Journal of Chromatography A, 1513 (2017) 27–34

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Full length article

Determination of volatile organic compounds in pen inks by a

dynamic headspace needle trap device combined with gas
chromatography–mass spectrometry
Xiaohuan Zang, Weiqian Liang, Qingyun Chang, Tong Wu, Chun Wang, Zhi Wang ∗
Department of Chemistry, College of Science, Hebei Agricultural University, Baoding 071001, China

a r t i c l e i n f o a b s t r a c t

Article history: Some harmful volatile organic compounds (VOCs), such as methylbenzene, ethylbenzene, xylene,
Received 8 January 2017 chlorobenzene and bromobenzene, are the commonly found chemicals in pen inks. In this work, a
Received in revised form 16 June 2017 dynamic headspace needle-trap device (D-HS-NTD) with a ZIF-8-derived nanoporous carbon (ZIF-8-NPC)
Accepted 9 July 2017
as the adsorbent was developed for the extraction of some VOCs in different pen inks prior to GC–MS
Available online 11 July 2017
detection. The main important influencing experimental parameters including the flow rate of the purge
gas N2 , extraction temperature, extraction time, desorption temperature and desorption time for the
Keywords:
extraction were optimized to obtain a high extraction efficiency. Under the optimized conditions, a good
Gas chromatography–mass spectrometry
Volatile organic compounds
linearity was obtained in the concentration range of 0.1–400 ␮g kg−1 with the correlation coefficients (r)
Pen inks ranging from 0.9911 to 0.9990 for the eleven VOCs. The LODs at a signal-to-noise ratio of 3 (S/N = 3) were
Dynamic headspace needle-trap device measured to be 10–20 ng kg−1 for the VOCs. The developed method was applied to determine the VOCs
from 20 pen inks. The recoveries of the VOCs for the method at the spiking levels of 0.5 and 20.0 ␮g kg−1
fell in the range from 80.0% to 108%.
© 2017 Published by Elsevier B.V.

1. Introduction
As an important kind of stationery, pen is commonly used in
people’s daily life and work. Pen inks are a complex mixture of some
chemical compounds, including various colorants and other addi-
tives such as surfactants, antioxidants, viscosity adjusters, resins,
glycol and glycerol [1]. Most of the colorants are composed of
volatile organic compounds (VOCs), such as methylbenzene, ethyl-
benzene, bromobenzene, and m-xylene. As a result, the VOCs in pen
inks can be easily breathed in by pen users. It has been reported that
a small amount of the VOCs can cause drowsiness, headache and
vertigo symptoms, and long-term exposure to the VOCs at high con-
centration levels may be responsible for respiratory diseases and
cancer [2]. Therefore, the effective detection of the VOCs in pen inks
is necessary for the assessment of their possible harm to pen users.
Currently the most commonly used methods for the analysis
of VOCs are chromatographic techniques combined with proper
sample pretreatment processes [3]. Several sample pretreatment
techniques have been reported for the determination of VOCs,
∗ Corresponding author at: College of Science, Hebei Agricultural University, Baod-
ing, 071001, Hebei, China.
E-mail addresses: wangzhi@hebau.edu.cn, zhiwang2013@aliyun.com (Z. Wang).
such as static headspace [4], direct aqueous injection (DAI) [5],
membrane extraction [6], hollow-fiber liquid phase microextrac-
tion (HF-LPME) [7], solid-phase microextraction (SPME) [8–10],
and purge and trap [11,12] methods. Recently in-tube or needle-
trap sorptive extraction methods [13–15] have been reported for
sample preparation. In needle-trap sorptive extraction method,
a needle-trap device (NTD), which is commonly composed of a
blunt stainless-steel needle packed with an appropriate sorbent,
is needed [16] and an efficient adsorbent in NTD plays a key role
for the extraction.
Due to their tunable pore size, high surface area and good
thermal stability, metal−organic frameworks (MOFs) have been
actively explored in recent years for the use in the field of sam-
ple preparation. Some of them, such as MIL-101[17–19], ZIF-8
[20,21] and UiO-67 [22] have been reported as the adsorbent for the
extraction of organochlorine pesticides, acidic drugs, nitrobenzene
compounds and polycyclic aromatic hydrocarbons. More recently,
the nanoporous carbons derived from MOFs have intrigued great
research interest [23] and the MOFs-derived nanoporous carbons
have been used in the fields of electrochemical capacitance [24], gas
adsorption [25], catalysis [26], and solid phase extraction [27,28]
due to their excellent adsorptive properties.
In this work, a ZIF-8-derived nanoporous carbon (ZIF-8-NPC)
was synthesized by the direct carbonization of ZIF-8 without any

