Research Article

Received: 4 May 2010 Accepted: 15 June 2010 Published online in Wiley Interscience: 16 July 2010

(www.interscience.com) DOI 10.1002/jms.1784

Potential of atmospheric pressure chemical ionization source in GC-QTOF MS for pesticide residue analysis
a A. Newtonb and P. Hancockb a J. V. Sancho,a F. Hernandez, T. Portoles,
The potential applications of a new atmospheric pressure source for GC-MS analysis have been investigated in this work. A list of around 100 GC-amenable pesticides, which includes organochlorine, organophosphorus and organonitrogenated compounds, has been used to evaluate their behavior in the new source. Favoring the major formation of the molecular ion in the source has been the main goal due to the wide-scope screening possibilities that this fact brings in comparison with the traditional, highly fragmented electron ionization spectra. Thus, the addition of water as modier has been tested as a way to promote the generation of protonated molecules. Pesticides investigated have been classied into six groups according to their ionization/fragmentation behavior. Four of them are characterized by the abundant formation of the protonated molecule in the atmospheric pressure source, mostly being the base peak of the spectrum. These results show that wide-scope screening could be easily performed with this source by investigating the presence of the protonated molecule ion, MH+. The developed procedure has been applied to pesticide screening in different food samples (nectarine, orange and spinach) and it has allowed the presence of several pesticides to be conrmed such as chlorpyriphos ethyl, deltamethrin and endosulfan sulfate. The availability of a quadrupole time-of-ight instrument made it feasible to perform additional MS/MS experiments for both standards and samples to go further in the conrmation of the identity of the detected compounds. Results shown in this paper have been obtained using a prototype source which exhibits promising features that could be applied to other analytical problems apart from those illustrated in this work. Copyright c 2010 John Wiley & Sons, Ltd. Keywords: atmospheric pressure chemical ionization; GC; quadrupole time-of-ight mass spectrometry; pesticides; screening

Introduction
The increasing use of high-resolution full spectrum acquisition techniques in the last decade, such as time-of-ight mass spectrometry (TOF MS), has allowed a huge amount of chemical information on sample composition to be obtained, widening the number of analytes that can be investigated in a single experiment. This analyzer provides the selectivity and sensitivity required for efcient, wide-scope screening, as it combines high full-spectral sensitivity with high mass resolution. The useful and accurate mass data obtained can be processed in both target and/or non-target way, which gives the instrument an interesting versatility depending on the aim of the analysis and allows the user to look at an analytical problem from a different point of view.[1 5] Additionally, the full spectrum dataset remains and offers the analyst the possibility of performing a retrospective analysis, i.e. to make a careful examination of old raw data looking for the presence of any compound that becomes interesting later. All these characteristics make this technique ideal to perform screening analysis that provides greater analytical information and allows the user to efciently discriminate samples with no detectable residues from those with contaminants or residues at a relevant level. In the pesticide residue analysis eld, there is a clear trend toward liquid chromatography coupled to mass spectrometry (LC-MS) as new pesticides are more polar, less volatile and thermostable, and consequently, less GC amenable. However, high usage pesticides (in Europe or in developing countries) are still volatile and thermostable, therefore GC-MS methods cannot be abandoned yet.

Regarding the huge amount of full spectrum with accurate mass data generated by TOF instruments, choosing the right strategy to get the maximum information from the data is one of the major keys to success. The way to proceed is mainly driven by the kind of mass spectrum delivered by the system. Normally, in LC-TOF MS analysis, the ionization occurs at atmospheric pressure by electrospray ionization (ESI), atmospheric pressure chemical ionization (APCI), atmospheric pressure photo ionization[3,6,7] or ambient ionization, e.g. direct analysis in real time or desorption ionization.[8,9] These types of ionization processes promote the formation of protonated or deprotonated molecules with very little fragmentation, yielding typically the [M+H]+ or [MH] ions as the base peak of the spectrum. Recent screening applications have been developed using LC-(ESI)TOF MS technique.[3,10,11] The predictable presence of the (de)protonated molecule in LC-TOF data has allowed an automatic and rapid post-target searching[12] for many LC amenable compounds (including pesticides, antibiotics, veterinary drugs and banned dyes, among others, as well as several metabolites) by extracting chromatograms, with narrow-mass windows, at the exact mass

Correspondence to: J. V. Sancho, Research Institute for Pesticides and Water, University Jaume I, Castell on 12071, Spain. E-mail: sanchoj@qfa.uji.es

a Research Institute for Pesticides and Water, University Jaume I, Castell on 12071, Spain b Waters Corporation MS Technologies Centre, Atlas Park, Simonsway, Manchester M22 5PP, UK

926

J. Mass. Spectrom. 2010, 45, 926936

Copyright c 2010 John Wiley & Sons, Ltd.

