ABSTRACT
Abstract
Since pesticides are potentially harmful to the environment and consequently to human beings through the consumption of pesticide contaminated food and water, food commodities have to be controlled to assure the non-violation of the maximum residue levels (MRLs) set by domestic and European regulations. In this work, a multiresidue GC-MS method for screening of 49 common pesticide residues in various foods of plant origin, which uses the QuEChERS method for sample extraction and cleanup, has been validated. Recovery values, linearity and reporting limits (RL) in various food commodities were determined for the analyzed compounds, and instrument performance was assessed by the use of mixture of internal standards. Eight pesticides did not show satisfactory recovery values or colud not be detected in levels required by the regulations, and were therefore excluded from the final scope of validation.
INTRODUCTION
More and more different pesticides are used nowadays in agriculture. Since pesticides are potentially harmful to the environment and consequently to human beings through the consumption of pesticide contaminated food and water, the European Community established maximum residue levels (MRLs), based on the assumption that good agricultural practice is applied at the use of pesticides in farming, for pesticide residues in water (Commission of the European Communities, 2000) and foodstuff (Commission of the European Communities, 1990). As a consequence, food commodities have to be controlled to assure the non-violation of the MRLs. For apolar and middle polar pesticides, the detection of pesticide residues is commonly achieved through analysis with gas chromatography (GC) coupled to single quadropole (SQ) and, less frequently, triple quadropole (QQQ) mass spectrometers (MS) (Lesueur et al, 2008). The determination of pesticides in fruits and vegetables has been simplified by a new sample preparation method, QuEChERS (Quick, Easy, Cheap, Effective, Rugged and Safe), and published recently as an AOAC Method 2007.01 (Lehotay, 2007).
In this work, a multiresidue GC-MS method for screening of 49 common pesticide residues in various foods of plant origin, which uses the QuEChERS method for sample extraction and cleanup (Anasta-ssiades, 2006), has been validated. Recovery values, linearity and reporting limits (RL) in various food commodities were determined for the analyzed compounds, and instrument performance was assessed by the use of mixture of internal standards.
EXPERIMENTAL
Materials
Pesticide standards were purchased either from Dr. Ehrenstorfer or from Sigma–Aldrich with the highest available purity. QUECHERS kits for sample extraction and dispersive cleanup were purchased from Agilent Technologies. Ultra-residue reagent n-heptane and HPLC grade acetonitrile were purchased from Sigma-Aldrich, and p.a. grade sodium-hydroxide was purchased from J.T. Baker. Internal standards were purchased from Sigma-Aldrich with the highest available purity.
Standard mixes and single pesticide standards were used for preparation of standard solutions. Standard stock solutions of pesticides were prepared by dissolving the standard mixes with acetonitrile and diluting them to obtain 20 µg/ml concentrations. Stock solutions of single standards were prepared by dissolving 10 mg of each standard in 10 ml of acetonitrile and further diluted with acetonitrile to get 20 µg/ml concentrations. Working standard solutions were prepared by diluting the stock solutions with acetonitrile to get 1, 0.5, 0.25, 0.125, 0.0625 and 0.05 µg/ml concentrations. Internal standard solutions were prepared by dissolving 11.8 mg of antracene, 5.50 mg of tris (1,3-dichloroizo-propyl) phosphate and 10 mg of PCB138 with acetonitrile to get 118. 5 and 100 µg/ml of each compound, respectively (ISTD 1). 100 µl of ISTD 1 was further diluted with n-heptane to obtain 1.18, 0.55 and 1.00 µg/ml concentrations of antracene, tris (1,3-di-chloroizopropyl)phosphate and PCB138, respecttively (ISTD 2).
Blank samples used for fortification experiments were fruits, vegetables and grains, bought at local supermarkets, and presented no pesticide residues when analyzed by GC-MS.
Sample extraction and cleanup
The fruit and vegetable samples were prepared with the modified QuEChERS method (CEN/TC 275, N 236, 2006). Around 10.0 g of previously homogenized sample was weighed into the extraction tube, and distilled water was added so that the total amount of water (water in sample plus added water) was about 10 ml. A pre-packaged mixture of MgSO4 and sodium-cytrate buffers was added and the extraction tube was shaken vigorously for few seconds to dissolve the salts. pH of the mixture was checked and, if needed, set to 5-5.5 value with 20% NaOH. After that, 10 ml of acetonitrile was added to the extraction tube, and shaken vigorously during one minute, after which it was centrifuged at 3000 rpm for 10 minutes. After centrifuging, 6 ml of the acetonitrile layer was transferred to the cleanup tube containing the needed sorbents for each comodity type, shaken vigorously during one minute, and centrifuged at 3000 rpm for 10 minutes. 1 ml of extract was then transferred into 2 ml glass vial and the solvent was exchanged by evaporation under a gentle stream of nitrogen to dryness and dissolving the dry residue in 200 µl of n-heptane. The extract dissolved in n-heptane was then ready for analysis by GC-MS.
