Talanta 132 (2015) 197–204
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Determination of pesticide residues in samples of green minor crops by gas chromatography and ultra performance liquid chromatography coupled to tandem quadrupole mass spectrometry Stanisław Walorczyk n, Dariusz Drożdżyński, Roman Kierzek Institute of Plant Protection-National Research Institute, Władysława Węgorka 20, 60-318 Poznań, Poland
art ic l e i nf o
a b s t r a c t
Article history: Received 21 June 2014 Received in revised form 19 August 2014 Accepted 28 August 2014 Available online 8 September 2014
A method was developed for pesticide analysis in samples of high chlorophyll content belonging to the group of minor crops. A new type of sorbent, known as ChloroFiltr, was employed for dispersive-solid phase extraction cleanup (dispersive-SPE) to reduce the unwanted matrix background prior to concurrent analysis by gas chromatography and ultra-performance liquid chromatography coupled to tandem quadrupole mass spectrometry (GC–MS/MS and UPLC–MS/MS). Validation experiments were carried out on green, unripe plants of lupin, white mustard and sorghum. The overall recoveries at the three spiking levels of 0.01, 0.05 and 0.5 mg kg 1 fell in the range between 68 and 120% (98% on average) and 72–104% (93% on average) with relative standard deviation (RSD) values between 2 and 19% (7% on average) and 3–16% (6% on average) by GC–MS/MS and UPLC–MS/MS technique, respectively. Because of strong enhancement or suppression matrix effects (absolute values 420%) which were exhibited by about 80% of the pesticide and matrix combinations, acceptably accurate quantification was achieved by using matrix-matched standards. Up to now, the proposed method has been successfully used to study the dissipation patterns of pesticides after application on lupin, white mustard, soya bean, sunflower and field bean in experimental plot trials conducted in Poland. & 2014 Elsevier B.V. All rights reserved.
Keywords: Pesticide residue analysis Chromatography Tandem mass spectrometry Dispersive solid phase extraction Minor crops
1. Introduction A particular crop may be called “minor” if such small amounts are grown that it will provide a limited market for pesticides. If the crop is considered to be of low economic importance at a national level, the pesticide manufacturers have little interest to do expensive research and development work needed for the registration of pesticides for use on the crop, and as a consequence of such a situation, the crop will have limited options for protection against pests and pathogens. Therefore, there are little data, or effectively no data available, about efficacy and residue behaviour of candidate pesticides having potential to be applied for the protection of minor crops. For authorization of pesticides on minor crops, or for minor use, it is preferable to explore other possibilities for determining the efficacy and crop safety of pesticides than those based on the amount of data required for authorization on major crops [1,2]. Although minor crops are grown in relatively small amounts compared with major crops, they are of substantial economic
n
Corresponding author: Tel.: þ 48 618649181; fax: þ 48 618679180. E-mail addresses: [email protected], [email protected] (S. Walorczyk). http://dx.doi.org/10.1016/j.talanta.2014.08.073 0039-9140/& 2014 Elsevier B.V. All rights reserved.