http://dx.doi.org/10.1016/j.chroma.2017.07.030
0021-9673/© 2017 Published by Elsevier B.V.
28 X. Zang et al. / J. Chromatogr. A 1513 (2017) 27–34
Fig 1. (A) Stainless-steel needle; (B) Needle-trap device (NTD); (C) Dynamic headspace needle-trap device (D-HS- NTD); (D) Thermal desorption at the injection port of
GC–MS.
additional carbon sources and a dynamic headspace needle-trap
device (D-HS-NTD) was prepared with the ZIF-8-NPC as an adsor-
bent. The D-HS-NTD was successfully used for the extraction of
eleven VOCs in pen inks prior to their GC–MS detection.
2. Experimental
2.1. Reagents and materials
The standards of the eleven VOCs (methylbenzene, chloroben-
zene, ethylbenzene, m-xylene, o-xylene, bromobenzene, m-
chlortoluene, mesitylene, m-dichlorobenzene, o-dichlorobenzene
and p-bromotoluene) were purchased from Agricultural Environ-
mental Protection Institution of Tianjin (Tianjin, China). A mixture
stock solution containing each of the eleven VOCs at 500.0 ␮g L−1
was prepared in acetone and stored at 4 ◦ C.
Dimethylformamide (DMF), dichloromethane (CH2 Cl2 ),
hydrofluoric acid (HF, 40%, w/w), trifluoroacetic acid (TFA,
99%), methyltrimethoxysilane (MTMOS) and hydroxy terminated
polydimethylsiloxane (HO-PDMS) were purchased from the
Boaixin Chemical Reagents Company (Baoding, China). Methanol,
Zn(NO3 )2 ·6H2 O, 2-methylimizole, acetonitrile, acetone, hydrochlo-
ric acid (HCl), sodium hydroxide (NaOH), and all other reagents
were obtained from Tianjin Tianda Chemical Reagent Company
(Tianjin, China). The water used in this study was prepared on a
SZ-93 automatic double-distiller purchased from Shanghai Yarong
Biochemistry Instrumental Factory (Shanghai, China). The vacuum
pump (Jinteng GM-0.33A) and temperature-controlled magnetic
stirrer (DF-101S) were both bought from Baoding Sunshine Sci-
ence Instrument Company (Baoding, China). The screw-capped
vials were purchased from Supelco (Bellefont, CA, USA). The pen
samples were bought from a local supermarket (Baoding, China).
The simulated pen ink matrix consisting of 30 wt% benzyl alco-
hol, 20 wt% ethylene glycol monophenyl ether, 15 wt% metlyl Violet
5BN, 10 wt% epoxy resin, 8 wt% alkyd resin, 7 wt% basic brilliant
blue B, 5 wt% phthalocyanine blue BX, 3 wt% triethanolamine oleate
and 2 wt% polyvingpyrrolidone (PVP) [29] was used to prepare the
matrix-matched calibration samples.
2.2. Preparation of the ZIF-8-NPC
ZIF-8-NPC was prepared according to the previous report with
some modifications [27]. First, Zn(NO3 )2 ·6H2 O (0.2933 g) and 2-
methylimizole (0.6489 g) (molar ratio 1:8) were dissolved in
methanol and then stirred for 6 h at room temperature. The result-
ing solid (ZIF-8) was collected by filtration. After it was washed
with methanol for 3 times and dried at 100 ◦ C in an oven for 12 h,
the obtained ZIF-8 was transferred into a quartz boat. After the air
was removed with a gentle flow of N2 for 2 h, the carbonization
of the ZIF-8 was performed at 900 ◦ C (from room temperature to
900 ◦ C at a heating rate of 5 ◦ C min−1 ) in a N2 flow for 8 h to get a raw
ZIF-8-NPC. The prepared raw ZIF-8-NPC was then treated with HCl
aqueous solution to remove ZnO, followed by being washed with
double-distilled water until its pH became 7 to obtain the desired
ZIF-8-NPC.
The surface morphology of the ZIF-8-NPC was observed by scan-
ning electron microscope (SEM) on a Hitachi SU8020 (Tokyo, Japan).
The X-ray diffraction (XRD) patterns were measured on a Tongda
TD-3500X X-ray diffractometer (Dandong, China) using Cu K˛ radi-
ation at a power of 40 kV and 150 mA in the 2␪ range of 5–80◦ . The
surface areas were determined from the N2 adsorption at 300 K
using a V- Sorb 2800P (Jinaipu, China).
2.3. Preparation of the needle-trap device
The NTD was prepared from a stainless-steel needle (0.8 mm
o.d., 0.6 mm i.d. and 50 mm in length). As shown in Fig. 1A, the
stainless-steel needle was sequentially filled with deactivated glass
wool and 5 mg of ZIF-8-NPC at the indicated position. A small quan-
tity of epoxy resin was then adhered to the needle tip to immobilize
the adsorbent particles. Nitrogen gas was passed through the nee-
dle until the epoxy resin was cured to prevent the resin from
blocking the needle. The prepared NTD was conditioned at 200 ◦ C
for 1 h and 280 ◦ C for 2 h in a N2 atmosphere in the GC injector.
Finally, the needle tip end was sealed with an inlet septum of gas
chromatography and the upper part of the needle was assembled
with a syringe valve (as shown in Fig. 1B) in order to prevent the
adsorbent from being contaminated during the storage.
 X. Zang et al. / J. Chromatogr. A 1513 (2017) 27–34 29

Table 1
Retention time and SIM parameters for each analyte.

