Tạp chí phân tích Hóa, Lý và Sinh học Tập 24, Số 4B/2019 DETERMINATION OF MULTI CLASS PESTICIDES IN SURFACE WATER BY ULTRAHIGH PERFORMANCE LIQUID CHROMATOGRAPHY IN COMBINATION WITH HIGH RESOLUTION MAS[.]
Trang 1Tạp chí phân tích Hóa, Lý và Sinh học - Tập 24, Số 4B/2019
DETERMINATION OF MULTI-CLASS PESTICIDES IN SURFACE WATER BY ULTRAHIGH PERFORMANCE LIQUID CHROMATOGRAPHY IN COMBINATION WITH HIGH RESOLUTION MASS SPECTROMETRY (UPLC-Orbitrap MS)
Đến tòa soạn 10-4-2019 Thuy Le Minh, Trung Nguyen Quang, Tung Nguyen Ngoc, Nam Vu Duc, Hung Nguyen Xuan Centre for Research and Technology Transfer, Vietnam Academy of Science and Technology
Hai Anh Vu, Lan Anh Nguyen, Tri Tran Manh, Minh Tu Binh Faculty of Chemistry, VNU University of Science, Vietnam National University-Hanoi
Dinh Binh Chu School of Chemical Engineering, Hanoi University of Science and Technology
Lan Anh Phung Thi School of Environmental Science and Technology, Hanoi University of Science and Technology
TÓM TẮT NGHIÊN CỨU XÁC ĐỊNH CÁC HÓA CHẤT BẢO VỆ THỰC VẬT
TRONG MẪU NƯỚC MẶT SỬ DỤNG THIẾT BỊ SẮC KÝ LỎNG
SIÊU HIỆU NĂNG CAO GHÉP NỐI KHỐI PHỔ PHÂN GIẢI CAO
(UPLC-Orbitrap MS) Trong nghiên cứu này, phương pháp sắc ký lỏng ghép nối khối phổ phân giải cao (UPLC-Orbitrap MS) được giới thiệu để xác định đa nhóm các hóa chất bảo vệ thực vật trong mẫu nước mặt Quá trình sắc
ký sử dụng cột tách Thermo Hypersil Gold PFP (2.1 mm x 100 mm x 3 µm) và hệ pha động có chứa 5
mM HCOONH4, 0.1% FA trong nước deion và methanol Phương pháp quét toàn dải (fullscan) ở độ phân giải 70000 FWHM được áp dụng cho cả chế độ ion dương và âm để phát hiện và định lượng các chất phân tích Các điều kiện sắc ký như nhiệt độ cột tách, pha động và chương trình dung môi, tốc dộ dòng cũng như các thông số về nguồn ion hóa đã được khảo sát và tối ưu Sự có mặt của các chất phân tích trong mãu thực được định tính dựa trên thời gian lưu, khối lượng phân tử, độ lệch khối, mảnh phổ con tương ứng tại năng lượng phân mảnh Các chất phân tích được tách ra khỏi nền mẫu bằng kỹ thuật chiết pha rắn sử dụng cột Water Oasis HLB Các thí nghiệm với mẫu thêm chuẩn được thực hiện để đánh giá phương pháp phân tích thông qua độ lặp lại, hiệu suất thu hồi, khoảng tuyến tính… Phương pháp xây dựng được giới hạn phát hiện nằm trong khoảng từ 0,0002 (Azoxystrobin) – 0,493 ngmL-1
(Parathion-methyl) Cũng trong nghiên cứu này, phương pháp đã tối ưu để phân tích đa nhóm chất bảo
vệ thực vật trong 20 mẫu nước sử dụng trong nông nghiệp, được thu thập tại 5 vùng trồng rau trên địa bàn Hà Nội
Key words: thuốc bảo vệ thực vật, nước mặt, UPLC, Orbitrap MS, chiết pha rắn
1 INTRODUCTION
“Pesticides” play an important role in
protecting crop yields [1]-[3] They are
synthesized chemicals to limit, inhibit and
prevent the increase of harmful animals, pestilent insects, weed and fungi [4]-[5] Besides, they are also used to improve dramatically quality of agricultural products
Trang 2[2] However, uncontrolled use of pesticides
has been increasing the presence of their
residues in various environment They can be
dispersed to the atmosphere, released to
surrounding soil or transported to surface
water Most of pesticides are known as toxic
compounds so, their exposure at high level
may have the risk of being carcinogenic and
mutagenic properties in humans and other
animals Facts have shown that presence of
pesticides in environment and food is
controlled more and more strictly in numerous
countries [2], [6] In EU, the legal limits with
0.1 µg L-1 per individual pesticide and 0.5 µg
L-1 for the sum of all pesticides was
recommended to protect human health and