Atmospheric pressure chemical ionization source in GC-QTOF MS of the protonated molecule. Standards of the most frequently detected compounds were also injected, so that the information of retention time and in-source fragmentation helped to increase the condence in assigning the identity of the detected compound. When using GC-TOF MS, ionization normally occurs under vacuum conditions either by EI or CI. EI is the most widely applied ionization technique as it is a robust source, which produces standardized mass spectra that can be compared with those from commercially available MS libraries. However, its extensive fragmentation can be a drawback for some applications when wide-scope screening is pursued. As we have recently shown, GC-(EI)TOF MS allows a rapid and automatic accurate mass screening of target analytes using extracted ion chromatograms with narrow-mass window (nw-XICs). In this way, the chemical background is notably reduced and the selectivity improved for the analysis of complex matrices.[1,2] However, the low probability of obtaining the molecular ion (M+) in most of the EI spectra forces the user to know in advance the exact masses of the main analyte fragments in order to perform the nw-XIC at those m/z ions. To obtain this information it is common to rely on the injection of standards into the GC-MS system or retrieving EI spectra from commercial libraries. This is a time-consuming procedure especially when no reference standard is available and the information must be taken from commercial libraries where the data are registered in nominal mass and the elucidation of the possible structure of the fragment ions becomes more difcult. Soft ionization techniques in GC-MS would overcome this issue, in those applications where the extensive EI fragmentation and the absence of the molecular ion in the spectrum could be considered a disadvantage. The high probability of the presence of the molecular ion would enhance the screening reliability, as the information on the molecular ion can be used to limit the number of potential candidates for a given compound, even more if accurate masses are acquired. Additionally, nal identication would be supported by fragmentation experiments, which could be easily achieved using hybrid quadrupole TOF MS (QTOF MS). CI source is a long-term available soft ionization technique for GC instruments, although it has been less frequently used than EI. Positive or negative CI gives better selectivity for several pesticides compared to EI. This fact results in chromatograms with reduced matrix interference,[13 16] but with higher signal intensity variation of different pesticides compared to EI ionization. Preferentially, GC-MS with CI is focused on special substance classes only, e.g. organohalogen pesticides,[16 18] pyrethroids,[19] polybrominated diphenyl ethers[13,14] and organophosphates.[20] It is rarely used in multi-residue methods, because it is not as universal ionization technique as EI and requires several injections of the sample to cover a wider analytes range.[21,22] Atmospheric pressure ionization has primarily been used to interface an MS with LC, as mentioned earlier. However, this attractive ionization interface can also be applied to GC. The ionization mechanism employed by the APCI source is lowenergy (soft), which generates spectral data typically rich in molecular or quasi-molecular ion information and ideal for compound conrmation. The rst few developments with APCI were carried out by Horning et al.[23,24] in the 1970s who were also the rst to interface a GC instrument to an APCI ion source. Since these initial publications, a series of papers in the late 1980s have been published by Korfmacher and coworkers[25 27] in which the efuent from a gas chromatograph is ionized at atmospheric pressure. However, GC/APCI was never fully commercialized, probably because of the high costs of the specialized instrumentation needed for these analyses at that time. Nowadays, new APCI sources are commercially available and capable to be interfaced with both GC and LC instruments.[28,29] This fact adds versatility and extends analytical capabilities providing exibility to determine volatile and semivolatile compounds of low and intermediate polarity, traditionally analyzed by dedicated vacuum GC-MS instruments. Very recently, GC/APCI in combination with TOF MS has been studied for its applicability to metabolic proling.[30] The aim of this work is to study the applicability of a new soft ionization APCI source, waters atmospheric pressure (APGC), using a GC-QTOF MS instrument to perform rapid and wide-scope screening and identication of pesticides residues in food samples. The predictable presence of the molecular ion/protonated molecule in the GC-(APCI)TOF MS spectrum allowed the rapid application of a screening method in a post-target way, easily generating a list of compounds to be monitored making use of their molecular formulae, i.e. exact masses of their protonated molecules. The behavior of around 100 pesticides using GC-(APCI)QTOF MS instrument has been studied. The developed procedure has been applied to the screening of pesticides in fruit and vegetable samples such as nectarine, orange and spinach. MS/MS experiments have been performed to go further in the conrmation of the identity of the compounds detected; thanks to the product ion spectra at accurate mass provided by the QTOF MS instrument.

Experimental
Reagents Reference standards of pesticides were purchased from Dr Ehrenstorfer (Augsburg, Germany). From solid reference standards, stock solutions (around 500 g/ml) were prepared by dissolving the standard in acetone and stored in a freezer at 20 C. Working solutions were prepared by diluting stock solutions in hexane. Instrumentation GC-(EI)TOF MS For the GC instrumentation, an Agilent 6890N GC system (Palo Alto, CA, USA) equipped with an Agilent 7683 autosampler was coupled to a TOF mass spectrometer, GCT (Waters Corporation, Manchester, UK), operating in EI mode. The GC separation was performed using a fused silica HP-5 MS capillary column with a length of 30 m 0.25 mm i.d. and a lm thickness of 0.25 m (J&W Scientic, Folson, CA, USA). The oven temperature was programmed as follows: 90 C (1 min); 5 C/min to 300 C (2 min). Splitless injections of 1 l sample were carried out. Helium was used as carrier gas at 1 ml/min. The interface and source temperatures were both set to 250 C and a solvent delay of 3 min was selected. TOF MS was operated at 1 spectrum/s acquiring the mass range m/z 50650 using a multi-channel plate voltage of 2800 V. TOF MS resolution was about 8500 (FWHM) at m/z 614. The application manager ChromaLynx XS, a module of MassLynx software, was used to investigate the presence of target compounds in samples. GC-(APCI)(Q)TOF MS For the GC instrumentation, an Agilent 7890A GC system (Palo Alto, CA, USA) equipped with an Agilent 7683 autosampler was