2.3. Analyses
The GC/MS analyses were performed on Agilent Technologies GC-MS Model 7890 A Series gas chromatograph coupled to 5975 C mass selective detector. A HP 5 MS (30 m × 0.25 mm i.d.) (J & W Scientific, USA) fused silica capillary column with a 0.25 µm film thickness was used with helium as carrier gas at a constant pressure (chlo-rpyrifos-methyl RT relocked to 16.596 min). 2.0 µl of the sample was injected in the splitless mode at 280 oC. The GC oven was operated with the following temperature program: initial temperature 70 oC held for 2 min, ramped at 25 oC /min to 150 oC not held, followed by a ramp of 3 oC /min to 200 oC not hold, followed by another ramp of 8 oC /min to 280 oC held for 10 min. The total run time was 41.86 min. The interface was kept at 250 oC, the quadropole at 150 oC and the mass spectra were obtained at an electron energy of 70 eV. The Agilent Chemstation Software version D.02.00 was used for data analysis and the analyses were operated on the principle of a simultaneous full scan/SIM mode method presented elsewhere (Wylie, 2006). Several distinct scan/SIM methods were derived for each group of pesticides by modifying the principal analysis method (TRI_PEST.M), to increase the sensitivity of the pesticide screening and to minimize the effects of matrix.
Method validation
We selected 57 pesticides for the GC/MS analysis, based on their relevance in food samples and chromatographic properties, for the validation of the QuEChERS method according to the European SANCO Guide-line (Commission of the European Communities, 2006) in 5 reference matrices: wheat (high starch content), rapeseed (high oil content), lettuce (high water content), tangerine (high acidity) and green tea (“difficult or unique commodities”). Due to a lack of immediate availability of all pesticide standards, the mentioned commodities were spiked only with organochlorine pesticide standard mix in three replicates at 0.1 and 1 mg/kg levels, and their recovery was determined. Linearity was determined from the calibration curve constructed for single standards. Reporting limits were determined by spiking lemon sample with mixture of the standards at 0.01 mg/kg levels, and 0.005 mg/kg level for cis- and trans-chlordane, as required by domestic and European regulations. Since SANCO Guideline also outlines requirements for GC-MS instrument performance, mixture of internal standards was used to assess the stability and repeatability of retention times and peak areas (Commission of the Euro-pean Communities, 2006).
RESULTS AND DISCUSSION
According to the European SANCO Guidline, recovery of pesticides in quantitative analytical methods should be between 70-120%, although lower and higher values might be acceptable if the analyzed compound shows a good repeatability (Commission of the European Communities, 2006). Organochlorine pesticides in standard mix all showed a satisfactory values of recovery by the above mentioned criteria. When different matrices, spiked with organochlorine pesticides mix at 0.1 and 1 mg/kg levels were analyzed, recovery values were generally satisfactory, with most values above 70%. Lower recovery values were shown for endrin-aldehyde (35% in green tea, 44% in lettuce, and 44% in tangerine). Higher than recommended recovery values were found for α-BHC (189% in green tea). The obtained recovery values fulfll the purpose of showing that the QuEChERS extraction and cleanup procedure gives satisfactory recovery values for different pesticides, and does not cause significant losses of desired compounds. This is in accordance with other published works which assessed the suitability of the QuEChERS method for pesticide analysis (Butler et al, 2008; Lesueur et al, 2008; Anastassiades, 2006).
Although the scope of the work vas to validate the analysis method for pesticide screening, it is possible to validate the same procedure as a confirmatory quantitative method for pesticide analysis, employing selected ion monitoring (SIM) GC-MS methods for quantification (Butler et al, 2008; Lesueur et al, 2008). This was also shown in results of spiking different food commodities with organochlorine pesticides at various levels (results not shown in this paper).
Linearity was calculated from calibration curves made for single pesticides, and the results are shown in Table 1. All pesticides showed a good linearity, with R2 value above 0.97, except 4,4'-DDT which showed R2 value of 0.95.