importance in many countries. Which crops are minor largely depends on the specific country and region. In Poland, the list of the plants classified as minor crops is published by the Ministry of Agriculture and Rural Development, and it comprises various crop groups including vegetables, fruit and berry plants, industrial plants and cereals, herbaceous plants and forest nurseries plants [3]. In this work, we focus on the analysis of pesticide residues in green crops of various plants including lupin, white mustard, soya, sunflower, and field bean by using gas chromatography and ultra performance liquid chromatography coupled with tandem quadrupole mass spectrometry (GC–MS/MS and UPLC–MS/MS)-based methods. It must be highlighted, that the majority of up to now published pesticide residue methods were mainly focused on analysis of less complicated matrices such as fruit and vegetables [4]. The main reason might be the fact that fresh fruits and vegetables are consumed in larger amounts than other crops, so there is a risk for high intake of pesticide residues, especially when they are present above their legislative maximum residue levels (MRLs) [5]. However, another reason is that multiresidue pesticide analysis in matrices of high chlorophyll content is more difficult owing to matrix interferences and complicated extraction procedures [6]. In pesticide analysis, green matrices, high in chlorophyll, represent a particular challenge due to a massive load of co-extractives in the extract. The co-extracted
198
S. Walorczyk et al. / Talanta 132 (2015) 197–204
chlorophyll is one of the most problematic matrix interferences in pesticide residue analysis because of its non-volatile characteristics. It may hinder identification of the pesticides of interest by contamination of the chromatographic systems, causing downtime, and reducing the overall analytical performance [7,8]. The use of graphitized carbon black (GCB) as a sorbent has been shown to be effective for removal of co-extracted chlorophyll and other pigments from extract of plant crops [9–15]. On the other hand, GCB is well-known to adsorb pesticides with planar functionality leading to unsatisfactory recoveries of a number of pesticides susceptible to this adsorption, e.g. chlorothalonil, cyprodinil, fenazaquin, mepanipyrim, pirymethanil, prochloraz, quinoxifen, quintozene and thiabendazole [16–18]. The main objective of this work was to investigate the possibility of application of a new type of sorbent, known as ChloroFiltr, in order to reduce the content of chlorophyll from extracts of green plants belonging to the category of minor crops, and thereby reduce the unwanted background while not affecting the recovery of the target pesticides. For the final determination, concurrent analyses were carried out by using a programmable temperature vaporization injector (PTV) gas chromatography–tandem quadrupole mass spectrometry (GC–MS/MS) and ultra-performance liquid chromatography–tandem quadrupole mass spectrometry (UPLC–MS/MS) to achieve improved selectivity and high accuracy. Comprehensive method validation was carried out to evaluate fitness for the intended application, and it involved determination of recovery, precision, linearity, assessment of matrix effects and estimation of measurement uncertainty. The developed and validated method was applied to the study of dissipation patterns of pesticides in experimental plot trials after application to minor crops including lupin, white mustard, soya bean, field bean and sunflower.
2. Material and methods 2.1. Chemicals and reagents Acetonitrile and acetone (for residue analysis) were obtained from Witko (Łódź, Poland). Anhydrous magnesium sulphate (reagent grade) and Supel Que Citrate (EN) tubes containing 4 g magnesium sulphate, 1 g sodium chloride, 0.5 g sodium citrate dibasic sesquihydrate, 1 g sodium citrate tribasic dehydrate, water with 0.1% formic acid (LC–MS Chromasolv) and methanol with 0.1% ammonium acetate were obtained from Sigma-Aldrich Sp.z o. o. (Poznań, Poland). Enviro Clean extraction tubes containing 900 mg magnesium sulphate, 300 mg PSA and 150 mg ChloroFiltr were obtained from UCT (Bristol, PA, USA). Deionized water of resistivity 18.2 MΩ cm was prepared with a Milli-Q Plus system from Millipore (Billerica, USA). 2.2. Pesticide analytical standards Certified pesticide standards and internal standards triphenylphosphate (TPP) and simazine-d10 were obtained from Dr. Ehrenstorfer (Augsburg, Germany). Stock solutions (approximately 1000 μg mL 1) were prepared in acetone. Purities of the standards were accounted for when calculating the concentration of each stock solution. A blend stock solution of all pesticides at a concentration of 10 μg mL 1 was prepared in acetone. The working standards at 0.005, 0.01, 0.02, 0.05, 0.2, 0.5 and 1.0 μg mL 1 were prepared by diluting this stock solution with acetonitrile and 0.1% ammonium acetate in methanol/0.1% formic acid in water (1:1, v/v) for the GC–MS/MS and UPLC–MS/MS analysis, respectively. Matrix-matched standards were prepared differently for GC–MS/MS and UPLC–MS/MS analysis. For the GC–MS/MS