927

J. Mass. Spectrom. 2010, 45, 926936

Copyright c 2010 John Wiley & Sons, Ltd.

www.interscience.wiley.com/journal/jms

 et al. T. Portoles

Figure 1. (A) Extracted ion chromatogram at M+ of dieldrin and TOF MS spectrum in EI source. (B) Extracted ion chromatogram at M+ of dieldrin and TOF MS spectrum in APCI source.

coupled to a quadrupole TOF mass spectrometer, Xevo QTOF (Waters Corporation, Manchester, UK), operating in APCI mode. The GC separation was performed using a fused silica DB-5 MS capillary column with a length of 30 m 0.25 mm i.d. and a lm thickness of 0.25 m (J&W Scientic). The oven temperature was programmed as follows: 70 C (1.5 min); 25 C/min to 180 C (3 min); 5 C/min to 310 C (5.1 min). Split injections (10 : 1) of 1 l sample were carried out at 280 C. Helium was used as carrier gas at 1.2 ml/min. In charge transfer mode, the interface temperature was set to 310 C using N2 as an auxiliary gas at 400 l/h, a make-up gas at 400 ml/min and a cone gas at 50 l/h. The APCI corona pin was operated at 0.4 A with a cone voltage of 30 V. In proton transfer mode using water, the interface temperature was set to 310 C using N2 as an auxiliary gas at 400 l/h, a make-up gas at 400 ml/min and a cone gas at 20 l/h. The APCI corona pin was operated at 1.7 A with a cone voltage of 20 V. The ionization process occurred within an enclosed ion volume, which enabled control over the protonation/charge transfer processes. Xevo QTOF MS was operated at 1 spectrum/s acquiring the mass range m/z 20700 using a multi-channel plate voltage of 2300 V. TOF MS resolution was approximately 10 000 (FWHM) at m/z 614. Sample treatment Orange, nectarine and spinach samples were bought from local markets in Barcelona (Spain). The samples were chopped, homogenized and stored in a freezer at 20 C until sample treatment. Sample treatment can be found elsewhere.[31] Briey, 10 g of sample was subjected to accelerated solvent extraction procedure

with ethyl acetate. In the case of spinach samples, a gel-permeation chromatography clean-up step was also applied before injection into the GC-MS system.

Results and Discussion

Preliminary experiments The soft ionization behavior of the new interface was tested using standards in solvent on molecules with a well-known high degree of fragmentation in EI spectra. As illustrative example, we show the MS spectra for dieldrin (Fig. 1(A)). In this particular case, when an nw-XIC at the M+ was extracted, no signal was observed in the chromatogram, as the molecular ion was practically absent in the EI spectrum. However, very different results were obtained using the softer ionization produced by the APCI interface (Fig. 1(B)). In this case, an important peak was located (tR = 16.56 min) when extracting an nw-XIC at the M+ of dieldrin as can be seen in Fig. 1(B). When looking at the two MS spectra, very different fragmentation patterns were observed. In contrast with the EI spectrum, where the molecular ion was almost absent, in the APCI spectrum M+ has become the base peak. This behavior, when using N2 as make-up gas, could be explained using the following molecule reactions where M represents the analyte. The corona discharge needle creates a nitrogen plasma, N2 + and N4 + , which in the case of charge transfer reacts directly with analyte molecules (Fig. 2(A)). These satisfactory results encouraged us to test this source on pesticides from the endosulfan family, which are known to have signicantly fragmentated EI spectra. In the case of -endosulfan, no chromatographic peak (tR = 15.69 min) was observed when

928

www.interscience.wiley.com/journal/jms

Copyright c 2010 John Wiley & Sons, Ltd.

J. Mass. Spectrom. 2010, 45, 926936

Atmospheric pressure chemical ionization source in GC-QTOF MS

(A)
Corona discharge needle

(B)
Corona discharge needle

facilitate the formation of the protonated molecule. The modier was placed in an uncapped vial, which was located within a specially designed holder placed in the source door. In this case, the results obtained were even more satisfactory as fragmentation decreased and the protonated molecule was converted in the base peak of the spectrum (Fig. 3(B)). This proton transfer behavior could be explained because the nitrogen plasma reacts with water or any proton source, and indirectly transfers protons to the analyte of interest (Fig. 2(B)). Systematic study At this point, a wider study of the APCI possibilities for screening purposes was carried out. Around 100 GC-amenable pesticides of different families (organochlorine, organophosphorus, organonitrogenated and others) were selected to study their APCI behavior. Standards in solvent were used for this purpose, which were injected into the GC(APCI)TOF MS instrument under generic chromatographic conditions. APCI-TOF MS data were evaluated as regards the presence/absence of the M+ in the spectrum obtained under N2 charged transfer conditions. The presence/absence of the MH+ was also investigated as a consequence of the observed behavior for some compounds to be protonated during preliminary experiments due to water traces present in the source. In a second step, water was added as a modier and the presence/absence and/or improvement on the signal of the protonated molecule was evaluated. All gathered information was nally used to classify the

