovery experiments were 10, 50, and 200 ng/g. Mean recoveries ranged between 76 and 117% (95.4% on average), with RSD below 15% (4.3% on average).
Introduction
Multi-residue analysis of pesticides in fruits, vegetables, and other foods is the primary function of many regulatory, industrial, and contract laboratories throughout the world. Because of the wide variety of pesticides and complexity of food matrices, the sample must be initially cleaned up using a sample preparation technique prior to analysis. Without question, the most efficient approach to pesticide analysis involves the use of multiclass, multi-residue methods. Once the preliminary analytical quality requirements, including accuracy, precision, sensitivity, selectivity and dynamic range, have been met to suit the needs of a particular analysis, other considerations should be evaluated. These additional considerations include sample throughput, ruggedness, ease of use, cost of materials and labor, toxic solvent usage, and waste generation. The QuEChERS method was introduced first by USDA scientists in 2003. [1] The method was then modified to address some problematic pesticides by using a buffered extraction system. [2] After a full validation for more than 200 pesticides, this improved method was formalized and adopted as AOAC Official Method 2007.01. [3] In summary, the method uses a single-step buffered acetonitrile (1% HAc) extraction while salting out water from the sample using anhydrous magnesium sulfate (MgSO4) to induce liquid-liquid partitioning. After removing an aliquot from the organic layer, for further cleanup a dispersive solid phase extraction (dispersive SPE) is conducted using a combination of primary secondary amine (PSA) to remove fatty acids from other components and anhydrous MgSO4 to reduce the remaining water in the extract. After mixing and centrifugation, the upper layer is ready for analysis. In this study, 16 pesticides were used for evaluating the performance of the Agilent AOAC Buffered Extraction kit (p/n 5982-5755) and SampliQ QuEChERS AOAC dispersive SPE kit for General Fruits and Vegetables (p/n 5982-5022 and 5982-5058), suitable for common fruit and vegetable applications. Apple was selected as the fruit matrix for the evaluation. Most of the pesticides are from the original representative pesticides list [2]. According to their experience, a method working well for these representative pesticides should work equally well for nearly all of the other pesticides
that are routinely monitored in multiclass, multi-residue methods. These pesticides are from 9 different pesticide classes, including acidic, basic, neutral, base-sensitive and acid-labile pesticides. Furthermore, the selected pesticides are suitable for LC/MS/MS analysis. The MRLs of these pesticides are a function of both the pesticide class and food matrix and have been set at 10 ng/g or higher. Table 1 shows the chemical and regulatory information for these multiple class pesticides in apple.
Experimental
All reagents and solvents were HPLC or analytical grade. Acetonitrile (ACN), methanol (MeOH) were from Honeywell (Muskegon, MI, USA). Dimethyl sulfoxide (DMSO) and acetic acid (HAc) were from Sigma-Aldrich (St Louis, MO, USA). Ammonium acetate (NH4OAc) was from Fisher Chemicals (Fair Lawn, NJ, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard, triphenyl phosphate (TPP), were purchased from Sigma-Aldrich (St Louis, MO, USA), ChemService (West Chester, PA, USA), Ultra Scientific (North Kingstown, RI, USA), or AlfaAesar (Ward Hill, MA, USA).

A stock solution of 1M ammonium acetate pH 5 was made by dissolving 19.27 g NH4OAc powder in 250 mL Milli-Q water. The pH was adjusted to 5 with HAc monitored with a pH meter. The solution was stored at 4 C. MeOH/H2O (20:80) containing 5 mM NH4OAc pH 5 was made by combining 200 mL MeOH and 800 mL Milli-Q water, adding 5 mL of 1M NH4OAc pH 5 stock solution. 5 mM NH4OAc in ACN was prepared by adding 5 mL of 1M NH4OAc pH 5 stock solution to 1 L ACN, mixing well and sonicating 5 min. 1% HAc in ACN was prepared by adding 10 mL of acetic acid to 1 L of ACN. Standard and internal standard (IS) stock solutions (2.0 mg/mL for all, except 0.5 mg/mL for carbendazim) were made in MeOH, 0.1% FA in ACN, or DMSO, respectively, and stored at 20 C. Three QC spiking solutions of 1.5, 7.5, and 30 g/mL were made fresh daily in 1:1 ACN/H2O (0.1% FA).
Table 1.
Pesticides Chemical and Regulatory Information [46] MRLs in apple (ng/g)*

O
Name Acephate
Class Organophosphate
Log P 0.89
pKa 8.35
Structure
O O P N S H
20
Carbaryl
Carbamate
2.36
10.4
O
NH O
50
Carbendazim
Benzimidazole
1.48
4.2
H N N
O OCH 3 NH
100
Cyprodinil
Anilinopyrimidine
4.44
N N
H N
50
Dichlofluanid
Sulphamide
3.7
NA
O N S O N S F O CI CI O P O O CI CI
5000
Dichlorvos
Organophosphate
1.9
NA
10
Imidacloprid
Neonicotinoid
0.57
NA
N CI N
NO2
500
NH
Methamidophos
Organophosphate
-0.79
NA
H3CO
O P NH2 SCH3
10
(Continued)
Table 1.
Hormones Used in this Study MRLs in apple (ng/g)*

Cl N N N
Name Penconazole
Class Triazole
Log P 3.72
pKa 1.51
CI
Structure
50
Propoxur
Carbamate
0.14
NA
O O O H N
1000
Pymetrozine
Pyridine
-0.19
4.06
N N N N O H N N N O NH NH S O NH O CH3 O CH3 NH
20
Thiabendazole
Benzimidazole
2.39
4.73 12.00
50
S
Thiophanate-methyl
Benzimidazole
1.45
7.28
HN
100
Tolylfluanid
Sulphamide
3.9
NA
CI
F C S N CI O S O N CH3 CH3
3000
H3C
Ethoprophos
Organophosphate
2.99
NA
H3C H3C S O
O P S CH3
Kresoxim-methyl
Strobilurin
3.4
NA
CH3 O CH3O O N OCH 3
50
A 10 g/mL standard solution in 1:1 ACN/H2O (0.1% FA) was made for preparation of calibration curves in the matrix blank extract by appropriate dilution. A 15 g/mL of TPP in 1:1 ACN/H2O (0.1% FA) was used as an internal standard (IS).
HPLC conditions Column: Agilent ZORBAX Solvent Saver Plus Eclipse Plus Phenyl-Hexyl, 3.0 x 150 mm, 3.5 m (p/n 959963-312) 0.3 mL/min 30 C 10 L A: 5mM NH4OAc, pH 5.0 in 20:80 MeOH/H2O B: 5 mM NH4OAc, pH 5.0 in ACN 1:1:1:1 ACN/MeOH/IPA/H2O (0.2% FA) Flow rate Time %B (mL/min) 0 0.5 8.0 10.0 10.01 12.0 13.0 4 min 17 min 20 20 100 100 20 100 STOP 0.3 0.3 0.3 0.3 0.5 0.5

Agilent 1200 HPLC with Diode Array Detector (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent 6410 Triple Quadrupole LC/MS/MS system with Electrospray Ionization (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent SampliQ Buffered QuEChERS AOAC Extraction kit, p/n 5982-5755, and SampliQ QuEChERS AOAC Dispersive SPE kit for General Fruits and Vegetables, p/n 5982-5022 and 5982-5058 (Agilent Technologies Inc., Wilmington, DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, So. Plainfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA) Grinder (St. Joseph, MI, USA)
Flow rate: Column Temperature: Injection volume: Mobile Phase:
Post run: Total cycle time: MS conditions Positive mode Gas Temperature: Gas Flow: Nebulizer: Capillary:
350 C 10 L/min 40 psi 4000 V
Other MS conditions relating to the analytes are listed in Table 2.
Table 2. Analyte Acephate
Instrument Acquisition Data Used for the Analysis of 16 Pesticides by LC/MS/MS MRM channels (m/z) 1) 184.0 > 94.9 2) 184.0 > 111.0 1) 142.0 > 94.0 2) 142.0 > 124.9 1) 218.1 > 105.0 2) 218.1 > 78.0 1) 192.1 > 160.0 2) 192.1 > 105.0 1) 221.0 > 109.0 2) 221.0 > 95.0 1) 343.1 > 151.0 2) 343.1 > 117.9 1) 210.1 > 111.0 2) 210.1 > 92.9 1) 202.0 > 145.0 2) 202.0 > 115.0 1) 226.1 > 93.0 2) 226.1 > 108.0 1) 333.0 > 123.0 2) 333.0 > 223.9 1) 243.1 > 130.9 2) 243.1 > 172.9 1) 284.1 > 158.9 2) 284.1 > 172.9 1) 347.0 > 136.9 2) 347.0 > 238.0 1) 202.1 > 175.0 2) 202.1 > 131.0 1) 256.1 > 209.1 2) 256.1 > 175.0 1) 327.1 > 77.0 2) 327.1 > 151.9 1) 314.0 > 222.1 2) 314.0 > 235.0 Fragmentor (V) 60 60 115 95 110 105 50 50 120 85 80 80 60 110 60 70 70 CE (V) 3 15 8 8 20 50 18 40 13 40 17 65 12 15 3 40 35 35 28 5 15 15 32 32 25 3 27 38 12 18 45 45 10 10 RT (min) 2.55 2.54 2.97 5.07 6.57 7.08 6.89 7.30 9.23 9.40 8.50 8.95 9.73 5.65 5.53 9.49 9.44
Methamidophos Pymetrozine Carbendazim Dichlorvos Thiophanate methyl Propoxur Carbaryl Cyprodinil Dichlorfluanid Ethoprophos Penconazole Tolyfluanid Thiabendazole Imidacloprid TPP Kresoxim methyl 1) Quantifier transition channel 2) Qualifier transition channel
Sample preparation
Sample comminution In order to get the most reliable statistical results, it is important to spend the necessary effort and time on conducting proper sampling and homogenization procedures. Organically grown, pesticide-free apples were purchased from a local grocery store. Approximately three pounds of apples were chopped into small, bean-sized cubes. Skin was included, but pit was discarded. The chopped apple cubes were put into a clean plastic bag and frozen at 20 C overnight. The bag was massaged occasionally to make sure the cubes were frozen loosely, to avoid clumping. The following day, a portion of
6
frozen apple cubes were removed and thoroughly blended. Certain precautions were exercised while blending the sample. First, the chopped apple cubes remained in the freezer until the point of blending. Only the portion of apple cubes necessary for homogenizing were removed; the rest were kept in the freezer until the next comminution. Dry ice was added, when possible, while comminuting to keep the temperature low. Second, the blender container was kept dry to prevent clumping. In between blending, the container was rinsed and dried. Third, samples were comminuted thoroughly to obtain the best sample homogeneity. No pieces of apple were visible in the final sample.
Extraction/Partitioning A 15 g (0.05 g) previously homogenized sample was placed into a 50 mL centrifuge tube from the SampliQ QuEChERS Extraction kit. QC samples were fortified with 100 L of appropriate QC spiking solution yielding QC samples with concentrations of 10, 50, and 200 ng/g. One hundred microliters of IS spiking solution (15 g/mL of TPP) were added to all samples except the control blank to yield a 100 ng/g concentration in each sample. Tubes were capped and vortexed for 1 min. Fifteen milliliters of 1% HAc in ACN were added to each tube using the dispenser. To each tube, an Agilent AOAC Buffered Extraction packet from the kit (p/n 5982-5755) containing 6 g of anhydrous MgSO4 and 1.5 g of anhydrous NaOAc, was added directly to the tubes. No powders were left in the threads or rims of the tubes. Tubes were sealed tightly and shaken vigorously for 1 min by hand to ensure that the solvent interacted well with the entire sample and crystalline agglomerates were broken up. Sample tubes were centrifuged at 4000 rpm for 5 min. Dispersive SPE Cleanup A 1 mL aliquot of the upper ACN layer was transferred into a SampliQ QuEChERS AOAC 2 mL dispersive SPE tube (p/n 5982-5022) or 8 mL aliquot were transferred into an SampliQ QuEChERS AOAC 15 mL dispersive SPE tube (p/n 5982-5058). The 2 mL tube contained 50 mg of PSA and 150 mg of anhydrous MgSO4; while the 15 mL tube contained 400 mg of PSA and 1200 mg of anhydrous MgSO4. The tubes were tightly capped and vortexed for 1 min. The 2 mL tubes were centrifuged with a micro-centrifuge at 13000 rpm for 2 min, and the 15 mL tubes were centrifuged in a standard centrifuge at 4000 rpm for 5 min. Two hundred microliters of extract were transferred into an autosampler vial. Then 800 L of water or another appropriate standard solution (prepared in water) were added. The samples were capped and vortexed thoroughly. The samples were then ready for LC/MS/MS analysis. The flow chart in Figure 1 illustrates the sample preparation procedure.
Accurately weigh 15 g homogenized sample (0.05 g) in 50 mL centrifuge tubes
Spike samples with 100 L of IS solution and vortex for 1 min
Add 15 mL of 1% acetic acid in ACN, shake vigorously for 1 min
Add SampliQ QuEChERS AOAC salt packet, cap tubes and shake vigorously for 1 min
Centrifuge at 4000 rpm for 5 min
Transfer upper ACN layer to SampliQ QuEChERS dispersive-SPE tube, 1 mL/2 mL tube or 8 mL/15 mL tube
Vortex 1 min then centrifuge
Transfer 200 L extract to autosampler vial, dilute with 800 L appropriate solution if necessary
Samples are ready for LC/MS/MS analysis
Figure 1.
QuEChERS AOAC sample preparation procedures flow chart.

In addition to being fast, easy, cheap, effective, rugged and safe, an additional key feature of the QuEChERS method is the potential for the simultaneous analysis of multi-pesticide residues. With the new design of SampliQ QuEChERS kits, the whole procedure is even faster, easier, and offers more time and labor savings, while ensuring consistency. An analyst can process 4050 samples in just a few hours. Adding a food
7
sample with a high percentage of water directly to the salts may create an exothermic reaction that can affect analyte recovery. Agilent's SampliQ salts and buffers are uniquely prepared in anhydrous packages. This allows addition AFTER adding solvent to the sample, as specified in the QuEChERS methodology. The final QuEChERS sample may contain food matrix impurities because it is a very simple sample extraction and cleanup procedure. The final apple extract appeared light green. But with the powerful selectivity of LC/MS/MS multiple reaction monitoring (MRM) mode, the extracted apple blank appeared to be clean and free of impurities, indicating the blank apple extract did not contribute any interferences with the target compounds. Figure 2 shows the chromatograms of a blank apple extract and a 10 ng/g fortified apple extract.
x10 3 1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0
+ MRM (226.10001 & 93.00000) control blank -1.d 1 1 2 2 3 3 4
0.5
1.5
2.5
3.5
4.5
5.5 6 6.5 7 7.5 8 8.5 Counts (%) vs. acquisition time (min)
9.5
10
10.5
11
11.5
12 12.5
13
Figure 2a
Chromatograms of apple extract blank. No interference was found in the blank.

x10 2 1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 Counts (%) vs. acquisition time (min) 9 9.5 10 10.5 11 11.5 12 12.5 13
1 + MRM (192.10001 & 160.00000) low QC -1.d
1 2
2 3 8
3 4
4 11 6
10 13
2 1
15 3 5 7 9 12 14 16
Figure 2b. Chromatogram of 10 ng/g fortified apple extract. Peak identification: 1. Methamidophos, 2. Acephate, 3. Pymetrozine, 4. Carbendazim, 5. Imidacloprid, 6. Thiabendazole, 7. Dichlorvos, 8. Propoxur, 9. Thiophanate methyl, 10. Carbaryl, 11. Ethoprophos, 12. Penconazole, 13. Cyprodinil, 14. Dichlofluanid, 15. Kresoxim methyl, 16, Tolyfluanid.

The linear calibration range for all the pesticides was 5 250 ng/g. Since two different dispersive SPE volumes (1 mL and 8 mL) were used for evaluation and comparison, two sets of calibration curves were generated respectively. Matrix blanks were prepared for each size. Calibration curves, spiked in matrix blanks, were made at levels of 5, 10, 50, 100, 200, and 250 ng/g. The TPP (IS) was used at 100 ng/g level.
The calibration curves were generated by plotting the relative responses of analytes (peak area of analyte/peak area of IS) to the relative concentration of analytes (concentration of analyte/concentration of IS). Table 1 shows that the 5 ng/g quantification limits LOQ (5 ppb) established for all of the pesticides is lower than the MRLs of these pesticides in fruit and vegetables. Table 3 shows the regression equation and correlation coefficient (R2) for both 1 mL and 8 mL dispersive SPE volumes.
Table 3. Analytes
Linearity of Pesticides in Apple Extract 1 mL dispersive SPE Regression equation Y = 0.2349X - 0.0013 Y = 0.1118X - 0.0012 Y = 0.2671X - 0.0016 Y = 0.9441X + 0.0063 Y = 0.0513X - 0.0009 Y = 0.7049X + 0.0044 Y = 0.0265X + 0.0001 Y = 2.0348X - 0.0091 Y = 0.2024X - 0.0054 Y = 0.4984X - 0.0002 Y = 0.8203X - 0.0064 Y = 0.1775X - 0.0006 Y = 0.3529X - 0.0023 Y = 0.0453X - 0.0004 Y = 0.2498X - 0.0024 Y = 0.0718X - 0.0016 R2 0.9949 0.9881 0.9950 0.9895 0.9905 0.9868 0.9884 0.9951 0.9307 0.9965 0.9952 0.9903 0.9960 0.9869 0.9932 0.9823 8 mL dispersive SPE Regression equation Y = 0.2300X - 0.0007 Y = 0.1094X - 0.0014 Y = 0.2290X - 0.0014 Y = 0.8583X + 0.0006 Y = 0.0500X - 0.0007 Y = 0.6198X + 0.0043 Y = 0.0247X + 0.0006 Y = 2.0264X - 0.0090 Y = 0.5090X - 0.0041 Y = 0.4889X - 0.0029 Y = 0.8536X - 0.0076 Y = 0.1783X - 0.0019 Y = 0.3528X - 0.0022 Y = 0.0460X - 0.0006 Y = 0.2490X - 0.0013 Y = 0.0755X - 0.0006 R2 0.9981 0.9980 0.9975 0.9968 0.9933 0.9961 0.9439 0.9965 0.9682 0.9976 0.9971 0.9848 0.9958 0.9954 0.9927 0.9788
Methamidophos Acephate Pymetrozine Carbendazim Imidacloprid Thiabendazole Dichlorvos Propoxur Thiophanate methyl Carbaryl Ethoprophos Penconazole Cyprodinil Dichlorfluanid Kresoxim methyl Tolyfluanid

The recovery and reproducibility were evaluated by spiking pesticide standards in homogeneous apple samples at levels of 10, 50, and 200 ng/g. These QC samples were quantitated against the matrix spiked calibration curve. The analysis was performed in replicates of six (n = 6) at each level. The recovery and reproducibility (RSD) data of 1 mL and 8 mL dispersive SPE sample volumes are shown in Tables 4 and 5, respective-
ly. It can be seen from the results that all of the pesticides give acceptable recoveries (average of 97.5% for 1 mL and 93.3% for 8 mL) and precision (average of 4.5% RSD for 1 mL and 4.1% RSD for 8 mL). The notoriously base-sensitive pesticides such as dichlorfluanid and tolyfluanid showed excellent recovery and precision. Acid labile pesticide, pymetrozine, also showed acceptable recovery and precision.
Table 4.
Recovery and Repeatability of Pesticides in Fortified Apple With 2 mL Dispersive SPE Tube (p/n 5982-5022) 10 ng/g fortified QC Recovery RSD (n=6) 83.6 106.8 78.3 101.0 107.0 106.2 78.2 106.3 79.0 93.4 95.8 117.0 106.9 92.5 98.2 96.6 5.6 5.8 11.4 6.5 6.5 6.6 11.4 0.8 15.4 1.9 4.5 4.8 4.0 6.5 9.3 9.5 50 ng/g fortified QC Recovery RSD (n=6) 81.3 95.6 76.6 98.5 97.6 103.7 94.2 105.7 76.7 98.4 96.1 111.9 102.0 96.3 101.9 105.1 2.6 2.3 11.6 4.3 3.4 2.6 7.2 1.2 15.4 2.2 1.8 2.3 2.8 2.2 2.7 1.8 200 ng/g fortified QC Recovery RSD (n=6) 83.4 97.3 108.1 91.0 107.4 95.5 95.8 101.2 102.2 97.5 94.7 111.0 102.4 99.4 104.1 102.2 1.4 2.0 5.3 2.6 3.0 2.0 1.8 1.6 8.1 1.1 1.3 1.6 1.8 2.6 1.8 1.7
Analytes Methamidophos Acephate Pymetrozine Carbendazim Imidacloprid Thiabendazole Dichlorvos Propoxur Thiophanate methyl Carbaryl Ethoprophos Penconazole Cyprodinil Dichlorfluanid Kresoxim methyl Tolyfluanid
10
Table 5.
Recovery and Repeatability of Pesticides in Fortified Apple With 15 mL Dispersive SPE Tube (p/n 5982-5058) 10 ng/g fortified QC Recovery RSD (n=6) 80.6 94.6 88.8 85.9 101.8 92.5 73.7 96.2 81.4 86.5 89.6 102.1 93.9 81.7 91.8 94.1 9.3 7.0 12.1 3.9 3.5 6.4 14.8 1.6 4.9 2.6 2.9 2.5 3.7 8.7 5.8 7.9 50 ng/g fortified QC Recovery RSD (n=6) 79.4 93.7 87.7 90.4 99.3 92.2 91.8 98.2 78.2 90.3 92.1 106.0 97.4 96.9 93.9 95.2 2.9 3.4 10.1 2.7 3.7 2.6 7.3 0.6 13.4 1.4 1.0 3.0 0.9 5.6 2.0 4.0 200 ng/g fortified QC Recovery RSD (n=6) 83.1 95.1 118.4 85.5 106.0 89.5 95.5 97.2 102.3 91.1 94.1 111.0 99.7 98.1 98.3 97.5 2.5 2.5 5.5 2.2 0.9 1.5 2.0 1.2 5.8 1.2 1.1 1.6 2.0 2.6 1.2 2.6
11
Figure 3 shows the recovery and precision results for 1 mL and 8 mL dispersive SPE. The two different dispersive SPE clean-ups were performed by using 1 mL or 8 mL of ACN extract from the same sample tube after the extraction step. In order to simplify the comparison, the average recovery and precision of three fortification concentrations were used for all pesticides. The results of two dispersive SPE clean-up approaches appeared to be independent of volume used. Both approaches provided efficient sample clean-up, and generated relatively equivalent results.
120
1 mL
8 mL
100
80
Recovery (%)
60
40
20
0 Thiophanate methyl Methamidophos Acephate Dichlorvos Ethoprophos Propoxur Thiabendazole Dichlorfluanid Carbendazim Penconazole Pymetrozine Imidacloprid Carbaryl Cyprodinil Kresoxim methyl Tolyfluanid
Figure 3.
Results comparison of 1 mL and 8 mL dispersive SPE sample volume.
12
Conclusions
Agilent SampliQ AOAC Buffered Extraction kit and SampliQ AOAC dispersive SPE kit for General Fruits and Vegetables provided a simple, fast, and effective method for the purification of representative pesticides in apple. The recovery and reproducibility, based on matrix spiked standards, were acceptable for multiclass, multi-residue pesticide determination in apple. The impurities and matrix effects from apple were minimal and did not interfere with the quantitation of any target compound. The LOQs of the pesticides were significantly lower than their regulated MRLs in apple. As the selected pesticides represented a broad variety of different classes and properties, the Agilent SampliQ QuEChERS AOAC Extraction and Dispersive kit for General Fruits and Vegetables can be used for other pesticides in similar fruit matrices.
References
1. M. Anastassiades, S. J. Lehotay, Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and Dispersive Solid-Phase Extraction for the Determination of Pesticide Residues in Produce, 2003, 86, 412- 431. S. J. Lehotay, et al; Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005, 88, 615-629. S. J. Lehotay, et. al.; Determination of Pesticide Residues in Foods by Acetonitrile Extraction and Partitioning with Magnesium Sulfate: Collaborative Study, J. AOAC Int., 2007, 90, 485-520. http://sitem.herts.ac.uk/aeru/footprint/en/index.htm http://www.m5.ws001.squarestart.ne.jp/foundation/ search.html http://www.mrldatabase.com/?selectvetdrug=0
2.
3.
4. 5. 6.