MS SIM parameter

Compound Retention time/min Quantification ion m/z Identification ion m/z

methylbenzene 4.03 91 92, 65

chlorobenzene 4.73 112 114, 77
ethylbenzene 5.26 91 106, 92
m-xylene 5.55 91 106, 105
o-xylene 5.86 91 106, 105
bromobenzene 6.32 156 158, 77
m-chlortoluene 6.62 91 126, 125
mesitylene 6.98 105 120, 119
m-dichlorobenzene 7.13 146 148, 111
o-dichlorobenzene 7.43 146 148, 111
p-bromotoluene 7.66 91 170, 172

Fig. 2. scanning electron micrograph of the ZIF-8-NPC at a magnification of 80,000 (A); the XRD spectrum of the ZIF-8-NPC (B); The N2 adsorption and desorption isotherms
of the ZIF-8-NPC (C); the pore size distribution of the ZIF-8-NPC (D).

2.4. Samples
A total of 20 pen samples of 4 different kinds made by five differ-
ent manufacturers were purchased from a local supermarket. The
pen inks were taken directly from the pen refills. For the measure-
ments of the VOCs, 1.0 g of the pen ink taken from the pen refill
was diluted to 15.0 mL with water in a screw-capped vial for the
following sampling process.
2.5. Sampling by the needle-trap device
As shown in Fig. 1C, 15 mL of the sample solution was put into a
22-mL screw-capped vial. The vial was kept at 30 ◦ C in a water bath.
The VOCs in the sample was purged into the NTD with a nitrogen
30 X. Zang et al. / J. Chromatogr. A 1513 (2017) 27–34
Fig. 3. Effect of different extraction conditions on the extraction efficiency: (A) the flow rate of N2 ; (B) extraction temperature; (C) extraction time; (D) desorption temperature.
flow at 4 mL min−1 for the adsorption by the ZIF-8-NP adsorbent
inside the NTD. After 35 min of purging, the NTD was inserted into
the injection port of the GC–MS for thermal desorption at 270 ◦ C
under a helium flow at 4 mL min−1 for GC–MS analysis (Fig. 1D).
The syringe valve of the NTD was turned on for 4 min for desorption
and then turned off.
2.6. GC–MS conditions
The adsorbed VOCs by the NTD were analyzed by using the gas
chromatography-mass spectrometry (GCMS-QP2010SE, Shimadzu,
Kyoto, Japan) equipped with a fused silica capillary column (30 m
long, 0.25 mm i.d.) coated with 0.25 ␮m film of TG-5MS (Scientific,
Thermo, www.thermo.com/columns). The injection port tempera-
ture was maintained at 270 ◦ C in the split mode (1:10). The oven
temperature program was set as follows: starting oven temper-
ature at 60 ◦ C (held for 1 min), then ramped at 10 ◦ C min−1 to
150 ◦ C (held for 1 min), and finally increased to 240 ◦ C at a rate of
30 ◦ C min−1 . The MS analyses were carried out in full-scan mode in
the scaning range of 50–400 m/z at an electron multiplier voltage
of 70 eV. The interface temperature and the ion source tempera-
ture were maintained at 250 ◦ C and 200 ◦ C, respectively. The ions,
one for quantification and the other two for identification (shown in
Table 1), were monitored for each of the VOCs based on the full-scan
results.
3. Results and discussion
3.1. Characterization of the ZIF-8-NPC
The morphology of the ZIF-8-NPC was investigated by scanning
electron microscope (SEM) at the magnification of 80 000. Fig. 2A
shows that the surface of the ZIF-8-NPC was uniform, granulated
and rough. It can be observed that the particle size was in the range
of 70–100 nm.
The XRD pattern for the ZIF-8-NPC (Fig. 2B) gives two broad
diffraction peaks at the 2␪ values of 26◦ and 44◦ , respectively, which
corresponded to the diffractions of the carbon. The signals for ZnO
species were not observed, indicating that no ZnO was left in the
ZIF-8-NPC.
The surface area and the pore size distribution of the ZIF-8-NPC
were investigated by N2 sorption isotherms. As shown in Fig. 2C,
the ZIF-8-NPC exhibits a typical type IV adsorption–desorption
isotherm and a discernible H1-type hysteresis loop, suggesting the
existence of both micropores and mesopores. As can be seen from
the pore size distribution plot in Fig. 2D, the micropore size of
the ZIF-8-NPC was mainly between 0.6 nm and 0.7 nm while the
mesopores had the sizes from 2 to 10 nm. The surface area (Stotal )
and pore volume (Vtotal ) of the ZIF-8-NPC were 1233 m2 g−1 and
1.17 cm3 g−1 , respectively. The above results indicate that the ZIF-
8-NPC had a good hierarchically porous structure.
 X. Zang et al. / J. Chromatogr. A 1513 (2017) 27–34 31

Table 2
Linear range (LR), limits of detections (LODs) and correlation coefficients (r) for eleven VOCs.