Table 1. Linearity of pesticide calibration curves
|
method
|
compound
|
range (mg/kg)
|
equation
|
R2
|
|
TRI_PESTSIMOP.M
|
dichlorvos
|
0.05-1
|
y=815408.84x+1294.73
|
0.98
|
|
TRI_PESTSIMOP.M
|
etoprophos
|
0.05-1
|
y=1052437.82x-31402.06
|
0.99
|
|
TRI_PESTSIMOP.M
|
disulfoton
|
0.05-1
|
y=2454390.68x-90075.89
|
1
|
|
TRI_PESTSIMOP.M
|
parathion-methyl
|
0.25-1
|
y=333993.84x-57029.07
|
0.98
|
|
TRI_PESTSIMOP.M
|
fenchlorphos
|
0.05-1
|
y=1835925.44x-87540.35
|
0.99
|
|
TRI_PESTSIMOP.M
|
chlorpyrifos
|
0.05-1
|
y=958808.29x-40622.64
|
0.99
|
|
TRI_PESTSIMOP.M
|
prothiofos
|
0.05-1
|
y=1264454.52x-39839.8
|
1
|
|
TRI_PESTSIMOCP.M
|
alpha-BHC
|
0.05-0.5
|
y=2428088.97x-63199.63
|
0.98
|
|
TRI_PESTSIMOCP.M
|
beta-BHC
|
0.05-0.5
|
y=658968.53x-19685.74
|
0.99
|
|
TRI_PESTSIMOCP.M
|
lindane
|
0.05-0.5
|
y=788128.61x-17003.63
|
0.99
|
|
TRI_PESTSIMOCP.M
|
heptachlor
|
0.05-0.5
|
y=432559.09x-10924.27
|
0.97
|
|
TRI_PESTSIMOCP.M
|
delta-BHC
|
0.05-0.5
|
y=222655.29x-10902.67
|
0.97
|
|
TRI_PESTSIMOCP.M
|
aldrin
|
0.05-0.5
|
y=987587.84x-2275.66
|
0.99
|
|
TRI_PESTSIMOCP.M
|
heptachlor-epoxyde
|
0.05-0.5
|
y=1211214.56x-21075.22
|
1
|
|
TRI_PESTSIMOCP.M
|
gamma-chlordane
|
0.05-0.5
|
y=1841353.57x-25764.88
|
1
|
|
TRI_PESTSIMOCP.M
|
alpha-chlordane
|
0.05-0.5
|
y=1893775.33x-31127.97
|
1
|
|
TRI_PESTSIMOCP.M
|
endosulfan I
|
0.0625-0.5
|
y=389662.34x-6791.99
|
1
|
|
TRI_PESTSIMOCP.M
|
4,4'-DDE
|
0.05-0.5
|
y=3816183.14x-68457.48
|
1
|
|
TRI_PESTSIMOCP.M
|
dieldrin
|
0.05-0.5
|
y=1864006.53x-30407.08
|
1
|
|
TRI_PESTSIMOCP.M
|
endrin
|
0.125-0.5
|
y=256920.99x-15041.87
|
0.99
|
|
TRI_PESTSIMOCP.M
|
4,4'-DDD
|
0.05-0.5
|
y=4431876.45x-153368.77
|
0.98
|
|
TRI_PESTSIMOCP.M
|
endosulfan II
|
0.05-0.5
|
y=403778.11x-1102.61
|
1
|
|
TRI_PESTSIMOCP.M
|
4,4'-DDT
|
0.05-0.5
|
y=723885.17x-37441.72
|
0.95
|
|
TRI_PESTSIMOCP.M
|
endrin-aldehyde
|
0.05-0.5
|
y=1050819.35x-24969.61
|
1
|
|
TRI_PESTSIMOCP.M
|
endrin-ketone
|
0.05-0.5
|
y=546932.25x+370.47
|
0.98
|
|
TRI_PESTSIMOCP.M
|
metoxychlor
|
0.05-0.5
|
y=1253695.08x-64067.56
|
0.97
|
|
TRI_PESTSIMTRIAZIN.M
|
simazine
|
0.05-1
|
y=1435149.27x-33292.71
|
0.99
|
|
TRI_PESTSIMTRIAZIN.M
|
prometon
|
0.05-1
|
y=1422293.67x-28250.21
|
0.99
|
|
TRI_PESTSIMTRIAZIN.M
|
atrazine
|
0.05-1
|
y=2558749.57x-73638.6
|
0.99
|
|
TRI_PESTSIMTRIAZIN.M
|
propazine
|
0.05-1
|
y=2439743.42x-92297.1
|
1
|
|
TRI_PESTSIMTRIAZIN.M
|
ametryn
|
0.05-1
|
y=3020996.61x-73263.13
|
1
|
|
TRI_PESTSIMTRIAZIN.M
|
prometryn
|
0.05-1
|
y=3208392.7x-98344.21
|
0.99
|
|
TRI_PESTSIMTRIAZIN.M
|
terbutryn
|
0.05-1
|
y=3075216.91x-106111.61
|
0.99
|
|
TRI_PESTSIMTRIAZIN.M
|
propiconazole
|
0.05-1
|
y=806332.19x-17901.88
|
0.98
|
|
TRI_PESTSIMPIRET.M
|
cyfluthrin
|
0.05-1
|
y=2747945.85x-11293.45
|
0.99
|
|
TRI_PESTSIMPIRET.M
|
cypermethrin
|
0.05-1
|
y=1373394.75x-78039.87
|
1
|
|
TRI_PESTMIX1.M
|
pirimiphos-methyl
|
0.25-1
|
y=2025807.55x-194942.79
|
0,99
|
|
TRI_PESTMIX1.M
|
malathion
|
0.5-1
|
y=1324626.09x-191174.49
|
1
|
|
TRI_PESTMIX1.M
|
dimethoate
|
0.5-1
|
y=1339398.05x-275515.27
|
1
|
|
TRI_PESTMIX2.M
|