Figure 2. Molecule reactions when using APCI source. (A) N2 transfer conditions and (B) proton transfer conditions. M, analyte.

doing an nw-XIC at its M+ on APCI data (Fig. 3(A)). However, when looking at the APCI spectrum, the expected isotopic pattern of hexachlorinated -endosulfan was clearly observed at MH+ instead of M+. An explanation of this fact might be the presence of water vapor traces in the source, which readily promote the formation of the protonated molecule instead of the molecular ion. This marked tendency of the molecule to be protonated led us to repeat the experiment but introducing water as a modier to

Figure 3. (A) Extracted ion chromatogram at M+ of -endosulfan and TOF MS spectrum when using APCI source. (B) Extracted ion chromatogram at MH+ of -endosulfan and spectrum in the APCI source using water as modier.

929

J. Mass. Spectrom. 2010, 45, 926936

Copyright c 2010 John Wiley & Sons, Ltd.

www.interscience.wiley.com/journal/jms

 et al. T. Portoles

Does M+ exist under N2-transfer conditions?

NO

YES

Does MH+ exist?

Does MH+ exist?

NO

YES

NO

YES

Does MH+ appear under proton transfer conditions?

Does MH+ improve its sensitivity under proton transfer conditions?

Does MH+ appear under proton transfer conditions?

Does MH+ improve its sensitivity under proton transfer conditions?

NO

YES

YES

NO

YES

YES

Group A

Group B

Group C

Group D

Group E

Group F

Figure 4. Scheme of the different behaviors of pesticides when using the APCI source.

compounds into six different groups according to their ionization behavior in the APCI source (Fig. 4). The rst group (A) included those compounds for which no presence of M+ or MH+ was observed in the APCI data under N2 - or proton transfer conditions (Fig. 5(A)). This behavior was observed for labile organochlorine insecticides that do not have an acidic group in their structure capable to be protonated (HCH isomers, p,p - DDT and mirex). As regard Group B, this was formed by those compounds that, although no presence of M+ or MH+ was observed in the APCI data with N2 , showed an abundant formation of MH+ when introducing water as a modier (Fig. 5(B) shows the MS spectra for phosmet, one of the pesticides included within Group B). For some compounds of this group, the addition of water as a modier was so favorable that the MH+ ion was the base peak of the spectrum. Within the compounds that did not show the presence of M+ in the N2 -transfer APCI spectrum, we might consider another group (C) characterized because MH+ was present in their spectrum. We proved that the formation of the protonated ion was highly favored when adding water as a modier with MH+ becoming the base peak of the spectrum (Fig. 5(C) shows propiconazole as an illustrative example). The rest of the pesticides, which presented M+ ion in the N2 APCI spectrum, were classied into another three groups. Group D was formed by those compounds that did not show MH+ in the N2 APCI spectrum or after adding water. This happened for organochlorine insecticides where no acid group capable of being protonated was present in their structure (pentachlorobenzene, isodrin, HCB, p,p -DDE, trans-chlordane, p,p -DDD, heptachlor and uvalinate). In these cases, when adding water as a modier, little change was observed in the spectrum (Fig. 6(D) shows HCB as an example). Another group (E) included those compounds that did not show MH+ in the N2 APCI spectrum, but protonating after water addition (Fig. 6(E) shows mevinphos as an example). The last group of compounds (F) was the most numerous. Both M+ and MH+ were present in the N2 APCI spectrum. Obviously, when

adding water as a modier, the formation of MH+ was favored and it nally became the base peak of the spectrum in most cases (Fig. 6(F) shows deltamethrin as an example). Groups of pesticides according to their ionization/fragmentation behavior in the APCI source are shown in Table 1. In light of our results, it seems that an APCI source using water as a modier gives us condence that the protonated molecule will be observed for the majority of the pesticides. Additionally, in most cases the protonated molecule is the base peak of the spectrum and very low fragmentation of the molecule occurs in APCI ionization compared to the well known EI spectra. This fact allows us to perform a wide-scope screening of pesticides in a different way to the traditional approach based on EI spectra. The low probability of observing the intact molecular ion (M+) in conventional EI spectra forces the analyst to know in advance the main fragment ions in order to select the appropriate nw-XIC at their exact masses. This information could be taken from the injection of standards into the GC-MS system or from commercial EI spectra libraries, resulting in a time-consuming procedure especially when no standard is available and the information has to be taken from commercial nominal mass libraries. However, using the new APGC source, the predictable presence of the protonated molecule in GC-(APCI)TOF MS data would allow an automatic and rapid post-target search of around 100 pesticides by only extracting the chromatogram at the exact mass of the MH+ ion. Then, we were able to use the same approach developed for UHPLC(ESI)QTOFMS using the ChromaLynx XS application manager, building a list of compound names and molecular formulae and performing fast, sensitive and wide-scope screening for GC-amenable pesticides. In addition, favoring the presence of the protonated molecule, the sensitivity improves due to the higher abundance of the ion monitored. Thus, the best option to perform wide-scope screening would be to use water as a modier, and investigate the presence of the MH+ ion as it was present in the APCI spectra of 90% of the pesticides studied in this work.