environment [1] In Vietnam, according to a
circular no 03/2018/TT-BNNPTNT, more
than 30 different pesticides were banned in
agriculture [7] Pesticide usage and residue
level of not many pesticides in soil, food were
also monitored Based on chemical structure,
pesticides can be classified into some groups:
organophosphate, carbamate, neonicotinoid,
strobilurin, triazole, etc Recently, many
researches have been published about pesticide
analysis in both GC-MS [8]-[9] and LC-MS
[10]-[15] Among them, LC-MS/MS technique
is the most popular for multi-residue
quantitative method of pesticide analysis
Moreover, the high resolution mass
spectrometry- a method with several
advantages, such as excellent mass accuracy,
high sensitivity… has been introduced for
screening of residual pesticides [10] Also,
several sample preparation procedures have
been frequently applied such as: liquid-liquid
extraction, solid phase extraction, etc., in order
to enhance sensitivity of analytical methods
The purpose of this study was to investigate
the analytical method of multi-residue
pesticides in surface water samples by using of
ultrahigh liquid chromatography in
combination with high resolution mass
spectrometry (UPLC-Orbitrap MS) In this
work, pesticide presence was isolated from the
matrix by solid-phase extraction (SPE) before
analysis by the UPLC-Orbitrap MS Also, the parameters such as: limit of detection (LOD), limit of quantitation (LOQ), linearity, recovery… have been investigated to evaluate method through spiking experiments Finally, the developed method was applied to screening multi-residue of pesticides in real samples which were collected from several villages in Hanoi
2 MATERIALS AND METHODS 2.1 Chemicals and reagents HPLC-grade methanol (MeOH), acetonitrile (ACN), ammonium formate (HCOONH4) and Formic acid (FA) with high purity were purchased from Merck (Germany) HPLC-grade water was obtained by purifying demineralized water in a Milli-Q Integral 3 from Merck Millipore (France) Twelve investigated pesticides (high purities, > 90%) were acquired from Dr Ehrenstorfer GmH (Germany) Standard solution of 10 ppm was prepared by mixing and diluting the 1000 ppm stock individual solutions in ACN Solution of
10 ppm of surrogates (IS) including dimethoate-d6, dichlorvos-d6 (Dr Ehrenstorfer) and malathion-d10 (Toronto Research Chemicals) were prepared in MeOH The working standard solutions were built at concentration 5, 10, 25, 50, 100, 250, 500 ppb with 250 ppb surrogates in MeOH solvent 2.2 Instrumentation and analytical method
An UPLC system (Ultimate 3000) coupled to a Q-Exactive Focus Orbitrap MS system (Thermo Scientific) was used to determine pesticides in surface water samples All of parameters of UPLC-HRMS system were monitored by the Thermo X-calibur software ver 4.0
For LC, a Thermo Hypersil gold PFP column (150 x 2.1 mm, 3 µm) was used for separation
of target analytes at temperature of 40 0C The binary mobile phases were 0.1% FA + 5 mM HCOONH4 in H2O (A) and in MeOH (B) Their gradient elution was started with 2% B for 0.25 mins, raised to 30% B for 0.75 mins, linearily increased to 100% B in 24 mins (held for 5 mins) then restored to the 2% B in 0.5
Trang 3mins and held for 7 mins to re-equilibrate the
column for the next injection The flow rate
was constantly kept at 0.3 mL.min-1 during
chromatographic separation of time (37.5
mins) 5 µL of each sample was injected to
LC-Orbitrap MS system via autosampler
Needle and sample loop in autosampler were
washed triplicate by using mixture of
methanol: deionized water (1:9, v:v)
A system of Orbitrap MS using electrospray