analysis, a volume (1.5 mL) of the standard at appropriate concentration was evaporated under nitrogen and reconstituted in acetonitrile sample extract at a concentration of 0.5 g mL 1. While for the UPLC–MS/MS analysis, a volume (1 mL) of the standard at appropriate concentration and 2 mL acetonitrile sample extract at a concentration of 0.5 g mL 1 were evaporated under nitrogen then reconstituted in 1 mL methanol with 0.1% ammonium acetate/water with 0.1% formic acid (1:1, v/v). 2.3. GC–MS/MS conditions The GC–MS/MS analysis of was carried out using a Varian CP3800 gas chromatograph coupled with a Varian 1200 triple quadrupole mass spectrometer (Varian Inc., Middelburg, The Netherlands). The analyte separation was obtained on a DB-5 30 m 0.25 mm 0.5 μm capillary column, protected by a deactivated guard column (2 m 0.53 mm). Helium (purity 99.9999%) at a flow rate of 1.2 mL min 1 was used as the carrier gas. The column oven temperature programme was as follows: 80 1C (held for 3 min), programmed at 30 1C min 1to 150 1C, then programmed at 10 1C min 1 to 300 1C (held for 10 min). Large volume injection (LVI) with programmed temperature vaporization (PTV) was used. The injection volume was 10 μL of sample extract in acetonitrile using a 100 μL syringe. The injector temperature programme was as follows: 70 1C (held for 0.5 min), programmed at 200 1C min 1to 300 1C (held for 15 min). The injector split ratio was initially set at 100:1, the splitless mode was enabled at 0.5 min, at 4 min the split ratio was set at 50:1, and it was reduced to 20:1 at 10 min. The mass spectrometer was operated in electron impact ionization mode (EI, 70 eV) with the filament current of 50 μA and electron multiplier voltage at 300 V above the voltage determined by automatic tuning with perfluorotribulylamine (PFTBA). The manifold ion source and transfer line temperatures were 40, 270 and 290 1C, respectively. The collision gas (argon, 99.9998% purity) was set at the collision cell pressure of 1.7 mTorr. Multiple reaction monitoring (MRM) transitions and other acquisition parameters can be found in supplementary data included with this article. Instrument control, data acquisition and evaluation was performed by using a Varian MS Workstation software, version 6.6. 2.4. UPLC–MS/MS conditions The UPLC–MS/MS analysis was carried out using a Waters ACQUITY UPLC ultra-performance liquid chromatography system (Milford, USA) coupled with a triple quadrupole mass spectrometer (Waters Inc., Micromass, Quattro Premier XE). The nebulizer and desolvation gas was obtained from a nitrogen generator model NM30-LA (Peak Scientific, Renfrewshire, Scotland, UK). The analyte separation was achieved using a BEH C18 100 mm 2.1 mm 1.7 μm UPLC column protected by a VanGuard PreColumn 5 mm 2.1 mm 1.7 μm. The temperature of the column was thermostated at 40 1C. The column was eluted with the mobile phase: water with 0.1% formic acid (A) and methanol with 0.1% ammonium acetate (B) at the flow rate of 0.3 mL min 1 using gradient mode. Gradient was programmed to increase the amount of B from an initial content of 10–100% in 6 min, held for 1 min and returned to the initial conditions (10% B) in 1 min (from 7 to 8 min), held for 1 min. The total run time was 9 min. Sample extract volumes of 5 μL were injected into the system. The temperature of the autosampler was thermostated at 25 1C. For the MS/MS data acquisition, the interface conditions were optimized for maximum intensity of the precursor ions: nebulizer and desolvation gas (nitrogen) flows were 100 L h 1 and 700 L h 1, respectively, source block and desolvation temperatures were 120 1C
S. Walorczyk et al. / Talanta 132 (2015) 197–204