13
Analysis of Pesticide Residues in Apple using Agilent SampliQ QuEChERS European Standard EN Kits by LC/MS/MS Detection
Application Note
Food Safety
Authors
Limian Zhao, David Schultz, and Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19809 USA
Abstract
This application note describes the use of a quick, easy, cheap, effective, rugged, and safe (QuEChERS) sample preparation approach described in the European Committee Standard (EN) for extraction and cleanup of 16 multiple class pesticide residues of interest in apple. The method employed involves initial extraction in an aqueous/acetonitrile system, an extraction/partitioning step after the addition of salt, and then a cleanup step utilizing dispersive solid phase extraction (dispersive SPE). The two different dispersive SPE clean-up approaches (1 mL and 6 mL sample volume) are evaluated simultaneously after sample extraction. The target pesticides in the apple extracts are then determined by liquid chromatography coupled to an electrospray ionization tandem mass spectrometer (LC-ESI-MS/MS) operating in positive ion multiple reaction monitoring (MRM) mode. The method is validated in terms of recovery and reproducibility. The 5 ng/g of limits of quantitation (LOQ) for pesticides in apple established in this application is well below their regulatory maximum residue limits (MRLs). The spiking levels for the recovery experiments are 10, 50, and 200 ng/g. Excluding pymetrozine, recoveries of the pesticides ranged between 73 and 111% (87% on average), and RSDs below 20% (5.8% on average).
Introduction
Multi-residue analysis of pesticides in fruits, vegetables, and other foods is a primary function of many regulatory, industrial, and contract laboratories throughout the world. Because of the wide variety of pesticides and complexity of food matrices, the sample must be initially cleaned up using a compatible sample preparation technique before injection into the detection system. It is unquestionable that the most efficient approach to pesticide analysis involves the use of multiclass, multi-residue methods. Once the preliminary analytical quality requirements of accuracy, precision, sensitivity, selectivity, and dynamic scope have been met to suit the need for a particular analysis, other practice considerations should be evaluated. These additional considerations include high sample throughput, ruggedness, ease of use, low cost, labor, minimal toxic solvent usage, and waste generation. The QuEChERS method was introduced first by USDA scientists in 2003 [1]. The method was then modified to address some problematic pesticides by using a buffered extraction system. [2] There is also a European variation, the prEn method 15662: 2007 [3], [4]. In summary, the method uses acetonitrile extraction, followed by the salting out of water from the sample using anhydrous magnesium sulfate (MgSO4), NaCl and buffering citrate salts to induce liquid-liquid partitioning. For cleanup, a dispersive solid-phase extraction (dispersive SPE) is conducted using a combination of primary secondary amine (PSA) to remove fatty acids from other components and anhydrous MgSO4 to reduce the remaining water in the extract. After mixing and centrifugation, the upper layer is ready for analysis. The EN methodology is similar in principal to the AOAC method, but has several differences. First, the extraction buffered system in the EN method uses sodium chloride, sodium citrate dehydrate, and disodium citrate hydrogenate sesquihidrate instead of sodium acetate in the extraction step. Second, in the dispersive SPE step, the EN method uses 25 mg PSA per mL of extract rather than 50 mg PSA per mL of extract used by the AOAC method. In this study, 16 pesticides are used to demonstrate the performance of Agilent SampliQ QuEChERS EN Buffered
Extraction kit (p/n 5982-5650) and EN dispersive SPE kit (p/n 5982-5021 and 5982-5056) for General Fruits and Vegetables, suitable for common fruit and vegetable applications. Most of the pesticides are from the original 'representative pesticides' list [1]. According to their experience, a method working well for these representative pesticides should work equally well for nearly all of the other pesticides that are routinely monitored in multiclass, multi-residue methods. These pesticides are from nine different pesticide classes, including acidic, basic, neutral, base-sensitive, and acid-labile pesticides. Furthermore, the selected pesticides are suitable for LC/MS/MS analysis. The MRLs of these pesticides have been set for 10 ng/g or higher. Table 1 shows the chemical and regulatory information of these pesticides.
Experimental
All reagents and solvents were HPLC or analytical grade. Acetonitrile (ACN), methanol (MeOH) were from Honeywell (Muskegon, MI, USA). Dimethyl sulfoxide (DMSO) was from Sigma-Aldrich (St Louis, MO, USA). Ammonium acetate (NH4OAc) was from Fisher Chemicals (Fair Lawn, NJ, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard (triphenyl phosphate, TPP) were purchased from Sigma-Aldrich (St Louis, MO, USA), ChemService (West Chester, PA, USA), Ultra Scientific (North Kingstown, RI, USA), or AlfaAesar (Ward Hill, MA, USA).

A 1M NH4OAc pH 5 stock solution was made by dissolving 19.27 g NH4OAc powder in 250 mL Milli-Q water. The pH was adjusted to 5 with HAc and monitored with a pH meter. The solution was stored at 4 C. 20:80 MeOH/H2O containing 5 mM NH4OAc pH 5 was made by combining 200 mL MeOH and 800 mL Milli-Q water, adding 5 mL of 1M NH4OAc pH 5 stock solution. A 5 mM NH4OAc in ACN solution was prepared by adding 5 mL of 1M NH4OAc pH 5 stock solution to 1 L ACN, mixing well and sonicating 5 min. A 1% FA in ACN solution was prepared by adding 1 mL of FA to 100 mL of ACN.
Table 1.

O
Name Acephate
Log P 0.89
pKa 8.35
Structure
O O P N S H
20
Carbaryl
Carbamate
2.36
10.4
O
NH O
50
Carbendazim
Benzimidazole
1.48
4.2
H N N
O OCH 3 NH
100
Cyprodinil
Anilinopyrimidine
4.44
N N
H N
50
Dichlofluanid
Sulphamide
3.7
NA
O N S O N S F O CI CI O P O O CI CI
5000
Dichlorvos
Organophosphate
1.9
NA
10
Imidacloprid
Neonicotinoid
0.57
NA
N CI N
NO2
500
NH
Methamidophos
Organophosphate
0.79
NA
H3CO
O P NH2 SCH3
10
(Continued)
Table 1. Name
Pesticides Chemical and Regulatory Information [57] Class Triazole Log P 3.72 pKa 1.51 Structure
CI Cl N N N
MRLs in apple (ng/g)* 50
Penconazole
Propoxur
Carbamate
0.14
NA
O O O H N
1000
Pymetrozine
Pyridine
-0.19
4.06
N N N N O H N N N O NH NH S O NH O CH3 O CH3 NH
20
Thiabendazole
Benzimidazole
2.39
4.73 12.00
50
S
Thiophanate-methyl
Benzimidazole
1.45
7.28
HN
100
Tolylfluanid
Sulphamide
3.9
NA
CI
F C S N CI O S O N CH3 CH3
3000
H3C
Ethoprophos
Organophosphate
2.99
NA
H3C H3C S
O P O S CH3
Kresoxim-methyl
Strobilurin
3.4
NA
50
Standard and internal standard (IS) stock solutions (2.0 mg/mL for all, except 0.5 mg/mL for carbendazim) were made in MeOH, 0.1% FA in ACN, or DMSO, respectively, and stored at 20 C. Three QC spiking solutions of 1,5, and 20 g/mL were made fresh daily in 1:1 ACN/H2O with 0.1% FA. A 10 g/mL standard spiking solution in 1:1 ACN/H2O with 0.1% FA was made for preparation of calibration curves in the matrix blank extract by appropriate dilution. A 10 g/mL IS spiking standard of TPP was made in 1:1 ACN/H2O (0.1% FA).
HPLC conditions Column: Agilent ZORBAX Solvent Saver Plus Eclipse Plus Phenyl-Hexyl, 3.0 x 150 mm, 3.5 m (p/n 959963-312) 0.3 mL/min 30 C 10 L A: 5 mM ammonium acetate, pH 5.0 in 20:80 MeOH/H2O B: 5 mM ammonium acetate, pH 5.0 in ACN 1:1:1:1 ACN/MeOH/IPA/H2O (0.2% FA.) Flow rate Time %B (mL/min) 0 0.5 8.0 10.0 10.01 12.0 13.0 4 min 17 min 20 20 100 100 20 100 STOP 0.3 0.3 0.3 0.3 0.5 0.5
Flow rate: Column temperature: Injection volume: Mobile phase:

Agilent 1200 HPLC with Diode Array Detector (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent 6410 Triple Quadrupole LC/MS/MS system with Electrospray Ionization (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent SampliQ QuEChERS extraction kit, p/n 5982-5650, and dispersive SPE tubes, p/n 5982-5021 and 5982-5056 (Agilent Technologies Inc., Wilmington, DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, So. Plainfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA) Grinder (St. Joseph, MI, USA)
Post run: Total cycle time: MS conditions Positive mode Gas temperature: Gas flow: Nebulizer: Capillary:
350 C 10 L/min 40 psi 4000 V
Other conditions relating to the analytes are listed in Table 2.
Instrument Acquisition Data Used for the Analysis of 16 Pesticides by LC/MS/MS MRM channels (m/z) 1) 184.0 > 94.9 2) 184.0 > 111.0 1) 142.0 > 94.0 2) 142.0 > 124.9 1) 218.1 > 105.0 2) 218.1 > 78.0 1) 192.1 > 160.0 2) 192.1 > 105.0 1) 221.0 > 109.0 2) 221.0 > 95.0 1) 343.1 > 151.0 2) 343.1 > 117.9 1) 210.1 > 111.0 2) 210.1 > 92.9 1) 202.0 > 145.0 2) 202.0 > 115.0 1) 226.1 > 93.0 2) 226.1 > 108.0 1) 333.0 > 123.0 2) 333.0 > 223.9 1) 243.1 > 130.9 2) 243.1 > 172.9 1) 284.1 > 158.9 2) 284.1 > 172.9 1) 347.0 > 136.9 2) 347.0 > 238.0 1) 202.1 > 175.0 2) 202.1 > 131.0 1) 256.1 > 209.1 2) 256.1 > 175.0 1) 327.1 > 77.0 2) 327.1 > 151.9 1) 314.0 > 222.1 2) 314.0 > 235.0 Fragmentor (V) 60 60 115 95 110 105 50 50 120 85 80 80 60 110 60 70 70 CE (V) 3 15 8 8 20 50 18 40 13 40 17 65 12 15 3 40 35 35 28 5 15 15 32 32 25 3 27 38 12 18 45 45 10 10 RT (min) 2.55 2.54 2.97 5.07 6.57 7.08 6.89 7.30 9.23 9.40 8.50 8.95 9.73 5.65 5.53 9.49 9.44
Methamidophos Pymetrozine Carbendazim Dichlorvos Thiophanate methyl Propoxur Carbaryl Cyprodinil Dichlorfluanid Ethoprophos Penconazole Tolyfluanid Thiabendazole Imidacloprid TPP Kresoxim methyl 1) Quantifier transition channel 2) Qualifier transition channel
Sample preparation
Sample comminution In order to get the most reliable statistical results, it is important to spend the necessary effort and time on conducting proper sampling and homogenization procedures. Organically grown, pesticide free apples were purchased from a local grocery store. Approximately three pounds of apples were chopped into small, bean sized cubes. Skin was included, but the pit was discarded. Then, the chopped apple cubes were put into a clean plastic bag and frozen at 20 C overnight. The bag was massaged occasionally to make sure the cubes were frozen loosely to avoid clumping. The following day, a
6
portion of frozen apple cubes were removed and thoroughly blended. Certain precautions were exercised while blending the sample. First, the chopped apple cubes remained in the freezer until the point of blending. Only the portion of apple cubes necessary for homogenizing were removed; the rest was kept in the freezer until the next comminution. Dry ice was added while comminuting to keep the temperature low. Second, the blender container was kept dry to prevent clumping. In between blending, the container was rinsed and dried. Third, samples were comminuted thoroughly to get the best sample homogeneity. There were not any pieces of apple visible in the final sample.
Extraction/Partitioning A 10 g (0.05g) previously homogenized sample was placed into a 50 mL centrifuge tube. QC samples were fortified with 100 L of appropriate QC spiking solution. A 100 L of IS spiking solution (10 g/mL of TPP) were added to all of samples except the control blank to yield a 100 ng/g concentration in sample. Tubes were capped and vortexed for 1 min. Ten milliliters of ACN was added to each tube using the dispenser. Tubes were then capped and shaken by hand for 1 min. To each tube, an Agilent SampliQ QuEChERS EN extraction salt packet (p/n 5982-5650), containing 4 g anhydrous MgSO4, 1g NaCl, 1g sodium citrate, and 0.5 g disodium citrate sesquihydrate, was added directly. No powders were left in the threads or rims of the tubes. Tubes were sealed tightly and shaken vigorously for 1 min by hand to ensure that the solvent interacted well with the entire sample and crystalline agglomerates were broken up sufficiently. Sample pH was checked with pH paper, and 5N NaOH was added to adjust the pH to 5-5.5. Sample tubes were centrifuged at 4000 rpm for 5 min. Dispersive SPE Cleanup A 1 mL aliquot of the upper ACN layer was transferred into an Agilent SampliQ QuEChERS EN dispersive SPE 2 mL tube (p/n 5982-5021); or 6 mL of aliquot was transferred into an Agilent SampliQ QuEChERS EN dispersive SPE 15 mL tube (p/n 5982-5056). The 2 mL tube contained 25 mg of PSA and 150 mg of anhydrous MgSO4; while the 15 mL tube contained 150 mg of PSA and 900 mg of anhydrous MgSO4. The tubes were capped tightly and vortexed for 1 min. The 2 mL tubes were centrifuged with a micro-centrifuge at 13,000 rpm for 2 min, and the 15 mL tubes with a standard centrifuge at 4000 rpm for 5 min. Ten microlitres of extract were transferred into an autosampler vial. A 10 L of a 1% FA in ACN solution was added immediately, in addition to 800 L of water or appropriate standard solution (prepared in water). The samples were capped and vortexed thoroughly, to prepare for LC/MS/MS analysis. The flow chart in Figure 1 illustrates the sample preparation procedure.
Spike samples with 100 L of IS solution and vortex for 1 min
Add 10 mL of ACN, and shake 1 min
Add SampliQ EN extraction packet, and shake vigorously by hand for 1 min
Transfer 1 mL of upper ACN layer to SampliQ EN dispersive SPE 2 mL tube, or 6 mL to SampliQ EN dispersive SPE 15 mL tube
Vortex 1 min, centrifuge at 13,000 rpm for 2 min for 2 mL tubes or at 4000 rpm for 5 min for 15 mL tubes
Transfer 200 L extract to autosampler vial, add 10 L of 1% FA in ACN, and dilute with 800 L water
Samples are ready for LC/MS/MS analysis
Figure 1.
QuEChERS EN sample preparation procedures flow chart.

In addition to being fast, easy, cheap, effective, rugged and safe, an additional key feature of the QuEChERS method is the potential for the simultaneous analysis of multi-pesticide residues. With the new design of Agilents SampliQ QuEChERS kits, the whole procedure is even faster, easier, and offers more time and labor savings, while ensuring con7
sistency. An analyst can process 4050 samples in just a few hours. Adding a food sample with a high percentage of water directly to the salts may create an exothermic reaction that can affect analyte recovery. Agilent's SampliQ salts and buffers are uniquely prepared in anhydrous packages. This allows addition AFTER adding solvent to the sample, as specified in the QuEChERS methology. The final QuEChERS sample still contains food matrix impurities because it is a very simple sample extraction and cleanup procedure. The final apple extract appeared light green. But with the powerful selectivity of LC/MS/MS multiple reaction monitoring mode, the extracted apple blank appeared to be clean and free of coeluting impurities, indicating that the cleaned-up apple extract did not contribute any interferences with the target compounds. Figure 2 shows the chromatograms of a blank apple extract and a 10 ng/g fortified apple extract.
x10 3 1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0
+ MRM (226.10001 & 93.00000) control blank -1.d 1 1 2 2 3 3 4
0.5
1.5
2.5
3.5
4.5
5.5 6 6.5 7 7.5 8 8.5 Counts (%) vs. acquisition time (min)
9.5
10
10.5
11
11.5
12 12.5
13
Figure 2a
Chromatograms of apple extract blank. No interference was found in the blank.

x10 2 1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 Counts (%) vs. acquisition time (min) 9 9.5 10 10.5 11 11.5 12 12.5 13
1 + MRM (192.10001 & 160.00000) low QC -1.d
1 2
2 3 8
3 4
4 11 6
10 13
2 1
15 3 5 7 9 12 14 16
Figure 2b. Chromatogram of 10 ng/g fortified apple extract. Peak identification: 1. Methamidophos, 2. Acephate, 3. Pymetrozine, 4. Carbendazim, 5. Imidacloprid, 6. Thiabendazole, 7. Dichlorvos, 8. Propoxur, 9. Thiophanate methyl, 10. Carbaryl, 11. Ethoprophos, 12. Penconazole, 13. Cyprodinil, 14. Dichlofluanid, 15. Kresoxim methyl, 16, Tolyfluanid.

The linear calibration range for all of the pesticides was 5 250 ng/g. Since two different dispersive SPE sizes (1 mL and 6 mL sample volume) were used for evaluation and comparison, two sets of calibration curves were generated respectively. Matrix blanks were prepared for each size. Calibration curves, spiked in matrix blanks, were made at levels of 5, 10, 50, 100, 200, and 250 ng/g. The TPP was used as an internal standard (IS) at 100 ng/g level. The calibration
curves were generated by plotting the relative responses of analytes (peak area of analyte/peak area of IS) to the relative concentration of analytes (concentration of analyte/concentration of IS). Table 1 shows that the 5 ng/g quantification limits LOQ (5 ppb) established for all of the pesticides is significantly lower than the MRLs of these pesticides in fruit and vegetables. Table 3 shows the regression equation and correlation coefficient (R2) for both 1 mL and 6 mL dispersive SPE.
Table 3. Analytes
Linearity of Pesticides in Apple Extract 1 mL dispersive SPE Regression equation Y = 0.3203X 0.0005 Y = 0.1373X 0.0021 Y = 0.4688X 0.0009 Y = 1.4253X + 0.0126 Y = 0.0647X 0.0004 Y = 0.9014X + 0.0127 Y = 0.0364X + 0.0002 Y = 2.4398X 0.0001 Y = 0.3171X 0.0015 Y = 0.6378X + 0.0017 Y = 1.0897X 0.0030 Y = 0.2334X 0.0012 Y = 0.4805X + 0.0008 Y = 0.0552X 0.0003 Y = 0.2958X 0.0005 Y = 0.0860X 0.0011 R2 0.9972 0.9975 0.9961 0.9931 0.9944 0.9922 0.9884 0.9989 0.9965 0.9989 0.9984 0.9978 0.9992 0.9970 0.9978 0.9918 6 mL dispersive SPE Regression equation Y = 0.3255X 0.0018 Y = 0.1375X 0.0010 Y = 0.3821X + 0.0007 Y = 1.3379X + 0.0045 Y = 0.0636X 0.0006 Y = 0.8600X + 0.0050 Y = 0.0362X + 0.0002 Y = 2.4272X + 0.0029 Y = 0.2869X 0.0020 Y = 0.6363X + 0.0003 Y = 1.0628X 0.0001 Y = 0.2186X 0.0003 Y = 0.4697X 0.0017 Y = 0.0562X 0.0012 Y = 0.2762X 0.0003 Y = 0.0845X 0.0008 R2 0.9957 0.9953 0.9782 0.9903 0.9974 0.9942 0.9892 0.9994 0.9904 0.9988 0.9992 0.9979 0.9985 0.9946 0.9966 0.9968
Methamidophos Acephate Pymetrozine Carbendazim Imidacloprid Thiabendazole Dichlorvos Propoxur Thiophanate methyl Carbaryl Ethoprophos Penconazole Cyprodinil Dichlorfluanid Kresoxim methyl Tolyfluanid

The recovery and reproducibility were evaluated by spiking pesticides standards in comminuted apple sample at levels of 10, 50, and 200 ng/g. These QC samples were quantitated against the matrix spiked calibration curve. The analysis was performed in replicates of six (n = 6) at each level. The recovery and reproducibility (RSD) data of 1 mL and 6 mL dispersive SPE are shown in Tables 4 and 5, respectively. It can be seen from the results, that all of the pesticides but pymetrozine
give acceptable recoveries (average of 85.7% for 1 mL and 88.2% for 6 mL) and precision (average of 6.0% RSD for 1 mL and 5.7% RSD for 6 mL). The notorious base-sensitive pesticides such as dichlorfluanid and tolyfluanid showed excellent recovery and precision. Pymetrozine, an acid labile pesticide, shows poor recovery using the European method when compared to the AOAC method [2]. With the AOAC method, an average recovery of 88% with 9.4% average RSD for pymetrozine was obtained [8].
Table 4.
Recovery and Repeatability of Pesticides in Fortified Apple With 2 mL EN Dispersive SPE Tube (p/n 5982-5021) 10 ng/g fortified QC Recovery RSD (n=6) 73.0 92.8 27.1 85.1 91.0 84.8 83.1 97.8 79.9 89.3 93.7 109.2 98.9 85.1 90.4 98.3 5.6 4.2 18.2 5.9 3.3 6.8 13.9 2.6 8.5 2.8 1.6 6.7 6.9 7.8 4.8 13.7 50 ng/g fortified QC Recovery RSD (n=6) 75.6 87.2 24.9 89.5 102.7 90.4 92.2 100.2 79.9 92.5 93.5 108.1 101.2 92.2 99.6 102.0 3.1 5.6 10.5 3.4 5.4 3.5 5.2 2.9 2.9 3.6 2.9 5.7 2.7 4.4 3.9 4.0 200 ng/g fortified QC Recovery RSD (n=6) 84.3 95.6 28.1 84.1 107.2 86.7 93.1 100.7 85.5 95.8 95.7 110.6 102.9 99.4 103.7 106.0 5.3 5.8 12.3 4.7 4.9 4.0 4.6 3.8 5.7 4.1 3.4 4.4 4.3 5.5 4.5 4.4
10
Table 5.
Recovery and Repeatability of Pesticides in Fortified Apple With 15 mL EN Dispersive SPE Tube (p/n 5982-5056) 10 ng/g fortified QC Recovery RSD (n=6) 77.6 86.6 28.1 89.9 105.3 89.8 97.8 99.5 87.4 92.9 94.8 106.8 102.7 99.7 102.6 92.0 4.9 7.6 24.3 6.8 10.8 4.8 14.5 3.8 5.8 6.1 5.5 4.9 4.5 18.9 12.0 9.3 50 ng/g fortified QC Recovery RSD (n=6) 77.8 87.8 27.5 88.9 105.2 87.6 98.2 104.0 88.3 93.7 99.2 111.2 105.7 97.4 106.1 105.5 6.4 5.5 12.2 3.3 4.8 3.1 5.6 2.6 4.7 3.0 3.0 3.0 3.5 4.5 2.0 3.3 200 ng/g fortified QC Recovery RSD (n=6) 81.2 91.5 29.1 81.8 106.6 84.2 98.1 100.9 89.0 93.6 98.8 109.0 102.4 98.9 106.1 105.1 2.1 1.5 10.4 3.6 5.0 1.4 2.5 3.3 7.6 2.4 3.8 4.1 2.6 5.5 5.6 4.3
11
Figure 3 shows the recovery and precision results comparison of 1 mL dispersive SPE and 6 mL dispersive SPE. The two different dispersive SPE clean-ups were performed by transferring 1 mL or 6 mL of ACN extract from the sample tube after the extraction step. To simplify the comparison, the average recovery and precision of three fortification concentrations were used for all pesticides. The results of two dispersive SPE clean-up approaches appear to be independent of volume used. There was < 10% difference in recovery and < 5% difference for RSD. Both approaches provided efficient sample clean-up, and generated relatively equivalent results.
120
1 mL
6 mL
100
80 Recovery (%)
60
40
20
0 Thiophanate methyl Methamidophos Ethoprophos Acephate Dichlorvos Propoxur Thiabendazole Dichlorfluanid Carbendazim Penconazole Pymetrozine Imidacloprid Carbaryl Cyprodinil Kresoxim methyl Tolyfluanid
Figure 3.
Results comparison of 1-mL dispersive SPE and 6-mL dispersive SPE.
12
Conclusions
The Agilent SampliQ QuEChERS EN fruit and vegetable kit provides a simple, fast and effective method for the purification and enrichment of selective representative pesticides in apple. The recovery and reproducibility results, based on matrix spiked standards, were acceptable for selected pesticide residue determination in apple. The impurities and matrix effect from apple were minimal and did not interfere with the quantitation of target compounds. The LOQs of the pesticides were lower than their MRLs in fruits and vegetables. As the selected pesticides represented a broad variety of different classes and properties, the Agilent SampliQ QuEChERS EN kit for General Fruits and Vegetables can be used for other pesticides in similar food matrices.
5. 6. 7. 8.
http://sitem.herts.ac.uk/aeru/footprint/en/index.htm http://www.m5.ws001.squarestart.ne.jp/foundation/ search.html http://www.mrldatabase.com/?selectvetdrug=0 L. Zhao, D. Schultz, and Joan Stevens Evaluation of the QuEChERS AOAC Sample Preparation Kit for the Analysis of Pesticide Residues in Apples with LC/MS/MS Detection, Agilent Technologies publication 59903937EN

References
1. M. Anastassiades, S. J. Lehotay, Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and Dispersive Solid-Phase Extraction for the Determination of Pesticide Residues in Produce, 2003, 86, 412- 431. S. J. Lehotay, et al; Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005, 88, 615-629. European Committee for Standardization/Technical Committee CEN/TC 275 (2007), Foods of plant origin: Determination of pesticide residues using GC-MS and/or LC-MS/MS following acetonitrile extraction/partitioning and cleanup by dispersive SPE-QuEChERS method. European Committee for Standardization, Brussels P. Pay, M. Anastassiades, Analysis of Pesticide Residues Using the Quick Easy Cheap Effective Rugged and Safe (QuEChERS) Pesticide Multiresidue Method in Combination With Gas and Liquid Chromatography and Tandem Mass Spectrometric Detection, Anal Bioanal Chem., 2007, 389, 1697-1714.
2.
3.
4.
13
Optimizing Recoveries of Planar Pesticides in Spinach Using Toluene and Agilent SampliQ AOAC QuEChERS Kits with Graphitized Carbon
Application Note
Food Safety
Authors
Limian Zhao, Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19809-1610 USA
Abstract
This application note describes the impact of toluene addition in the dispersive solid phase extraction (SPE) step on the analysis of pesticides in spinach using Agilent SampliQ QuEChERS AOAC kits for highly pigmented fruits and vegetables. Graphitized carbon black (GCB) is required in the dispersive SPE kits in order to remove high levels of pigments from the matrix. However, it also retains pesticides with planar structures resulting in poor recovery and precision. The eight problematic pesticides found in the original AOAC method, by either LC/MS/MS or GC/MS, generated poor results with about 20% to 60% recovery with >15% relative standard deviation (RSD). In the modified AOAC method, an aliquot of toluene was added to the dispersive SPE cleanup tube, in a ratio of 8:3 (acetonitrile (ACN) extracts/toluene). It significantly improved the extraction efficiency of the problematic planar pesticides. With the modified AOAC method, the eight problematic pesticides generated substantially improved recoveries, 50% to 100%, and < 10% RSD. However, the addition of toluene also introduced more matrix impurities into the final sample, and caused problems for some pesticides which gave good results originally. Therefore, the modified AOAC method cannot be considered a "drop in" replacement for the original AOAC method; but it can be a very useful alternative for the problematic pesticides affected by GCB in the pesticides analysis of highly pigmented matrix.
Introduction
The AOAC quick, easy, cheap, effective, rugged, safe (QuEChERS) method has been widely applied in the analysis of pesticides in food since it was introduced by USDA scientists. [1-3] In general, it contains two major steps: extraction and dispersive SPE clean-up. In the extraction step, the method uses a single-step buffered acetonitrile (1% HAc) extraction while simultaneously salting out water from the sample using anhydrous magnesium sulfate (MgSO4) to induce liquid-liquid partitioning. For cleanup, a dispersive solid phase extraction (dispersive SPE) step is employed using a combination of primary secondary amine (PSA) to remove polar organic acids as well as other components, and anhydrous MgSO4 to reduce the remaining water in the extract. After mixing and centrifugation, the upper layer is ready for analysis. Various food matrices require modifications to the dispersive SPE clean-up step. For general fruits and vegetables, 50 mg PSA and 150 mg MgSO4, per mL of ACN extracts are used for clean-up to remove polar organic acids, some sugars and lipids, and excess water. Pigmented fruits and vegetables kits, besides PSA and MgSO4, include 50 mg GCB per mL of ACN extracts to remove pigments like chlorophyll and carotinoides. For fruits and vegetables with fats and waxes, 50 mg C18 per mL of ACN extracts is added with PSA and MgSO4 removing lipids and sterols. Therefore, according to the food matrix, analysts need to select a suitable dispersive SPE kit in order to analyze pesticides of interest. Previously, we demonstrated the excellent performance of SampliQ QuEChERS AOAC buffered AOAC extraction kits and dispersive SPE kits for general fruits and vegetables on a representative group of pesticides in apple by LC/MS/MS and GC/MS. [4, 5] For the SampliQ QuEChERS AOAC kits for pigmented fruits and vegetables, spinach was selected as the matrix in order to evaluate the extraction and performance of the dispersive kit. GCB was added to the dispersive SPE kit to remove the high level of pigments, such as chlorophyl and carotinoides, which can cause more matrix effect and introduce more interferences. Conversely, GCB can cause a significant loss of planar pesticides, for example, thiobendazole, chlorothalonil, coumaphos, cyprodinil. [3, 6] Therefore, the use of GCB is recommended when planar pesticides are not being analyzed; greatly limiting the usefulness of GCB to the clean-up of pigmented matrix. In previous GCB SPE column extractions [7], solvent mixtures containing toluene were commonly used to elute pesticides through GCB columns. ACN/toluene (3:1) mixtures have been used for the multiclass multiresidue method (MRM) elution of pesticides through
tandem GCB-NH2 [8], GCB-PSA [9], and GCB SAX-PSA. [10] In this study, toluene was added into the ACN extracts in the second step of QuEChERS, the dispersive SPE clean-up. We determined that the ratio of 8:3 (ACN extract toluene) generated higher recoveries (50% to 300% higher), and substantially better precision (< 10% RSD) for the eight GCB retained pesticides. However, it was noted that the addition of toluene caused adverse affects, such as additional matrix impurities in the final extracted samples, lower recovery and higher imprecision for certain pesticides which originally produced good results without the addition of toluene.
Experimental
All reagents and solvents were HPLC or analytical grade. Methanol (MeOH), and toluene were from Honeywell (Muskegon, MI, USA). Acetonitrile (ACN), dimethyl sulfoxide (DMSO) and acetic acid, glacial (HAc) were from SigmaAldrich (St Louis, MO, USA). Ammonium acetate (NH4OAc) was from Fisher Chemicals (Fair Lawn, NJ, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard triphenyl phosphate, (TPP) were purchased from Sigma-Aldrich (St Louis, MO, USA), ChemService (West Chester, PA, USA), Ultra (North Kingstown, RI, USA), or AlfaAesar (Ward Hill, MA, USA).