ionization (ESI), with 70000-FWHM
resolution (at 200 Da) was operated in both
positive and negative ionization modes The
mass spectrometer was calibrated before each
batch of measurement by Pierce
Positive/negative ion mass calibration solution
(Thermo Fisher, CAS number 88324) This
system was optimized for each pesticide by
direct infusion experiment Optimum operating
conditions followed parameters: Sheath gas
pressure at 32 psi, auxiliary gas flow rate at 7
L.min-1, sweep gas flow rate at 0 L.min-1,
Spray voltage +2.8 kV and -2.5 kV for positive
and negative mode, respectively Capillary and
vaporizer temperature at 320 0C and 295 0C,
respectively; S-lens RF level at 50 V The
FullMS mode (resolution 70000-FWHM) was
conducted in mass range 80-1000 to measure
target ions of precursors The FullMS/ddMS2
(Full-scan data dependent MS/MS) mode could
simultaneously record the fragmentation
spectra for the precursors Besides, the
FullMS/confirmation mode (with inclusion
list) was also used to confirm fragments of the
selected precursors Both the dd-MS2 and
confirmation mode conditions followed
parameters: resolution 17500-FWHM, mass
isolation window 1.0 m/z, maximum and
minimum automatic gain control (AGC) target
8x104 and 5x103, respectively; normalized
collision energy (NCE) 30%; spectrum data
format for confirmation: centroid A software
of Thermo TraceFinder ver 3.3 was applied
for data evaluation and reporting
2.3 Sample collection and preparation
Twenty surface water samples were collected
from 5 local villages including Van Noi (coded
N1), Tien Duong (N3), Tam Xa (N5) (Dong Anh district), Van Duc (Gia Lam district) (N4) and Yen My (Thanh Tri district) (N2) in Hanoi then acidified to pH 3 by FA and stored at 4 0C until analysis For method evaluation, the experiment was conducted with pooled samples, which were the mixture of equal volume of individual sample SPE technique using Waters OASIS HLB 6 cc column was applied for sample preparation Cartridges were firstly conditioned by 10 mL MeOH followed by 10 mL deionized water adjusted with pH about 3 by FA Each 100 mL acidified surface water sample with 250 ppb IS, was loaded up to cartridge at 1 drop.sec-1 flow rate
by gravity Target compounds were eluted by
10 mL MeOH The SPE eluent was dried under
a gentle nitrogen flow to 0.5 mL, then add 2
mL mobile phase before removing solvent to exact 1 mL Samples were filtered then injected into UPLC-HRMS system The real samples were followed above procedure for determination of pesticides
3 RESULTS AND DISCUSSION 3.1 Optimization chromatographic separation for all target analytes
In this work, Thermo Hypersil GOLD PFP column was used for chromatographic separation of multi-class pesticides based on retaining of polar compounds The working standard solution and matrix-match were prepared and injected on the LC-Orbitrap MS followed the optimized operating conditions which have shown in 2.2
As clearly shown on the Table 1, the repeatability of both retention time and precursor of all twelve selected compounds in standard solution and in matrix match were very good
Trang 4Table 1: The retention time and mass accuracy of investigated analytes in neat
solvent and match- matrix
Precursor (Polarity)
RT± SD (n=6), min
[M+H] or [M-H] experiment
m/z ± SD (n=6)
Mass accuracy (ppm) m/z ± SD (n=6)
Mass accuracy (ppm) Dimethoate C5H12NO3PS2
Organophosphate
230.0069 (+) 5.48 ± 0.03 230.0069 ± 0.0001 0.11 230.0069 ± 0.0001 0.14
261.9945 ±
261.9944 ±
Dichlorvos C4H7Cl2O4P 220.9532 (+) 7.81 ± 0.03 220.9533 ± 0.0001 0.23 220.9532 ± 0.0000 -0.08 Methomyl C5H10N2O2S