930

www.interscience.wiley.com/journal/jms

Copyright c 2010 John Wiley & Sons, Ltd.

J. Mass. Spectrom. 2010, 45, 926936

931

J. Mass. Spectrom. 2010, 45, 926936

Atmospheric pressure chemical ionization source in GC-QTOF MS

Copyright c 2010 John Wiley & Sons, Ltd.

www.interscience.wiley.com/journal/jms

Figure 5. APCI spectra of selected pesticides, belonging to Groups A, B and C: mirex (A), phosmet (B) and propiconazole (C). Spectra without adding water as modier (top) and spectra using water as modier (bottom).

932
et al. T. Portoles

www.interscience.wiley.com/journal/jms

Copyright c 2010 John Wiley & Sons, Ltd.

J. Mass. Spectrom. 2010, 45, 926936

Figure 6. APCI spectra of selected pesticides belonging to Groups D, E and F: hexachlorobenzene (D), mevinphos (E) and deltamethrin (F). Spectra without adding water as modier (top) and spectra using water as modier (bottom).

Atmospheric pressure chemical ionization source in GC-QTOF MS

Table 1. Groups of pesticides according to its ionization/fragmentation behavior in APCI source Group A -HCH -HCH -HCH -HCH p,p -DDT Mirex Group B Malathion Phosmet Terbufos Disulfoton Methidathion Azinphos methyl Group C Methamidophos Dichlorvos Ethoprophos Dimethoate Fonofos Tolclofos methyl Fenchlorphos Chlorfenvinphos -Endosulfan Tetrachlorvinphos Imazalil -Endosulfan Propiconazole I Tebuconazole Pyriproxyfen Iprodione Metoxychlor Group D Pentachlorobenzene HCB p,p -DDE Trans-chlordane p,p -DDD Heptachlor Fluvalinate Group E Isodrin Dieldrin Endosulfan sulfate Mevinphos Methacrifos Heptenophos Chlorpropham Phorate Dichlouanid Tolyluanid Group F Tecnazene Diphenylamine Triuralin Simazine Atrazine Terbuthylazine Propyzamide Pyrimethanil Diazinon Chlorthalonil Etrimfos Endosulfan ether Pirimicarb Fosfamidon Metribuzin Parathion methyl Metalaxil Fenitrothion Pirimiphos methyl Aldrin Fenthion Chlorpyrifos ethyl Parathion ethyl Pirimiphos ethyl Cyprodinil Heptachlor epox A Chlozolinate Quinalphos Procymidone Profenofos Myclobutanil Bupirimate Endrin Oxadixyl Ethion Triazophos Fosalone Lambda-Cyhalothrin I Fenarimol Pyrazophos Pyridaben Coumaphos Cypermethrin Deltamethrin Azoxystrobin Quintozene Chlorpyrifos methyl Alachlor Buprofezin Bifenthrin Permethrin I Fenvalerate Esfenvalerate

A second injection without using water would be carried out in order to investigate the presence of M+ for the rest of compounds, such as HCH isomers, p,p -DDE, p,p -DDD, p,p -DDT, mirex, pentachlorobenzene, HCB, trans-chlordane, heptachlor and uvalinate. Application to real samples The developed procedure was applied to the screening of pesticides in incurred food samples. The analysis of orange, nectarine and spinach samples was carried out in a post-target way performing an accurate mass screen of around 100 pesticides after full MS acquisition using ChromaLynx XS application manager. Briey, this application manager automatically processes MS data and obtains XICs with a user-dened mass window (0.02 Da in this study) at selected m/z ions, typically those corresponding to the exact masses of the protonated molecules, based on a preselected list of exact masses and retention times, if available. Besides, the software allows visualization of the complete TOF MS spectrum of the positive ndings. This facilitates a rapid and simple review by cataloging pollutants by colors, as a function of their attainment of retention time and mass errors. Following this approach, the insecticides chlorpyriphos ethyl and deltamethrin were detected and identied in a nectarine sample. As an example, Fig. 7 illustrates the detection and reliable identication of chlorpyriphos ethyl in incurred nectarine at a concentration level of 0.01 mg/kg. A mass error of 0.9 mDa was obtained for the MH+ ion, and the expected isotopic pattern associated to the presence of three chlorine atoms was also observed.