RSD (%)

Compound LR (␮g kg−1 ) LOD (ng kg−1 ) r one NTD NTD to NTD

methylbenzene 0.1–400 10.0 0.9931 9.7 8.9

chlorobenzene 0.2–400 15.0 0.9967 8.6 8.8
ethylbenzene 0.2–400 15.0 0.9925 9.6 10
m-xylene 0.1–400 10.0 0.9911 8.8 9.2
o-xylene 0.1–400 10.0 0.9990 9.1 8.7
bromobenzene 0.2–400 20.0 0.9935 9.5 9.1
m-chlortoluene 0.1–400 10.0 0.9927 9.8 10
mesitylene 0.1–400 10.0 0.9925 7.9 8.2
m-dichlorobenzene 0.1–400 10.0 0.9940 8.5 7.7
o-dichlorobenzene 0.1–400 10.0 0.9955 9.6 9.4
p-bromotoluene 0.1–400 10.0 0.9963 8.5 8.1

Table 3
Recoveries and precisions (n = 5) of the eleven VOCs in four samples obtained by current method.

Marker Pen Fluorescent Pen Ball Pen Gel Pen

Compound Spiked Found ± SD R, RSD Found ± SD R, RSD Found ± SD R, RSD Found ± SD R, RSD
(␮g kg−1 ) (␮g kg−1 ) (%) (␮g kg−1 ) (%) (␮g kg−1 ) (%) (␮g kg−1 ) (%)

methylbenzene 0.0 0.50 ± 0.011 1.20 ± 0.034 0.32 ± 0.021 0.51 ± 0.082
0.5 0.98 ± 0.043 96.0, 8.9 1.73 ± 0.14 106, 7.5 0.77 ± 0.061 90.0, 9.7 0.97 ± 0.087 92.0, 9.8
20.0 20.7 ± 1.2 101, 5.8 21.2 ± 1.5 100, 6.9 19.9 ± 2.1 97.9, 11 19.2 ± 1.8 93.5, 10
chlorobenzene 0.0 1.12 ± 0.019 0.58 ± 0.016 1.20 ± 0.16 0.06 ± 0.0076
0.5 1.57 ± 0.13 90.0, 9.1 0.98 ± 0.056 80.0, 7.2 1.65 ± 0.13 90.0, 8.9 0.51 ± 0.038 90.0, 8.3
20.0 19.8 ± 1.7 93.4, 9.0 20.7 ± 1.7 101, 8.3 19.9 ± 1.7 93.5, 9.3 20.4 ± 2.1 102, 9.9
ethylbenzene 0.0 3.05 ± 0.063 2.12 ± 0.14 1.02 ± 0.022 1.62 ± 0.074
0.5 3.55 ± 0.31 100, 8.6 2.52 ± 0.17 80.0, 8.2 1.46 ± 0.13 88.0, 9.8 2.06 ± 0.13 88.0, 7.3
20.0 21.7 ± 1.8 93.3, 9.1 22.0 ± 2.1 99.4, 9.5 18.7 ± 1.6 88.4, 9.6 19.1 ± 1.6 87.5, 9.5
m-xylene 0.0 2.12 ± 0.19 1.75 ± 0.092 0.33 ± 0.014 0.17 ± 0.022
0.5 2.62 ± 0.26 100, 10 2.18 ± 0.17 86.0, 9.3 0.83 ± 0.052 100, 6.3 0.62 ± 0.058 94.0, 10
20.0 21.7 ± 1.2 97.9, 5.8 22.3 ± 1.4 103, 5.9 19.1 ± 1.8 93.9, 10 18.2 ± 1.5 90.2, 8.9
o-xylene 0.0 2.33 ± 0.044 2.88 ± 0.23 0.85 ± 0.039 0.96 ± 0.73
0.5 2.81 ± 0.17 96.0, 6.2 3.35 ± 0.26 94.0, 8.2 1.30 ± 0.13 90.0, 11 1.47 ± 1.3 102, 9.0
20.0 21.0 ± 2.2 93.4, 11 24.0 ± 2.0 106, 7.7 19.1 ± 1.7 91.3, 9.9 17.6 ± 1.2 83.2, 8.1
bromobenzene 0.0 2.68 ± 0.21 7.71 ± 0.66 0.22 ± 0.029 0.74 ± 0.36
0.5 3.12 ± 0.24 88.0, 6.7 8.19 ± 0.68 96.0, 8.7 0.68 ± 0.059 92.0, 9.5 1.19 ± 0.12 90.0, 11
20.0 20.9 ± 1.8 91.1, 9.5 27.9 ± 2.8 101, 9.9 17.5 ±1.3 86.4, 8.5 18.6 ± 1.0 89.3, 6.3
m-chlortoluene 0.0 2.50 ± 0.23 0.21 ± 0.010 0.47 ± 0.042 0.07 ± 0.0046
0.5 2.99 ± 0.32 98.0, 11 0.75 ± 0.067 108, 8.3 0.92 ± 0.063 90.0, 7.6 0.54 ± 0.042 94.0, 8.2
20.0 21.1 ± 1.9 93.0, 10 21.0 ± 1.7 104, 7.9 17.3 ± 1.1 84.2, 7.3 20.7 ± 1.6 103, 7.7
mesitylene 0.0 1.72 ± 0.055 0.23 0.13 ± 0.010 nd
±0.037
0.5 2.15 ± 0.18 86.0, 9.9 0.68 ± 0.050 90.0, 8.1 0.60 ± 0.050 94.0, 8.8 0.45 ± 0.028 90.0, 6.9
20.0 19.5 ± 1.4 88.9, 8.1 19.7 ± 1.8 97.4, 9.4 19.2 ± 1.2 95.4, 6.5 17.0 ± 1.6 85.0, 11
m-dichlorobenzene 0.0 2.12 ± 0.087 2.45 ± 0.36 0.10 ± 0.011 0.11 ± 0.013
0.5 2.63 ± 1.9 102, 6.9 2.89 ± 0.17 88.0, 6.5 0.53 ± 0.040 86.0, 8.7 0.58 ± 0.029 94.0, 5.3
20.0 21.1 ± 1.0 94.9, 5.2 19.9 ± 1.2 87.3, 7.1 20.7 ± 1.9 103, 8.8 20.1 ± 1.9 100, 9.4
o-dichlorobenzene 0.0 2.31 ± 0.26 0.68 ± 0.081 1.05 ± 0.063 nd
0.5 2.80 ± 0.27 98.0, 10 1.10 ± 0.077 84.0, 8.3 1.58 ± 0.15 106, 9.0 0.48 ± 0.039 96.0, 8.5
20.0 20.5 ± 1.8 91.0, 9.5 20.5 ± 1.5 99.1, 7.5 17.6 ± 1.5 82.8, 10 19.3 ± 1.7 96.5, 8.9
p-bromotoluene 0.0 1.55 ± 0.044 0.55 ± 0.047 0.11 ± 0.014 nd
0.5 2.00 ± 1.6 90.0, 8.9 0.99 ± 0.078 88.0, 8.9 0.55 ± 0.032 88.0, 6.6 0.46 ± 0.032 92.0, 7.5
20.0 20.1 ± 1.5 92.8, 7.8 20.1 ± 1.5 97.8, 7.8 17.3 ± 1.0 86.0, 6.9 19.8 ± 2.2 99.0, 11