difenoconazole
|
0.05-1
|
y=5458032.21x-378786.55
|
0.98
|
|
TRI_PESTMIX2.M
|
vinclozolin
|
0.05-1
|
y=728810.39x-24575.08
|
1
|
|
TRI_PESTMIX2.M
|
flutriafol
|
0.05-1
|
y=3648430.64x-300878.61
|
0.99
|
|
TRI_PESTMIX2.M
|
fenoxycarb
|
0.05-1
|
y=1992001.62x-126621.71
|
0.99
|
|
TRI_PESTMIX3.M
|
metalaxyl
|
0.25-1
|
y=1457691.4x-329830.59
|
0.99
|
|
TRI_PESTMIX3.M
|
hexaconazole
|
0.25-1
|
y=2423345.71x-679576.47
|
0.97
|
|
TRI_PESTMIX3.M
|
fenarimol
|
0.25-1
|
y=2782758.82x-678786.92
|
0.98
|
Reporting limit represents the lowest calibration point at which an analyzed compound can be determined with high certainty. Reporting limits were determined by spiking lemon sample with mixture of the standards at 0.01 mg/kg levels, and 0.005 mg/kg level for cis- and trans-chlordane, as required by domestic and European regulations, and the determined values are shown in Table 2. The obtained values were in accordance with the lowest maximum residue levels (MRLs) permitted by domestic and European regulations. Of the 57 pesticides analyzed, eight compounds (azin-phosmethyl, parathion-methyl, indoxacarb, acetampirid, cymoxanil, chlorothalonil, folpet and captafol) showed either very low responses, or could not be detected by GC-MS at the concentration levels required by the regulations.
Table 2. Reporting limits determined for single pesticide standards
|
pesticide
|
reporting limit (mg/kg)
|
MRL in Serbia (mg/kg)
|
EU MRL (mg/kg)
|
|
aldrin
|
0.01
|
0.01
|
0.01
|
|
α-BHC
|
0.01
|
0.01
|
0.01
|
|
β-BHC
|
0.01
|
0.01
|
0.01
|
|
lindane
|
0.01
|
0.05
|
0.01
|
|
δ-BHC
|
0.01
|
0.01
|
0.01
|
|
α-chlordane
|
0.005
|
0.005
|
0.01
|
|
γ-chlordane
|
0.005
|
0.005
|
0.01
|
|
4,4'-DDD
|
0.01
|
0.05
|
0.05
|
|
4,4"-DDE
|
0.01
|
0.05
|
0.05
|
|
4,4′-DDT
|
0.01
|
0.05
|
0.05
|
|
dieldrin
|
0.01
|
0.01
|
0.01
|
|
α-endosulfan
|
0.05
|
0.1
|
0.05
|
|
β-endosulfan
|
0.05
|
0.1
|
0.05
|
|
endosulfan sulfate
|
0.05
|
0.1
|
0.05
|
|
endrin
|
0.01
|
0.01
|
0.01
|
|
endrin aldehyde
|
0.05
|
-
|
-
|
|
endrin ketone
|
0.01
|
-
|
-
|
|
heptachlor
|
0.01
|
0.01
|
0.01
|
|
heptachlor epoxide isomer B
|
0.01
|
0.01
|
0.01
|
|
methoxychlor
|
0.01
|
-
|
0.01
|
|
chlorpyrifos (Dursban)
|
0.01
|
0.05
|
0.05
|
|
dichlorvos
|
0.01
|
0.05
|
0.01
|
|
disulfoton
|
0.01
|
-
|
0.02
|
|
ethoprophos (MOCAP)
|
0.01
|
0.02
|
0.02
|
|
fenchlorphos (Ronnel)
|
0.01
|
-
|
0.01
|
|
prothiofos (Tokuthion)
|
0.01
|
-
|
-
|
|
pirimiphos-methyl
|
0.05
|
0.05
|
0.05
|
|
malathion
|
0.01
|
0.5
|
0.02
|
|
dimethoate
|
0.05
|
0.05
|
0.02
|
|
fenitrothion
|
0.01
|
0.05
|
0.01
|
|
ametryn
|
0.01
|
0.05
|
-
|
|
atrazine
|
0.05
|
0.1
|
0.05
|
|
prometon
|
0.01
|
-
|
-
|
|
prometryn
|
0.01
|
0.05
|
-
|
|
propazine
|
0.01
|
-
|
-
|
|
simazine
|
0.01
|
0.05
|
0.05
|
|
terbutryn
|
0.01
|
0.05
|
-
|
|
propiconazole
|
0.05
|
0.05
|
0.05
|
|
difenoconazol
|
0.01
|
-
|
0.05
|
|
flutriafol
|
0.01
|
0.02
|
0.05
|
|
cyfluthrin
|
0.02
|
20
|
0.02
|
|
α-cypermethrin
|
0.02
|
0.05
|
0.05
|
|
deltamethrin
|
0.05
|
0.01
|
0.05
|
|
fenoxycarb
|
0.01
|
-
|
0.05
|
|
vinclozolin
|
0.01
|
0.1
|
0.05
|
|
metalaxyl