The availability of a QTOF instrument made it feasible to perform MS/MS experiments for both standards and samples to go further in the conrmation of the identity of the compounds detected. Collision energy of 10 eV was applied for this purpose. Thus, when comparing the relative ion abundances in the suspected positive sample of chlorpyriphos with those of a reference standard, all deviations were within the limits established by general guidelines in residue analysis, as the European Decision 2002/657/EC for contaminants and residues in food of animal origin legislation or Document SANCO/10684/2009 for pesticides control in food and feed.[32,33] Chemical structures for the most abundant product ions were suggested based on the elemental compositions proposed accordingly to the accurate mass measurements given by the instrument for these ions. All mass errors for chlopyriphos product ions in the nectarine sample were below 2.7 mDa as depicted in Fig. 8. Finally, retention times for the reference standard and sample peak were also compared, obtaining a deviation lower than 0.5%. Therefore, this sample was unequivocally conrmed by QTOF to be positive for chlorpyriphos ethyl. Similarly, positive ndings of deltamethrin in nectarine and endosulfan sulfate in orange could be reported. In all cases, MS/MS experiments, selecting MH+ as precursor ion, led to the reliable conrmation of their identity.

Conclusions
The usefulness of a new atmospheric pressure source for GC-MS systems is reported in this work. The predictable presence of the protonated molecule when using this prototype APGC source in a GC-TOF MS system has allowed us to perform rapid and

933

J. Mass. Spectrom. 2010, 45, 926936

Copyright c 2010 John Wiley & Sons, Ltd.

www.interscience.wiley.com/journal/jms

 et al. T. Portoles

(A)

(B)

(C)

Figure 7. Detection and identication of chlorpyriphos ethyl by GC-(APCI)TOF MS in a nectarine sample. (A) List of compounds investigated together with their molecular formulae, exact mass and retention time when available. (B) Extracted ion chromatogram at m/z 349.9341 (MH+). (C) APCI mass spectrum of chlorpyriphos ethyl in the nectarine sample.

wide-scope screening of around 100 pesticides in food samples. The use of water as a modier favored the formation of the protonated molecule, which was present in the TOF MS spectra of 90% of the pesticides investigated. Additionally, in most cases the protonated molecule was the base peak of the spectrum, and very low fragmentation was observed in APCI ionization when compared with the well-known EI spectra. This facilitates rapid and sensitive screening, searching for the protonated pesticide molecule, which is typically absent in the highly fragmented EI spectra. The ChromaLynx XS application manager allowed a list of compound names, molecular formulae, and exact masses to be built making it easier to perform fast and wide-scope screening of GC-amenable pesticides. The developed procedure has been applied to the analysis of fruit and vegetable samples with the result of detecting and identifying several pesticides. Additionally, the availability of a QTOF instrument made it feasible to perform MS/MS experiments to go further in the conrmation of the identity of the compounds detected in samples. The results shown in this paper have been obtained using a prototype source, which exhibits promising features that might also be applied to other analytical approaches. For instance, to facilitate the detection and conrmation step in only one injection,

a collision cell fragmentation might be performed. In this case, two acquisition functions would be monitored: the rst one, at low collision energy, to detect the protonated molecule and the second one, at high collision energy, as a conrmatory function obtaining a similar fragmentation to that of MS/MS experiments. This acquisition mode, known as MSE , has shown excellent results in LC-QTOF MS applications investigated by our group,[34] and provides reproducible fragmented spectra without the need for ion pre-selection in the quadrupole. The reduced fragmentation generated by the APGC source will surely have a signicant impact on target analysis at trace levels. The high degree of fragmentation for many compounds when using EI constitutes a problem when selecting the precursor ion in tandem MS experiments. Often, this fact reduces selectivity and sensitivity of analysis. With the reduced fragmented spectrum given by the APGC source, the precursor ion selection no longer requires a compromise between selectivity and sensitivity allowing more convenient tandem MS experiments. Preliminary results shown in this paper are promising, but further research is required to have a better knowledge on the possibilities of this source in different elds, e.g. environmental MS or food safety.

934

www.interscience.wiley.com/journal/jms

Copyright c 2010 John Wiley & Sons, Ltd.

J. Mass. Spectrom. 2010, 45, 926936

Atmospheric pressure chemical ionization source in GC-QTOF MS

Figure 8. Product ion spectra (collision energy, 10 eV) for chlorpyriphos ethyl (precursor ion, m/z 350) from a nectarine positive sample (top) and the reference standard (bottom). Chemical structures proposed for the most abundant product ions.

Acknowledgments The authors are very grateful to the Serveis Centrals Cientca (SCIC) of the University Jaume I for dInstrumentacio the use of GC-TOF MS (GCT) and to the MS Technologies Center (Waters Corporation, Manchester, UK) for using the GC-APGCXevo QTOF. The authors acknowledge the nancial support of Generalitat Valenciana, as research group of excellence PROMETEO/2009/054. The authors also acknowledge the Agencia de Salud Publica (Barcelona) for providing the samples and Ines is very pleased to the Cervera for sample treatment. T. Portoles University Jaume I for her post-doctoral grant.