R: recovery; nd: not detection.

3.2. Optimization of sampling conditions 3.2.2. Effect of extraction temperature

3.2.1. Effect of the N2 flow rate on the sampling
Since the VOCs were purged into the NTD by a N2 flow, the
flow rate of the N2 was an influencing factor for the extraction.
As shown in Fig. 3A, the extraction efficiency was increased with
the N2 flow rate being increased from 2.0 to 3.5 mL min−1 , then
remained almost unchanged from 3.5 to 4.5 mL min−1 , and after
that, there was a decrease for the extraction efficiency when the
N2 flow exceeded 4.5 mL min−1 . The reason for this could be that
the contact time between the analytes and the adsorbent may be
too short for the VOCs to be effectively adsorbed by the NTD at a too
high flow rate of N2 . In order to get a stable extraction, a flow rate of
the N2 at 4.0 mL min−1 was chosen for the following experiments.
Extraction temperature often plays an important role on the
extraction. In this study, the extraction temperature was inves-
tigated in the range from 20 to 45 ◦ C. Fig. 3B shows that the
extraction efficiencies for most of the VOCs increased slowly with
the extraction temperature being increased from 20 to 35 ◦ C and
then decreased when the temperature was further increased from
35 to 45 ◦ C. The extraction temperature influences the analytes
adsorption both dynamically and thermodynamically. From the
dynamic point of view, the higher the extraction temperature is,
the higher the diffusion rate for analytes from the sample solution
to the adsorbent is. While thermodynamically, the amount of the
adsorbed analytes would decrease with the increase of the extrac-
tion temperature since the adsorption process is exothermic. On
32 X. Zang et al. / J. Chromatogr. A 1513 (2017) 27–34

Fig. 4. The typical GC–MS chromatograms for the pen inks of marker pen (A), fluorescent pen (B), ball pen (C) and gel pen (D). (a) the sample spiked at 0.5 ␮g kg−1 each of
the eleven VOCs; (b) sample blank.
Peak identification: (1) methylbenzene, (2) chlorobenzene, (3) ethylbenzene, (4) m-xylene, (5) o-xylene, (6) bromobenzene, (7) m-chlortoluene, (8) mesitylene, (9) m-
dichlorobenzene, (10) o-dichlorobenzene, (11) p-bromotoluene.

Table 4
Method comparisons for the analysis of VOCs.

