ANALYTICAL SCIENCES MAY 2003, VOL.

19 701
2003 The Japan Society for Analytical Chemistry

Analysis of Acetamiprid in Vegetables Using Gas

ChromatographyTandem Mass Spectrometry
Manuel MATEU-SNCHEZ, Mercedes MORENO, F. Javier ARREBOLA, and
Jos Luis MARTNEZ VIDAL

Department of Analytical Chemistry, University of Almera, 04120 Almera, Spain

A new analytical method has been validated for determining the insecticide acetamiprid in vegetables using gas
chromatography (GC) and different mass spectrometric detection techniques, such as full-scan mass spectrometry (MS),
and tandem mass spectrometry (MS/MS). For this purpose, a previous extraction of the vegetable sample was carried out
with ethyl acetate. In GC-MS/MS, the lowest detectable concentration was 0.001 mg kg1, the average recovery rates at
various fortification levels (0.015 and 0.030 mg kg1) ranged between 82.4 and 85.7% and the relative standard deviations
were lower than 12.2% in all cases.

(Received March 18, 2002; Accepted February 24, 2003)

Introduction
Acetamiprid ((E)-N1-[(6-chloro-3-pyridyl)methyl]-N2-cyano-N1-
methylacetamidine) is a systemic insecticide for soil and foliar
applications. It is used to control Hemiptera, especially aphids,
Thysanoptera and Lepidoptera on a wide range of crops,
especially vegetables, fruits and tea.1 It acts on the central
nervous system, causing irreversible blocking of the
postsynaptic nicotinergic acetylcholine receptors.2 Its structure
is represented in Fig. 1B.
Acetamiprid is a new insecticide used in agricultural practices.
The maximum residue level (MRL) in Japan Legislation ranges
between 1 mg kg1 (orange) and 50 mg kg1 (tea). In South
Africa MRL is 0.5 mg kg1 for citric. There is no legislation for
regulating its use in EU. Acetamiprid is an insecticide
recommended as a substitute for some worldwide-used
organophosphate pesticides. This fact makes it very important
to determine acetamiprid in vegetables with validated analytical
methodologies in order to efficiently control its applications to
crops.
The determination of low concentrations of this pesticide in
matrices, such as vegetables, requires the application of an
effective extraction of the vegetable, followed by a final
chromatographic determination in order to separate as much as
possible the analyte from the matrix interference substances.
Gas chromatography (GC) combined with selective and
sensitive electron capture detector (ECD) and nitrogen
phosphorus detector (NPD) has become a routine technique in
the analysis of pesticides in vegetables. Until now, some
analytical methods have been published for determining
acetamiprid in crops and soils by GC-ECD,35 by high-
performance liquid chromatography (HPLC)6 and by enzyme-
linked immunosorbent assays (ELISA).7 However, the high
sensitivity of these detectors contrasts with their lack of Fig. 1 A, Full-scan EI-MS spectrum for acetamiprid; B, full-scan
CI-MS spectrum and structural formula for acetamiprid; C, CI-
To whom correspondence should be addressed.
MS/MS spectrum for acetamiprid; D, CI-MS/MS spectrum for
E-mail: jlmartin@ual.es caffeine.
702
identification power. Relative retention-time-based
identifications and determinations of pollutants are no longer
sufficient. An additional confirmatory technique is necessary,
and detection by mass spectrometry (MS) is frequently used
because of its identification capability. Recently, tandem mass
spectrometry (MS/MS) was introduced into routine analysis
laboratories. It is being applied to the determination of trace
amounts of contaminants in complex samples, such as
biological fluids,810 waters samples,1113 and vegetables.14
In this work, acetamiprid residues in vegetables were analyzed
by using an extraction with solvents and a subsequent
determination by GC-MS and -MS/MS. In this way, these
mentioned techniques were evaluated with respect to the
advantage required for routine analysis. The limitations and
advantages of using the GC-MS or -MS/MS techniques for the
quantitative and qualitative determinations of acetamiprid are
compared and discussed.