The 1 M ammonium acetate pH 5 stock solution was made by dissolving 19.27 g NH4OAc powder in 250 mL Milli-Q water, and the pH adjusted to 5 with glacial acetic acid. The solution was stored at 4 C. Methanol/H2O (20:80) containing 5 mM ammonium acetate pH 5 was made by combining 200 mL MeOH and 800 mL Milli-Q water, adding 5 mL of 1M ammonium acetate pH 5 stock solution and mixing well. A 5 mM ammonium acetate in ACN solution was prepared by adding 5 mL of 1 M ammonium acetate pH 5 stock solution to 1 L ACN, mixing well and sonicating 5 min. 1% HAc in ACN was prepared by adding 10 mL of glacial acetic acid to 1 L of ACN, and mixing well. Standard and internal standard (IS) stock solutions (2.0 mg/mL for all, except 0.5 mg/mL for carbendazim) were made in MeOH, 0.1% FA in ACN, or DMSO, respectively, and stored at 20C. Three QC spiking solutions of 1.5, 7.5, and 30 g/mL were made fresh daily in 1:1 ACN/H2O containing 0.1% FA. A 10 g/mL standard spiking solution in 1:1 ACN/H2O containing 0.1% FA was made for preparation of LC/MS/MS calibration curves in the matrix blank extract by appropriate dilution.
A 2.5 g/mL standard solution in ACN containing 0.1% FA was used to prepare the GC/MS calibration curves in the matrix blank extract by appropriate dilution. A 15 g/mL IS spiking standard of TPP in 1:1 ACN/H2O containing 0.1% FA was made.
GC conditions Inlet: Inlet liner: Carrier gas: Inlet pressure: Inlet temperature: Injection volume: Purge flow to split vent: Oven temperature program: Splitless Helix double taper, deactivated (p/n: 5188-5398) Helium 19.6 psi (constant pressure mode) during run 1.0 psi during back flush 250 C 1.0 L 30 mL/min at 0.75 min 70 C (1 min), 50 C/min to 150 C (0 min), 6 C /min to 200 C (0 min), 16 C/min to 280 C (6 min) 3 min Purged Ultimate Union (p/n: G3186B) used for backflushing the analytical column and inlet. Helium plumbed to Purged Ultimate Union 4.0 psi during run, 80.0 psi during backflush Agilent J&W HP-5ms Ultra Inert 15 m 0.25 mm, 0.25 m (p/n: 19091S-431UI) Between inlet and Purged Ultimate Union (p/n: G3186B) 65 cm x 0.15 mm, 0.15 m DB-5 ms Ultra Inert. Between the Purged Ultimate Union and the MSD.
Equipment and material

Agilent 1200 Series HPLC with Diode Array Detector (Agilent Technologies Inc., CA, USA). Agilent 6410 triple quadrupole MS/MS system with Electrospray Ionization (Agilent Technologies Inc., CA, USA). Agilent Gas Chromatograph (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent 5975C Mass Spectrometer (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent SampliQ QuEChERS AOAC Extraction kits, p/n 5982-5755, and SampliQ QuEChERS AOAC dispersive SPE kits for Pigmented Fruits and Vegetables, p/n 5982-5222 and 5982-5258 (Agilent Technologies Inc., DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, South Painfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA)
Post run: Capillary flow technology:
Aux EPC gas: Aux EPC pressure: Column:
Connections: Restrictor: Connections:
HPLC conditions Column: Agilent ZORBAX Solvent Saver Plus Eclipse Plus Phenyl-Hexyl, 3.0 150 mm, 3.5 m (p/n: 959963-312) 0.3 mL/min 30C 10 L A: 5 mM ammonium acetate, pH 5.0 in 20:80 MeOH/H2O; B: 5 mM ammonium acetate, pH 5.0 in ACN 1:1:1:1 ACN/MeOH/IPA/H2O w/0.2% FA. Time 0 0.5 8.0 10.0 10.01 13.0 Post run: Total cycle time: 4 min 17 min. %B 20 20 100 100 20 STOP Flow rate (mL/min) 0.3 0.3 0.3 0.3 0.5
For the instrument acquisition data of MS/MS in LC/MS/MS and MS in GC/MS relating to the analytes, please refer to the acquisition data table in the previous Agilent publications. [4, 11]
Sample Preparation
The sample preparation procedure includes sample comminution, extraction/partitioning and dispersive SPE clean-up. The QuEChERS method employing spinach as the vegetable matrix is similar to the method described in detail in previous application notes [4,5], with the exception of the dispersive SPE step which includes a toluene addition. The frozen chopped organic spinach was homogenized thoroughly. Fifteen grams ( 0.1g) of homogenized sample was placed into 50 mL centrifuge tubes. Samples were fortified with appropriate QC spiking solutions (100 L) if necessary, and then 100 L of IS spiking solution (15 g/mL of TPP). After vortexing the samples for 30 s, 15 mL of 1% HAc in ACN was added to each tube. An Agilent SampliQ QuEChERS AOAC extraction salt packet (p/n 5982-5755) was added directly to each tube. Sample tubes were sealed tightly, and hand-shaken vigorously for 1 min. Tubes were centrifuged at 4,000 rpm for 5 min.
3
Flow rate: Column temperature: Injection volume: Mobile phase:
A 1 mL aliquot of the upper ACN layer was transferred into an Agilent SampliQ QuEChERS dispersive SPE 2 mL tube (p/n 5982-5222); or an 8 mL aliquot was transferred into an Agilent SampliQ QuEChERS dispersive SPE 15 mL tube (p/n 5982-5258). The 2 mL tube contained 50 mg of PSA, 50 mg of GCB and 150 mg of anhydrous MgSO4, while the 15 mL tube contained 400 mg of PSA, 400 mg of GCB and 1200 mg of anhydrous MgSO4. Subsequently, 375 L of toluene were added to the 2 mL tubes, and 3 mL of toluene were added to the 15 mL tubes. The tubes were tightly capped and vortexed for 1 min. The 2 mL tubes were centrifuged with a micro-centrifuge at 13,000 rpm for 2 min, and the 15 mL tubes centrifuged in a standard centrifuge at 4,000 rpm for 5 min. An 825 L amount of extract was transferred into a 15 mL centrifuge tube and dried by N2 flow. Samples were reconstituted with 600 L of ACN containing 0.1% FA, vortexed and sonicat-
ed. A 200 L aliquot of the extract was transferred into an autosampler vial, and 800 L of water or appropriate standard solutions (prepared in water) were added. The samples were capped and vortexed thoroughly prior to LC/MS/MS analysis. For samples analyzed by GC/MS, a 600 L reconstituted sample was either transferred directly to an autosampler vial or used to prepare the calibration curves. In order to determine toluene's affect on the dispersive SPE procedure, another aliquot of ACN extracts was processed following the original dispersive SPE clean-up procedure. Figure 1 shows the dispersive SPE procedure scheme according to the original method (w/o toluene) and the modified method (w/ toluene).
ACN extracts after first extraction/partitioning step

Original method Modified method
Transfer 1 mL of ACN extracts to 2 mL dispersive SPE tube
Add 325 L of Toluene Vortex 30 sec Vortex 30 sec
Transfer 825 L of upper ACN layer to another tube
Dry with N2 flow at 30C
Reconstitute into 600 L of 0.1%FA in ACN
Vortex and sonicate to completely dissolve the sample Transfer certain volume for LC/MS/MS or GC/MS analysis Transfer certain volume for LC/MS/MS or GC/MS analysis
Figure 1.
Dispersive SPE procedures of original method (w/o toluene) and modified method (w/toluene).

Impact on the Clean-up of Matrix
The QuEChERS methodology for pesticide residue analysis provided high-quality results with a fast, easy, inexpensive approach. For pigmented fruits and vegetables, the addition of GCB in the dispersive SPE tube can greatly remove pigments and sterols. This was clearly shown by the color of the extracts. The spinach ACN extract after the first salt extraction step was very dark green in color. When a dispersive SPE kit for pigmented produce (with GCB) was employed for dispersive SPE clean-up, the upper ACN extract layer became
clear with an almost colorless to very light yellow color. On the contrary, when a dispersive SPE kit for general fruits and vegetables was used without GCB, the upper layer was still a dark green to black color. The dispersive SPE extracts modified by the addition of toluene gave a bright yellow color after vortexing and centrifuging. The increase of color for the extracts suggested that the addition of toluene either reduced the affinity of GCB for those pigment molecules, or backextracted those molecules from the GCB. The addition of toluene resulted in more impurities in the final extracted sample which is demonstrated by the comparison of the UV chromatograms ( = 254 nm) for the two matrix blanks as shown in Figure 2.
Comparisons of Matrix Blanks for UV with and without the Addition of Toluene
10 1 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 DAD1-A:Sig = 250,4 1mL zero blank-organic2.d 10 1 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 Response units (%) vs. acquisition time (min) 12 10 1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 11 Response units (%) vs. acquisition time (min) 12 1 2 3 4 5 6 7 8 9 10 11 Response units (%) vs. acquisition time (min) 12 DAD1-A:Sig = 250,4 1mL zero blank-organic2.d
10 1 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
DAD1-A:Sig = 250,4 1mL zero.d
DAD1-A:Sig = 250,4 1mL zero.d
3 4 5 6 7 8 9 10 11 Response units (%) vs. acquisition time (min)
12
Figure 2.
UV chromatogram ( = 254 nm) comparison of matrix blank obtained with original method without toluene (A) and modified method with addition of toluene (B). Left chromatograms shown in small scale for detail comparison, and right chromatograms shown full scale for big interference peaks comparison. In both cases, the same scale was used for blank A and B chromatograms.
However, the increase in matrix impurities didn't affect the LC/MS/MS or GC/MS pesticide analysis. Figure 3 shows the spinach matrix blank LC/MS/MS chromatograms processed by the modified method with addition of toluene (A) and the original method without addition of toluene (B). With the enhanced selectivity of LC/MS/MS, the two blank samples
(A and B) showed similarly clean chromatograms. Figure 4 shows the spinach matrix blank GC/MS chromatograms processed by the modified method with the addition of toluene (A) and the original method without the addition of toluene (B). The two blank chromatograms show some minor differences, but similarities are confirmed.
Comparison of Matrix Blanks for LC/MS/MS and the Negligible Affect of Toluene Addition
x10 2 3 1 2.5 2 1.5 1 0.5 0 x10 2 3 1 2.5 2 1.5 1 0.5 0 1 2 3 4 5 6 7 8 9 10 11 12 12 2 3 3 4 4
12
23
34
Figure 3.
Spinach matrix blank LC/MS/MS chromatogram. A. Spinach matrix blank processed by modified method (w/toluene); B. Spinach matrix blank processed by original method (w/o toluene).
Comparison of Matrix Blanks for GC/MS and the Negligible Affect of Toluene Addition
4000 3500 3000 2500 2000 1500 1000 500 0 4.00 4000 3500 3000 2500 2000 1500 1000 500 0 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00
Figure 4.
Spinach matrix blank GC/MS chromatograms. A. Spinach matrix blank processed by modified method (w/toluene); B. Spinach matrix blank processed by original method (w/o toluene).
Significant Improvement Made on Some Pesticides

The improvements made by the addition of toluene on certain pesticides was very significant (50% to 300% increase in recovery). Because GCB adsorbs planar compounds, the method produced very low recoveries (20% to 60%) of pesticides with planar compounds and poor precision (>14% RSD). These problematic pesticides determined by the original method included carbendazim, thiabendazole, pymetrozin, cyprodinil, chlorthalonil, coumaphous, dichlorobenzophenone, and folpet. The first four pesticides were analyzed by LC/MS/MS, and the second four pesticides by GC/MS. The optimum volume of toluene addition was determined by parallel spinach samples spiked at the same concentration level and subjected to buffered salt extraction. An 8 mL
7
aliquot of ACN extract was transferred into a 15 mL dispersive tube. Different volumes of toluene were added according to the following ratios: 8:1, 8:2 and 8:3 (ACN extracts/ toluene, n = 3). Samples without the addition of toluene were also processed for comparison. The final samples were analyzed by LC/MS/MS, and an average of analyte responses (peak area of analyte/peak area of IS) were used for response comparison. As shown in Figure 5, the addition of toluene increased the extraction efficiency, as indicated by a 200% to 300% higher analyte response. In general, the more toluene added, the higher the responses obtained. Therefore, the addition of toluene at a ratio of 8:3 was selected for both the LC/MS/MS and GC/MS experiments. This ratio is comparable to the ratio of 3:1 ACN/toluene that Schenck recommended. [7]
Results Comparison of Different Toluene Addition Ratios and their Increase of Recovery for Certain Pesticides
40
35
No toluene 8:1 toluene addition
Analytes responses (analyte peak area/IS peak area)
30
8:2 toluene addition 8:3 toluene addition
25
20
15
10
0 Carbendazim Pymetrozine Thiobendazole Cyprodinil
Figure 5.
Results comparison of different toluene addition volumes. First column: results generated with no toluene addition; second column: results generated with toluene addition at ratio of 8:1 (ACN extracts/toluene); third column: results generated with toluene addition at ratio of 8:2; fourth column: results generated with toluene addition at ratio of 8:3.
The two different sizes of dispersive SPE (1 mL and 8 mL) were also compared for toluene addition. According to the ratio of 8:3, 3 mL of toluene were added to the 8 mL tubes; while 375 L of toluene were added to the 1 mL tubes. The results obtained by the modified method were also compared to those from the original method. As shown in Figure 6, both dispersive SPE volumes incorporating the modified method significantly increased the recovery of the difficult pesticides by 200-300%, and gave a substantial improvement in precision. The 1 mL volume dispersive SPE provided slightly higher recovery compared to the 8 mL volume dispersive SPE, espe-
cially for pymetrozine and thiabendazole. Processing a single sample with the buffered salt extraction and partitioning step produced about 14 mL of ACN extract, which is enough to process dispersive SPE by both the original and modified methods at a 1 mL volume simultaneously. Additionally, a smaller amount of toluene was required. Therefore, the use of the1 mL volume dispersive SPE kits with the modified method is recommended for problematic pesticides. This eliminates the need for another buffered salt extraction, saving analyst time, labor and additional sample and solvents.
Results Comparison of Different Sizes of Dispersive SPE and the Drastic Increase in Recovery for Certain Pesticides upon Toluene Addition
120
1 mL w/ toluene 1 mL w/o toluene

100 Analytes responses (analyte peak area/IS peak area)
8 mL w/ toluene 8 mL w/o toluene
80
60
40
20
0 Carbendazim Pymetrozine Thiobendazole Cyprodinil
Figure 6.
Results comparison of 1 mL and 8 mL dispersive SPE with the modified method (w/ toluene) and the original method (w/o toluene).
Impact on Other Pesticides

The impact of toluene addition on other pesticides was monitored and the results used to classify these pesticides into three groups. The first group of pesticides showed the same recovery and precision from both the original method and modified method. The second group of pesticides were those in which the addition of toluene generated about 10% to15% less recovery, but still showed acceptable precision. The third group included only one pesticide, dichlorvos, from the 34 pesticides screened by LC/MS/MS or GC/MS. For this pesti-
cide, the addition of toluene adversely affected the analysis of dichlorvos producing much lower recovery and unacceptable precision. In general, these negative impacts were observed more on GC amenable pesticides than LC amenable pesticides, and may be linked to the additional drying step in the modified method. Table 1 shows the impact the addition of toluene made on the modified dispersive SPE analysis of representative pesticides.
Table 1.
The Impact on Certain Pesticides by the Modified Dispersive-SPE with Addition of Toluene Original method (w/o toluene) Recovery RSD (n=6) 38.9 21.8 27.6 29.6 21.1 30.1 53.7 62.0 88.8 88.6 94.9 103.9 81.4 95.5 108.0 97.0 14.6 19.7 21.2 23.4 16.4 24.0 4.5 14.6 6.0 4.6 5.9 4.5 7.2 5.6 2.5 3.1 Modified method (w/ toluene) Recovery RSD (n=6) 98.5 69.7 65.2 63.1 47.3 87.9 77.7 88.2 20.4 73.7 81.3 101.3 83.3 99.8 109.1 96.7 2.5 2.7 3.7 3.2 5.9 6.1 6.1 6.3 89.8 7.4 4.0 4.5 5.1 4.7 1.9 2.5 Impact with modified method Positive Positive Positive Positive Positive Positive Positive Positive Greatly negative Slightly negative Slightly negative None None None None None
Analytes Carbendazim Thiabendazole Pymetrozine Cyprodinil Chlorthalonil Coumaphos Dichlorobenzophenone Folpet Dichlorvos -Phenylphenol Diazinon Chlordane Permethrin Acephate Carbaryl Propoxur
Detection method LC/MS/MS LC/MS/MS LC/MS/MS LC/MS/MS GC/MS GC/MS GC/MS GC/MS GC/MS GC/MS GC/MS GC/MS GC/MS LC/MS/MS LC/MS/MS LC/MS/MS
Conclusion
This application note discusses the impact of the addition of toluene on the AOAC QuEChERS method for the analysis of multiclass pesticide residues using Agilent SampliQ AOAC buffered extraction kits and SampliQ AOAC dispersive SPE kits for pigmented fruits and vegetables. The addition of toluene at a ratio of 8:3 (ACN extracts/toluene) to the dispersive SPE step can significantly increase the recovery of problematic pesticides with planar structure by 50% to 300% and improve precision. The addition of toluene can also generate some negative effects, by introducing more matrix impurities, and reducing the recovery of certain pesticides. Therefore, the modified method should not be considered a direct replace-
ment for the original method. It does provide an option for problematic pesticides affected by GCB in the analysis of a highly pigmented matrix. The extraction will not have to be repeated from the beginning. The ACN extracts after the first buffered salt extraction step can be processed by both the original and modified AOAC methods simultaneously with Agilent SampliQ 2 mL dispersive SPE kits for pigmented matrix, saving the analyst additional sample preparation and solvent usage. By combining the results from the original and modified methods, analysts can obtain extremely impressive results and analyze a greater variety of multiclass pesticides in pigmented fruits and vegetables relative to the original method.
10
Reference
1. Anastassiades M., Lehotay S.J.; Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and "Dispersive Solid-Phase Extraction" for the Determination of Pesticide Residues in Produce, J. AOAC Int., 2003, 86, 412- 431. Lehotay S.J., et al; Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005, 88, 615-629. Lehotay S.J., et. al.; Determination of Pesticide Residues in Foods by Acetonitrile Extraction and Partitioning with Magnesium Sulfate: Collaborative Study, J. AOAC Int., 2007, 90, 485-520. L. Zhao, D. Schultz, Evaluation of the QuEChERS AOAC Sample Preparation Kit for the Analysis of Pesticide Residues in Apples with LC/MS/MS Detection, Agilent Technologies publication 5990-3937EN. L. Zhao, D. Schultz, J. Stevens, Analysis of Pesticide Residues in Apple Using Agilent SampliQ QuEChERS AOAC Kits by GC/MS, Agilent Technologies publication 5990-4068EN. S. J. Lehotay, Quick, Easy, Cheap, Effective, Rugged, and Safe Approach for Determining Pesticide Residues, Methods in Biotechnology, Vol. 19, Pesticide Protocols, Edited by Martnez Vidal J.L. and Garrido Frenich A., Humana Press Inc., Totowa, NJ, 2006.
7.
F. J. Schenck, J. W. Wong, Determination of Pesticides in Food of Vegetal Origin, Analysis of Pesticides in Food and Environmental Samples, Chapter 6, edited by J. L. Tadeo, CRC Press Inc., Boca Raton, FL, 2008. G. F. Pang, et al.; Simultaneous Determination of 446 Pesticide Residues in Fruits and Vegetables by Three Cartridge Solid Phase Extraction/Gas ChromatographyMass Spectrometry and Liquid Chromatography-Tandem Mass Spectrometry, J. AOAC Int., 2005, 89, 740-771. M. Okihashi, et al., Rapid Method for the Determination of 180 Pesticide Residues in Foods by Gas Chromatography/Mass Spectrometry and Flame Photometric Detection, J. Pestic. Sci., 2005, 30, 368-377.
8.
2.
9.
3.
4.
10. R. S. Sheridan, and J. R. Meola, Analysis of Pesticide Residues in Fruits, Vegetables and Milk by Gas Chromatography/Tandem Mass Spectrometry, J. AOAC Int., 1999, 82, 982. 11. L. Zhao, P. L. Wylie, J. Stevens, Analysis of Pesticide Residues in Apple Using Agilent SampliQ QuEChERS EN Kits by GC/MS, Agilent Technologies publication 5990-4073EN.
5.