and 350 1C, respectively. Argon (99.9998% purity) was used as the collision gas at the pressure of 6.9 10 3 mbar. The optimization of multiple reaction monitoring (MRM) transitions was performed individually for each analyte on the instrument used in this work. All the compounds were analysed using positive electrospray ionization mode (ESIþ ) except for bentazone which was analysed in the negative electrospray ionization mode (ESI ). Multiple reaction monitoring (MRM) transitions and other acquisition parameters can be found in supplementary data included with this article. The instrument was controlled using Waters MassLynx 4.0 software and data evaluation was carried out using Waters TargetLynx software. 2.5. Sample preparation procedure A 5 g homogenized sample was weighted into a polypropylene centrifuge tube (50 mL), internal standards (50 μL TPP at 150 μg mL 1 and 100 μL simazine-d10 at 87 μg mL 1) were added and the sample was and extracted with 5 mL deionized water (Milli-Q) and 10 mL acetonitrile on a laboratory shaker (Heidolph Promax 2020, Schwabach, Germany) for 10 min. Then 0.5 g disodium hydrogencitrate sesquehydrate, 1 g trisodium citrate dihydrate, 4 g anhydrous magnesium sulphate, and 1 g sodium chloride were added and immediately shaken for 1 min, and then centrifuged for 2.5 min at 4500 rpm. A 6 mL aliquot of the acetonintrile supernatant was transferred to a polypropylene centrifuge tube (15 mL) containing 0.9 g anhydrous magnesium sulphate, 0.3 g PSA, and 0.3 g ChloroFiltr. The contents were vortexed for 1 min and centrifuged for 2.5 min at 4500 rpm. For the GC–MS/MS analysis, an aliquot ( 0.4 mL) of the supernatant was transferred into an autosampler vial with a fixed insert. For the UPLC–MS/MS, 2 mL of the supernatant were evaporated to dryness under nitrogen, reconstituted with 1 mL 0.1% ammonium acetate in methanol/0.1% formic acid in water (1:1, v/v), and filtered using a 13 mm, 0.2 m GHP membrane into an autosampler vial. The final concentration of sample was 0.5 g mL 1 and 1 g mL 1 in the case of GC–MS/MS and UPLC–MS/MS analysis, respectively. 2.6. Validation study The scope of the developed method comprises 25 pesticides (including 15 herbicides, nine fungicides and two herbicides). These are candidate pesticides having potential to be applied for control of the most important pests and pathogens affecting the production of the minor crops under study (i.e. field bean, lupin, sorghum, soya bean, sunflower and white mustard). Data on pesticides, matrices and conducted field experiments are detailed in Supplementary information included with this article. The developed method was subjected to validation study using green, unripe plants of lupin, white mustard and sorghum samples (previously checked to be free of the target pesticides) in order to determine linearity, matrix induced effects, recovery, precision, limit of quantification (LOQ) and measurement uncertainty. The calibration was carried out by internal standard method with reference to TPP and simazine-d10 which served as the internal standard for the GC–MS/MS and UPLC–MS/MS analytes, respectively. Linearity was studied by injecting standards at pesticide concentrations 0.005–1.0 μg mL 1 and 0.01–1.0 μg mL 1 in acetonitrile (GC–MS/MS) or 1 mL methanol with 0.1% ammonium acetate/water with 0.1% formic acid, 1:1, v/v (UPLC–MS/MS), and in matrix extracts of lupin, white mustard and sorghum. These concentration ranges corresponded to pesticide content in real samples between 0.01 and 2.0 mg kg 1 (GC–MS/MS) and 0.01 and 1.0 mg kg 1 (UPLC–MS/MS).
199
The percentage of matrix effect (%ME) was obtained for each analyte from the slopes of the calibration curves and was determined by comparing solvent and matrix-matched (lupin, white mustard and sorghum) slopes by using the formula: %ME¼100% (1slopesolvent/slopematrix) [19]. The recoveries were studied at the three spiking levels of 0.01, 0.05 and 0.5 mg kg 1 (n ¼6) on lupin and additional recovery tests at 0.05 mg kg 1 (n¼ 6) were carried out on white mustard and sorghum. Repeatability precision was expressed in terms of relative standard deviation (RSD) and calculated for each spiking level (n ¼6). The samples were spiked with the pesticides before proceeding with the extraction and the results from the recovery study were assessed for compliance with the European Union guidelines SANCO/12571/2013, according to which the average recovery should fall in the range of 70–120% with an associated RSD less or equal 20% [20]. The LOQ was defined as the lowest spiking level validated with satisfactory values of recovery (70–120%) and RSD (r20%). The measurement uncertainty was estimated based on the data obtained in the validation study. The major uncertainty sources included in the uncertainty budget were the repeatability of recoveries from spiked samples and uncertainty of the average recovery calculated from rectangular distribution. The relative expanded uncertainty was calculated by using the coverage factor k¼ 2 at the confidence level of 95% [21].