Experimental
Chemicals and reagents
Pesticide-quality solvents (dichloromethane, cyclohexane,
hexane, acetone and methanol) were purchased from Scharlau
(Barcelona, Spain). A standard of the pesticide acetamiprid was
obtained certified from Dr. Ehrenstorfer (Ausburg, Germany),
and caffeine (GC-MS and -MS/MS internal standard) was
supplied by Riedel-de-Han (Seelze-Hanover, Germany). All
of the standards had a purity higher than 99%. Stock standard
solutions of 200 g mL1 were prepared by exact weighting and
dissolving them in acetone. Working standard solutions were
prepared by appropriate dilutions with the same solvent and
stored in a refrigerator (4C). Anhydrous sodium sulfate
(quality for analysis of pesticides) was obtained from Panreac
(Barcelona, Spain).
Apparatus
A Saturn 2000 ion-trap mass spectrometer from Varian
Instrument (Walnut Creek, CA, USA) was used. The integrated
gas chromatograph was a GC 3800 Model fitted with an
autosampler (8200) prepared for injecting large volumes, a
split/splitless programmed temperature injector (1079), an
electronic flow control system, a fused-silica untreated capillary
column (2 m 0.25 mm i.d.; Supelco, Bellefonte, PA, USA)
used as a guard column and a DB-5MS (30 m 0.25 mm i.d.
0.25 m film thickness) capillary column (J & W Scientific,
Folsom, CA, USA). The injection port had a carbofrit (Restek
Corporation, Bellefonte, PA, USA) inserted in a large-volume
injection glass liner. The ion-trap mass spectrometer was
operated in the chemical ionization (CI) mode with the MS/MS
option for both acetamiprid and caffeine. The computer which
controlled the system had an MS/MS library specially created
for the target analytes under our experimental conditions. The
carrier gas used was helium (purity 99.999%).
A chopper from Hamilton Beach (Washington, USA) and a
Polytron PT2100 from Kinematica AG (Littan/Luzern,
Switzerland) were used. A rotary evaporator (R-114) with a
thermostated water bath was purchased from Buchi (Flawil,
Switzerland).
Chromatographic conditions
GC-MS. Ten microliter volumes were injected by an
autosampler at a flow rate of 10 L s1. The injector
temperature was programmed from 70C (hold 0.5 min) to
310C at 100C min1 (hold 10 min). The split valve was
ANALYTICAL SCIENCES MAY 2003, VOL. 19
initially opened for 0.5 min, and was then closed for 3 min, and
finally opened until the next analysis. The oven temperature
was modified from 70C (hold 3.5 min) to 300C at 50C min1
(hold 7 min). The column flow was set at 1 mL min1 at 150C
oven temperature. Transfer-line, manifold and trap
temperatures were 280, 50, and 200C, respectively. The MS
conditions were: a solvent delay of 5 min; 70 eV of electron
impact energy; scan rate of 0.8 scans s1 and scanned-mass
range (m/z) 80 300. The automatic gain control (AGC) was
switched on with the target fixed at 5000 counts; the emission
current was 30 A and methanol was used as the CI gas. The
mass spectrometer was calibrated weekly. The electron impact
(EI) and CI full-scan spectrum obtained for acetamiprid in the
selected experimental conditions are shown in Figs. 1A and 1B,
respectively. They were stored in a full-scan laboratory-made
library.
GC-MS/MS. For GC-MS/MS, the sample was injected under
the gas-chromatographic conditions described for GC-MS. The
characteristic parameters of the compounds were activated from
7 to 8.25 min (caffeine) and from 8.25 to 10 min (acetamiprid).
The m/z ranges were 80 210 (caffeine) and 80 235
(acetamiprid). The emission current was set at 30 A to ionize
the isolated m/z 195 (caffeine) and 223 (acetamiprid), which
were isolated from the other ions with an isolation window of 5
a.m.u. The AGC target was 2000 counts. The collision-induced
dissociation process was carried out in a non-resonant
waveform with excitation storage levels of 85 (caffeine) and 90
m/z (acetamiprid) and excitation amplitudes of 80 (caffeine) and
70 V (acetamiprid). The scan time was 0.8 s scan1. The
MS/MS spectrum obtained in the selected conditions is
represented in Figs. 1C (acetamiprid) and 1D (caffeine). They
were stored in an MS/MS laboratory-made library.