6.
11
Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS AOAC Kit by LC/MS/MS Detection
Application Note
Food Safety
Authors
Limian Zhao, Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19809-1610 USA
Abstract
This application note describes the use of a quick, easy, cheap, effective, rugged, and safe (QuEChERS) AOAC sample preparation approach for the extraction and cleanup of 13 pesticide residues representing various pesticide classes in spinach. The original AOAC method employed involves initial extraction in a buffered aqueous/acetonitrile system, an extraction/partitioning step after the addition of salt, and a cleanup step using dispersive solid-phase extraction (dispersive SPE). In order to address the significant loss of planar pesticides caused by graphitized carbon black (GCB) in dispersive SPE, a modified method with the addition of toluene was employed. The presence of the target pesticides in the spinach extracts were then determined by liquid chromatography coupled to an electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) operating in positive ion multiple reaction monitoring (MRM) mode. With the combination of original and modified dispersive SPE, the method was validated in terms of recovery and reproducibility for all of the analytes of interest. The 5 ng/g limit of quantitation (LOQ) for pesticides in spinach shown in this application was well below the maximum residue limits (MRLs). The spiking levels for the recovery experiments were 10, 50, and 200 ng/g. Mean recoveries ranged between 64% and 108% (average of 91.9%), with RSD below 10% (average of 3.3%).
Introduction
The AOAC QuEChERS method has been widely applied in the analysis of pesticides in food since it was introduced by USDA scientists. [1-3] In general, it contains two major steps: extraction and dispersive SPE cleanup. In the extraction step, the method uses a single step buffered 1% acetonitrile (ACN) extraction while simultaneously salting out water from the sample using anhydrous magnesium sulfate (MgSO4) to induce liquid-liquid partitioning. For cleanup, a dispersive solid phase extraction (dispersive SPE) step is employed using a combination of primary secondary amine (PSA) to remove fatty acids as well as other components, and anhydrous MgSO4 to reduce the remaining water in the extract. According to different food matrices, other ingredients may be added in this step, such as graphitized carbon black (GCB) to remove pigments and sterol, or C18 to remove more lipids and waxes. Spinach is considered to be a highly pigmented vegetable since it contains high levels of chlorophyll. Therefore, the dispersive SPE kits with GCB were selected for further cleanup. In these kits, 50 mg of GCB per mL of ACN extracts are added to 50 mg of PSA and 150 mg of MgSO4. GCB adsorbs planar molecules such as pigments and sterols. Therefore, it is helpful in the cleanup of pigmented matrices such as spinach. However, GCB also adsorbs pesticides with planar structure such as carbendazim, and thiabendazole. As a result, this type of dispersive SPE kit is not recommended for use with planar pesticides. This limitation will have a negative impact on the analysis of planar pesticides from pigmented matrices. In the previous Application Note, [4] we discussed the impact of toluene addition to the dispersive SPE tube on the analysis of pesticides in pigmented matrix. This Application Note illustrated that this modification can greatly increase the extraction efficiency for problematic pesticides. GCB was employed for the analysis of planar pesticides in pigmented matrices such as spinach with the addition of toluene. In this study, 13 pesticides were used for evaluating the performance of the Agilent AOAC Buffered Extraction kit (p/n 5982-5755) and SampliQ QuEChERS AOAC Dispersive SPE kits for Pigmented Fruits and Vegetables (p/n 5982-5222 and 5982-5258). With the combination of original and modified dispersive SPE, the
method was validated in terms of recovery and reproducibility. Table 1 shows the chemical and regulatory information for these pesticides in spinach.
Experimental
All reagents and solvents were HPLC or analytical grade. Methanol (MeOH), and toluene were from Honeywell (Muskegon, MI, USA). Acetonitrile (ACN), dimethyl sulfoxide (DMSO) and glacial acetic acid (HAc) were from SigmaAldrich (St Louis, MO, USA). Ammonium acetate (NH4OAc) was from Fisher Chemicals (Fair Lawn, NJ, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard, triphenyl phosphate, (TPP) were purchased from Sigma-Aldrich (St Louis, MO, USA), ChemService (West Chester, PA, USA), Ultra (Kingstown, RI, USA), or AlfaAesar (Ward Hill, MA, USA).

The 1M NH4OAc pH 5 stock solution was made by dissolving 19.27 g NH4OAc powder in 250 mL Milli-Q water. The pH was adjusted to 5 with HAc monitored with a pH meter. The solution was stored at 4 C. A 5 mM NH4OAc solution in 20:80 MeOH/H2O, pH 5, was made by combining 200 mL MeOH and 800 mL Milli-Q water, adding 5 mL of 1 M NH4OAc pH 5 stock solution and mixing well. A 5 mM NH4OAc in ACN solution was prepared by adding 5 mL of 1 M NH4OAc pH 5 stock solution to 1 L ACN, mixing well and sonicating 5 min. A 1% HAc in ACN solution was prepared by adding 10 mL of HAc to 1 L of ACN, and mixing well. Standard and internal standard (IS) stock solutions (2.0 mg/mL for all, except 0.5 mg/mL for carbendazim) were made in MeOH, 0.1% FA in ACN, or DMSO, respectively, and stored at 20 C. Three QC spiking solutions of 1.5, 7.5 and 30 g/mL were made fresh daily in 1:1 ACN/H2O containing 0.1% FA. A 10 g/mL standard spiking solution in 1:1 ACN/H2O containing 0.1% FA was made for the preparation of calibration curves in the matrix blank extract by appropriate dilution. A 15 g/mL IS spiking standard of TPP was made in 1:1 ACN/H2O containing 0.1% FA.
Table 1.

O O P N S H O
Name Acephate
Log P 0.89
pKa 8.35
Structure
20
Carbaryl
Carbamate
2.36
10.4
O
NH O
50
Carbendazim
Benzimidazole
1.48
4.2
H N N
O OCH 3 NH
100
Cyprodinil
Anilinopyrimidine
4.44
N N
H N
500
Imazalil
Imidazole
3.82
6.53
Cl O Cl N N CH2
20
Imidacloprid
Neonicotinoid
0.57
NA
N CI N
NO2
1000
NH
Methamidophos
Organophosphate
0.79
NA
H3CO
O P NH2 SCH3
10
Penconazole
Triazole
3.72
1.51
CI
Cl N N N
50
(Continued)
Table 1.
Pesticides Chemical and Regulatory Information [57] MRLs in apple (ng/g)* 2000
O O O H N
Name Propoxur
Class Carbamate
Log P 0.14
pKa NA
Structure
Pymetrozine
Pyridine
0.19
4.06
N N N N O H N N O H3C H3C S P O S CH3 N NH
600
Thiabendazole
Benzimidazole
2.39
4.73 12.00 0
50
S
Ethoprophos
Organophosphate
2.99
NA
Kresoxim-methyl
Strobilurin
3.4
NA
50
*The MRLs numbers listed in the table are for spinach or other vegetables. They could be different in different commodities.

Agilent 1200 Series HPLC with Diode Array Detector (Agilent Technologies Inc., CA, USA). Agilent 6410 triple quadrupole LC/MS/MS system with Electrospray Ionization (Agilent Technologies Inc., CA, USA). Agilent SampliQ QuEChERS AOAC Extraction kits, p/n 5982-5755, and SampliQ QuEChERS AOAC dispersive SPE kits for Pigmented Fruits and Vegetables, p/n 5982-5222 and 5982-5258 (Agilent Technologies Inc., DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, So. Painfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA)
Sample Preparation
The sample preparation procedure includes sample comminution, extraction/partitioning and dispersive SPE cleanup. It has been described in detail in previous Application Notes. [8] The procedure used in spinach was similar except for the addition of toluene to the dispersive SPE cleanup step. Frozen chopped organic spinach was homogenized thoroughly. A 15 g ( 0.1g) amount of homogenized sample was placed into a 50 mL centrifuge tube. Samples were fortified with appropriate QC spiking solutions (100 L) when necessary, and then 100 L of IS spiking solution (15 g/mL of TPP) were added. After vortexing sample for 30 s, 15 mL of 1% HOAc in ACN was added to each tube using the dispenser. To each tube, an Agilent SampliQ QuEChERS AOAC extraction salt packet (p/n 5982-5755) was added directly. Sample tubes were capped tightly, and hand-shaken vigorously for 1 min. Tubes were centrifuged at 4000 rpm for 5 min. The ACN extracts were separated into two samples for both original and modified dispersive SPE methods. The modified dispersive SPE method follows a different procedure, therefore it is described below in detail. The volume of ACN extracts (~14 mL) was enough for simultaneously processing samples with original and modified dispersive SPE when using 2 mL size dispersive SPE tubes. When 15 mL size tubes were used, 14 mL of ACN extracts from one sample was not enough for processing dispersive SPE by two methods simultaneously. Therefore, another sample was extracted from the beginning. A 1 mL aliquot of the upper ACN layer was transferred into an Agilent SampliQ QuEChERS dispersive SPE 2 mL tube (p/n 5982-5222); or 8 mL aliquot into an Agilent SampliQ QuEChERS dispersive SPE 15 mL tube (p/n 5982-5258). The 2 mL tube contained 50 mg of PSA, 50 mg of GCB and 150 mg of anhydrous MgSO4; while the 15 mL tube contained 400 mg of PSA, 400 mg of GCB and 1200 mg of anhydrous MgSO4. Next, 375 L of toluene were added to the 2 mL tube, or 3 mL of toluene was added to 15 mL tube. The tubes were tightly capped and vortexed for 1 minute. The tubes were vortexed for a few seconds before sample addition, to prevent agglomerates. The 2 mL tubes were centrifuged with a microcentrifuge at 13,000 rpm for 2 min, and the 15 mL tubes in a standard centrifuge at 4000 rpm for 5 min. An 825 L amount
The previous LC/MS/MS method was directly used. [8] HPLC conditions Column: Agilent ZORBAX Solvent Saver Plus Eclipse Plus Phenyl-Hexyl, 3.0 x 150 mm, 3.5 m (p/n 959963-312) 0.3 mL/min 30 C 10 L A: 5 mM NH4OAc, pH 5.0 in 20:80 MeOH/H2O B: 5 mM NH4OAc, pH 5.0 in ACN 1:1:1:1 ACN/MeOH/isopropyl alcohol (IPA)/H2O w/0.2% FA. Flow rate Time %B (mL/min) 0 0.5 8.0 10.0 10.01 13.0 4 min 17 min 20 20 100 100 20 STOP 0.3 0.3 0.3 0.3 0.5
Post run: Total cycle time: MS conditions Positive mode Gas temperature: Gas flow: Nebulizer: Capillary:
350 C 10 L/min 40 psi 4000 V
Instrument Acquisition Data for the Analysis of 13 Pesticides by LC/MS/MS MRM channels (m/z) 1) 184.0 > 94.9 2) 184.0 > 111.0 1) 142.0 > 94.0 2) 142.0 > 124.9 1) 218.1 > 105.0 2) 218.1 > 78.0 1) 192.1 > 160.0 2) 192.1 > 105.0 1) 256.1 > 209.1 2) 256.1 > 175.0 1) 202.1 > 175.0 2) 202.1 > 131.0 1) 210.1 > 111.0 2) 210.1 > 92.9 1) 202.0 > 145.0 2) 202.0 > 115.0 1) 243.1 > 130.9 2) 243.1 > 172.9 1) 297.1 > 158.9 2) 297.1 > 200.9 1) 284.1 > 158.9 2) 284.1 > 172.9 1) 226.1 > 93.0 2) 226.1 > 108.0 1) 314.0 > 222.1 2) 314.0 > 235.0 1) 327.1 > 77.0 2) 327.1 > 151.9 Fragmentor (V) 60 60 115 95 60 110 50 50 80 80 80 120 70 70 CE (V) 3 15 8 8 20 50 18 40 12 18 27 38 12 15 3 40 15 15 22 15 32 32 35 35 10 10 45 45 RT (min) 2.55 2.54 2.97 5.07 5.53 5.65 6.89 7.30 8.50 8.52 8.95 9.23 9.44 9.49
Methamidophos Pymetrozine Carbendazim Imidacloprid Thiabendazole Propoxur Carbaryl Ethoprophos Imazalil Penconazole Cyprodinil Kresoxim methyl TPP (IS) 1) Quantifier transition channel 2) Qualifier transition channel
of extract was then transferred into another tube, and dried by N2 flow. Samples were reconstituted into 600 L of ACN containing 0.1%FA. After vortexing and sonicating, 200 L of extract were transferred into an autosampler vial, and then 800 L of water or other appropriate standard solution (prepared in water) were added. The samples were capped and vortexed thoroughly for LC/MS/MS analysis. Another aliquot of ACN extracts was processed following the original dispersive SPE clean-up procedure. Figure 1 shows the flow chart of the whole extraction procedure (original and modified dispersive SPE) for a spinach sample.
Weigh 15 g spinach sample (0.1 g) in 50 mL centrifuge tube
Spike 100 L of IS and QC spike solution (if necessary), vortex 1 min.
Add 15 mL of 1% HAc in ACN, and SampliQ AOAC QuEChERS extraction kit
Cap and shake vigorously by hand for 1 min, centrifuge at 4000 rpm for 5 min
Centrifuge at 13,000 rpm for 2 min
Dry with N 2 flow at 30C Reconstitute into 600 L of 0.1% FA in ACN
Vortex and sonicate
Transfer certain volume for LC/MS/MS or GC/MS analysis
Transfer certain volume for LC/MS/MS or GC/MS analysis
Figure 1.
Flow chart of the QuEChERS AOAC extraction procedure (original and modified dispersive SPE, 2 mL size) for a spinach sample.

QuEChERS method for pesticide residues analysis provides high-quality results in a fast, easy, inexpensive approach. For the pigmented fruits and vegetables, the addition of GCB in the dispersive SPE tube can improve the removal of pigments and sterols. Toluene was added to increase the extraction efficiency of planar pesticides. Previously it was established that the addition of toluene produces a yellow final sample, indicating that matrix impurities are retained. [4] However, with the powerful selectivity provided by LC/MS/MS, there have not been any chromatographic differences found between the samples processed with the original and modified methods. Figures 2 and 3 show the LC/MS/MS chro-
matograms of matrix blank (IS spiked) and 50 ng/g fortified spinach extract processed by original and modified dispersive SPE method. Four pesticides with planar structure showed significant loss by the original dispersive SPE method. The modified method with toluene addition increased the recovery of those four pesticides by two to three times, from 20% to 40% and 60% to 100%. In addition, the repeatability improved from >15% to <5% RSD. The addition of toluene had no affect on the quantitation results of other pesticides. Therefore, the results from the original method for high recovered pesticides were combined with the results from modified method for planar pesticides. The method was validated in terms of recovery and reproducibility, and the quantitation results are discussed.
Comparison of LC/MS/MS Chromatograms Representing Identical Chromatograms with or without Toluene Addition
10 2 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 10 2 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 11 12
1 1
1 2
2 3
3 4
IS
10
11
12
1 2
2 3
3 4
IS
Figure 2.
LC/MS/MS chromatograms of spinach matrix blank processed by original dispersive SPE (A) and modified dispersive SPE (B). IS: Internal Standard TPP.
Comparison of LC/MS/MS Chromatograms Representing Improved Planar Pesticide Recovery with Toluene Addition
10 2 1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.5 10 2 1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.5 1 1.5 2 1 2 6 3 3 3.5 4 4.5 5 5 5.5 6 6.5 7 7.5 8 8.5 4 8 9 10 11 9 12 13 10 10.5 11 11.5 12 12.5 1 1.5 2 1 2 3 3 3.5 4 4.5 5 4 6 6 6.5 7 7.5 8 8.5 8 9 10 11 9 12 9.5 13 10 10.5 11 11.5 12 12.5
12
23
34
IS
2.5
5.5
1 2
2 3
3 4
IS
2.5
9.5
Figure 3.
LC/MS/MS chromatograms of 50 ng/g fortified spinach sample extracts processed by original dispersive SPE (A) and modified dispersive SPE (B). Peak identification: 1. Methamidophos, 2. Acephate, 3. Pymetrozine, 4. Carbendazim, 5. Imidacloprid 6. Thiabendazole, 7. Propoxur, 8. Carbaryl, 9. Ethoprophos, 10. Imazalil, 11. Penconazole, 12. Cyprodinil, 13. Kresoxim methyl IS: Internal Standard, TPP.

The linear calibration range for all of the pesticides was 5250 ng/g. For samples processed by original and modified methods, the corresponding matrix blanks were used to prepare the calibration curves respectively. Calibration curves, spiked in matrix blanks, were made at levels of 5, 10, 50, 100, 200, and 250 ng/g. The TPP was used as an internal standard at 100 ng/g. The calibration curves were generated by
plotting the relative responses of analytes (peak area of analyte / peak area of IS) to the relative concentration of analytes (concentration of analyte / concentration of IS). The 5 ng/g quantification limits LOQ (5 ppb) established for all of the pesticides was lower than or equal to the MRLs of these pesticides in fruits and vegetables. Table 3 shows the linear regression equation and correlation coefficient (R2) for both 1 mL and 8 mL dispersive SPE.
Table 3. Analytes
Linearity of Pesticides in Spinach Extract 1 mL dispersive-SPE Regression equation Y = 0.2358X 0.0008 Y = 0.0862X 0.0003 Y = 0.2073X 0.0002 Y = 0.8375X + 0.0032 Y = 0.0652X 0.0007 Y = 0.4081X - 0.0008 Y = 1.9253X 0.0042 Y = 0.4243X 0.0013 Y = 0.7859X 0.0012 Y = 0.4586X + 0.0002 Y = 0.1643X 0.0014 Y = 0.3274X 0.0024 Y = 0.1809X 0.0015 R2 0.9976 0.9975 0.9995 0.9915 0.9905 0.9995 0.9995 0.9979 0.9983 0.9954 0.9923 0.9904 0.9975 8 mL dispersive-SPE Regression equation Y = 0.2164X 0.0014 Y = 0.0804X 0.0006 Y = 0.2034X 0.0013 Y = 0.8383X + 0.0002 Y = 0.0620X 0.0011 Y = 0.4102X 0.0011 Y = 1.8253X 0.0037 Y = 0.3993X 0.0019 Y = 0.7420X 0.0012 Y = 0.4229X + 0.0005 Y = 0.1468X 0.0003 Y = 0.3067X 0.0013 Y = 0.1659X 0.0008 R2 0.9983 0.9942 0.9978 0.9982 0.9742 0.9975 0.9996 0.9946 0.9985 0.9903 0.9944 0.9978 0.9928
Methamidophos Acephate Pymetrozine * Carbendazim * Imidacloprid Thiabendazole * Propoxur Carbaryl Ethoprophos Imazalil Penconazole Cyprodinil * Kresoxim methyl
* Results from modified dispersive SPE method.

The recovery and reproducibility were evaluated by spiking pesticides standards in communited spinach sample at levels of 10, 50 and 200 ng/g. These QC samples were quantitated against the matrix spiked calibration curve. The analysis was performed six times at each level. The recovery and reproducibility (shown as RSD) data of 1 mL and 8 mL volume dispersive SPE are shown in Tables 4 and 5, respectively. The results show that the nine pesticides processed with the orig-
inal method resulted in excellent recoveries (average of 97.8% for 1 mL and 103.4% for 8 mL) and precision (average of 3.6% RSD for 1 mL and 4.3% RSD for 8 mL). Although the four pesticides processed with the modified method resulted in lower recoveries (average of 78.5% for 1 mL and 69.7% for 8 mL) but high precision (average of 2.7% RSD for 1 mL and 3.3% RSD for 8 mL). The results from the modified method were much better than the results obtained by original methods. Please refer to previous Application Note [4] for a detailed discussion.
10
Table 4.
Excellent Recovery and Reproducibility of Pesticides in Fortified Spinach with a 1 mL volume, 2 mL Dispersive SPE Tube (p/n 5982-5222) 10 ng/g fortified QC Recovery RSD (n=6) 91.8 93.4 74.0 105.3 98.2 79.0 100.0 110.8 98.8 84.0 103.1 69.1 104.4 4.2 3.3 2.9 4.0 4.5 2.7 1.7 3.2 1.6 3.8 5.4 4.7 4.8 50 ng/g fortified QC Recovery RSD (n=6) 93.3 91.3 71.1 109.1 100.4 76.6 98.1 108.1 98.2 89.6 98.4 62.0 101.2 3.7 5.6 3.2 2.5 3.7 2.3 3.5 1.0 3.3 2.5 3.5 2.9 5.0 200 ng/g fortified QC Recovery RSD (n=6) 93.8 101.9 70.3 88.9 100.0 75.5 93.0 105.1 95.1 89.8 97.2 61.3 102.6 5.7 7.8 2.9 1.7 2.7 1.8 4.0 3.2 3.1 1.7 1.9 1.1 3.0
Analytes Methamidophos Acephate Pymetrozine * Carbendazim * Imidacloprid Thiabendazole * Propoxur Carbaryl Ethoprophos Imazalil Penconazole Cyprodinil * Kresoxim methyl
Table 5.
Excellent Recovery and Reproducibility of Pesticides in Fortified Spinach with a 8 mL volume, 15 mL Dispersive SPE Tube (p/n 5982-5258) 10 ng/g fortified QC Recovery RSD (n=6) 98.6 95.5 62.4 95.7 112.7 58.0 104.9 116.9 105.3 86.3 103.5 63.1 111.2 3.8 8.9 4.3 1.6 4.2 3.5 1.4 2.2 2.5 3.9 10.4 2.8 4.5 50 ng/g fortified QC Recovery RSD (n=6) 94.2 91.5 53.9 98.6 107.6 62.1 103.3 114.6 105.7 94.9 106.9 60.6 106.6 7.1 6.3 3.4 1.9 7.7 3.3 3.7 2.4 2.8 4.3 3.6 4.8 3.2 200 ng/g fortified QC Recovery RSD (n=6) 97.8 105.6 59.3 93.3 110.4 66.8 99.0 110.8 103.0 93.9 99.2 62.7 112.0 2.9 5.7 5.4 2.9 3.7 2.8 3.3 2.1 2.3 3.4 6.4 2.9 3.0
11
Figure 4 shows the recovery and precision results obtained by 1 mL and 8 mL volume dispersive SPE. To simplify the comparison, the average recovery and precision of three fortification concentrations were used for all pesticides. The results of two dispersive SPE cleanup approaches appeared to be independent of volume used. Apparently, both approaches provided efficient and similar sample cleanup, and thus generated relatively equivalent results. However, if 8 mL size dis-
persive SPE volume is used, two duplicated extractions must be performed initially to complete both original and modified dispersive SPE. If 1 mL size dispersive SPE is used, only one extraction is needed to provide enough volume to perform both original and modified dispersive SPE simultaneously. This is more cost effective saving time, sample amount, and labor. The extractions can be performed according to the users requirements and regulations.
Exceptional Recoveries and Precision for 1 and 8 mL volume Dispersive SPE

120
1 mL
8 mL
100
80
Recovery (%)
60
40
20
0 Methamidophos Ethoprophos Propoxur Acephate Carbaryl Carbendazim Thiabendazole Penconazole Pymetrozine Imidacloprid Cyprodinil Kresoxim methyl Imazalil
Figure 4.
The recovery and precision results for 1 mL dispersive SPE and 8 mL dispersive SPE.
12
Conclusions
Agilent SampliQ QuEChERS AOAC buffered extraction kits and dispersive SPE kits for pigmented fruits and vegetables provide a simple, fast and effective method for the purification of representative pesticides in spinach. The modified dispersive SPE method with the addition of toluene provides a very useful way to limit the loss of planar pesticides caused by GCB in dispersive SPE. The recovery and reproducibility of this method, based on matrix spiked standards, were acceptable for multiclass, multi-residue pesticide determination in spinach. The impurities and matrix effects from spinach were minimal and did not interfere with the quantitation of any target compound. As the selected pesticides represented a broad variety of different classes and properties, the Agilent SampliQ QuEChERS AOAC Buffered Extraction and Dispersive kits for Pigmented Fruits and Vegetables can be used for other pesticides in other similar pigmented matrices.
3.
S. J. Lehotay, et. al., Determination of Pesticide Residues in Foods by Acetonitrile Extraction and Partitioning with Magnesium Sulfate: Collaborative Study, J. AOAC Int., 2007, 90, 485-520. L. Zhao, J. Stevens, The Impact of Toluene Addition in the Dispersive SPE to the Analysis of Pesticides in Spinach Using Agilent SampliQ QuEChERS AOAC Kits for Highly Pigmented Fruits and Vegetables, Agilent Technologies publication, EN xxxx-xxxx. http://sitem.herts.ac.uk/aeru/footprint/en/index.htm http://www.m5.ws001.squarestart.ne.jp/foundation/ search.html http://www.mrldatabase.com/?selectvetdrug=0 L. Zhao, D. Schultz, J. Stevens, Evaluation of the QuEChERS AOAC Sample Preparation Kit for the Analysis of Pesticide Residues in Apples with LC/MS/MS Detection. Agilent Technologies publication 5990-3937EN.
4.
5. 6. 7. 8.
References
1. M. Anastassiades, S. J. Lehotay, Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and "Dispersive Solid-Phase Extraction" for the Determination of Pesticide Residues in Produce, J. AOAC Int., 2003, 86, 412- 431. S. J. Lehotay, et al., Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005, 88, 615-629.