References
[1] F. Hern andez, T. Portol es, E. Pitarch, F. J. Lopez. Searching for anthropogenic contaminants in human breast adipose tissues using gas chromatography-time-of-ight mass spectrometry. J. Mass Spectrom. 2009, 44, 1. [2] F. Hern andez, T. Portol es, E. Pitarch, F. J. Lopez. Target and nontarget screening of organic micropollutants in water by solidphase microextraction combined with gas chromatography/highresolution time-of-ight mass spectrometry. Anal. Chem. 2007, 79, 9494. [3] M. Ib anez, J. V. Sancho, F. Hern andez, D. McMillan, R. Rao. Rapid non-target screening of organic pollutants in water by ultraperformance liquid chromatography coupled to time-of-light mass spectrometry. Trends Anal. Chem. 2008, 27, 481.

[4] T. Portol es, M. Ib anez, J. V. Sancho, F. J. Lopez, F. Hern andez. Combined use of GC-TOF MS and UHPLC-(Q)TOF MS to investigate the presence of nontarget pollutants and their metabolites in a case of honeybee poisoning. J. Agric. Food Chem. 2009, 57, 4079. [5] T. Portol es, E. Pitarch, F. J. Lopez, J.V Sancho, F. Hern andez. Methodical approach for the use of GC-TOF MS for screening and conrmation of organic pollutants in environmental water. J. Mass Spectrom. 2007, 42, 1175. [6] H. Farwanah, J. Wirtz, T. Kolter, K. Raith, R. H. H. Neubert, K. Sandhoff. Normal phase liquid chromatography coupled to quadrupole time of ight atmospheric pressure chemical ionization mass spectrometry for separation, detection and mass spectrometric proling of neutral sphingolipids and cholesterol. J. Chromatogr. B 2009, 877, 2976. [7] N. Yoshioka, Y. Akiyama, T. Matsuoka, T. Mitsuhashi. Rapid determination of ve post-harvest fungicides and metabolite in citrus fruits by liquid chromatography/time-of-ight mass spectrometry with atmospheric pressure photoionization. Food Control 2010, 21, 212. [8] E. Jagerdeo, M. Abdel-Rehim. Screening of cocaine and its metabolites in human urine samples by direct analysis in real-time source coupled to time-of-ight mass spectrometry after online preconcentration utilizing microextraction by packed sorbent. J. Am. Soc. Mass Spectrom. 2009, 20, 891. [9] N. A. Hagan, T. J. Cornish, R. S. Pilato, K. A. Van Houten, M. D. Antoine, T. P. Lippa, A. F. Becknell, P. A. Demirev. Detection and identication of immobilized low-volatility organophosphates by desorption ionization mass spectrometry. Int. J. Mass Spectrom. 2008, 278, 158. C. Guerrero. Antibiotic residue [10] F. Hern andez, J. V. Sancho, M. Ib anez, determination in environmental waters by LC-MS. Trends Anal. Chem. 2007, 26, 466.

935

J. Mass. Spectrom. 2010, 45, 926936

Copyright c 2010 John Wiley & Sons, Ltd.