Method Instrument Sorbent Matrix Extraction time LOD Linearly RSD Ref.
(Extraction solvent) (min) (␮g L−1 ) (␮g L−1 ) (%)

D-HS-NTDa GC–MS ZIF-8-NPC Pen inks 35 10–20 ng kg−1 0.1–400 ␮g <9.8 Present work
kg−1
HS-SPME b GC–MS Ionic liquid Paints 30 0.1-0.8 ␮g 0.4–120 ␮g 3.5–10 [9]
mL−1 mL−1
HS-NTD c GC–MS Carboxen 1000 Blood 25–45 0.020–0.080 0.1–0.4 <15 [13]
Static HS-NTD d GC–MS Tenax TA Natural and 30–60 0.01–0.06 0.01–50 [14]
Waste waters
HS-K-HFME e GC–MS 1-Octanol River and 20 0.2–10 1–80000 4.5–11.6 [30]
Wastewater
M-SPME f GC-FID Polydimethylsiloxane Tap water and 30 0.011–0.031 0.5–100 <14 [31]
and polyethylene Seawater
glycol
DHS-OSFME g GC-FID 1-Dodecane Wastewater 5.4 0.18–0.35 0.5–200 4.5–11.6 [32]
a
Dynamic headspace needle-trap device.
b
Headspace solid-phase microextraction.
c
Headspace needle-trap device.
d
Static headspace needle-trap device.
e
Headspace knotted hollow fiber microextraction.
f
Membrane solid-phase microextraction.
g
Dynamic headspace organic solvent film microextraction.

the basis of the above experimental result, 30 ◦ C was chosen for ensure a good reproducibility for the extraction, the extraction for
the extraction. 35 min was chosen.