2.
13
Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS EN Kit by LC/MS/MS Detection
Application Note
Food Safety
Authors
Limian Zhao, Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19808 USA
Abstract
This application note describes the use of a quick, easy, cheap, effective, rugged, and safe (QuEChERS) EN sample preparation approach for extraction and cleanup of 13-pesticide residues representing various classes in spinach. Because spinach is considered a highly pigmented matrix, the EN dispersive SPE kit for highly pigmented fruits and vegetables is selected. Graphitized carbon black (GCB) in the amount of 7.5 mg/mL of ACN extract is added to the kit. The target pesticides in the spinach extracts are then determined by liquid chromatography coupled to an electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) operating in positive ion multiple reaction monitoring (MRM) mode. GCB is reported to have a significantly negative impact on the extraction of pesticides with planar structure. However, with the small amount of GCB addition in the EN dispersive SPE kit, our results show that the impact of GCB on planar pesticides is negligible and acceptable quantitation results are obtained. The 5 ng/g limit of quantitation (LOQ) for pesticides in spinach shown in this application is well below the maximum residue limits (MRLs). The spiking levels for the recovery experiments are 10, 50, and 200 ng/g. Mean recoveries range between 60 and 99% (85.4% on average), with an RSD below 11% (5.5% on average).
Introduction
The EN QuEChERS method has been widely employed in the analysis of pesticides in food, especially in Europe. [1-2] The method uses acetonitrile extraction, followed by salting out water from the sample using anhydrous magnesium sulfate (MgSO4), NaCl and buffering citrate salts to induce liquid-liquid partitioning. For cleanup, a dispersive solid phase extraction (dispersive SPE) is conducted using a combination of primary secondary amine (PSA) to remove fatty acids from among other components, and anhydrous MgSO4 to reduce the remaining water in the extract. According to different food matrices, other ingredients may be added in this step, such as graphitized carbon black (GCB) to remove pigments and sterol, or C18 to remove more lipids and waxes. Spinach is considered to be a highly pigmented vegetable since it contains high levels of chlorophyll. Therefore, the EN dispersive SPE kits for highly pigmented commodities were selected for this application. In these kits, besides 25 mg of PSA and 150 mg of MgSO4, 7.5 mg of GCB is added per mL of ACN extracts. GCB adsorbs planar molecules like pigments and sterols; hence it is very helpful in cleaning up pigmented matrices like spinach. The efficiency of cleanup is dependent upon the amount of GCB used. The more GCB used, the more planar molecules are absorbed, and therefore, a cleaner sample matrix is obtained. The main difference between the EN method and AOAC method for cleaning up the highly pigmented matrix is the amount of GCB used in the dispersive SPE step. Instead of the relatively high amount of GCB used in AOAC method (50 mg of GCB per mL of ACN extracts), a much lower amount of GCB was used in the EN methods (2.5 mg of GCB per mL of ACN extracts for pigmented produce, or 7.5 mg of GCB per mL of ACN extracts for highly pigmented produce). The GCB impacted the extraction of planar pesticides differently, depending upon the method used. The AOAC method generated much cleaner final sample matrix but caused significant loss of planar pesticides; the EN method, on the contrary, caused little to no loss of planar pesticides but generated a more complicated sample matrix. Previously, we described that a modified AOAC method with toluene addition in the dispersive SPE step greatly increased the extraction efficiency of planar pesticides in a pigmented matrix such as spinach. [3] Subsequently, we demonstrated the performance of SampliQ QuEChERS AOAC kit for the analysis of pesticides in spinach using combination of the modified (with toluene addition) and the original AOAC
method (without toluene addition). [4, 5] In this study, 13 pesticides were used for evaluating the performance of the Agilent EN Buffered Extraction kit (p/n 5982-5650) and SampliQ QuEChERS EN Dispersive SPE kits for Highly Pigmented Fruits and Vegetables (p/n 5982-5321 and 59825356). The method was validated in terms of recovery and reproducibility. Table 1 shows the chemical and regulatory information for these pesticides in spinach.
Experimental
All reagents and solvents were HPLC or analytical grade. Methanol (MeOH), and toluene were from Honeywell (Muskegon, MI, USA). Acetonitrile (ACN), dimethyl sulfoxide (DMSO) and glacial acetic acid (HAc) were from SigmaAldrich (St Louis, MO, USA). Ammonium acetate (NH4OAc) was from Fisher Chemicals (Fair Lawn, NJ, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard (triphenyl phosphate, TPP) were purchased from Sigma-Aldrich (St Louis, MO, USA), ChemService (West Chester, PA, USA), Ultra (Kingstown, RI, USA), or AlfaAesar (Ward Hill, MA, USA).

A 1 M ammonium acetate pH 5 stock solution was made by dissolving 19.27 g NH4OAc powder in 250 mL Milli-Q water, pH adjusted to 5 with acetic acid monitored with a pH meter. The solution was stored at 4 C. A 5 mM ammonium acetate in 20:80 MeOH/H2O solution, pH 5, was made by combining 200 mL MeOH and 800 mL Milli-Q water, adding 5 mL of 1 M ammonium acetate pH 5 stock solution and mixing well. A 5 mM ammonium acetate in ACN was prepared by adding 5 mL of 1 M ammonium acetate pH 5 stock solution to 1 L ACN, mixing well and sonicating 5 min. A 1% formic acid in ACN solution was prepared by adding 1 mL of formic acid to 100 mL of ACN, and mixing well. Standard and internal standard (IS) stock solutions (2.0 mg/mL for all except 0.5 mg/mL for carbendazim) were made in MeOH, 0.1% FA in ACN, or DMSO, respectively, and stored at -20 C. Three QC spiking solutions of 1, 5, and 20 g/mL, were made fresh daily in 1:1 ACN/H2O with 0.1% FA. A 10 g/mL standard spiking solution in 1:1 ACN/H2O with 0.1% FA was made also for the preparation of a calibration curve in the matrix blank extract by appropriate dilution. A 15 g/mL of TPP in 1:1 ACN/H2O with 0.1% FA was made as an IS spiking solution.
Table 1.
Pesticides Chemical and Regulatory Information [68] MRLs in spinach (ng/g)*

O O P N S H O
Name Acephate
Log P 0.89
pKa 8.35
Structure
20
Carbaryl
Carbamate
2.36
10.4
O
NH O
50
Carbendazim
Benzimidazole
1.48
4.2
H N N
O OCH 3 NH
100
Cyprodinil
Anilinopyrimidine
4.44
N N
H N
500
Imazalil
Imidazole
3.82
6.53
Cl O Cl N N CH2
20
Imidacloprid
Neonicotinoid
0.57
NA
N CI N
NO2
1000
NH
Methamidophos
Organophosphate
-0.79
NA
CH3O
O P NH2 SCH3
10
Penconazole
Triazole
3.72
1.51
CI
Cl N N N
50
(Continued)
Table 1.
Pesticides Chemical and Regulatory Information [68] MRLs in spinach (ng/g)* 2000
O O O H N
Name Propoxur
Class Carbamate
Log P 0.14
pKa NA
Structure
Pymetrozine
Pyridine
0.19
4.06
N N N N O H N N N O H3C H3C S P O S CH3 NH
600
Thiabendazole
Benzimidazole
2.39
4.73 12.00 0
50
S
Ethoprophos
Organophosphate
2.99
NA
Kresoxim-methyl
Strobilurin
3.4
NA
CH3 O CH3O N OCH 3
50
*The MRLs numbers list in the table are for spinach or other vegetables. They could be higher in different commodities.

Agilent 1200 Series HPLC with Diode Array Detector (Agilent Technologies Inc., CA, USA). Agilent 6410 triple quadrupole LC/MS system with Electrospray Ionization (Agilent Technologies Inc., CA, USA). Agilent SampliQ QuEChERS EN Extraction kits, p/n 59825650, and SampliQ QuEChERS EN dispersive SPE kits for Highly Pigmented Fruits and Vegetables, p/n 5982-5321 and 5982-5356 (Agilent Technologies Inc., DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, So. Painfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA)
Sample Preparation
The sample preparation procedure includes sample comminution, extraction/partitioning and dispersive SPE cleanup. It was described in detail in the previous application notes. [9] The procedure used in spinach was similar to the one used in apple, except that the dispersive SPE kit was for highly pigmented produce rather than general fruits and vegetables. Briefly, the frozen chopped organic spinach was homogenized thoroughly. A 10 g (0.1g) of homogenized sample was placed into a 50 mL centrifuge tube. Samples were fortified with appropriate QC spiking solutions (100 L) when necessary, and then 66.7 L of IS spiking solution (15 g/mL of TPP). After vortexing sample for 30 s, 10 mL of ACN was added to each tube using the dispenser. Tubes were then capped and shaken by hand for 1 min. To each tube, an Agilent SampliQ QuEChERS EN extraction salt packet (p/n 5982-5650), containing 4 g anhydrous MgSO4, 1 g NaCl, 1 g Na3Citrate, and 0.5 g Na2HCitrate sesquihydrate, was added directly. Sample tubes were capped tightly, and hand-shaken vigorously for 1 min. Tubes were centrifuged at 4000 rpm for 5 min. A 1 mL aliquot of upper ACN layer was transferred into Agilent SampliQ QuEChERS EN dispersive SPE 2 mL tube (p/n 5982-5321); or 6 mL aliquot into Agilent SampliQ QuEChERS EN dispersive SPE 15 mL tube (p/n 5982-5356). The 2 mL tube contains 25 mg of PSA, 150 mg of anhydrous MgSO4 and 7.5 mg of GCB; while the 15 mL tube contains 150 mg of PSA, 900 mg of anhydrous MgSO4 and 45 mg of GCB. The tubes were capped tightly and vortexed for 1 min. The 2 mL tubes were centrifuged with a micro-centrifuge at 13,000 rpm for 2 min, and the 15 mL tubes in a standard centrifuge at 4000 rpm for 5 min. A 200 L aliquot of extract was transferred into an autosampler vial. An aliquot of 10 L 1% FA in ACN was added immediately. Then 800 L of water or appropriate standard solutions (prepared in water) were added. The samples were capped and vortexed thoroughly for LC/MS/MS analysis.
The previous LC/MS/MS method was used. [9] HPLC conditions Column: Agilent ZORBAX Solvent Saver Plus Eclipse Plus Phenyl-Hexyl, 3.0 x 150 mm, 3.5 m (p/n 959963-312) 0.3 mL/min 30 C 10 L A, 5 mM ammonium acetate, pH 5.0 in 20:80 MeOH/H2O B, 5 mM ammonium acetate, pH 5.0 in ACN 1:1:1:1 ACN/MeOH/IPA/H2O w/0.2% FA. Flow rate Time % Acetonitrile (mL/min) 0 0.5 8.0 10.0 13.0 4 min 17 min 20 20 100 100 STOP 0.3 0.3 0.3 0.3
Post run: Total cycle time: MS conditions Positive mode Gas temp.: Gas flow: Nebulizer: Capillary:
350 C 10 L/min 40 Psi 4000 V
Instrument Acquisition Data Used for the Analysis of 13 Pesticides by LC/MS/MS MRM channels (m/z) 1) 184.0 > 94.9 2) 184.0 > 111.0 1) 142.0 > 94.0 2) 142.0 > 124.9 1) 218.1 > 105.0 2) 218.1 > 78.0 1) 192.1 > 160.0 2) 192.1 > 105.0 1) 256.1 > 209.1 2) 256.1 > 175.0 1) 202.1 > 175.0 2) 202.1 > 131.0 1) 210.1 > 111.0 2) 210.1 > 92.9 1) 202.0 > 145.0 2) 202.0 > 115.0 1) 243.1 > 130.9 2) 243.1 > 172.9 1) 297.1 > 158.9 2) 297.1 > 200.9 1) 284.1 > 158.9 2) 284.1 > 172.9 1) 226.1 > 93.0 2) 226.1 > 108.0 1) 314.0 > 222.1 2) 314.0 > 235.0 1) 327.1 > 77.0 2) 327.1 > 151.9 Fragmentor (V) 60 60 115 95 60 110 50 50 80 80 80 120 70 70 CE (V) 3 15 8 8 20 50 18 40 12 18 27 38 12 15 3 40 15 15 22 15 32 32 35 35 10 10 45 45 RT (min) 2.55 2.54 2.97 5.07 5.53 5.65 6.89 7.30 8.50 8.52 8.95 9.23 9.44 9.49
Methamidophos Pymetrozine Carbendazim Imidacloprid Thiabendazole Propoxur Carbaryl Ethoprophos Imazalil Penconazole Cyprodinil Kresoxim methyl TPP (IS) 1) Quantifier transition channel 2) Qualifier transition channel

The QuEChERS method for pesticide residue analysis provides high-quality results in a fast, easy, inexpensive approach. For the pigmented fruits and vegetables, the addition of GCB in the dispersive SPE tube can improve the removal of pigments and sterols. The cleaning efficiency of the method with GCB is related to the amount of GCB added. The more GCB used, the cleaner the matrix after treatment and less matrix interferences remaining in the final sample. Since GCB can also cause the removal of planar pesticides during the extraction procedure, smaller amounts of GCB used in the EN dispersive SPE step has less of an effect on the planar pesticides. Compared to the AOAC method, the EN method for pigmented produce uses much less GCB in the dispersive SPE step. For normal pigmented commodities like carrots and romaine lettuce, 2.5 mg of GCB can be used per mL of ACN extract;
6
and for highly pigmented commodities like spinach or red sweet pepper, 7.5 mg of GCB can be used per mL of ACN extract. [1] According to the recommendation, the EN dispersive SPE kit for highly pigmented products was used for spinach in our study. Given the highly pigmented kit, the amount of GCB used in the EN method is still much lower than that used in AOAC method, which is 50 mg of GCB per mL of ACN extract. Therefore, visually, the efficiency of matrix cleanup provided by the EN method was much weaker than that provided by AOAC method. The final sample processed by EN method still appeared dark green in color; while the previous final sample processed by AOAC method showed almost colorless transparency. The matrix blank differences are also shown in the UV chromatogram at = 254 nm shown in Figure 1. More interference peaks appear in the matrix blank processed by the EN method. Also more impurities may have accumulated
in the column or ionization source, which can have negative effects on the column and MS instrument. However, with the powerful selectivity provided by LC/MS/MS, the MRM chromatogram of matrix blank did not show any interference peaks to the target analytes. Figure 2 shows the LC/MS/MS chromatograms of matrix blank (IS spiked) and 50 ng/g fortified spinach extract processed by EN dispersive SPE method. Four pesticides including Carbendazim, Thiabendazole, Cyprodinil, and Pymetrozine, with planar structure showed significant loss by the original AOAC dispersive SPE method. In addition, the modified method with toluene in the disper-
sive SPE step increased the extraction efficiency. [3,4] In order to investigate the impact of GCB on the planar pesticides, a comparison experiment with and without toluene addition in the dispersive SPE step for spinach samples fortified with the same level of pesticide standard (50 ng/g) was performed. The results showed little to no loss of planar pesticides caused by the small amount of GCB used in the EN method, and no significant improvement obtained by the addition of toluene. Therefore, the original EN method was employed for subsequent experiments. The method was validated in terms of recovery and reproducibility, and the quantitation results are discussed subsequently.
Differences Observed between the AOAC and EN Method, relative to GCB content
10 2 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 DAD1 - A:Sig=250,4 1mL zero blank-organic 2.d
0.5
1.5
2.5
3.5
4.5
5.5
6.5
7.5
8.5
9.5
10 10.5 11 11.5 12 12.5
Response units (%) vs. acquisition time (min) 10 2 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 12.5 Response units (%) vs. acquisition time (min) DAD1 - A:Sig=250,4 1mL zero - 1.d
Figure 1.
UV chromatogram ( = 254nm) of spinach matrix blank processed by AOAC method (A) and EN method (B).
LC/MS/MS Selectivity Observed with the EN QuEChERS Approach to Highly Pigmented Fruits and Vegetables
10 2 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 12.5
IS
Response units (%) vs. acquisition time (min) 10 2 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.5 1 1.5 2
1 2
2 3
3 4
IS
9 4 2 1
2.5
8 5, 6 10 11 12 13
6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 12.5
3
3 3.5 4 4.5 5
5.5
Response units (%) vs. acquisition time (min)
Figure 2.
MRM chromatograms of spinach matrix blank (A) and 50 ng/g fortified sample (B) processed by EN method. Peak identification: 1. Methamidophos, 2. Acephate, 3. Pymetrozine, 4. Carbendazim, 5. Imidacloprid 6. Thiabendazole, 7. Propoxur, 8. Carbaryl, 9. Ethoprophos, 10. Imazalil, 11. Penconazole, 12. Cyprodinil, 13. Kresoxim methyl IS: Internal Standard, TPP.
Linearity and limit of quantification (LOQ)

The linear calibration range for all of the pesticides tested was 5250 ng/g. Calibration curves, spiked in matrix blanks, were made at levels of 5, 10, 50, 100, 200, and 250 ng/g. The TPP was used as an internal standard at 100 ng/g. The calibration curves were generated by plotting the relative responses of analytes (peak area of analyte/peak area of IS) to the relative concentration of analytes (concentration of analyte/concentration of IS). The 5 ng/g quantification limits LOQ (5 ppb) established for all of the pesticides is lower than the MRLs of these pesticides in fruits and vegetables. Table 3 shows the linear regression equation and correlation coefficient (R2) for both 1 mL and 6 mL dispersive SPE.
Table 3. Analytes
Linearity of Pesticides in Spinach Extract 1 mL dispersive SPE Regression equation Y = 0.2220X + 0.0005 Y = 0.0814X + 0.0008 Y = 0.2063X + 0.0009 Y = 0.9015X + 0.0164 Y = 0.0630X + 0.0001 Y = 0.3028X + 0.0059 Y = 1.3721X + 0.0018 Y = 0.3459X + 0.0009 Y = 0.7588X - 0.0011 Y = 0.4644X + 0.0007 Y = 0.1647X 0.0010 Y = 0.2575X + 0.0010 Y = 0.1175X 0.0003 R2 0.9950 0.9972 0.9559 0.9945 0.9814 0.9539 0.9983 0.9968 0.9979 0.9889 0.9937 0.9884 0.9976 6 mL dispersive SPE Regression equation Y = 0.2244X + 0.0003 Y = 0.0797X + 0.0005 Y = 0.1544X - 0.0006 Y = 0.8526X + 0.0008 Y = 0.0682X - 0.0002 Y = 0.2315X + 0.0007 Y = 1.3304X + 0.0003 Y = 0.3224X 0.0003 Y = 0.7211X 0.0023 Y = 0.4203X + 0.0002 Y = 0.1595X 0.0008 Y = 0.2272X + 0.0007 Y = 0.1779X 0.0008 R2 0.9893 0.9974 0.9946 0.9917 0.9952 0.9968 0.9981 0.9963 0.9984 0.9990 0.9979 0.9987 0.9962
Methamidophos Acephate Pymetrozine Carbendazim Imidacloprid Thiabendazole Propoxur Carbaryl Ethoprophos Imazalil Penconazole Cyprodinil Kresoxim methyl

The recovery and reproducibility were evaluated by spiking pesticides standards in comminuted spinach sample at levels of 10, 50 and 200 ng/g. These QC samples were quantitated against the matrix spiked calibration curve. The analysis was performed in replicates of six at each level. The recovery and reproducibility (shown as RSD) data of 1 mL and 6 mL dispersive SPE are shown in Tables 4 and Table 5, respectively. It can be seen from the results that the nine pesticides with non-planar structure give excellent recoveries (average of 90.4% for 1 mL and 94.3% for 6 mL) and precision (average of 4.7% RSD for 1 mL and 5.3% RSD for 6 mL). The four pesticides with planar structure give lower but still acceptable
recovery (average of 71.8% for 1 mL and 79.8% for 6 mL) but good precision (average of 5.8% RSD for 1 mL and 4.8% RSD for 6 mL). The impact of GCB on planar pesticides is visible and varies with different compounds. Cyprodinil gave excellent recovery and precision. Carbendazim gave excellent recovery and precision for low and mid level QCs, but poorer recovery for high level QC. Pymetrozine and thiabendazole gave lower recovery but still acceptable precision. The data in Table 6 show that the results of planar pesticides generated by EN method and AOAC modified method (with toluene addition) are comparable.
Table 4.
Recovery and Reproducibility of Pesticides in Fortified Spinach with 1 mL Dispersive SPE Tube (p/n 5982-5321) 10 ng/g fortified QC Recovery RSD (n=6) 85.5 83.7 60.0 78.0 96.5 64.3 93.7 93.8 97.1 86.6 107.8 89.6 101.5 4.1 8.3 6.4 7.1 6.2 7.0 4.7 5.6 4.6 5.7 4.9 4.4 3.8 50 ng/g fortified QC Recovery RSD (n=6) 84.4 84.6 57.8 87.7 91.1 71.5 92.0 89.4 89.8 80.6 94.4 88.6 94.6 3.8 5.9 4.7 3.9 4.6 6.5 4.1 3.6 2.6 4.9 3.2 4.5 1.4 200 ng/g fortified QC Recovery RSD (n=6) 87.5 91.6 61.4 49.8 94.6 71.5 86.7 91.4 83.7 84.2 81.2 80.8 92.8 6.2 5.8 9.1 6.8 4.6 5.8 4.3 4.1 4.1 4.8 3.7 3.9 3.8
* Pesticides with planar structure.
Table 5.
Recovery and Reproducibility of Pesticides in Fortified Spinach with 6 mL Dispersive SPE Tube (p/n 5982-5356) 10 ng/g fortified QC Recovery RSD (n=6) 85.0 88.6 68.7 94.0 102.0 77.2 98.2 98.5 102.3 88.8 104.5 101.5 99.7 8.3 5.1 3.7 5.4 8.9 4.4 5.7 3.6 6.0 6.4 2.5 4.2 6.1 50 ng/g fortified QC Recovery RSD (n=6) 87.7 84.6 65.7 91.4 85.4 77.6 96.3 94.0 95.3 86.8 96.4 92.2 97.4 2.7 3.1 1.5 2.7 6.1 2.4 1.8 1.7 1.7 2.8 2.0 2.4 1.6 200 ng/g fortified QC Recovery RSD (n=6) 95.0 94.6 71.9 53.5 100.1 79.2 93.9 97.4 91.0 93.5 84.6 86.8 95.3 9.4 9.3 10.8 9.3 7.7 9.7 7.2 7.2 6.8 7.7 5.5 7.6 6.9
* Pesticides with planar structure.
10
Table 6.
Results Comparison of Planar Pesticides Generated by EN Method and Modified AOAC Method (With Toluene Addition)* EN method for highly pigmented matrix Mean recovery (%) Mean RSD (%) 75.7 89.9 64.3 73.2 5.9 4.5 6.0 4.9 Modified AOAC method by toluene addition Mean recovery (%) Mean RSD (%) 98.5 63.1 65.2 69.7 2.5 3.2 3.7 2.7
Analytes Carbendazim Cyprodinil Pymetrozine Thiabendazole
*The data can be found in reference [4].
Conclusions
Agilent SampliQ QuEChERS EN buffered extraction kits and dispersive SPE kits for highly pigmented fruits and vegetables provide a simple, fast and effective method for the purification of representative pesticides in spinach. The small amount of GCB used in dispersive SPE does not impact the extraction of planar pesticides significantly, which makes the extraction procedure in this highly pigmented matrix as simple as the one used in general fruit and vegetables. The recovery and reproducibility, based on matrix spiked standards, are acceptable for multiclass, multi-residue pesticide determination in spinach. However, the final extract matrix contains more impurities, which may result in more negative impacts on the column and MS instrument. The selected pesticides represent a broad variety of different classes and properties; therefore, the Agilent SampliQ QuEChERS EN Buffered Extraction and Dispersive kits for Highly Pigmented Fruits and Vegetables can be used for other pesticides in similar highly pigmented matrices.
2.
P. Pay, M. Anastassiades; Analysis of Pesticide Residues Using the Quick Easy Cheap Effective Rugged and Safe (QuEChERS) Pesticide Multiresidue Method in Combination with gas and Liquid Chromatography and Tandem Mass Spectrometric Detection, Anal Bioanal Chem., 2007, 389, 1697-1714. L. Zhao, J. Stevens, Optimizing Recoveries of Planar Pesticides in Spinach Using Toluene and Agilent SampliQ AOAC QuEChERS Kits with Graphized Carbon, Agilent Technologies publication 5990-4247EN. L. Zhao, J. Stevens, Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS AOAC Kit by LC/MS/MS Detection. Agilent Technologies publication 5990-4248EN. L. Zhao, J. Stevens, Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS AOAC Kit by GC/MS. Agilent Technologies publication 5990-4305EN. http://sitem.herts.ac.uk/aeru/footprint/en/index.htm http://www.m5.ws001.squarestart.ne.jp/foundation/ search.html http://www.mrldatabase.com/?selectvetdrug=0
3.
4.
5.
6. 7.
References
1. European Committee for Standardization/Technical Committee CEN/TC 275 (2007), Foods of plant origin: Determination of pesticide residues using GC-MS and/or LC-MS/MS following acetonitrile extraction/partitioning and cleanup by dispersive SPE-QuEChERS method. European Committee for Standardization, Brussels 8.

11
Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS AOAC Kits by GC/MS
Application Note
Food Safety
Authors
Limian Zhao, Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19809 USA
Abstract
This application note describes the use of a quick, easy, cheap, effective, rugged, and safe (QuEChERS) AOAC sample preparation approach for extraction and cleanup of 18 GC-amenable multiple pesticide class residues in spinach. The method employed involves initial extraction in a buffered aqueous/acetonitrile system, an extraction/ partitioning step after the addition of salt, and a cleanup step using dispersive solid phase extraction (dispersive SPE). In order to address the significant loss of planar pesticides caused by graphitized carbon black (GCB) in dispersive SPE, a modified method with addition of toluene was employed for the planar pesticides. The target pesticides in the spinach extracts were then analyzed by gas chromatography/mass spectrometry (GC/MS) operating in selective ion monitoring (SIM) mode. The method was validated in terms of recovery and reproducibility. The limit of quantitation (LOQ) for most pesticides is 10 ng/g; however folpet has an LOQ of 50 ng/g in spinach. This application, employing SampliQ QuEChERS kits, produced results well below the maximum residue limits (MRLs) for all pesticides screened. The spiked levels for the recovery experiments were 10, 50, and 200 ng/g.
Introduction
The AOAC QuEChERS method has been widely applied for the analysis of pesticides in food since it was introduced by USDA scientists. [1-3] In summary, the method uses a singlestep buffered acetonitrile (1% HAc) extraction while simultaneously salting out water from the sample using anhydrous magnesium sulfate (MgSO4) to induce liquid-liquid partitioning. For cleanup, a dispersive SPE step is employed using a combination of primary secondary amine (PSA) to remove fatty acids as well as other components, and anhydrous MgSO4 to reduce the remaining water in the extract. After mixing and centrifugation, the upper layer is ready for analysis. According to different food matrices, other ingredients may be added in this step, such as graphitized carbon black (GCB) to remove pigments and sterol, or C18 to remove more lipids and waxes. Spinach is considered to be a highly pigmented matrix since it contains large amounts of chloryophyll. Therefore, the dispersive SPE kits with GCB were selected for further clean-up. GCB adsorbs planar molecules such as pigments and sterols; therefore it is very helpful in cleaning-up pigmented matrix like spinach. However, GCB also adsorbs pesticides with planar structure, such as carbendazim, chlorothalonil, and coumaphos. As a result, this kind of dispersive SPE kit is not recommended for the analysis of planar pesticides. Previously, we discussed the impact of toluene addition to the dispersive SPE tube on the analysis of pesticides in pigmented matrices [4]. It turned out that this modification can greatly increase the extraction efficiency of those problematic pesticides. With the combination of the original (w/o toluene) and modified (w/toluene) dispersive SPE, the performance of SampliQ AOAC Buffered Extraction Kits and SampliQ AOAC Dispersive SPE kits for pigmented produce was demonstrated to be excellent for the analysis of LC amenable pesticides in spinach. [5] In this study, the performance of the SampliQ AOAC Buffered Extraction kit (p/n 5982-5755) and SampliQ AOAC DispersiveSPE kits for Pigmented Fruits and Vegetables (p/n 5982-5222 and 5982-5258) was evaluated for the extraction of volatile and semi-volatile pesticides. Analysis was performed by GC/MS. Seventeen GC-amenable pesticides were selected which represent multiple classes, including non-polar organochlorine pesticides (OCs), certain organophosphorus pesticides (OPs) and organonitrogen pesticides (ONs). Table 1 shows the chemical and regulatory information for these pesticides in spinach.
Experimental
All reagents and solvents were high-performance liquid chromatography (HPLC) or analytical grade. Methanol (MeOH) and toluene were from Honeywell (Muskegon, MI, USA), acetonitrile (ACN) and glacial acetic acid (HAc) were from SigmaAldrich (St Louis, MO, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard (triphenyl phosphate, TPP) were purchased from Sigma-Aldrich (St Louis, MO, USA), Chem Service (West Chester, PA, USA), or Ultra Scientific (North Kingstown, RI, USA).