Sampling and storage
For each vegetable, the samples weighed 2.5 kg were
immediately stored in polyethylene bags for transporting to the
laboratory. Vegetable pieces were chopped, blended,
thoroughly mixed and each divided into subsamples. The
samples were analyzed on the same day that they were picked in
order to avoid possible stability problems of acetamiprid in the
matrices during the homogenization and storage processes.
Extraction procedure
An aliquot of 15 g of the chopped vegetable sample was
mixed with 50 mL of ethyl acetate with a Polytron for 2 min.
Then, 50 g of anhydrous sodium sulfate was slowly added with
agitation. The mixture was filtered through a paper filter and
the liquid was collected in a separating funnel. The sample
container and filter were washed with 20 mL of ethyl acetate,
that were collected in a separating funnel. This process was
repeated again with another 10 mL of the same solvent. The
extract was passed through a conical funnel with 5 g of
anhydrous sodium sulfate to eliminate any possible water. The
extract was evaporated in a rotary evaporator (water bath
temperature = 60C) and re-dissolved in 10 ml of cyclohexane
containing the internal standard (0.1 mg kg1). The extract was
ready to be injected into the gas chromatograph.
Results and Discussion
Optimization of the GC-MS and -MS/MS variables
A GC-MS/MS chromatogram of a clean vegetable fortified
with 0.05 mg kg1 of acetamiprid is shown in Fig. 2. The
pesticide was eluted in less than 10 min. The capability of the
ANALYTICAL SCIENCES MAY 2003, VOL. 19
Fig. 2 GC-MS/MS chromatogram of a vegetable containing 0.05
mg kg1 of acetamiprid and selecting the quantification ions; A, m/z
110 caffeine; B, m/z 196 acetamiprid.
MS detector to monitor selective ions of the target analyte
discriminated the analytical signal, while reducing the influence
of the interferences of coeluted substances. It was not able to
discriminate with conventional GC detectors, such as an
electron-capture detector, which requires a previous GC
resolution of peaks before detection in order to avoid false
positives and quantification errors. The injection volume was
set at 10 L in order to increase the sensitivity. Nevertheless,
due to the absence of a previous clean-up step in the extraction
procedure, the use of carbofrit into the glass liner and guard
column is recommended in order to reduce the amount of low-
volatile compounds that could reach the analytical column and
the detector. The EI and CI modes were tested for the MS
analysis of acetamiprid. The highest sensitivity was obtained
using CI with methanol. The use of this solvent to generate the
reagent gas has the advantages of simpler handling (liquid) and
safer use instead of other frequently used gases (methane, iso-
butane or ammonia). Additionally, this ionization technique
was more selective and reduced the interferent peaks observed
in the chromatograms. AGC was switched on in order to
optimize the sensitivity by completely filling the trap with the
target ions. In the full-scan mode, the mass range (m/z 80 300)
and background mass (m/z 80) were selected to optimize the
sensitivity ejecting as much as possible of the matrix and
solvent ions. The compounds were characterized by their full-
scan spectra under these experimental conditions. In the
MS/MS mode, m/z 223 was chosen for acetamiprid and m/z 195
for caffeine as the precursor ion for the CI-MS/MS conditions.
The breakages proposed for the main ions of acetamiprid in an
EI-full scan and CI-MS/MS are summarized in Table 1.
Acetamiprid presented different spectra under the above
operating conditions, as can be expected from the difference in
the ionization energy. The presence of m/z 196 (base peak) in
the CI-MS/MS spectrum by the loss of the CN group must be
mentioned, which is less favored under the EI conditions. They
were the main ions in the full-scan spectra. The AGC target
was set at 2000 counts because higher values caused
electrostatic interactions between the ions in the ion-trap
chamber. A non-resonant waveform was selected for the
collision-induced dissociation (CID) process. The objective
was to produce ion spectra with the precursor ion as their
molecular peaks (between 10 and 30% of relative abundance).
The excitation amplitude was studied for this proposes. The
target analyte was searched in its relative time window (RTW)
and identified by comparing with the MS and MS/MS libraries.