www.interscience.wiley.com/journal/jms

 et al. T. Portoles
[11] I. Ferrer, E. M. Thurman, J. A. Zweigenbaum. Screening and conrmation of 100 pesticides in food samples by liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2007, 21, 3869. [12] F. Hern andez, O. J. Pozo, J. V. Sancho, F. J. Lopez, J. M. Marn, M. Ib anez. Strategies for quantication and conrmation of multiclass polar pesticides and transformation products in water by LC-MS2 using triple quadrupole and hybrid quadrupole time-ofight analyzers. Trends Anal. Chem. 2005, 24, 596. [13] C. M. Medina, E. Pitarch, F. J. Lopez, C. V azquez, F. Hern andez. Determination of PBDEs in human breast adipose tissues by gas chromatography coupled with triple quadrupole mass spectrometry. Anal. Bioanal. Chem. 2008, 390, 1343. [14] C. M. Medina, E. Pitarch, T. Portol es, F. J. Lopez, F. Hern andez. GC-MS/MS multi-residue method for the determination of organochlorine pesticides, polychlorinated biphenyls and polybrominated diphenyl ethers in human breast tissues. J. Sep. Sci. 2009, 32, 2090. [15] M. D. Hernando, A. Aguera, A. R. Fern andez-Alba, L. Piedra, M. Contreras. Gas chromatographic determination of pesticides in vegetable samples by sequential positive and negative chemical ionization and tandem mass spectrometric fragmentation using an ion trap analyser. Analyst 2001, 126, 46. s, F. J. Lopez, [16] E. Pitarch, C. Medina, T. Portole F. Hern andez. Determination of priority organic micro-pollutants in water by gas chromatography coupled to triple quadrupole mass spectrometry. Anal. Chim. Acta 2007, 583, 246. [17] F. Artigas, E. Martnez, E. Gelpi. Organochlorine pesticides by negative ion chemical ionization. Brain metabolites of lindane. Biomed. Environ. Mass Spectrom. 1988, 16, 279. [18] R. Chaler, R. Vilanova, M. Santiago-Silva, P. Fern andez, J. O. Grimalt. Enhanced sensitivity in the analysis of trace organochlorine compounds by negative-ion mass spectrometry with ammonia as reagent gas. J. Chromatogr. A 1998, 823, 73. [19] A. Ramesh, P. E. Ravi. Negative ion chemical ionization-gas chromatographic-mass spectrometric determination of residues of different pyrethroid insecticides in whole blood and serum. J. Anal. Toxicol. 2004, 28, 660. [20] M. V. Russo, L. Campanella, P. Avino. Determination of organophosphorus pesticide residues in human tissues by capillary gas chromatography-negative chemical ionization mass spectrometry analysis. J. Chromatogr. B 2002, 780, 431. [21] L. Alder, K. Greulich, G. Kempe, B. Vieth. Residue analysis of 500 high priority pesticides: better by GC-MS or LC-MS/MS? Mass Spectrom. Rev. 2006, 25, 838. [22] J. L. Martnez Vidal, F. J. Arrebola Li ebanas, M. J. Gonz alez Rodrguez, A. Garrido Frenich, J. L. Fern andez Moreno. Validation of a gas chromatography/triple quadrupole mass spectrometry based method for the quantication of pesticides in food commodities. Rapid Commun. Mass Spectrom. 2006, 20, 365. [23] E. C. Horning, M. G. Horning, D. I. Carroll, I. Dzidic, R. N. Stillwell. New picogram detection system based on a mass spectrometer with an external ionization source at atmospheric pressure. Anal. Chem. 1973, 45, 936. E. C. Horning, D. I. Carroll, I. Dzidic. Development and use of analytical systems based on mass spectrometry. Clin. Chem. 1977, 23, 13. W. A. Korfmacher, L. G. Rushing, R. J. Engelbach, J. P. Freeman, Z. Djuric, E. K. Fifer, F. A. Beland. Analysis of three aminonitropyrene isomers via fused silica gas chromatography combined with negative ion atmospheric pressure ionization mass spectrometry. J. High Resolut. Chromatogr., Chromatogr. Commun. 1987, 10, 43. W. A. Korfmacher, L. G. Rushing, J. Arey, B. Zielinska, J. N. Pitts Jr. Identication of mononitropyrenes and mononitrouoranthenes in air particulate matter via fused silica gas chromatography combined with negative ion atmospheric pressure ionization mass spectrometry. J. High Resolut. Chromatogr., Chromatogr. Commun. 1987, 10, 641. T. Kinouchi, A. T. L. Miranda, L. G. Rushing, F. A. Beland, W. A. Korfmacher. Detection of 2-aminouorene at femtogram levels via high-resolution gas chromatography combined with negative ion atmospheric pressure ionization mass spectrometry. J. High Resolut. Chromatogr. 1990, 13, 281. C. N. McEwen, R. G. McKay. A combination atmospheric pressure LC/MS:GC/MS ion source: advantages of dual AP-LC/MS:GC/MS instrumentation. J. Am. Soc. Mass Spectrom. 2005, 16, 1730. R. Schiewek, M. Lorenz, R. Giese, K. Brockmann, T. Benter, S. G ab, O. J. Schmitz. Development of a multipurpose ion source for LC-MS and GC-API MS. Anal. Bioanal. Chem. 2008, 392, 87. A. Carrasco-Pancorbo, E. Nevedomskaya, T. Arthen-Engeland, T. Zey, G. Zurek, C. Baessmann, A. M. Deelder, O. A. Mayboroda. Gas chromatography/atmospheric pressure chemical ionization-time of ight mass spectrometry: analytical validation and applicability to metabolic proling. Anal. Chem. 2009, 81, 10071. M. I. Cervera, C. Medina, T. Portol es, E. Pitarch, J. Beltr an, E. Serrahima, L. Pineda, G. Munoz, F. Centrich, F, Hern andez. Multiresidue determination of 130 multiclass pesticides in fruits and vegetables by gas chromatography coupled to triple quadrupole tandem mass spectrometry. Anal. Bioanal. Chem. 2010, DOI: 10.1007/s00216-010-3597-8. European Commission Decision 2002/657/EC, implementing Council Directive 96/23/EC concerning the performance of analytical methods and the interpretation of results. Off. J. Eur. Commun. 2002, L221, 8. Method validation and quality control procedures for pesticide residues analysis in food and feed. Document No. SANCO/10684/2009. Implemented by 01/01/2010. Rapid wideF. Hern andez, L. Bijlsma, J. V. Sancho, R. Daz, M. Ib anez. scope screening of drugs of abuse, prescription drugs with potential for abuse and their metabolites in inuent and efuent urban wastewater by UHPLC-QTOF MS. Environ. Sci. Technol. 2010, (in press).

[24] [25]

[26]

[27]

[28] [29] [30]

[31]

[32]

[33] [34]

936
www.interscience.wiley.com/journal/jms
Copyright c 2010 John Wiley & Sons, Ltd. J. Mass. Spectrom. 2010, 45, 926936