A 1% acetic acid in ACN solution was prepared by adding 10 mL of HAc to 1 L of ACN. Standard and internal standard (IS) stock solutions (2 mg/mL) were made in MeOH, respectively, and stored at 20 C. Three QC spiking solutions of 1.5, 7.5 and 30 g/mL were made fresh daily in 1:1 ACN/H2O containing 0.1% FA. A 2.5 g/mL standard solution in ACN containing 0.1% FA was used to prepare the calibration curves in the matrix blank extract by appropriate dilution. A 15 g/mL of TPP spiking solution in 1:1 ACN/H2O containing 0.1% FA was used as the internal spiking standard (IS).

Agilent Gas Chromatograph (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent 5975C Mass Spectrometer (Agilent Technologies Inc., Santa Clara, CA, USA). SampliQ QuEChERS AOAC Extraction kits, p/n 5982-5755 (Agilent Technologies Inc., Wilmington, DE, USA). SampliQ QuEChERS AOAC dispersive SPE kits for Pigmented Fruits and Vegetables, p/n 5982-5222 and 5982-5258 (Agilent Technologies Inc., Wilmington, DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, So Painfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA)
Table 1.
Pesticides Chemical and Regulatory Information [69] MRLs in spinach (ng/g)*
Name s-Phenylphenol
Category Phenol
Log P 3.18
pKa 9.4
Structure
OH
2000
Dichlorvos
Organophosphate
1.9
NA
O Cl Cl O P O O
10
Lindane
Organochlorine
3.69
NA
Cl Cl Cl Cl Cl Cl
10
Diazinon
Organophosphate
3.69
2.6
100
N O
S P O O
Chlorothalonil
Chloronitrile
2.94
NA
N Cl Cl Cl Cl
4000
N
30
Chlorpyrifosmethyl
Organophosphate
4.00
NA
Cl N Cl Cl
S O
O O
CH3 CH3
Dichlorobenzophenone
Organochlorine
4.44
NA
O Cl C Cl N O H3C O P S O CH3 Cl
500
Chlorpyrifos
Organophosphate
4.7
NA
Cl Cl
10
(Continued)
Table 1. Name
Pesticides Chemical and Regulatory Information [69] Category Organochlorine Log P 5.83 pKa NA Structure MRLs in spinach (ng/g)*
Heptachlor epoxide
Cl
Cl
Cl Cl Cl
30
Cl Cl O
Folpet Phthalimide 3.02 NA
O N S Cl O Cl
Cl Cl Cl
Cl Cl Cl Cl
2000
Cl
Chlordane
2.78
NA
20
Cl
DDE
Organochlorine
6.55
NA
Cl
Cl
50
Cl
Dieldrin Chlorinated hydrocarbon 3.7 NA
Cl
Cl Cl Cl Cl
10
Cl
Cl
300
O
Ethion
Organophosphate
5.07
NA
O
S P O
S P O
(Continued)
Table 1. Name
Pesticides Chemical and Regulatory Information [69] Category Log P pKa Structure MRLs in spinach (ng/g)*
Endosulfan sulfate
Organochlorine
3.13
NA
CI Cl Cl Cl
Cl Cl O O S O O
50
Permethrins
Pyrethroid
6.1
NA
50
Cl Cl O
Coumaphos Organothio phosphate 3.86 NA 100
O Cl
O P O S
GC conditions Inlet: Inlet liner: Carrier gas: Inlet pressure: Inlet temperature: Injection volume: Purge flow to split vent: Splitless Helix double taper, deactivated (p/n 5188-5398) Helium 19.6 psi (constant pressure mode) during run 1.0 psi during backflush 250 C 1.0 L 30 mL/min at 0.75 min
Sample Preparation
The sample preparation procedure includes sample comminution, extraction and partitioning and dispersive SPE clean-up. This process has been described in detail in previous application notes. [8] The procedure used for spinach was similar with the exception of the dispersive SPE clean-up step which includes toluene addition. The frozen chopped organic spinach was homogenized thoroughly. A 15 g (0.1 g) amount of homogenized sample was placed into a 50 mL centrifuge tube. Samples were fortified with appropriate QC spiking solutions (100 L) when necessary, then fortified with 100 L of IS spiking solution (15 g/mL of TPP). After vortexing the sample for 30s, 15 mL of 1% HAc in ACN was added to each tube using the dispenser. To each tube, an Agilent SampliQ QuEChERS AOAC extraction salt packet (p/n 5982-5755) was added directly. Sample tubes were capped tightly, and hand-shaken vigorously for 1 min. Tubes were centrifuged at 4000 rpm for 5 min. Next, the ACN extracts were separated into two parts for both original and modified dispersive SPE methods. The modified dispersive SPE method has a different procedure; therefore, it is described below in detail. The volume of ACN extracts (about 14 mL) will be enough for simultaneously processing samples with original and modified dispersive SPE when using the 2 mL size dispersive SPE tube. If you are using the 15 mL size tube, then 14 mL of ACN extracts from the one sample will not be enough for processing dispersive SPE by the two methods (since 8 mL are required for each dispersive SPE method). Therefore, another sample must be extracted from the beginning.
Oven temperature program: 70 C (1 min), 50 C/min to 150 C (0 min), 6 C /min to 200 C (0 min), 16 C/min to 280 C (6 min) Post run: Capillary flow technology: 3 min Purged Ultimate Union (p/n G3186B) - used for backflushing the analytical column and inlet. Aux EPC gas: Helium plumbed to Purged Ultimate Union Aux EPC pressure: Column: Connections: Restrictor: Connections: MS conditions Tune file Mode Source, quad, transfer line temperature Solvent delay Multiplier voltage Atune.u SIM (refer to Table 2 for settings in detail) 230 C, 150 C and 280 C respectively 2.30 min Autotune voltage 4.0 psi during run, 80.0 psi during backflush Agilent J&W HP-5MS Ultra Inert 15 m 0.25 mm, 0.25 m (p/n 19091S-431UI) Between inlet and Purged Ultimate Union (p/n: G3186B) 65 cm x 0.15 mm, 0.15 m DB-5MS Ultra Inert Between the Purged Ultimate Union and the MSD
Table 2. Analyte
Instrument Acquisition Data Used for the Analysis of 18 Pesticides by GC/MS SIM 184.9 170.1, 169.1 180.9, 182.9 137.1, 179.1 265.8, 263.8 285.9, 287.9 250.0,139.0 196.8, 198.8 352.8, 354.8 259.9, 261.9 372.8, 374.8 245.9, 317.9 372.8, 374.8 262.9, 264.9 230.9 273.8 325.1, 326.1 183.1 362.0, 225.9 Collection window (min) 2.3 4.0 4.0 5.0 5.0 6.9 6.9 7.7 6.9 7.7 7.7 8.6 8.6 10.0 8.6 10.0 10.0 10.4 10.4 10.85 10.85 11.6 10.85 11.6 10.85 11.6 11.0 12.3 12.3 13.6 12.3 13.6 13.6 15.0 15.0 23.0 15.0 23.0 RT (min) 2.88 4.35 6.67 7.19 7.34 8.25 9.55 9.57 10.31 10.75 10.97 11.21 11.50 11.89 12.97 13.35 13.84 15.69, 15.79 15.83
(1) Dichlorvos (2) -Phenylphenol (3) Lindane (4) Diazinon (5) Chlorothalonil (6) Chlorpyrifos-methyl (7) Dichlorobenzophenone (8) Chlorpyrifos (9) Heptachlor epoxide (10) Folpet (11) -Chlordane (12) DDE (13) -Chlordane (14) Dieldrin (15) Ethion (16) Endosulfan sulfate TPP (IS) (17) Permethrin (18) Coumaphos
A 1 mL aliquot of the upper ACN layer was transferred into an Agilent SampliQ QuEChERS dispersive SPE 2 mL tube (p/n 5982-5222); or an 8 mL aliquot was transferred into an Agilent SampliQ QuEChERS dispersive SPE 15 mL tube (p/n 5982-5258). The 2 mL tube contained 50 mg of PSA, 50 mg of GCB and 150 mg of anhydrous MgSO4; while the 15 mL tube contained 400 mg of PSA, 400 mg of GCB and 1200 mg of anhydrous MgSO4. Next, 375 L of toluene were added to the 2 mL tube, and 3 mL of toluene were added to the 15 mL tube. The tubes were tightly capped and vortexed for 1 min. We suggest vortexing the tubes for a few seconds before adding the sample, to prevent possible agglomerates. The 2 mL tubes were centrifuged
with a micro-centrifuge at 13,000 rpm for 2 min, and the 15 mL tubes were centrifuged in a standard centrifuge at 4000 rpm for 5 min. An 825 L amount of extract was then transferred into another tube, and dried by N2 flow. Samples were reconstituted into 600 L of ACN containing 0.1% FA. After vortexing and sonicating, the reconstituted samples were transferred directly into autosampler vials for GC/MS injection. The reconstituted blank samples were directly used to prepare the calibration curve. Another aliquot of ACN extracts was processed following the original dispersive SPE clean-up procedure. Figure 1 shows the flow chart of the whole extraction procedure (original and modified dispersive SPE, 2 mL size) for spinach samples.
Weigh 15 g spinach sample ( 0.1 g) in 50 mL centrifuge tube
Spike 100 L of IS and QC spike solution (if necessary), vortex 1 min.
Add 15 mL of 1% HAc in ACN, and SampliQ AOAC QuEChERS extraction kit
Cap and shake vigorously by hand for 1 min, centrifuge at 4000 rpm for 5 min
Dry with N 2 flow at 30C Reconstitute into 600 L of 0.1%FA in ACN
Vortex and sonicate
Transfer certain volume for GC/MS
Transfer certain volume for GC/MS
Figure 1.
Flow chart of the QuEChERS AOAC extraction procedure (original and modified dispersive SPE, 2 mL size) for spinach sample.

The QuEChERS method for pesticide residues analysis provides high-quality results in a fast, easy, inexpensive approach. For the pigmented fruits and vegetables, the addition of GCB in the dispersive SPE tube can greatly remove pigments and sterols. In order to address the significant loss of planar pesticides, toluene was added to increase the extraction efficiency of those pesticides. Previously we discussed that the addition of toluene retained more matrix impurities in the final sample. [4] In the application using LC/MS/MS, there's no chromatographic differences between the samples processed by the original and modified methods due to the powerful selectivity of LC/MS/MS. The selectivity of GC/MS (SIM mode) is not as powerful as that of
LC/MS/MS (MRM mode). In GC/MS there are interference peaks apparent in the blank chromatogram. Fortunately most of the pesticides tested are free of co-eluting interferences. There was also an interference eluting at a retention time very close to that of -phenylphenol, and this cannot be differentiated for quantitation. The response of this interferent within the blank was integrated to be less than 20% of the response of the -phenylphenol peak at the LOQ (10 ng/g) sample. Therefore, it was considered selectivity-acceptable for this compound. Finally, the GC/MS blank chromatograms showed minor differences from samples processed by the original and modified methods, but these differences did not affect the analysis of the target analytes. Figure 2 and Figure 3 show the GC/MS chromatograms of matrix blank (IS spiked) and 50 ng/g fortified spinach extract processed by the original and modified dispersive SPE methods.
GC/MS Chromatograms of Spinach Matrix Blank with Minors Differences Detected

4000 3500 3000 2500 2000 1500 1000 500 0 4.00 6.00 8.00 10.00 12.00 Time 14.00 16.00 18.00 20.00
IS
4000 3500 3000 2500 2000 1500 1000 500 0 4.00 6.00 8.00 10.00 12.00 Time 14.00 16.00 18.00 20.00
B
IS
Figure 2.
GC/MS chromatograms of spinach matrix blank processed by original dispersive SPE (A) and modified dispersive SPE (B). IS: Internal Standard TPP.
GC/MS Chromatograms of 50 ng/g Fortified Spinach Samples Implementing the Original and Modified AOAC Dispersive Method
2
4500 4000 3500 3000 2500 2000 1500 1000 500 0 4.00 6.00 8.00 10.00
7,8 6
12 IS 17 18 16
14.00 16.00 18.00 20.00
3 1
4 5
15 9 11 13 10 14
12.00 Time
2
4500 4000 3500 3000 2500 2000 1500 1000 500 0
IS 7,8 6 9 12 15 18 5 11 13 10 14 16 17
3 1
4.00 6.00
8.00
10.00
12.00 Time
14.00
16.00
18.00
20.00
Figure 3.
GC/MS chromatograms of 50 ng/g fortified spinach sample extracts processed by original dispersive SPE (A) and modified dispersive SPE (B). Peak identification: 1. Diachlorvos, 2. -Phenylphenol, 3. Lindane, 4. Diazinon, 5. Chlorothalonil 6. Chloropyrifos methyl 7. Dichlorobenzophenone, 8. Chlorpyrifos, 9. Heptachlor epoxide, 10. Folpet, 11. -Chlordane, 12. DDE, 13. -Chlordane, 14. Dieldrin, 15. Ethion, 16. Endosulfan sulfate, 17. Permethrin, 18. Coumaphos. IS: Internal Standard, TPP.

The linear calibration range for all of the pesticides was 10400 ng/g, except folpet, which was 50400 ng/g. For the sample processed by original and modified methods, the corresponding matrix blank was used to prepare the calibration curves respectively. Calibration curves, spiked in matrix blanks, were made at levels of 10, 20, 50, 100, 250, and 400 ng/g. The TPP was used as an internal standard at 100 ng/g. The calibration curves were generated by plotting
the relative responses of analytes (peak area of analyte/ peak area of IS) to the relative concentration of analytes (concentration of analyte / concentration of IS). The 10 ng/g quantification limits LOQ (10 ppb) and 50 ng/g LOQ for folpet (50 ppb) established for the pesticides are substantially lower than many MRLs of those pesticides in fruit and vegetables. The regression fit used for the calibration curves was the average response factor. Table 3 shows the linear regression equation and correlation coefficient (R2) for both 1 mL and 8 mL dispersive SPE.
10
Table 3. Pesticide Dichlorvos
Linearity of 17 Pesticides in Spinach Extract 1 mL dispersive SPE Linear Term RF Rel Std Dev (%) 5.55e-001 2.93e+000 8.34e-001 1.03e+000 7.67e-001 1.26e+000 3.03e+000 6.46e-001 4.36e-002 1.73e-001 2.98e+000 1.37e-001 3.41e-001 1.07e+000 3.06e-001 1.14e+000 3.45e-001 9.1 7.9 9.5 8.7 14.2 12.5 9.8 6.9 10.2 6.2 7.4 8.1 5.1 15.6 4.6 6.4 6.4 8 mL dispersive SPE Linear Term RF Rel Std Dev (%) 4.71e-001 2.30e+000 6.98e-001 9.25e-001 7.83e-001 1.20e+000 2.61e+000 5.99e-001 3.33e-002 1.38e-001 2.49e+000 1.07e-001 2.92e-001 1.42e+000 2.42e-001 1.22e+000 2.64e-001 6.8 9.3 8.1 11.2 13.7 10.3 12.2 11.5 11.4 6.4 6.2 8.1 10.6 12.2 4.2 9.7 11.3
-Phenylphenol Lindane Diazinon Chlorothalonil * Chlorpyrifos methyl Dichlorobenzophenone * Chlorpyrifos Folpet *,** -Chlordane DDE -Chlordane Dieldrin Ethion Endosulfan sulfate Permethrin Coumaphos *
* Results from modified dispersive SPE ** Calibration curve range: 50 400 ng/g.

The recovery and reproducibility were evaluated by spiking pesticides standards in comminuted spinach sample at levels of 10, 50 and 200 ng/g. These QC samples were quantitated against the matrix spiked calibration curve. The analysis was performed in replicates of six (n = 6) at each level. The recovery and reproducibility (shown as % RSD) data for 1 mL and 8 mL dispersive SPE are shown in Table 4 and Table 5, respectively. Since it was demonstrated that the dispersive SPE size (1 mL and 8 mL) didn't affect the results, the 8 mL size modified dispersive SPE test was not performed due to sample volume limitation. In the 18 GC-amenable pesticides we screened, four pesticides, chlorothalonil, dichlorobenzophenone, folpet, and coumaphos, were found to be adversely affected by the GCB in the dispersive SPE step. With the addition of toluene, the recoveries of those pesticides were increased from 50% to 200% with better precision. However, the modified method also reduced the recovery of certain pesticides that had generated good results originally.
Therefore, the quantitation results shown here are the combination of 14 pesticides from original dispersive SPE and four pesticides from the modified method. It can be seen from the results that the 14 pesticides processed by the original method give out good recoveries (average of 88.8% for 1 mL and 86.3% for 8 mL) and precision (average of 5.4% RSD for 1 mL and 4.8% RSD for 8 mL). Although the four pesticides processed by the modified method give lower recovery (average of 75.3% for 1 mL) but great precision (average of 6.1% RSD for 1 mL), the results were much better than the results obtained by original methods (average recovery of 41.7% with 14.9% average RSD). Please refer to the previous application note [4] for discussions in more detail. Folpet was quantified, but the LOQ was found to be 50 ng/g due to poor sensitivity.
11
Table 4.
Spinach AOAC Dispersive, 1 mL Sample Volume, 2 mL Tube, LC/MS/MS Results Low QC (10 ng/g) Recovery RSD 94.0 95.0 83.7 97.3 47.5 74.1 97.5 88.3 74.9 NA 106.0 80.3 107.6 99.7 91.4 93.7 84.7 98.4 3.0 2.2 3.1 4.3 6.8 4.6 7.6 3.0 1.9 NA 4.9 2.2 4.2 2.6 3.4 4.8 5.7 5.5 Mid QC (50 ng/g) Recovery RSD 91.7 92.0 93.9 95.6 44.9 71.7 66.8 79.6 81.6 98.8 112.2 86.8 108.4 93.7 100.0 97.3 74.8 84.2 10.5 7.9 12.2 9.9 6.6 4.5 3.9 3.5 11.7 6.0 3.3 9.6 3.5 9.6 5.0 8.8 9.9 9.5 High QC (200 ng/g) Recovery RSD 80.9 78.7 91.8 91.8 49.4 72.2 68.8 77.0 78.2 77.7 93.6 75.4 91.6 78.9 107.4 89.8 84.6 81.2 4.6 3.8 3.3 3.3 4.3 5.8 6.8 3.5 3.9 6.7 5.3 3.5 3.7 3.4 7.6 4.3 6.0 3.2
Pesticide Dichlorvos -Phenylphenol Lindane Diazinon Chlorothalonil * Chlorpyrifos methyl Dichlorobenzo Phenone * Chlorpyrifos Heptachlor epoxide Folpet * -Chlordane DDE -Chlordane Dieldrin Ethion Endosulfan sulfate Permethrin Coumaphos *
12
Table 5.
Spinach AOAC Dispersive 8 mL Volume, 15 mL Tube, Results by GC/MS Low QC (10 ng/g) Recovery RSD 93.7 87.9 83.1 85.8 21.1 76.4 93.3 77.8 78.6 NA 106.8 80.8 104.2 96.4 83.8 90.5 84.0 61.2 2.6 5.1 5.1 6.9 49.7 2.4 4.1 3.6 4.6 NA 5.7 4.2 6.2 6.2 4.0 8.8 4.9 22.2 Mid QC (50 ng/g) Recovery RSD 92.5 92.5 85.4 85.2 23.6 73.9 56.6 70.2 79.6 60.3 110.7 81.8 103.6 93.0 82.8 87.5 78.4 42.6 4.2 6.6 2.9 2.9 14.3 2.7 2.0 4.6 2.6 17.4 3.5 2.6 3.3 1.2 2.3 7.3 6.8 28.8 High QC (200 ng/g) Recovery RSD 86.2 95.2 84.5 87.3 23.2 73.8 61.4 69.0 85.3 53.7 100.4 81.3 95.8 79.3 85.3 84.5 79.5 35.3 5.9 6.3 5.2 5.5 14.0 3.6 4.5 3.1 5.1 10.7 4.9 4.9 5.3 5.3 4.9 6.4 10.7 20.6
Pesticide Dichlorvos -Phenylphenol Lindane Diazinon Chlorothalonil* Chlorpyrifos methyl Dichlorobenzophenone* Chlorpyrifos Heptachlor epoxide Folpet* -Chlordane DDE -Chlordane Dieldrin Ethion Endosulfan sulfate Permethrin Coumaphos*
*Poor results caused by GCB added in dispersive SPE, can be improved by addition of toluene in the dispersive SPE.
Conclusions
Agilent SampliQ QuEChERS AOAC buffered extraction kits and dispersive SPE kits for pigmented fruits and vegetables provide a simple, fast and effective method for the purification of representative volatile to semi-volatile pesticides in spinach. The modified dispersive SPE method with the addition of toluene provides a very useful option to improve the loss of planar pesticides caused by GCB in dispersive SPE
tubes. The recovery and reproducibility, based on matrix spiked standards, were acceptable for multiclass, multiresidue pesticide determination in spinach. The impurities and matrix effects from spinach did not interfere with the quantitation of target compounds. As the selected pesticides represented a broad variety of different classes and properties, the Agilent SampliQ QuEChERS AOAC Buffered Extraction and Dispersive kits for Pigmented Fruits and Vegetables can be used for other pesticides in similar pigmented matricies.
13
References
1. M. Anastassiades, S. J. Lehotay, Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and "Dispersive Solid-Phase Extraction" for the Determination of Pesticide Residues in Produce, J. AOAC Int., 2003, 86, 412- 431. S. J. Lehotay, et al., Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005, 88, 615-629. S. J. Lehotay, et. al., Determination of Pesticide Residues in Foods by Acetonitrile Extraction and Partitioning with Magnesium Sulfate: Collaborative Study, J. AOAC Int., 2007, 90, 485-520. L. Zhao, J. Stevens, Optimizing Recoveries of Planar Pesticides in Spinach Using Toluene and Agilent SampliQ AOAC QuEChERS Kits with Graphitized Carbon, Agilent Technologies publication, 5990-4247EN. L. Zhao, J. Stevens, Analysis of Pesticide Residues in Spinach Using Agilent SampliQ QuEChERS AOAC kit by LC/MS/MS Detection, Agilent Technologies publication, 5990-4248EN. http://sitem.herts.ac.uk/aeru/footprint/en/index.htm http://www.m5.ws001.squarestart.ne.jp/foundation/ search.html http://www.mrldatabase.com/?selectvetdrug=0
9.
http://www.ams.usda.gov/AMSv1.0/getfile?dDoc Name=PDP1995Summary
10. P. L. Wylie, C. K. Meng, A Method for the Trace Analysis of 175 Pesticides Using the Agilent Triple Quadrupole GC/MS/MS. Agilent Technologies publication 5990-3578EN
2.

3.
4.
5.
6. 7. 8.
Agilent Technologies, Inc., 2009 Printed in the USA July 16, 2009 5990-4305EN
Analysis of Pesticide Residues in Apple Using Agilent SampliQ QuEChERS EN Kits by GC/MS
Application Note
Food Safety
Authors
Limian Zhao, Philip L. Wylie and Joan Stevens Agilent Technologies, Inc. 2850 Centerville Road Wilmington, DE 19809-1610 USA
Abstract
This application note describes the use of a quick, easy, cheap, effective, rugged, and safe (QuEChERS) sample preparation approach described in the European Committee (EN) for extraction and cleanup of 17 GC-amenable multiple pesticide class residues in apple. The method involves initial extraction in an aqueous/acetonitrile system, an extraction/partitioning step after the addition of salt, and a cleanup step using dispersive solid phase extraction (dispersive SPE). The two different dispersive SPE cleanup approaches (1 mL and 6 mL aliquot volumes) were evaluated simultaneously after sample extraction. The target pesticides in the apple extracts were then analyzed by gas chromatography/mass spectrometry (GC/MS) operating in selective ion monitoring (SIM) mode. The method was validated in terms of recovery and reproducibility. The limit of quantitation (LOQ) for pesticides in apple is 10 ng/g. This application employed Agilent's SampliQ QuEChERS kit and produced results well below the maximum residue limits (MRLs) for all the pesticides screened. The spiked levels for the recovery experiments were 10, 50, and 200 ng/g. Recoveries ranged between 68 and 112% (86.0% on average), with RSD below 15% (4.7% on average).
Introduction
The QuEChERS method for pesticide analysis was first introduced by USDA scientists in 2003. [1] The method was modified to address some problematic pesticides by including a buffered extraction system. [2] The EN method 15662:2007 is a European variation to the QuEChERS method. [3, 4] The method uses acetonitrile extraction, followed by the salting out of water from the sample using anhydrous magnesium sulfate (MgSO4), NaCl and buffering citrate salts to induce liquid-liquid partitioning. A dispersive solid phase extraction (dispersive SPE) is conducted for cleanup using a combination of primary secondary amine (PSA) to remove fatty acids among other components and anhydrous MgSO4 to reduce the remaining water in the extract. After mixing and centrifugation, the upper layer is ready for analysis. Although the EN and AOAC are similar methods, they do have several differences. First, the extraction buffered system in the EN method uses sodium chloride, sodium citrate and disodium citrate sesquihidrate instead of sodium acetate in the AOAC extraction step. Second, in the dispersive SPE step, the EN method uses 25 mg PSA per mL of extract rather than 50 mg PSA per mL of extract as stated in the AOAC method. Gas Chromatography/Mass Spectrometry (GC/MS) has been widely used in pesticide analysis for many years. Many pesticides are volatile or semi-volatile, which makes them GCamenable compounds. Previously, we evaluated the performance of Agilent's SampliQ EN buffered extraction kit and SampliQ EN dispersive SPE kits for the analysis of polar pesticides in apple using LC/MS/MS for detection and quantification. [5] In this study, the performance of the SampliQ EN Buffered Extraction kit (p/n 5982-5650) and SampliQ EN dispersive SPE kit for General Fruits and Vegetables (p/n 59825021 and 5982-5056) was evaluated for the extraction of volatile and semi-volatile pesticides. Analysis was performed by GC/MS. Seventeen GC-amenable pesticides were selected which represent multiple classes, including non-polar organochlorine pesticides (OCs), certain organophosphorus pesticides (OPs) and organonitrogen pesticides (ONs). The MRLs of these pesticides are a function of both the pesticide class and food matrix and have been set at 10 ng/g or higher. Table 1 shows the chemical and regulatory information for these pesticides in apple.
Experimental
All reagents and solvents were HPLC or analytical grade. Acetonitrile (ACN), methanol (MeOH) were from Honeywell (Muskegon, MI, USA). Formic acid (FA) was from Fluka (Sleinheim, Germany). The pesticide standards and internal standard (triphenyl phosphate, TPP) were purchased from Sigma-Aldrich (St Louis, MO, USA), Chem Service (West Chester, PA, USA), or Ultra Scientific (North Kingstown, RI, USA).