Carbamate
163.0536 (+) 4.47 ± 0.13 163.0535 ± 0.0000 -0.46 163.0536 ± 0.0001 0.20 Carbaryl C12H11NO2 202.0862 (+) 9.05 ± 0.04 202.0863 ± 0.0001 0.25 202.0863 ± 0.0001 0.33 Azoxystrobin C22H17N3O5 Strobilurin 404.1241 (+) 13.19 ± 0.03 404.1238 ± 0.0001 -0.80 404.1236 ± 0.0001 -1.32 Dinotefuran C7H14N4O3 Neonicotinoid 201.0993 (-) 3.95 ± 0.13 201.0986 ±
201.0986 ±
Carbendazim C9H9N3O2 benzimidazole 192.0768 (+) 4.74 ± 0.04 192.0768 ±
192.0768 ±
Clomazone C12H14ClNO2 Isoxazolidinone 240.0786 (+) 11.06 ± 0.04 240.0786 ± 0.0001 -0.21 240.0784 ± 0.0001 -0.83 Tebufenozide C22H28N2O2 diacylhydrazine 351.2078 (-) 14.35 ± 0.02 351.2080 ± 0.0001 0.57 351.2080 ± 0.0001 0.43 Indoxacarb C22H17ClF3N3O7 Oxadiazine 528.078 (+) 17.86 ± 0.02 528.0771 ±
528.0773 ±
Pyridaben C19H25ClN2OS Pyridazinone 365.1449 (+) 19.29 ± 0.03 365.1446 ±
365.1445 ±
3.2 Mass accuracy of the target analytes
The mass accuracy of all target compounds
was assessed by injected 6 times independently
of 100 ngmL-1 neat solvent and sample with 25
ppb of all analytes on LC-Orbitrap MS system
The theorical mass was calculated by online
Envipat Web 2.2 From Table 1, it should also
be noted that the excellent mass accuracy was
achieved in both neat solvent and matrix-match
solutions Mass accuracies of all target were
below ± 3.5 ppm The identification of target
analytes was carried out by comparison of retention time, mass accuracies, fragments… with the standard at experimental conditions Fragments were searched in Mass bank and taken by Thermo Mass Frontier sofware ver 7.0 From table 2, mass accuracy of all product ions achieved below 5 ppm, except dinotefuran (-11.5 ppm) and pyridaben (-5.44 ppm)
Trang 5Table 2: Protonated mass and mass accuracy of product ions of target analytes
formula
Precursor (Polarity)
Formula m/z theor exp m/z accuracy Mass Formula m/z theor exp m/z accuracy Mass Dimethoate C5H12NO3PS2 230.0069 (+) C2H6O2PS 124.9826 124.9820 -4.80
Parathion-methyl C8H10NO5PS 261.9944 (-) C7H6O3N 152.0348 152.0343 -3.29 C6H4O3N 138.0191 138.0187 -2.90 Dichlorvos C4H7Cl2O4P 220.9532 (+) C2H8O4P 127.0160 127.0154 -4.72 CH4O2P 78.9949 78.9948 -1.27 Methomyl C5H10N2O2S 163.0536 (+) C3H6NS 88.0221 88.0219 -2.27 C3H8NOS 106.0326 106.0321 -4.72 Carbaryl C12H11NO2 202.0862 (+) C10H9O 145.0653 145.0647 -4.14 C9H9 117.0704 117.0700 -3.42 Azoxystrobin C22H17N3O5 404.1241 (+) C21H14N3O4 372.0984 372.0972 -3.22 C20H14N3O3 344.1035 344.1024 -3.20 Dinotefuran C7H14N4O3 201.0993 (-) HN2O2 61.0038 61.0031 -11.47
Carbendazim C9H9N3O2 192.0768 (+) C8H6N3O 160.0511 160.0505 -3.75
Clomazone C12H14ClNO2 240.0786 (+) C7H6Cl 125.0158 125.0153 -4.00
Tebufenozide C22H28N2O2 351.2078 (-) C9H9O2 149.0602 149.0598 -2.68
Indoxacarb C22H17ClF3N3O7528.0780 (+) C15H7O3N 249.0426 249.0421 -2.01 C13H10ClN2O4293.0329 293.0324 -1.71 Pyridaben C19H25ClN2OS 365.1449 (+) C11H15 147.1174 147.1166 -5.44 C15H18ClN2OS 309.0828 309.0814 -4.53
3.3 Validation of the developed method
In this work, the Waters Oasis HLB cartridge was
selected because it is a universal sorbent that has
been extensively used for extraction of organic
contaminants, and its hydrophilic /lipophilic
character makes it suitable for interaction with
analytes of a wide polarity range From Table 3,
recovery of target analytes was in ranged from
62% (azoxystrobin) to 125% (pyridaben) The
lower recovery was attributed by ionization
suppression in mass spectrometry, especially in
heated electrospray ionization mass spectrometry
The higher recovery was explained by
enhancement of analytical in sample matrix
However, both effects could be eliminated by