3.2.3. Effect of extraction time
The influence of the extraction time on the extraction was inves-
tigated in the range between 20 and 40 min. As shown in Fig. 3C,
the peak areas of the VOCs were first increased remarkably with
the extraction time being increased from 10 min to 30 min and then
remained almost unchanged after that, indicating that the extrac-
tion equilibrium for the analytes was achieved at about 30 min. To
3.3. Effect of desorption temperature and desorption time
The desorption temperature can influence the mass transfer rate
of the VOCs from the adsorbent of the NTD to the carrier gas. In the
present study, the desorption temperature was investigated in the
range from 255 to 280 ◦ C. According to Fig. 3D, there were increased
peak areas when the desorption temperature was increased from
 X. Zang et al. / J. Chromatogr. A 1513 (2017) 27–34
255 to 265 ◦ C, and then remained almost unchanged. Based on the
above result, 270 ◦ C was chosen for the desorption.
The desorption time was then investigated in the range from 1 to
8 min when the desorption temperature was kept at 270 ◦ C. Based
on the experimental results, 4.0 min of desorption was chosen since
a complete desorption was achieved when the desorption time was
longer than 4.0 min.
3.4. Evaluation of the method performance
The analytical parameters, including linear range (LR), correla-
tion coefficients (r), limits of detection (LOD) and relative standard
deviations (RSDs), for the developed method were evaluated under
optimized conditions. To reduce the possible sample matrix effect,
the simulated VOCs-free pen ink matrix [29] was used to prepare
the matrix-matched calibration samples. An appropriate volume of
the mixture stock solution was added into 1.0 g of the simulated pen
ink matrix in the screw-capped vials to get the final concentrations
of the VOCs at 0.1, 1.0, 10.0, 50.0, 100, 200, 300, 400 ␮g kg−1 , respec-
tively. Then, each sample was diluted to 15.0 mL with water. The
prepared calibration samples were subjected to the extraction and
analysis according to the procedures described in the Experimental
section. For each concentration level, five replicate analyses were
performed. The results are listed in Table 2. A good linearity was
observed in the concentration range of 0.1–400 ␮g kg−1 with the
correlation coefficients (r) ranging from 0.9911 to 0.9990 for the
VOCs. The LODs measured at a signal-to-noise ratio of 3 (S/N = 3)
were 10–20 ng kg−1 for the VOCs. The same NTD was used for eval-
uating the single-NTD repeatability by five replicate extractions,
and the five different NTDs prepared in the same way were used
for evaluating the reproducibility of different NTDs. The RSD val-
ues for the single-NTD repeatability of the developed method for
the analysis of the VOCs were less than 9.8% and the RSDs for the
reproducibility of five different NTDs were below 10%, indicating
that the prepared NTDs had a quite good reproducibility.
The reusability of the NTD was evaluated. The results showed
that no obvious changes in the extraction efficiency were found
under the present experimental conditions when an NTD was
reused about 100 times.
3.5. Determination of the VOCs in pen inks samples
The VOCs in 20 pen ink samples of 4 different kinds from five
manufacturers were determined with the developed method. As a
result, all of the pen ink samples were found to contain one or more
of the VOCs. Ethylbenzene, bromobenzene and m-xylene were the
top three among the detected VOCs. Although the total amount
of the VOCs detected in the pen inks is quite low, the pen users
still need to pay attention to the harmful effect of the VOCs from
pens. The four different kinds of pen inks made by the same com-
pany were used for the measurements of the recoveries of the VOCs
with the current method to evaluate the method accuracy. For the
measurements of the recoveries, the pen ink samples were spiked
with the standard of the VOCs to each concentration of the VOCs at
0.5 and 20.0 ␮g kg−1 , respectively, and five replicate analyses were
performed at each concentration level. The results in Table 3 show
that the recoveries of the VOCs from the pen ink samples fell in
the range from 80.0% to 108%, with RSDs between 5.2% and 11%,
demonstrating that both the accuracy and precision of the method
are acceptable. Fig. 4 shows the typical GC–MS chromatograms for
the pen inks.
3.6. Comparison with other methods
The developed method was compared with other relevant
reported methods in the literature for the determination of VOCs
33
in different samples. The data in Table 4 show that the current D-
HS-NTD method exhibits comparable linear ranges and RSDs with
and lower LODs than the other relevant methods.
4. Conclusions
A D-HS-NTD packed with a newly prepared ZIF-8-NPC as the
adsorbent was developed for the extraction of some VOCs from pen
inks prior to GC–MS detection. The ZIF-8-NPC material had a high
surface area and high porosity and showed efficient for the extrac-
tion of the VOCs. Under the optimized conditions, low LODs, good
recoveries and good reproducibility for the analytes were obtained
by the current method. The method provides an approach of choice
for the determination of VOCs in pen inks.
Conflict of intrest
The authors have declared no conflict of interest.
Acknowledgements
Financial support from the National Natural Science Founda-
tion of China (No. 31471643, 31571925 and 31671930), the Hebei
“Double First Class Discipline” Construction Foundation for the
Discipline of Food Science and Engineering of Hebei Agricultural
University (2016SPGCA18), the Youth Scientific and Technological
Research Foundation of the Department of Education of Hebei for
Hebei Provincial Universities (QN2017085), and the Natural Sci-
ence Foundation of Hebei Agricultural University (ZD201507) is
gratefully acknowledged.
References
[1] R.L. Brunelle, K.R. Crawford, Advances in the Forensic Analysis and Dating of
Writing Ink, Charles C Thomas Pub Limited, Springfield, 2003.
[2] S.K. Lim, H.S. Shin, K.S. Yoon, S.J. Kwack, Y.M. Um, J.H. Hyeon, H.M. Kwak, J.Y.
Kim, T.H. Kim, Y.J. Kim, T.H. Roh, D.S. Lim, M.K. Shin, S.M. Choi, H.S. Kim, B.M.
Lee, Risk assessment of volatile organic compounds benzene, toluene
ethylbenzene, and xylene (BTEX) in consumer products, J. Toxicol. Environ.
Health A 77 (2014) 1502–1521.
[3] R. Costa, Newly introduced sample preparation techniques: towards
miniaturization, Crit. Rev. Anal. Chem. 44 (4) (2014) 299–310.
[4] J.L. Perez Pavon, M. del Nogal Sanchez, M.E. Fernandez Laespada, B. Moreno