Standards and internal standard (IS) stock solutions (2 mg/mL) were made in MeOH, respectively, and stored at 20 C. Three QC spiking solutions of 1, 5 and 20 g/mL were made fresh daily in 1:1 ACN/H2O (0.1% FA). A 2.5 g/mL standard solution (17 pesticides) in ACN (0.1% FA) was used to prepare the calibration curves in the matrix blank extract by appropriate dilution. A 10 g/mL amount of TPP spiking solution in 1:1 ACN/H2O (0.1% FA) was used as the internal spiking standard (IS).
Equipment and Materials

Agilent Gas Chromatograph (Agilent Technologies Inc., Santa Clara, CA, USA). Agilent 5975C Mass Spectrometer (Agilent Technologies Inc., Santa Clara, CA, USA). SampliQ QuEChERS EN Extraction kit, p/n 5982-5650 (Agilent Technologies Inc., Wilmington, DE, USA). SampliQ QuEChERS EN dispersive SPE kits for General Fruits and Vegetables, p/n 5982-5021 and 5982-5056 (Agilent Technologies Inc., Wilmington, DE, USA). CentraCL3R Centrifuge (Thermo IEC, MA, USA) Bottle top dispenser (VWR, So Painfield, NJ, USA) Eppendorf microcentrifuge (Brinkmann Instruments, Westbury, NY, USA) Grinder (St Joseph, MI USA)
Table 1. Name
Pesticides Chemical and Regulatory Information [69] Category Phenol Log P 3.18 pKa 9.4 Structure MRLs in apple (ng/g)* 20
s-Phenylphenol
OH
Dichlofluanid
Sulphamide
3.7
NA
5000
S N S N S
Dichlorvos Organophosphate 1.9 NA
Cl S F
10
Cl
O Cl Cl
Diazinon Organophosphate 3.69 2.6
P O
O
100
N O
S P O O
10
Chlorothalonil
Chloronitrile
2.94
NA
N Cl Cl Cl Cl
N
500
O
Dichlorobenzophenone
Organochlorine
4.44
NA
Cl C Cl
Chlorpyrifosmethyl
Organophosphate
4.00
NA
Cl N Cl Cl S O P O CH3 O CH3
500
(Continued)
Table 1. Name Lindane
Pesticides Chemical and Regulatory Information [69] Class Organochlorine Log P 3.69 pKa NA Structure MRLs in apple (ng/g)* 10
Cl Cl Cl Cl Cl Cl
Chlordane
2.78
NA
Cl
Cl Cl
20
Cl
Cl Cl Cl
Cl
Cl
10
Dieldrin
3.7
NA
Cl Cl Cl
O
DDE Organochlorine 6.55 NA
Cl
Cl
50
Cl
Cl
Cl
Ethion Organophosphate 5.07 NA
Cl
300
S P O
S P O
Endosulfan sulfate
Organochlorine
3.13
NA
50
HO Cl Cl Cl
Cl Cl O O S O O
(Continued)
Table 1. Name
Pesticides Chemical and Regulatory Information [69] Class Organochlorine Log P 5.83 pKa NA Structure MRLs in apple (ng/g)*
Heptachlor epoxide
Cl
Cl
Cl Cl
10
Cl Cl O
Permethrins Pyrethroid 6.1 NA 50
Cl Cl O
Coumaphos Organothiophosphate 3.86 NA 100
O Cl
P O S
GC conditions Inlet Inlet liner Carrier gas Inlet pressure Inlet temperature Injection volume Purge flow to split vent Oven temperature program Splitless Helix double taper, deactivated (p/n 5188-5398) Helium 20.18 psi (constant pressure mode) during run 1.0 psi during backflush 250 C 1.0 L 30 mL/min at 0.75 min 70 C (1 min), 50 C/min to 150 C (0 min), 6 C /min to 200 C (0 min), 16 C/min to 280 C (6 min) 3 min
Purged Ultimate Union (p/n G3186B) used for backflushing the analytical column and inlet. Helium plumbed to Purged Ultimate Union 4.0 psi during run, 80.0 psi during backflush Agilent J&W HP-5ms Ultra Inert 15 m 0.25 mm 0.25 m (p/n 19091S-431UI) Between inlet and Purged Ultimate Union (p/n G3186B) 65 cm 0.15 mm 0.15 m DB-5MS Ultra Inert. Between the Purged Ultimate Union and the MSD Atune.u SIM (refer to Table 2 for settings in detail) 230 C, 150 C and 280 C respectively, 2.30 min Autotune voltage
Aux EPC gas Aux EPC pressure Column Connections Restrictor Connections MS conditions Tune file Mode Source, quad, transfer line temperatures Solvent delay Multiplier voltage
Post run
Table 2. Analyte
Instrument Acquisition Data Used for the Analysis of 17 Pesticides by GC/MS. SIM 184.9 170.1, 169.1 180.9, 182.9 137.1, 179.1 265.9, 263.9 285.9, 287.9 123, 167.0 139, 249.9 352.8, 354.8 372.8, 374.8 245.9, 317.9 372.374.8 262.9, 264.9 230.9 273.8 325.1, 326.1 183.1 362.0, 225.9 Collection window (min) 2.3 4.0 4.0 5.0 5.0 6.9 6.9 7.7 14.65 16.0 7.7 8.6 8.6 9.35 18.8 20.5 10.0 10.4 10.85 11.6 10.85 11.6 10.85 11.6 11.0 12.3 12.3 13.6 12.3 13.6 13.6 15.0 15.0 23.0 15.0 23.0 RT (min) 2.88 4.35 6.67 7.19 14.8 8.25 9.16 19.2 10.31 10.97 11.21 11.50 11.89 12.97 13.35 13.84 15.69, 15.79 15.83
(1) Dichlorvos (2) -Phenylphenol (3) Lindane (4) Diazinon (5) Chlorothalonil (6) Chlorpyrifos-methyl (7) Dichlofluanid (8) Dichlorobenzophenone (9) Heptachlor epoxide (10) -Chlordane (11) DDE (12) -Chlordane (13) Dieldrin (14) Ethion (15) Endosulfan sulfate TPP (IS) (16) Permethrin (17) Coumaphos
Sample preparation
Sample comminution Organically grown, pesticide free apples were purchased from a local grocery store. Approximately three pounds of apples were chopped into small, bean sized cubes. Skin was included, but the core was discarded. The chopped apple cubes were then placed into a clean plastic bag and frozen at 20 C overnight. The bag was massaged occasionally to make sure the cubes remained separate. The following day, only the required amount of frozen apple cubes was removed and thoroughly blended. Dry ice was added while comminuting, when possible. Samples were comminuted thoroughly to get the best sample homogeneity. It was verified that no pieces of apple were visible in the final sample. Extraction/Partitioning A 10 g ( 0.1 g) amount of previously homogenized sample was placed into a 50 mL centrifuge tube. QC samples were fortified with 100 L of appropriate QC spiking solution. 100 L of IS spiking solution (10 g/mL of TPP) was added to all the samples except the control blank to yield a 100 ng/g concentration in the samples. Tubes were capped and vortexed for 1 min. A 10 mL aliquot of ACN was added to each tube using the dispenser. Tubes were capped and shaken by hand for 1 min. An Agilent SampliQ QuEChERS EN extraction
6
salt packet (p/n 5982-5650), containing 4 g anhydrous MgSO4, 1 g NaCl, 1 g Na3Citrate, and 0.5 g Na2H Citrate sesquihydrate, was added directly to each tube. The salt bag was massaged carefully to loosen any clumped salts before pouring. No powders were left in the threads or rims of the tubes. Tubes were sealed tightly and shaken vigorously for 1 min by hand to ensure that the solvent interacted well with the entire sample and crystalline agglomerates were broken up sufficiently. Sample pH was checked and 5M NaOH solution was used to adjust the pH to 55.5, if necessary. Sample tubes were centrifuged at 4000 rpm for 5 min. Dispersive SPE Cleanup A 1 mL aliquot of upper ACN layer was transferred into Agilent SampliQ QuEChERS EN dispersive SPE 2 mL tube (p/n 5982-5021); or a 6 mL of aliquot was transferred into Agilent SampliQ QuEChERS EN dispersive SPE 15 mL tube (p/n 5982-5056). The 2 mL tube contained 25 mg of PSA and 150 mg of anhydrous MgSO4; while the 15 mL tube contained 150 mg of PSA and 900 mg of anhydrous MgSO4. The tubes were capped tightly and vortexed for 1 min. The 2 mL tubes were centrifuged with a micro-centrifuge at 13,000 rpm for 2 min, and the 15 mL tubes in a standard centrifuge at 4000 rpm for 5 min. A 500 L portion of the extract was transferred into an autosampler vial and 25 L of 1% FA in ACN was added immediately.
Figure 1 shows the flow chart for the QuEChERS EN sample extraction procedure.

Using the SampliQ QuEChERS kits, the entire procedure is fast, easy, and offers time and labor savings, while ensuring consistency. An analyst can process 40-50 samples in just a few hours. Agilent's SampliQ extraction salts are uniquely prepared in an anhydrous package. The addition of a food sample with a high content of water directly to the salts creates an exothermic reaction, which can affect analyte recoveries, especially for volatile pesticides. The unique SampliQ anhydrous salts packet allows salt addition AFTER the addition of organic solvent to the sample, as specified in the original QuEChERS methodology. Our previous study demonstrated good performance of Agilent's SampliQ QuEChERS EN kits on the extraction of a broad variety of semi-polar to polar pesticides analyzed by LC/MS/MS. [5] It is also advantageous to evaluate the performance of the EN kit for the analysis of volatile and semivolatile pesticides using GC/MS, since these classes of pesticides have been widely used for many years. The selectivity of GC/MS (SIM mode) is not as powerful as that of LC/MS/MS (MRM mode). Furthermore, the final QuEChERS prepared samples still contain some food matrix impurities, which can be observed in the GC/MS chromatogram of blank apple extract. Therefore, it is critical to carefully choose the selected ions of each compound for monitoring when setting up the SIM method. In general, the most abundant ions were selected in order to achieve the best sensitivity; however in a few instances the sensitivity was compromised to obtain better selectivity by using more unique but less abundant ions. Another potential issue with the use of GC/MS for the analysis of QuEChERS samples is the contamination of the ionization source and deterioration of the GC column. QuEChERS food samples usually still contain high-boiling indigenous impurities, which can accumulate on the head of the column, causing peak tailing and retention time shift. Over time, these impurities can migrate to the mass spectrometer (MS) source, causing contamination of the source. Decreased sensitivity and peak shape distortion, especially for the semipolar compounds, were observed when additional QuEChERS samples were injected into the GC/MS system. Therefore, column backflushing was employed to increase column life as well as preserve the MS source. Agilent's capillary flow technology makes column backflushing routine [1012]. Several different capillary flow devices can be used for this purpose.
Add 100 L of IS (TPP) solution, and QC spike solution if necessary, vortex 1 min
Add 10mL of ACN, shake for 1min by hand
Add SampliQ EN QuEChERS Extraction salt packet
Cap and shake vigorously for 1 min
Transfer 1 mL of upper ACN layer to SampliQ En Dispersive SPE 2 mL tube, or 6 mL to SampliQ EN Dispersive SPE 15 mL tube
Vortex 1 min, centrifuge at 13,000 rpm for 2 min for 2 mL tubes or at 4000 rpm for 5 min for 15 mL tubes
Transfer 500 L extract to autosampler vial, add 25 L of 1% FA in ACN, mix well
Analyze by GC/MS Figure 1. Flow chart of the Agilent SampliQ QuEChERS EN extraction procedure.
In this study, the GC/MS system used a Purged Ultimate Union. The analytical column was connected to the capillary flow device. A short restrictor (65 cm 0.15 mm 0.15 m of DB-5ms Ultra Inert column) was used to couple the capillary flow device to the mass spectrometer. In a previous application note [10], there are figures showing the backflush system, that was used in this study. Figure 2(a, b) shows the chromatograms of a blank apple extract and a 50 ng/g fortified apple extract. As shown in Figure 2a, interference peaks are found in the blank chromatogram; fortunately most pesticides are free of co-eluting interferences. There was an interference eluting at a retention time very close to that of -phenylphenol (peak 2), and cannot be differentiated for integration. The average response of this interference in the blank extract was 215 (n=4), while the average response of -phenylphenol in the LOQ (10 ng/g)
20000 18000 16000 14000 Abundance 12000 10000 8000 6000 4000 2000 0
GC/MS Chromatogram of Apple Extracts, Blank relative to Fortified Sample, 50 ng/g after Agilent's SampliQ QuEChERS Extraction and Dispersive SPE kits, for General Fruits and Vegetables
IS
4.00
6.00
8.00
10.00
12.00 Time (min)
14.00
16.00
18.00
20.00
2
18000 16000 14000 12000 Abundance 10000 8000 6000 4000 2000 0 4.00 6.00 8.00 10.00 12.00 Time (min) 14.00 16.00 18.00 20.00
IS
8 5 3 1 4 6 7 9
11
14 10 12 13 15
16 17
Figure 2.
GC/MS chromatogram of apple extract. (A) apple extract blank; (B) 50ng/g fortified apple extract. Peak Identification: 1. Dichlorvos, 2. -Phenylphenol, 3. Lindane, 4. Diazinon, 5. Chlorothalonil, 6. Chlorpyrifos-methyl, 7. Dichlofluanid, 8. Dichlorobenzophenone, 9. Heptachlor epoxide, 10. -Chlordane, 11. DDE, 12. -Chlordane, 13. Dieldrin, 14. Ethion, 15. Endosulfan sulfate, 16, Permethrin, 17. Coumaphos. IS. Triphenyl phosphate (TPP)
was 3196 (n=12). The interference response was less than 20% of the response of the -phenylphenol peak at the LOQ (10 ng/g) sample. Therefore the selectivity was considered acceptable for this compound.

The linear calibration range for all of the pesticides was 0400 ng/g. Two different dispersive SPE volumes (1 mL and 6 mL) were used for evaluation and comparison; therefore, two calibration curves were generated from matrix blanks prepared from each size. Calibration curves were made at levels of 10, 20, 50, 100, 250, and 400 ng/g. The TPP was the internal standard (IS) at 100 ng/g in all cases. The calibration
8
curves were generated by plotting the relative responses of analytes (peak area of analyte/peak area of IS) to the relative concentration of analytes (concentration of analyte/concentration of IS). Table 1 shows that the 10 ng/g quantification limits LOQ (10 ng/g or 10 ppb) established for the pesticides are substantially lower than many MRLs for the pesticides in fruit and vegetables. The regression fit used for the calibration curves was the average response factor. Table 3 shows the linear term and RF relative standard deviation (%) for both 1 mL and 6 mL dispersive SPE samples. The RF relative SD is an important parameter for the evaluation of the linearity of calibration. In general, the smaller the value the better linearity of the curve, and it is usually acceptable for less than
Table 3. Analytes Dichlorvos
Linearity of 17 Pesticides in Apple Extract 1 mL dispersive SPE Linear Term RF Rel Std Dev (%) 4.53e001 2.41e+000 6.79e001 8.35e001 1.39e+000 1.32e+000 1.03e+000 6.08e001 5.41e001 1.77e001 2.44e+000 1.34e001 2.85e001 7.06e001 2.95e001 8.73e001 2.36e001 7.9 7.5 11.5 15.0 14.1 14.7 11.7 10.0 12.4 9.3 10.7 10.0 9.8 27.8 11.2 11.8 19.0 6 mL dispersive SPE Linear Term RF Rel Std Dev (%) 5.52e001 2.82e+000 8.09e001 9.32e001 1.69e+000 1.31e+000 1.29e+000 7.13e001 5.58e001 1.83e001 2.67e+000 1.38e001 3.09e001 7.30e001 3.29e001 8.20e001 2.16e001 8.6 9.3 9.2 13.6 14.1 16.5 12.9 10.4 12.3 9.1 9.5 9.4 6.9 27.9 11.5 17.6 28.7
-Phenylphenol Lindane Diazinon Chlorothalonil Chlorpyrifos-mehyl Dichlofluanid Dichlorobenzophenone Heptachlor epoxide -Chlordane DDE -Chlordane Dieldrin Ethion Endosulfan sulfate Permethrin Coumaphos
20% RF relative SD. The data in Table 3 shows that the linearity of the calibration curve for most analytes is perfect with less than a 15% of RF relative SD value. However, the two standard curves of Ethion (1 mL and 6 mL) and one standard curve of Coumaphos generated RF relative SD values higher than 20%, possibly due to the matrix effect.

The recovery and reproducibility were evaluated by spiking pesticides standards in comminuted apple sample at levels of 10, 50 and 200 ng/g. These QC samples were quantitated against the matrix spiked calibration curve. The analysis was
performed in replicates of six (n=6) at each level. The recovery and reproducibility (shown as % RSD) data for 1 mL and 6 mL dispersive SPE sample are shown in Table 4 and 5, respectively. It can be seen from the results that all of the pesticides give good recoveries (average of 84.7% for 1 mL and 87.2% for 6 mL) and precision (average of 4.3% RSD for 1 mL and 5.1% RSD for 6 mL). Compared to the results of these pesticides extracted with AOAC QuEChERS method [13], the EN QuEChERS method gives slightly lower recovery (recovery 56% lower on average) but similar precision (RSD 45% on average for both methods). Variance may be possible due to a different buffering system and solvent volume used in the first extraction step.
Table 4.
Recovery and Repeatability of Pesticides in Fortified Apple With Agilents SampliQ Dispersive SPE Tube, 2 mL (p/n 5982-5021); Recovery 84.7%, RSD 4.3% (avg) 10 ng/g fortified QC Recovery RSD (n=6) 97.6 94.4 87.4 83.6 68.3 79.3 91.8 83.9 80.0 79.6 80.5 84.8 83.4 97.7 93.2 88.8 101.4 5.1 5.5 4.9 5.6 4.9 4.5 5.6 6.4 4.7 4.3 3.2 3.3 3.1 4.4 5.4 6.2 4.7 50 ng/g fortified QC Recovery RSD (n=6) 90.8 83.1 80.0 79.6 71.8 80.7 85.8 83.0 82.9 80.5 80.3 83.1 80.5 104.9 88.5 93.9 111.9 6.2 6.6 6.1 5.5 5.8 4.8 6.9 4.8 5.0 5.4 5.1 4.7 4.3 4.8 4.5 4.6 3.9 200 ng/g fortified QC Recovery RSD (n=6) 81.0 76.3 73.3 69.6 69.6 83.1 65.2 80.0 81.4 78.3 76.8 78.6 76.2 91.7 87.9 104.3 111.2 6.9 5.2 3.8 5.0 5.0 3.0 4.7 3.1 1.7 1.6 1.2 1.3 1.1 1.4 1.1 0.7 1.6
Analytes Dichlorvos -Phenylphenol Lindane Diazinon Chlorothalonil Chlorpyrifos-mehyl Dichlofluanid Dichlorobenzo phenone Heptachlor epoxide -Chlordane DDE -Chlordane Dieldrin Ethion Endosulfan sulfate Permethrin Coumaphos
10
Table 5.
Recovery and Repeatability of Pesticides in Fortified Apple With Agilents SampliQ Dispersive SPE Tube, 15 ML (p/n 5982-5056), Recovery 87.2%, RSD 5.1% (avg) 10 ng/g fortified QC Recovery RSD (n=6) 99.4 76.9 87.8 87.0 71.7 77.7 80.0 86.2 82.6 89.6 91.9 90.3 93.6 81.0 96.4 89.3 89.0 8.2 8.9 7.0 8.3 11.1 9.8 7.8 6.1 5.7 7.1 5.5 4.4 7.3 6.7 5.3 5.6 10.8 50 ng/g fortified QC Recovery RSD (n=6) 90.9 81.6 88.9 86.6 77.9 82.7 86.5 87.6 86.7 85.1 88.7 88.0 88.1 94.2 91.2 95.5 90.7 2.6 1.6 2.7 1.8 1.8 2.3 6.1 2.4 2.8 2.9 3.5 3.1 4.3 4.0 3.9 3.9 6.6 200 ng/g fortified QC Recovery RSD (n=6) 85.7 82.0 86.3 89.3 75.9 86.7 76.6 85.7 85.9 83.6 83.8 84.0 83.2 91.1 89.8 108.9 97.1 4.4 3.6 2.7 2.7 3.8 2.2 5.1 1.5 1.9 2.1 1.8 1.3 1.6 1.7 1.2 1.3 2.0
Analytes Dichlorvos -Phenylphenol Lindane Diazinon Chlorothalonil Chlorpyrifos-mehyl Dichlofluanid Dichlorobenzo phenone Heptachlor epoxide -Chlordane DDE -Chlordane Dieldrin Ethion Endosulfan sulfate Permethrin Coumaphos
Figure 3 shows the recovery and precision results for 1 mL dispersive SPE and 6 mL dispersive SPE. The two different dispersive SPE clean-ups were performed by transferring 1 mL or 6 mL of ACN extract from the same sample following the extraction step. In order to simplify the comparison, the average recovery and precision of three fortification concentrations were used for all of the pesticides. The results of each dispersive SPE clean-up appeared to be independent of volume used. Both approaches provided similar efficient sample clean-up and generated relatively equivalent results.
11
Exceptional Recoveries and Precision for 1 and 6 mL Sample Extract Volumes for Agilent's SampliQ Dispersive, 2 and 15 ml kits.
120
1 mL
6 mL
100
80
Recovery (%)
60
40
20
0 Dichlorfluanid Dichlorvos -Chlordane -Phenylphenol Heptachlor epoxide -Chlordane DDE Chlopyrifos-methyl Diazinon Dichlorobenzophenone Ethion Chlorothalonil Endosulfan sulfate Coumaphos Lindane Permethrin Dieldrin
Figure 3.
The recovery and precision results of 1 and 6 mL sample volumes employing Agilents SampliQ Dispersive SPE, 2 and 15 mL kits, respectively.
Conclusions
Agilent's SampliQ QuEChERS EN Extraction and Dispersive SPE kits for General Fruits and Vegetables provide a simple, fast and effective method for the purification and enrichment of representative volatile to semi-volatile pesticides in apple. The recovery and reproducibility, based on matrix spiked standards, were acceptable for multi-class, multi-residue pesticide
determination in apple. The impurities and matrix effects from apple did not interfere with the quantitation of target compounds. The LOQs of the pesticides were lower than regulated MRLs in apple. Since the selected pesticides represented a broad variety of different classes and properties, the Agilent SampliQ QuEChERS EN Extraction and Dispersive SPE kits for General Fruits and Vegetables is an excellent choice for other pesticides in similar food matricies
12
References
1. M. Anastassiades, S. J. Lehotay, Fast and Easy Multiresidue Method Employment Acetonitrile Extraction/Partitioning and "Dispersive Solid-Phase Extraction" for the Determination of Pesticide Residues in Produce, J. AOAC Int., 2003, 86, 412- 431. S. J. Lehotay, et al, Use of Buffering and Other Means to Improve Results of Problematic Pesticides in a Fast and Easy Method for Residue Analysis of Fruits and Vegetables, J. AOAC Int., 2005, 88, 615-629. European Committee for Standardization/Technical Committee CEN/TC 275 (2007), Foods of plant origin: Determination of pesticide residues using GC-MS and/or LC-MS/MS following acetonitrile extraction/partitioning and cleanup by dispersive SPE-QuEChERS method. European Committee for Standardization, Brussels P. Pay, M. Anastassiades, Analysis of pesticide residues using the Quick Easy Cheap Effective Rugged and Safe (QuEChERS) pesticide multiresidue method in combination with gas and liquid chromatography and tandem mass spectrometric detection, Anal Bioanal Chem., 2007, 389, 1697-1714. L. Zhao, D. Schultz, J. Stevens, Evaluation of the QuEChERS En Kits for Sample Preparation Following the European Standard for the Analysis of Pesticide Residues in Apple with LC/MS/MS Detection, Agilent Technologies publication 5990-3938EN. http://sitem.herts.ac.uk/aeru/footprint/en/index.htm http://www.m5.ws001.squarestart.ne.jp/foundation/ search.html http://www.mrldatabase.com/?selectvetdrug=0 http://www.ams.usda.gov/AMSv1.0/getfile?dDoc Name=PDP1995Summary
11. C. K. Meng, Improving Productivity and Extending Column Life with Backflush, Agilent Technologies publication 5989-6018EN. 12. P. L. Wylie, Direct Injection of Fish Oil for the GC-ECD Analysis of PCBs: Results Using a Deans Switch with Backflushing, Agilent Technologies publication, 5989-6095EN. 13. L. Zhao, D. Schultz, J. Stevens, Analysis of Pesticide Residues in Apple Using Agilent SampliQ QuEChERS AOAC Kits by GC/MS, Agilent Technologies publication 5990-4068EN.
2.
3.