3. Results and discussion 3.1. GC–MS/MS and UPLC–MS/MS conditions MRM-based data acquisition methods with two or three MS/MS transitions for each analyte were developed on both instruments in order to carry out reliable quantification and identification of the pesticides in samples. Optimization of the conditions was started by using pesticide standards with the mass spectrometer operating in full scan mode to obtain the single stage MS spectrum. Then the fragmentation pattern of the selected precursor ions was studied in product ion scan mode at different collision energies (5–50 eV). In the case of UPLC–MS/MS, the pseudo-molecular ion [MþH] þ (and [M H] þ for bentazone) was present as the peak base of the MS spectrum for all the compounds except for desmedipham and phenmedipham. The [MþH] þ ion was used as the precursor for the MS/MS fragmentation. In the case of desmedipham and phenmedipham, due to lack of sensitivity of [Mþ H] þ ion, the ammonium adduct [MþNH4] þ was chosen as the precursor ion. For each target compound, the most selective and sensitive transitions were determined and used for the detection and quantification of the target analytes. To avoid chances of matrix interferences, two or three transitions were used except for internal standards (TPP and simazine-d10). The GC–MS/MS and UPLC–MS/MS acquisition conditions including precursor and product ions, collision energies and other parameters are detailed in Supplementary information included with this article. 3.2. Sample preparation and method validation Plant pigments including chlorophyll may contribute to high background in chromatograms and occurrence of matrix effects and thereby affect identification and quantification of the compounds of interest. The pigments can thermally degrade in the injection liner or accumulate in the columns which can lead to increased maintenance of the instrument. Thus, routine analysis of pesticide residues in chlorophyll containing matrices might be difficult and the proper choice of sorbent(s) is critical to remove
200
S. Walorczyk et al. / Talanta 132 (2015) 197–204
co-extracted matrix interferences, while assuring good recoveries of the target analytes. To prepare samples before chromatographic analysis, we employed the well-known and currently widely used quick, easy, cheap, effective, rugged and safe (QuEChERS) strategy to sample extraction and cleanup as this is a very flexible approach and offers many options for analysis depending on the range of pesticides and matrices being tested [22]. But we applied mechanical shaking for an extended time of 10 min (compared with 30 s of manual shaking in some other versions of the QuEChERS) [23]. Also, we lowered the sample amount (to 5 g) and added water (5 ml) before proceeding with the acetonitrile extraction to enable the extraction solvent penetrate better the plant tissues and ensure complete transfer of the analytes from naturally contaminated samples [24]. The QuEChERS technique inherently involves a dispersive-solid phase extraction (dispersive-SPE) step, for which the base sorbent is primary and secondary amine (PSA) which primarily removes fatty acids and other organic acids, sugars, and some pigments. The matrix constituents that might adversely affect the chromatographic analysis. In addition, a new type of sorbent, known as ChloroFiltr, was evaluated to help the dispersive-SPE performance and complement the PSA cleanup by reduction of the content of chlorophyll from green plant extracts [25,26]. Once the sample preparation method was selected, it was subjected to comprehensive validation study to prove its fitness for purpose through the assessment of analytical performance parameters including linearity, matrix effects, recovery, precision and estimation of measurement uncertainty. 3.3. Study of linearity and matrix effect Generally, the linearity parameters were highly satisfactory. The calibration curves were studied by using internal standardization in solvent solutions and matrix-matched in extracts of lupin, white mustard and sorghum, and gave the coefficients of determination (R2) of the least squared linear calibration curves 40.99. The exceptions occurred only for desmedipham, flufenacet, phenmedipham and prochloraz in the case of solvent-only calibration (GC–MS/MS), and bentazone in the case of solvent-only, lupin- and sorghum-matched calibration (UPLC–MS/MS). Detailed linearity data (equations and R2 values) obtained by using solvent-only and matrix-matched standards can be found in Supplementary information included with this article. The