703
Table 1 Important mass spectral fragments and their tentative
ions
Ionization mode m/z Tentative ion
EI full scan 233 [M]+
207 [M]+ CH3
181 [M]+ CH3CNH
166 [ClC5H3NCH2NCN]+
152 [ClC5H3NCH2NC]+
141 [ClC5H3NCH2N]+ + H+
126 [ClC5H3NCH2]+
CI MS/MS 223 [M]+
206 [M]+ CH3
196 [M]+ CNH
187 [M]+ Cl
181 [M]+ CH3CNH
146 [C5H4NCH2NCNCH3]+
126 [ClC5H3NCH2]+
A positive analyte identification required a minimum spectral fit
of > 700 and a signal-to-noise ratio (S/N) > 3 (for the ion of
quantification). For quantification, S/N must be higher than 10.
Optimization of the extraction process
A partition with organic solvents was evaluated as an
extraction process of the vegetable samples because it is simple
and fast, and can be applied in routine analysis laboratories. For
that purpose, different aliquots of 15 g of clean cucumbers were
fortified with acetamiprid (0.2 mg kg1) being extracted with
several solvents (dichloromethane, ethyl acetate, and n-hexane).
The process was carried out with a polytron in order to improve
the contact between the sample and the organic solvent. The
best recovery rates were obtained when 50 mL of ethyl acetate
was used.
Validation of the methods
The limit of detection (LOD) and quantification (LOQ) were
calculated based on the values of the blanks (n = 10) obtained
by extracting clean vegetable samples at the retention time of
the analyte and following the IUPAC recommendations.15 The
obtained results are given in Table 2. The LOD and LOQ
obtained by the MS/MS technique were lower than those
obtained by full-scan mode. This can be attributed to a drastic
reduction of the background matrix that the MS/MS process
obtains. When the full-scan mode is used as a detector, a prior
clean-up step would improve the obtained results. Nevertheless,
it would increase the analysis time, cost, and uncertainty of the
final result.
In order to avoid any matrix effects, calibration curves were
prepared using a blank matrix extract to fill up to volume. An
instrument calibration for GC-MS was performed by injecting
extracts with concentrations of 0.03, 0.1, 0.2, and 0.3 mg kg1 of
acetamiprid (0.2 mg L1 of internal standard). The result
achieved using relative heights were better than using relative
areas. Good linearity was found in the concentration range
studied with a correlation coefficient higher than 0.98. On the
other hand, the instrument calibration for the GC-MS/MS
method was performed by injecting extracts fortified with 0.01,
0.03, 0.1, and 0.3 mg kg1 of acetamiprid. The calibration was
carried out while considering the area of the peaks and referring
data to the internal standard (0.2 mg L1 of caffeine). The
instrumental response can be considered to be linear in the
concentration range studied in full-scan and MS/MS modes, as
shown by the correlation coefficient (r2 > 0.99). Wider
704
Table 2 Relative time windows (RTW), limits of detection
(LOD) and quantification (LOQ), calibration curve, recovery
rates (R%), and precision (RSD%)
Parameter GC-MS GC-MS/MS
RTW (min) 9.48 9.63 9.49 9.63
LOD (mg kg1) 0.009 0.001
LOD (ng) 0.135 0.015
LOQ (mg kg1) 0.030 0.003
LOQ (ng) 0.450 0.045
Calibration curve y = 0.041 + 0.487x y = 0.006 + 0.750x
r2 0.981 0.992
R% (RSD%)
0.015 (mg kg1) 82.4(12.2)
0.030 (mg kg1) 30.4(20.3) 85.7(5.2)
0.200 (mg kg1) 54.2(10.1)
concentration ranges would generate a calibration curve that
requires a quadratic or cubic fit.
Recovery and repeatability studies of the proposed method
were also carried out. For GC-MS, ten aliquots of clean
vegetable samples were fortified with 0.03 and 0.2 mg kg1 of
acetamiprid, and were extracted and determined. The recovery
rate and precision (expressed as the relative standard deviation)
are given in Table 2. Similar studies were performed at two
different fortification levels for GC-MS/MS (0.015 and 0.030
mg kg1). The results are summarized in Table 2. The best
results were found when the recovery studies were carried out
using MS/MS, even at lower fortification levels.
The identification of the pesticide is based on its retention
time and the spectrum obtained, which is compared with
laboratory-made and commercial mass spectra libraries.