4.
5.
6. 7. 8. 9.
10. P. L. Wylie, C. K. Meng, A Method for the Trace Analysis of 175 Pesticides Using the Agilent Triple Quadrupole GC/MS/MS, Agilent Technologies publication 5990-3578EN.
13
Agilent Technologies, Inc., 2009 Printed in the USA June 30, 2009 5990-4073EN
NOTICE: This document contains references to Varian. Please note that Varian, Inc. is now part of Agilent Technologies. For more information, go to www.agilent.com/chem.
Application Note SI-01002

Multi-Residue Confirmation of Pesticides in Honey using Solid Supported Liquid Extraction
C. Pirard
Mass Spectrometry Laboratory (C.A.R.T), University of Liege, alle de la chimie, 3., B6c Sart-Tilman, B-4000 Liege, Belgium
Introduction
In recent years, several Belgian beekeepers were asked to shut down their hives due to withering of the hive. The cause of the withering was unknown. A multi-factoral study was initiated, which included a multi-class pesticide residue study from honey, to determine if pesticides were the cause of the decline. The extensive distribution of pesticides causes bees that have been fed on contaminated blossoms to transfer pesticide residues into honey. Multi-residue confirmation methods to identify and quantify widely used pesticides, which could have been the source of the bees decline, needed to be developed. Previously published papers already report determination methods for pesticides in honey, however, most of them analyze only one or two pesticide groups, such as organochlorine or organophosphorous residues. This application note shows the development and validation of 17 pesticides and metabolites of different chemical classes:
Insecticides: Carbofuran (Ca), Methiocarb (Mh), Pirimicarb (Pi), Dimethoate (Dm), Fipronil (Fi), Imidacloprid (Im) Herbicides: Amidosulfuron (Am), Rimsulfuron (Ri), Atrazine (At), Simazine (Si), Chlorotoluron (Ch), Linuron (Li), Isoxaflutole (Is), Metosulam (Mo) Fungicides: Diethofencarb (De) Metabolites: Methiocarb sulfoxide (Mhs), 2-Hydroxytertbutylazine (TOH)
is incorporated into the cartridge as a safeguard to ensure that organic eluents remain uncontaminated by aqueous matrix. The extraction on Chem Elut is carried out with gravity only - no vacuum is required (Figure 1).
Figure 1. Solid Supported Extraction on Chem Elut cartridges
Sample Preparation and Clean Up

SLE procedure on 5 mL Chem Elut cartridge (p/n 12198006)
Method: 1. Spike 1 g honey sample with 20 L of surrogate standard solution (concentrations listed in figure 2). 2. Mix with 1.25 mL water and 2.5 mL acetone. 3. Add 1.25 mL of NaCl solution (20 g/100 mL). 4. Apply sample to Chem Elut cartridge by gravity flow. 5. Allow 15 mins for complete adsorption to take place. 6. Elute twice with 10 mL ethyl acetate. 7. Evaporate at 30 C. 8. Reconstitute with 200 L acetonitrile/water (10/90). 9. Inject 20 L into LC-MS/MS.
The application described here is based on solid supported liquidliquid extraction method (SLE) followed by LCMS/MS. The results are compared to data from standard liquid-liquid extraction (LLE) to check extraction efficiency and appropriateness of the SLE method. More detailed information on the method development and validation has been published previously [1].
Analysis Conditions
The method is based on HPLC coupled to mass spectrometry (MS) operating in tandem mode (MS/MS) according to EU advice 2002/657/EC [2].
Column: Temperature: Mobile Phase A: Mobile Phase B: Linear Gradient Conditions: Flow Rate: Polaris C18-A 3 m, 2.0 x 150 mm (p/n A2001150X020) 40 C Water + 0.1 % acetic acid Acetonitrile + 0.1 % acetic acid held 10 % B for 1 min, to 80 % B in 14 mins, to 100 % B in 2 mins, held 100 % B for 2 mins 0.4 mL/min
Chem Elut - the solid support

The solid support consists of specially processed, widepore, diatomaceous earth packed into clean polypropylene cartridges. The aqueous sample is applied to the dry Chem Elut sorbent. The sample is distributed as a thin film over the chemically inert support, which acts as a stationary phase. Subsequently, elution takes place using immiscible organic solvents. The lipophilic substances are extracted from the aqueous into the organic phase, while the aqueous phase remains on the Chem Elut sorbent. A phase-separation filter Application Note SI-01002
1 of 2
www.varianinc.com

Figure 2 shows the comparison between recoveries obtained after SLE on Chem Elut and classical LLE. The LLE was performed by mixing 1 g honey with 2 mL water and 6.5 mL acetonitrile for 30 mins. After centrifugation the organic layer is evaporated to 100 L and 100 L of water is added prior to LC-MS. The comparison shows that the Chem Elut extraction provides similar or even higher extraction efficiency than LLE for most compounds. The greatest advantage of SLE is that the SLE technique avoids emulsion formation, which is standard in LLE, significantly easing the extraction procedure. Key advantages of Chem Elut cartridges are their ease of use and the wide range of compounds that can be extracted efficiently.
Recovery %
LLE SLE
Am
At
Ca
Ch
De
Dm
Fi
Im
Is
Li
Mh MhS Mo
Pi
Ri
Si
TOH
Figure 3. Chromatogram obtained for a honey blank matrix (a) and for a methanolic standard solution (b) using the Polaris C18-A column. Concentrations between 0.4 ng/mL and 20 ng/mL.
Pesticide (ng/mL): Am 0.4, At 0.4, Ca 0.4, Ch 20.0, De 2.0, Dm 2.0, Fi 10.0, Im 2.0, Is 2.0, Li 2.0, Mh 10.0, MhS 20.0, Mo 2.0, Pi 0.4, Ri 0.4, Si 2.0, TOH 1.0 Figure 2. Recovery comparison of pesticides between solid supported liquidliquid extraction (SLE) on Chem Elut and classical liquid-liquid extraction (LLE).
Conclusion
A rapid, reliable, time- and resource-saving analytical method is reported for the measurement of 17 pesticides of different chemical classes used in apiculture or in the surrounding agriculture in the context of a bee mortality study. The multi-residue analytical procedure developed in this study was based on a solid supported liquidliquid extraction step using diatomaceous earth as inert solid support. Extracts were analyzed without further purification by LCMS/MS in ESI mode. The SLE with Chem Elut cartridges has proven to be efficient for a wide range of pesticides, nearly independent of their polarity.
References [1] C. Pirard, J. Widart, B.K. Nguyen, C. Deleuze, L. Heudt, E. Haubruge, E. De Pauw, J.-F. Focant; Development and validation of a multi-residue method for pesticide determination in honey using on-column liquid-liquid extraction and liquid chromatography tandem mass spectrometry; Journal of Chromatography A. 1152, 116-123, (2007) [2] Commission Decision 2002/657/EC, Off. J. Eur. Commun, 12 August 2002
Figure 3 shows the chromatogram of a methanolic standard solution with pesticide concentration between 0.4 and 20 ng/mL and a blank honey matrix. The total analysis time was 23 mins. The Polaris C18-A column is based on ultra-pure silica with a polar group placed between the primary C18 chain and the silica surface. The resultant packing material contains a surface, which is easily wetted with polar eluents and shows unique selectivity for a broad range of chemically different compounds. Furthermore, the polar group shields reactive silanols from polar silanophilic compounds, which improves peak symmetry and minimizes the collapse of the C18 chains in high-aqueous eluents. Seventeen pesticides were separated by optimizing the LC gradient, and the coeluted pesticides with different masses were identified using MRM mode.
These data represent typical results. For further information, contact your local Varian Sales Office.
Chem Elut, Polaris, Varian and the Varian logo are trademarks or registered trademarks of Varian, Inc. in the U.S. and other countries.
2008 Varian, Inc.
Varian, Inc. www.varianinc.com North America: 800.926.3000 925.939.2400 Europe: The Netherlands: 31.118.67.1000 Asia Pacific: Australia: 613.9560.7133 Latin America: Brazil: 55.11.3238.0400
NOTICE: This document contains references to Varian. Please note that Varian, Inc. is now part of Agilent Technologies. For more information, go to www.agilent.com/chem.

Fast Analysis of Pesticides in Foods using QuEChERS with 320-MS LC/MS/MS
Miguel Angel Perez Varian, Inc. Introduction
Pesticides are widely used in many areas of modern agriculture as they are considered economically important for high yield agricultural production. In todays world of extensive importing and exporting of food goods, the analysis and monitoring of pesticides is essential, although challenging. The following application illustrates a complete solution approach for low level, multi-residue pesticide analysis. The data and conclusions summarize the results of calibration studies with RSD and recovery values from an orange matrix using the QuEChERS extract technique. Flow Rate: 200 L/m Injection Volume: 10 L MS Parameters Ionization: ESI (+/-) Drying Gas: N2, 35 psi, SelecTemp Nebulizing Gas: N2, 65 psi Needle: 5000 V (+), 4500 V (-) Shield: 600 V Detection: 1600 V (EDR max) CID Gas Pressure: Argon, 2 mTorr
Instrumentation

12.5
Varian 212 LC Pumps Varian 410 Autosampler Varian 320-MS LC/MS with SelecTemp Control
Materials and Reagents

Column: Varian Pursuit 3m, 50 x 2 mm (pn: A3001050x020) Varian Bulk Bondesil PSA for QuEChERS (pn: 12213024)
10
Sample Preparation
QuEChERS sample preparation was used to extract 48 pesticides (see p 4 of Complete Solutions for Pesticides Residue Analysis in Foods for additional details.)
7.5 MCounts 5 2.5 0 0 2.5 5 min Figure 1. Varian 320-MS LC/MS with SRM transitions offers high quality quantification with more than 10 data points per peak and an analysis of 48 pesticides in less than 10 minutes. 7.5 10
Conditions
HPLC Parameters Solvent A: Water with 0.2 % formic acid and 5mM ammonium formate Solvent B: 90 % MeOH with 10 % water with 0.2 % formic acid and 5 mM ammonium formate.
Time 0:00 0:10 8:00 12:00 12:01 17:00 %A 95 50 0 0 95 95 %B 5 50 100 100 5 5
1 of 3
www.varianinc.com
A patented feature of the 320 Varian API interface, SelecTemp software allows the temperature of the drying gas to be controlled on a time segment basis. This enables the analysis of various pesticide groups, including those which are thermally labile, using one method and potentially one injection. As demonstrated in Figure 2, by adjusting the temperature of the drying gas, thermal degradation is minimized while desolvation is optimized.
Peak Name
Retention Time (min) 4,479 4,259 4,571 4,861 4,929 5,168 5,214 5,430 5,531 5,595 5,605 6,002 6,125 6,202 6,388 6,396 6,679 6,747 7,166 7,237 7,247 7,701 7,799 8,187 8,639 8,680 8,722 9,098 9,215
Area
Analyzed Conc (ppb) 10.5 11.3 13.3 8.3 10.8 8.1 7.3 14.6 8.7 9.0 9.7 9.5 12.2 8.7 9.0 16.6 9.8 8.8 7.3 10.2 8.2 8.8 9.8 8.2 9.5 9.2 10.0 8.2 14.0
Spiked Conc (ppb) 14.6 11.3 23.7 11.8 11.9 11.4 11.3 14.1 11.4 12.2 11.7 11.9 11.9 10.5 12.1 24.2 11.9 10.7 8.8 11.3 11.4 11.3 11.7 12.0 12.4 10.7 11.6 12.2 11.7
Recovery % 71.8 100.7 56.3 70.4 90.6 70.8 64.5 103.7 76.6 73.7 83.3 79.4 102.8 82.8 73.8 68.7 82.5 82.2 82.5 90.4 71.9 78.1 83.8 68.2 76.5 86.3 86.1 66.9 119.6
Forchlorfenuron Pyrimetanil Azinphos methyl Azoxystrobin Dimethomorph Fludioxonil Methiocarb Mepanipyrim Triadimenol Fenhexamid Napropamide Tebufenozide Diflubenzuron Fipronil Benalaxyl Prochloraz Clofentezine Pyraclostrobin Triflumizole Spinosyn A Trifloxystrobin Spinosyn D Hexaflumuron Hexythiazox Emamectine Teflubenzuron Lufenuron Flufenoxuron Diafenthiuron
9.15E+07 1.64E+07 1.92E+08 3.79E+08 3.02E+07 1.97E+07 2.56E+07 8.96E+06 7.53E+06 6.51E+06 2.76E+07 7.22E+07 4.52E+06 9.20E+07 7.65E+07 8.83E+07 1.87E+07 3.38E+07 7.52E+07 5.47E+07 7.97E+07 1.13E+07 5.67E+07 3.10E+07 4.06E+07 3.70+E07 2.62E+07 7.41E+06 1.77E+07
MC Count
400C 40
300C 30 200C 20 10
2.5
5.0
7.5
10.0
Mins
12.5
Figure 2. Diagram showing minimal degradation and optimized desolvation can be achieved by adjusting the drying gas temperature.
Table 1. The table shows retention time and order and the combined effectiveness of the Pursuit C18 column, the 320MS LC/MS and SelecTemp software with the QuEChERS method using Bondesil bulk PSA sorbent for successful high recovery extraction and analysis from a complex food matrix.
Peak Name
Retention Time (min) 2,534 2,537 2,540 2,574 2,640 2,719 2,747 2,754 2,883 2,097 2,910 3,033 3,149 3,189 3,316 3,619 3,847 4,132 4,127
Area
Analyzed Conc (ppb) 9.8 8.4 11.2 8.9 19.0 16.1 21.1 20.5 15.1 17.8 17.9 10.3 14.9 9.7 19.5 7.9 9.7 9.9 16.8
Spiked Conc (ppb) 11.2 11.3 13.6 10.7 25.1 23.2 23.6 23.6 22.3 23.3 23.9 11.5 22.3 12.0 23.4 11.7 11.9 12.6 19.8
Recovery % 87.5 74.5 82.9 83.3 75.7 69.4 89.4 86.7 67.8 76.3 74.9 90.0 67.0 80.7 83.5 67.8 81.4 79.0 85.0
Pymetrozine Acephate Omethoate Propamocarb Oxamyl Methomyl Thiamethoxan Carbendazime Thiabendazole Imidacloprid Clothianidin Acetamiprid Cymoxanil Thiacloprid Aldicarb Propoxur Carbaryl Thiodicarb Imazalil
3.21E+06 8.21E+06 1.01E+07 2.45E+07 5.81E+07 1.72E+07 2.68E+07 3.55E+07 1.27E+07 8.39E+06 6.33E+06 1.57E+07 2.13E+07 1.69E+07 7.94E+07 1.87E+07 1.86E+07 3.89E+07 8.77E+06
Conclusions
The Varian 320-MS LC/MS/MS is capable of fast MS analysis with rapid polarity switching for +/- ions with short dwell times (<10 ms). With a single injection, all pesticides can be analyzed with three transitions per compound for confirmation purposes. SelecTemp controls allow the analysis of all pesticides, including thermolabile compounds in a single injection. Sensitivity offers LOQs at sub-ppb levels for most pesticides in a complex orange matrix also indicating the effectiveness of the sample preparation technique. Linearity shows a coefficient correlation >0.99 and RSDs <10 % for most pesticides.
2 of 3
www.varianinc.com
References
Anastassia des, M., Lehotay, S.J., Stajnbaher, D., and Schenck, F., Fast and Easy Multiresidue Method Employing Acetonitrile Extraction/Partitioning and Dispersive Solid Phase Extraction for the Determination of Pesticide Residues in Produce (QuEChERS method). J AOAC Int., 86, No. 2 (2003) 412-431.
Please visit http://www.varianinc.com/pesticides for additional applications on pesticides in food. For a copy of Complete Solutions for Pesticides Residue Analysis in Foods, please contact your local Varian office or sales representative.
These data represent typical results. For further information, contact your local Varian Sales Office. Varian, Inc. www.varianinc.com North America: 800.926.3000 925.939.2400 Europe: The Netherlands: 31.118.67.1000 Asia Pacific: Australia: 613.9560.7133 Latin America: Brazil: 55.11.3238.0400
SelecTemp, Pursuit, Bondesil, Varian and the Varian Logo are trademarks or registered trademarks of Varian, Inc. in the U.S. and other countries.
2008 Varian, Inc.
Index
GC/MS and LC/MS Analyzers and Application Kits
QQQ LC/MS Pesticide Application Kit QQQ GC/MS Pesticide Analyzer GC/MSD Pesticide Analyzer TOF & Q-TOF LC/MS Pesticide App Kit Analyzers & App Kits Brochure
GC/MS and GC/MS/MS

The GC/MS/MS Analyzer and the Pesticides and Environmental Pollutants MRM Database Organophosphorus Pesticides Analysis Using a J&W DB-5ms Ultra Inert Capillary GC Column Highly Sensitive & Rugged GC/MS/MS Tool for Multi-residue Analysis in Food Trace Analysis Using the QQQ GC/MS/MS Analysis of Complex Samples by GC/MS/MS Marine Biota Screen, Confirm, & Quantify by GC/MS & LC/MS GC/MS Solution for the Japanese Positive List Method Screen Pesticide Residues in TCM Using GC/MS Replace Multi 50-Min GC & GC/MS/SIM with One 15-Min Full-Scan GC/MS Analysis Screen Pesticides & Endocrine Disruptors by GC/MS with DRS & Pesticide Library Screen Haz Chems in Homeland Security & Environ Samples Using GC/MS/ECD/FPD & DRS Identify Pesticides with Full Scan, SIM, uECD, & FPD from a Single Injection Pesticide Residues in Spiked & Unspiked Fruit Extracts Using DRS Pesticide Screening by GC/MSD Using DRS Analysis of Components, Contaminants, & Impurities by GC/MS & LC/MS Identify & Quantitate in the PPT Range Using RTL & GC/MS Screen Pesticides & Endocrine Disrupters Using 6890/5973N GC/MSD, Part II Trace Level Analysis by GC/MS Using Large-Volume Injection Screen Pesticides & Endocrine Disrupters Using 6890/5973N GC/MSD, Part 1
Index
LC/MS and LC/MS/MS
Analysis using 1290 Infinity LC with 6140 Single Quadrupole LC/MS DMRM Compound Database to Screen & Identify Using LC/MS QQQ 6540 Accurate-Mass Q-TOF: High-Confidence Characterization of Unknowns Determine Pesticides in Baby Food by UHPLC/MS/MS using 1290 Infinity LC& 6460 QQQ LC/MS Multi-Residue Analysis with DMRM & QQQ LC/MS/MS Determine Acidic Herbicides using 6460 QQQ LC/MS with Jet Stream Technology & Direct Aqueous Injection App Kit for Multi-Residue Screening using LC/TOF or Q-TOF with a Pesticide PCD Pesticide PCD and LC (Q)TOF MS for Screening & Identification Q-TOF LC/MS Screen & Confirm Non-Targeted Pesticides in a Strawberry Extract Direct On-Column Aqueous Injection of Potable & Environ Samples with 6410BA LC/QQQ Multi-residue Analysis in Food Samples by LC/QQQ MS Screen, Confirm, & Quantify Organic Pesticides in Foods by GC/MS & LC/MS Determine Pesticides in H2O by SPE & LC/MS/MS in Pos & Neg Ion Modes Screen Pesticides in Food by LC/TOF MS with Molecular Feature-DB Multi-residue Analysis of Pesticides in Food by LC/QQQ MS Determine Pesticides in Foodstuffs by LC/MS/MS Determine Chlorinated Acid Herbicides in Soil by LC/MS/MS Analysis of Fungicides & Metabolites in Citrus Fruits & Juices by TOF & Ion Trap LC/MS Analysis of Terbuthylazine in Olive Oil by TOF & Ion Trap LC/MS Determine Fungicides in Fruits & Veg by TOF & Ion Trap LC/MS Identify Unknown Pesticides in Food Using LC/MSD TOF & Ion Trap MS(n) Analysis of Components, Contaminants, & Impurities in Fungicide by GC/MS & LC/MS Determine Acidic Herbicides in Ground & Potable Water by LC/MSD Using SIM Analysis of Glyphosate & Aminomethyl Phosphoric Acid by LC/MS Analysis of Simazine, Thiobencarb, & Thiuram by LC/MS Analysis of Organophosphate Pesticides by LC/MS Develop an LC/MS Method for the Analysis of Rodenticides
Index
Productivity Tools
Can Deconvolution Improve GC/MS Detectability? Maintain RTs with Backflush Controlled Tee Config for 7890A GCs w/ 5975 MSD & 7000 QQQ Achieving Lower Detection Limits with a MMI inlet CFT for GC/MS: Simple Tee Config for Trace Analysis w/Rapid Backflush RTL with MSD Productivity ChemStation Japanese Positive List Method in Food: Benefits of GC/MS w/ DRS & RTL MSD Reducing Analysis Time Using GC/MSD & DRS Direct Injection of Fish Oil for the GC-ECD Analysis of PCBs using CFT Backflush Improve Productivity & Extend Column Life with Backflush Using RTL & 3-Way Splitter to Identify Unknown in Strawberry Extract Improve Method Translation for Fast & High Res Separations Screen Haz Chems in Homeland Security & Environ Samples Using GC/MS/ECD/FPD & DRS New Tools for Rapid Analysis in High Matrix Samples Precise Time-Scaling of GC Methods w/ Method Translation & RTL RTL: Concepts & Applications Large Volume Injection for GC Using a PTV Inlet
Sample Preparation
Residues in Apple Using QuEChERS AOAC Kits by GC/MS Residues in H2O by Stir Bar Sorptive Exctraction-TD with GC/MS Residues in Apple by GC/MS using QuEChERS Kits for Preinjection Cleanup Residues in Apples using QuEChERS AOAC Kit by LC/MS/MS Detection Residues in Apple using QuEChERS EN Kits by LC/MS/MS Detection Planar Pesticides in Spinach Using QuEChERS AOAC Kits w/ Graphitized Carbon Residues in Spinach Using QuEChERS AOAC Kit by LC/MS/MS Detection Residues in Spinach Using QuEChERS EN Kit by LC/MS/MS Detection Residues in Spinach Using QuEChERS AOAC Kit by GC/MS Residues in Apple Using QuEChERS EN Kit by GC/MS Multi-Residue in Honey using Solid Supported Liquid Extraction Pesticides in Foods using QuEChERS with 320-MS LC/MS/MS

Agilent Pesticides App Compend

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Agilent Pesticides App Compend

Enviado por

Direitos autorais:

Formatos disponíveis

Solutions that meet your demands for

faster and more condent analysis

Agilent Pesticides Applications Compendium

GC/MS and LC/MS

A quick, cost-effective path through your toughest problems

> Return to Table of Contents > Search entire document

Agilent Analyzers and Application Kits

Triple Quadrupole LC/MS Pesticide Application Kit

Counts vs. acquisition time (min)

Ordering information: Agilents LC/MS Pesticide Application Kit (G1733AA)

Put your lab on the productivity fast track.

READY TO GO FOR PESTICIDES AND ENVIRONMENTAL POLLUTANTS

Agilent Analyzers and Applications Kits

medium matrix interference (overlapping peak gives incorrect ion ratio)

strong matrix interference (overlapping peak gives inaccurate quantitation result)

Two alternative transitions of Methamidophos

minimum matrix interference

minimum matrix interference

11.9 12 12.1 12.2 Aquisition Time (min)

Put your lab on the productivity fast track.

Agilent Analyzers and Application Kits

GC/MSD Pesticide Analyzer

Deconvolution peaks and spectra

Put your lab on the productivity fast track.

SP1 7890-0457: Japanese Positive List DRS Screening GC/MSD Analyzer

Agilent Analyzers and Application Kits

TOF and QTOF LC/MS Pesticide Application Kit

Ordering information: Agilents (Q)TOF LC/MS Pesticide Application Kit (G6854AA)

Put your lab on the productivity fast track.

Agilent Analyzers and Application Kits:

A quick, cost-effective path through your toughest problems

Actual kit contents may vary based on application.

Simplify your startup and reduce the time to technical mastery

Lower your method development costs by up to 90%

Analyzers also include:

Implement advanced technologies quickly and confidently

RTL databases for 927 pesticide compounds

Synchronous SIM/SCAN App. specific supplies

Dynamic MRM PCD or DMRM database

Sample Prep QuEChERS

Start your lab on the fast track to success.

For more information

Contact your Agilent Representative today to discuss your specific needs.

GC/MS and GC/MS/MS

> Return to Table of Contents > Search entire document

G3445A option 411 Setup

Constant pressure Post-column backflush MS/MS

1070+ compounds MRM Database

G3445A option 412 Setup

Constant flow Mid-column backflush EPC

1070+ compounds MRM Database

Provides ultimate performance and the shortest cycle time

Method Parameters Included with the Databases

Run time Column flow Feature

Locking compound and RT MS source temperature Quad temperature Backflush

Chlorpyrifos-methyl locked to 18.111 min 300 C Q1 = Q2 = 180 C Mid-column, post-run

Chlorpyrifos-methyl locked to 9.143 min 300 C Q1 = Q2 = 180 C Mid-column, post-run

Average and exact Molecular Weight

Each compound is classified in two ways

MassHunter format for building acquisition methods

The scale and relative intensities of transitions

Table 2 gives a breakdown of the compounds included in the database.

Results and Discussion