Cordero, Simultaneous determination of gasoline oxygenates and benzene
toluene, ethylbenzene and xylene in water samples using
headspace-programmed temperature vaporization-fast gas
chromatography-mass spectrometry, J. Chromatogr. A 1175 (2007) 106–111.
[5] B. Yu, Y. Song, L. Han, H. Yu, Y. Liu, H. Liu, Optimizations of packed sorbent and
inlet temperature for large volume-direct aqueous injection-gas
chromatography to determine high boiling volatile organic compounds in
water, J. Chromatogr. A 1356 (2014) 221–229.
[6] M.T. Garcia-Valverde, R. Lucena, S. Cardenas, M. Valcarcel, In-syringe
dispersive micro-solid phase extraction using carbon fibres for the
determination of chlorophenols in human urine by gas chromatography/mass
spectrometry, J. Chromatogr. A 1464 (2016) 42–49.
[7] A. Sarafraz-Yazdi, A.H. Amiri, Z. Es’haghi, BTEX determination in water
matrices using HF-LPME with gas chromatography-flame ionization detector,
Chemosphere 71 (2008) 671–676.
[8] H.H. Lim, H.S. Shin, Determination of volatile organic compounds including
alcohols in refill fluids and cartridges of electronic cigarettes by headspace
solid-phase micro extraction and gas chromatography-mass spectrometry,
Anal. Bioanal. Chem. 409 (5) (2017) 1247–1256.
[9] J.F. Liu, N. Li, G.B. Jiang, J.M. Liu, J.A. Jönsson, M.J. Wen, Disposable ionic liquid
coating for headspace solid-phase microextraction of benzene, toluene,
ethylbenzene, and xylenes in paints followed by gas chromatography–flame
ionization detection, J. Chromatogr. A 1–2 (2005) 27–32, 1066.
[10] Q. Li, X. Ma, D. Yuan, J. Chen, Evaluation of the solid-phase microextraction
fiber coated with single walled carbon nanotubes for the determination of
benzene, toluene, ethylbenzene, xylenes in aqueous samples, J. Chromatogr. A
1217 (15) (2010) 2191–2196.
[11] H.W. Liu, Y.T. Liu, B.Z. Wu, H.C. Nian, H.J. Chen, K.H. Chiu, J.G. Lo, Process
sampling module coupled with purge and trap-GC-FID for in situ
auto-monitoring of volatile organic compounds in wastewater, Talanta 80
(2009) 903–908.
[12] A. Lara-Gonzalo, J.E. Sanchez-Uria, E. Segovia-Garcia, A. Sanz-Medel, Critical
comparison of automated purge and trap and solid-phase microextraction for
34 X. Zang et al. / J. Chromatogr. A 1513 (2017) 27–34
routine determination of volatile organic compounds in drinking waters by
GC–MS, Talanta 74 (2008) 1455–1462.
[13] M. Alonso, M. Castellanos, E. Besalu, J.M. Sanchez, A headspace needle-trap
method for the analysis of volatile organic compounds in whole blood, J.
Chromatogr. A 1252 (2012) 23–30.
[14] M. Alonso, L. Cerdan, A. Godayol, E. Antico, J.M. Sanchez, Headspace
needle-trap analysis of priority volatile organic compounds from aqueous
samples: application to the analysis of natural and waste waters, J.
Chromatogr. A 1218 (2011) 8131–8139.
[15] C. Bicchi, C. Cordero, E. Liberto, B. Sgorbini, P. Rubiolo, Headspace sampling of
the volatile fraction of vegetable matrices, J. Chromatogr. A 1184 (2008)
220–233.
[16] H.L. Lord, W. Zhan, J. Pawliszyn, Fundamentals and applications of needle trap
devices: a critical review, Anal. Chim. Acta 677 (2010) 3–18.
[17] Z.Y. Gu, X.P. Yan, Metal-organic framework MIL-101 for high-resolution
gas-chromatographic separation of xylene isomers and ethylbenzene, Angew.
Chem. Int. Edit. 49 (8) (2010) 1477–1481.
[18] X. Zang, G. Zhang, Q. Chang, X. Zhang, C. Wang, Z. Wang, Metal organic
framework MIL-101 coated fiber for headspace solid phase microextraction of
volatile aromatic compounds, Anal. Methods 7 (3) (2015) 918–923.
[19] Z. Huang, H.K. Lee, Micro-solid-phase extraction of organochlorine pesticides
using porous metal-organic framework MIL-101 as sorbent, J. Chromatogr. A
1401 (2015) 9–16.
[20] D. Ge, H.K. Lee, Sonication-assisted emulsification microextraction combined
with vortex-assisted porous membrane-protected micro-solid-phase
extraction using mixed zeolitic imidazolate frameworks 8 as sorbent, J.
Chromatogr. A 1263 (2012) 1–6.
[21] D. Ge, H.K. Lee, Zeolite imidazolate frameworks 8 as sorbent and its
application to sonication-assisted emulsification microextraction combined
with vortex-assisted porous membrane-protected micro-solid-phase
extraction for fast analysis of acidic drugs in environmental water samples, J.
Chromatogr. A 1257 (2012) 19–24.
[22] X. Zang, X. Zhang, Q. Chang, S. Li, C. Wang, Z. Wang, Metal-organic framework
UiO-67-coated fiber for the solid-phase microextraction of nitrobenzene
compounds from water, J. Sep. Sci. 39 (14) (2016) 2770–2776.
[23] W. Chaikittisilp, K. Ariga, Y. Yamauchi, A new family of carbon materials:
synthesis of MOF-derived nanoporous carbons and their promising
applications, J. Mater. Chem. A 1 (2013) 14–19.
[24] G. Xu, B. Ding, L. Shen, P. Nie, J. Han, X. Zhang, Sulfur embedded in metal
organic framework-derived hierarchically porous carbon nanoplates for high
performance lithium–sulfur battery, J. Mater. Chem. A 1 (2013) 4490–4496.
[25] S. Shahid, K. Nijmeijer, High pressure gas separation performance of
mixed-matrix polymer membranes containing mesoporous Fe(BTC), J.
Membr. Sci. 459 (2014) 33–44.
[26] J. Kim, N.D. McNamara, T.H. Her, J.C. Hicks, Carbothermal reduction of
Ti-modified IRMOF-3: an adaptable synthetic method to support catalytic
nanoparticles on carbon, ACS Appl. Mater. Interfaces 5 (2013) 11479–11487.
[27] L. Hao, X. Liu, J. Wang, C. Wang, Q. Wu, Z. Wang, Use of ZIF-8-derived
nanoporous carbon as the adsorbent for the solid phase extraction of
carbamate pesticides prior to high-performance liquid chromatographic
analysis, Talanta 142 (2015) 104–109.
[28] L. Hao, C. Wang, Q. Wu, Z. Li, X. Zang, Z. Wang, Metal-organic framework
derived magnetic nanoporous carbon: novel adsorbent for magnetic
solid-phase extraction, Anal. Chem. 86 (2014) 12199–12205.
[29] M.C. Pan, Introduction to Chinese patent of ballpoint pen ink, China Writ.
Instrum. 2 (2013) 33–47.
[30] P.S. Chen, Y.H. Tseng, Y.L. Chuang, J.H. Chen, Determination of volatile organic
compounds in water using headspace knotted hollow fiber microextraction, J.
Chromatogr. A 1395 (2015) 41–47.
[31] A. Spietelun, L. Marcinkowski, A. Kloskowski, J. Namiesnik, Determination of
volatile organic compounds in water samples using membrane-solid phase
microextraction (M-SPME) (headspace version), Analyst 138 (17) (2013)
5099–5106.
[32] A. Mohammadi, N. Alizadeh, Automated dynamic headspace organic solvent
film microextraction for benzene, toluene, ethylbenzene and xylene.
Renewable liquid film as a sampler by a programmable motor, J. Chromatogr.
A 1107 (1–2) (2006) 19–28.