using isotopic labelled internal standards
It can be seen from table 3 that correlation (R2) between analytical signal and analytical concentration were achieved higher than 0.99 The method of detected limit (MDL) and quantitative limit (MQL) were investigated by spiking experiments at 25 ppb target analytes with 250 ppb mixture of three surrogate compounds Based on European Pharmacopoeia guideline, LOD and LOQ was calculated by 3*S/N and 10*S/N, respectively, where S/N refers for signal to noise ratio From Table 3, MDL ranged from 0.0002 ng mL-1 (azoxystrobin) to 0.493 ng mL-1 (parathion-methyl)
Table 3: Calibration curves, MDL, MQL of twelve native compounds in this study
Compound Precursor (Polarity) RT ± SD, min Regression equation R2 MDL
(ng/mL)
MQL (ng/mL)
Recovery (%) Dimethoate 230.0069 (+) 5.48 ± 0.03 Y = 1.512e-2*X + 1.03e-1 0.9995 0.001 0.004 93.68 Parathion-methyl 261.9944 (-) 6.40 ± 0.02 Y = 2.631e4*X + 1.886e5 0.9977 0.493 1.644 117.71 Dichlorvos 220.9532 (+) 7.81 ± 0.03 Y = 4.096e-3*X - 3.452e-2 0.9999 0.157 0.522 85.74 Methomyl 163.0536 (+) 4.47 ± 0.13 Y = 2.868e-4*X - 3.706e-3 0.9939 0.282 0.940 74.40 Carbaryl 202.0862 (+) 9.05 ± 0.04 Y = 1.024e-2*X + 8.166e-2 0.9995 0.012 0.042 100.33 Azoxystrobin 404.1241 (+) 13.19 ± 0.03 Y = 4.916e-2*X + 1.123e-1 0.9993 0.0002 0.001 62.47 Dinotefuran 201.0993 (-) 3.95 ± 0.13 Y = 2.623e5*X - 2.953e5 0.9999 0.137 0.456 73.46 Carbendazim 192.0768 (+) 4.74 ± 0.04 Y = 1.073e-3*X - 2.462e-2 0.9933 0.005 0.018 108.30 Clomazone 240.0786 (+) 11.06 ± 0.04 Y = 2.096e-2*X + 4.949e-2 0.9993 0.016 0.052 92.60 Tebufenozide 351.2078 (-) 14.35 ± 0.02 Y = 1.805e5*X + 2.073e6 0.9968 0.045 0.151 111.43 Indoxacarb 528.078 (+) 17.86 ± 0.02 Y = 3.352e-3*X - 1.168e-2 0.9959 0.047 0.157 72.00 Pyridaben 365.1449 (+) 19.29 ± 0.03 Y = 9.504e-3*X - 6.26e-3 0.9995 0.055 0.183 124.90
Trang 63.5 Analysis of real samples
In general, carbendazim and indoxacarb were
found in all of twenty investigated samples
Dinotefuran and clomazone were present with
lower frequencies (67% and 56%,
respectively) Meanwhile, insecticide namely
Tebufenozide was detected in 22% samples It
was hardly to detect organophosphate group
(dimethoate, parathion-methyl, dichlorvos),
carbaryl and pyridaben Comparison with
report taken by Rousis et.al., carbendazim was
observed with the highest frequency in all test
samples in both Spain and Italy Besides,
carbaryl, methomyl, parathion-methyl,
dichlorvos, dimethoate were not present in
collected samples in this study, resulting from
the regular recently banning pesticide uses in
TT03/2018/TT-BNNPTNN
Figure 2: The amount of individual analytes in
5 areas Figure 2 shows a summary of twelve pesticide
concentrations in five investigated areas
Indoxacarb displayed the highest concentration
(1010 ng mL-1 and 545 ng mL-1) in N3 and N5
areas, respectively Meanwhile, in both N2 and
N4 areas, carbendazim was occurred at the top
leading level (> 100 ng mL-1) It was
significant to pay attention that carbendazim is
one of pesticides are not approved by the EU,
according to the EU regulation 1107/2009
Carbendazim can be a metabolite of benomyl
and thiophanate Besides, clomazone was
found 24.1 ng mL-1- the highest value in N1
area
Figure 3: Pesticide distribution in individual
sample Figure 3 shows pesticide distribution in individual samples Generally, in N3 and N5 villages, Dinotefuran as recorded was found at the highest concentration in surface water samples The other samples were found the significant distribution of carbendazim