Nevertheless, when trace amounts of the target analytes are
determined in complex samples, such as vegetables by full-scan,
the spectral identification and sensitivity are not as good as they
should be. This is because of coelution problems between the
matrix and the target peaks at trace levels. With MS/MS, if a
coeluted interference has the same identification ion as the
analyte, such an interferent can be avoided using special
experimental conditions for the collision-induced dissociation
and quantifying with a specific ion from the MS/MS spectrum
of the analyte. MS/MS permits the determination of very low
concentrations of acetamiprid in the presence of important
amounts of matrix interferents. Additionally, MS/MS spectra
offer valuable qualitative information to confirm the results.
This information is higher than the Selected Ion Monitoring
(SIM) mode offers when other mass spectrometric detectors are
used.
Conclusions
A new method to determine acetamiprid in vegetables is
proposed. It is fast and simple, and can be used as a routine
pesticide residue control in a laboratory. The extraction process
is an organic solvent partition that does not include a clean-up
step, thus simplifying manipulation of the sample. The
determination is carried out by GC-CI-MS/MS, which is a very
ANALYTICAL SCIENCES MAY 2003, VOL. 19
selective and sensitive technique that considerably reduces the
matrix influence. The method was validated and the LOD,
LOQ, calibration curves, precision, and recovery data were
given.
Acknowledgements
The authors are grateful to Instituto Nacional de Investigacin y
Tecnologa Agraria y Alimentaria (INIA) (Project CAL00-064)
and the Council Town of El Ejido (Almera, Spain) for their
financial support.
References
1. T. Roberts and D. Hutson, Metabolic pathways of
agrochemicals. Part two: Insecticides and fungicides,
1999, The Royal Society of Chemistry, Cambridge, 111
120.
2. A. Okazawa, Y. Nakagawa, M. Akamatsu, T. Ueno, and K.
Nishimura, J. Pest. Science (Nihon Nouyaku Gakkaishi),
2000, 25, 40.
3. M. Tokieda, M. Ozawa, S. Kobayashi, and M. Gomyo, J.
Pest. Science (Nihon Nouyaku Gakkaishi), 1997, 22, 77.
4. M. Tokieda, M. Ozawa, S. Kobayashi, M. Gomyo, and M.
Takeda, J. Pest. Science (Nihon Nouyaku Gakkaishi), 1999,
24, 115.
5. M. Tokieda, M. Ozawa, and M. Gomyo, J. Pest. Science
(Nihon Nouyaku Gakkaishi), 1999, 24, 181.
6. M. Tokieda, T. Tanaka, M. Ozawa, and M. Gomyo, J. Pest.
Science (Nihon Nouyaku Gakkaishi), 1998, 23, 296.
7. S. Wanatabe, S. Itob, Y. Kamata, N. Omoda, T. Yamazaki,
H. Munakata, T. Kaneko, and Y. Yuasa, Anal. Chim. Acta,
2000, 427, 211.
8. J. L. Martnez-Vidal, F. J. Arrebola, A. Fernndez-
Gutirrez, and M. A. Rams, J. Chromatogr. B, 1998, 719,
71.
9. F. J. Arrebola, J. L. Martnez-Vidal, and A. Fernndez-
Gutirrez, Toxicol. Lett., 1999, 107, 15.
10. F. J. Arrebola, J. L. Martnez-Vidal, A. Fernndez-
Gutirrez, and M. H. Akhtar, Anal. Chim. Acta, 1999, 401,
45.
11. M. C. Pablos-Espada, F. J. Arrebola, A. Garrido-Frenich,
and J. L. Martnez-Vidal, Int. J. Environ. Anal. Chem.,
1999, 75, 165.
12. J. L. Martnez-Vidal, M. C. Pablos-Espada, A. Garrido-
Frenich, and F. J. Arrebola, J. Chromatogr. A, 2000, 867,
235.
13. A. Garrido-Frenich, J. L. Martnez-Vidal, M. C. Pablos-
Espada, M. D. Gil-Garca, and F. J. Arrebola,
Chromatographia, 2000, 52, 614.
14. F. J. Arrebola, F. J. Egea-Gonzlez, M. Moreno, A.
Garrido-Frenich, M. E. Hernndez-Torres, and J. L.
Martnez-Vidal, Pest. Management Sci., 2001, 57, 645.
15. A. Gonzlez-Casado, L. Cuadros-Rodrguez, E. Alonso-
Hernndez, and J. L. Vlchez, J. Chromatogr. A, 1996, 726,
133.