|
0.01
|
0.05
|
0.05
|
|
azoxystrobin
|
0.01
|
-
|
0.1
|
|
fenarimol
|
0.01
|
0.1
|
0.02
|
|
hexaconazol
|
0.01
|
0.05
|
0.02
|
The mixture of internal standards was analysed in triplicates on a monthly basis, to assess the the stability and repeatability of retention times and peak areas. The obtained results showed less than 5% of differrence in retention times, as well as less than 10% of difference in peak areas for single compounds, which is in accordance with the values required by SANCO for ensuring the correct instrument performance (Commission of the European Communities, 2006). Since a single internal standard compound represents the chromato-graphic properties of a larger group of pesticides, we have also spiked the mixture of internal standards into a representative matrix (wheat flour), which was then extracted and cleaned up by the QuEChERS procedure and analyzed by GC-MS. The results of the analysis of internal standards mixture and the spiked sample are shown in Table 3. The calculated recovery values were 93.49% for antracene, 120.64% for tris (1,3-dichloroizopropyl) phosphate and 95.02% for PCB138, which showed very good extraction and cleanup efficiency of the QuEChERS procedure. Chromatogram of the spiked sample is shown in the Fig. 1.
Table 3. Retention times and responses of internal standards mixture and the spiked sample
|
|
antracene
|
tris(1,3-dichloroizopropyl) phosphate
|
PCB138
|
|
ISTD2
|
|
|
|
|
R.T.
|
14.070
|
26.798
|
27.110
|
|
QIon
|
178
|
75
|
360
|
|
Response
|
20347458
|
1589948
|
5173550
|
|
ISTD2 spike
|
|
|
|
|
R.T.
|
14.071
|
26.799
|
27.112
|
|
QIon
|
178
|
75
|
360
|
|
Response
|
19023634
|
1918181
|
4915737
|
Fig. 1.Selected ion monitoring (SIM) chromatogram of the sample spiked with internal standards mixture
CONCLUSION
In this work, a screening method for determination of 49 different pesticides in various food commodities of plant origin has been validated. The obtained results fulfill the European SANCO Guideline requirements for screening analytical methods of pesticide traces in foods. Most of the analyzed pesticides showed good recovery values between 70-120%, which indicates that the QuEChERS extraction and cleanup method are well suited for analysis of pesticides in various, often difficult to analyze food commodities. Eight compounds (azinphos-methyl, parathion-methyl, indoxacarb, ace-tampirid, cymoxanil, chloro-thalonil, folpet and captafol) did not show satisfactory recovery values or colud not be detected in levels required by the regulations. These pesticides were excluded from the final scope of validation, and further possibility of analyzing them by high performance liquid chromatography coupled to triple quadric-pole mass detector (LC-QQQ) should be investigated.
ACKNOWLEDGEMENT
Original scientific paper was written as a result of work on a project TR20068 “Prehrambeni proizvodi za grupe potrošača sa specijalnim zahtevima i potrebama“, funded by Ministry of Science and Technological Development, Republic of Serbia.
Download full article PDF
DOWNLOAD PDF