4 CONCLUSION This study was developed method of multi-residue pesticide analysis in surface water samples collected from several villages in Hanoi by using SPE technique for preparation and LC-ESI-Orbitrap MS for analysis This procedure can analyze simultaneous difference pesticide groups by quantification in fullscan mode at 70000-FWHM resolution and confirmation of fragmentation based on data-dependent MSMS mode at 17500-FWHM resolution This method was applied to analyze twelve pesticides in twenty investigated samples It demonstrated clearly that the surface samples were contaminated with carbendazim, dinotefuran and indoxacarb ACKNOWLEDDGEMENT
This research is funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED) under grant number 104.04-2018.331
REFERENCES [1] J Casado, D Santillo, P Johnston, “Multi-residue analysis of pesticides in surface water
by liquid chromatography quadrupole-Orbitrap high resolution tandem mass spectrometry,” Analytica Chimica Acta, vol 1024, pp 1-17,
2018 [2] Water Corporation, “QuEChERS Procedure
Trang 7For Multi-Residue Pesticide Analysis -disQuE
dispersive Sample Preparation,” Waters - The
Science of What’s Possible pp 1–8, 2011
[3] Helen V Botitsi, Spiros D Garbis,
Anastasios Economou, and Despina F Tsipi,
“Current Mass spectrometry strategies for the
analysis of pesticides and their metabolites in
food and water matrices,” Wiley Online
Library, 2010
[4] Nikolaos I Rousis, Richard Bade, Lubertus
Bijlsma, Ettore Zuccato, Juan V Sancho, Felix
Hernandez, Sara Castiglioni, “Monitoring a
large number of pesticides and transformation
products in water samples from Spain and
Italy,” Environmental Research, vol 156, pp
31-38, 2017
[5] R Meffe and I de Bustamante, “Emerging
organic contaminants in surface water and
groundwater: A first overview of the situation
in Italy,” Sci Total Environ., vol 481, no 1,
pp 280–295, 2014
[6] Anastassiades et al, “Fast and Easy
Multiresidue Method Employing Acetonitrile,”
J AOAC Int., vol 86, no 2, pp 412–431,
2003
[7] Ministry of Agriculture and Rural
development, Circulars No.03, 2018
[8] N Ochiai et al., “Optimization of a
multi-residue screening method for the determination
of 85 pesticides in selected food matrices by
stir bar sorptive extraction and thermal
desorption GC-MS,” J Sep Sci., vol 28, no
9–10, pp 1083–1092, 2005
[9] Helvécio C Menezes, Breno P Paulo,
Maria Jose N Paiva, and Zenilda L Cardeal,
“A simple and quick method for the
determination of pesticides in environmental
water by HF-LPME-GCMS,” Journal of
Analytical methods in Chemistry, vol 2016
[10] Rajski, M M Gómez-Ramos, and A R
Fernández-Alba, “Application of
LC-Time-of-Flight and Orbitrap-MS/MS for Pesticide
Residues in Fruits and Vegetables,”
Comprehensive Analytical Chemistry, vol 71
pp 119–154, 2016
[11] A Kamel, “Refined methodology for the
determination of neonicotinoid pesticides and
their metabolites in honey bees and bee products by liquid chromatography-tandem mass spectrometry (LC-MS/MS),” J Agric Food Chem., vol 58, no 10, pp 5926–5931,
2010
[12] S J Lehotay, Y Sapozhnikova, and H G
J Mol, “Current issues involving screening and identification of chemical contaminants in foods by mass spectrometry,” TrAC - Trends Anal Chem., vol 69, pp 62–75, 2015
[13] J F García-Reyes, M D Hernando, A Molina-Díaz, and A R Fernández-Alba,
“Comprehensive screening of target, non-target and unknown pesticides in food by LC-TOF-MS,” TrAC - Trends in Analytical Chemistry, vol 26, no 8 pp 828–841, 2007
[14] California Department of Food and Agriculture, “Determination of Acephate and Methamidophos in SW by LCMS,” 2008 [15] “Determination of pesticide residues in
https://www.researchgate.net/publication/3038 43678