percentage of matrix effect (%ME) was calculated as the difference between the slope of the matrix-matched and solventonly calibration curves divided by the slope of solvent-only calibration curve. %ME in the range between 20 and 20% can be considered as insignificant because such variability is close to the repeatability RSD values [27,28]. But for most of the compounds under study, much higher values of %ME were present and both enhancement as well as suppression effects could be observed (Table 1). In the case of UPLC–MS/MS, the analyte responses produced by matrix-matched standards were nearly always substantially higher than the responses obtained for the equivalent standards in solvent, so the matrix enhancement effect occurred predominantly. Otherwise, in the case of GC–MS/MS analysis, the matrix effects occurred in a larger extent ( 116 to 82%) but no consistent pattern with respect to different analytes could be observed, i.e. enhancement, suppression, or negligible matrix effects o20% (absolute value) occurred for different compounds. The prevailing occurrence of enhancement effect in the case of UPLC–MS/MS technique was contrary to our expectation, because the ion suppression effect tend to be more common in positive electrospray mode (ESI þ) [29]. A possible explanation for the phenomenon we observed might be a specific suppression of the
Table 1 Matrix effects of the pesticides in the GC–MS/MS and UPLC–MS/MS analysis of lupin, white mustard and sorghum extracts, determined as 100% (1 slopesolvent/ slopematrix). Pesticides
Acetochlor Azoxystrobin Bentazone Boscalid Clomazone Cyproconazole_1 Cyproconazole_2 Deltamethrin 1 Deltamethrin 2 Desmedipham Dimethenamid-P Dimoxystrobin Fluazifop-P-butyl Flufenacet Linuron Metamitron Metconazole Metolachlor-S Pendimethalin Phenmedipham Picoxystrobin Prochloraz Propiconazole 1 Propiconazole 2 Prosulfocarb Quizalofop-P-ethyl Tebuconazole Thiacloprid
Lupine
White mustard
Sorghum
GC
UPLC
GC
UPLC
GC
UPLC
76 24 – 2 88 29 – 68 20 75 70 19 22 44 – – 31 45 39 79 26 11 40 45 – 4 23 –
43 42 62 46 38 38 43 – – 39 48 39 1 38 38 52 39 42 91 37 30 27 41 – 23 8 40 46
88 12 – 4 44 19 – 70 30 77 62 15 16 48 – – 25 48 39 82 20 116 25 26 – 4 23 –
53 51 57 53 48 43 51 – – 47 55 49 38 49 50 59 47 52 5 46 46 41 53 – 45 31 49 51
72 20 – 5 59 19 – 56 9 73 70 15 16 37 – – 28 48 33 77 20 2 30 36 – 4 14 –
43 43 64 43 32 37 40 – – 38 46 39 20 41 38 50 35 61 50 34 35 28 41 – 28 8 38 40
internal standard (simazine-d10) occurring due to co-eluting matrix compounds which resulted in analytes overestimation. However, this was not a major concern due to consistency of this effect among the matrices which led to sufficiently accurate quantification, provided that matrix-matched calibration would be used. By using either techniques (GC–MS/MS or UPLC–MS/MS), the matrix effects were relatively consistent among studied matrices for the majority of the pesticides, so application of matrixmatched standards in any matrix would give acceptably accurate results for quantification of pesticide residues in other matrices, as will be presented further in this section. The exceptions, in the case of GC–MS/MS technique, were flufenacet and prochloraz. Flufenacet exhibited positive effect in lupin and sorghum but negative in white mustard whereas prochloraz exhibited considerable stronger negative matrix effect in white mustard ( 116%) than in lupin ( 11%) and sorghum ( 2%). In the case of UPLC–MS/ MS technique, inconsistencies in the occurrence of matrix effects were observed for fluazifop-P-butyl, quizalofop-P-ethyl and pendimethalin. In the case of fluazifop-P-butyl and quizalofop-P-ethyl, stronger positive matrix effects was observed in white mustard than in lupin and sorghum. In contrast, in the case of pendimethalin, stronger negative matrix effects were observed in lupin ( 91%) and sorghum ( 50%) than in white mustard ( 5%). Whereas negligible matrix effects, with absolute values o20% in all matrices, were only observed in the case of boscalid, dimoxystrobin and quizalofop-P-ethyl when using the GC–MS/MS technique. Because of common occurrence of matrix effects in analysis of pesticide residues in chlorophyll containing matrices under this study, the use of matrix-matched calibration was necessary as a practical approach to compensate for matrix effects and achieve
S. Walorczyk et al. / Talanta 132 (2015) 197–204
201
Table 2 Average recoveries, RSDs and expanded uncertainties U, % (k ¼ 2, confidence level 95%) in the GC–MS/MS and UPLC-MS/MS analysis of pesticides in green lupin. Pesticides
Recovery, %
U, %
0.01
Acetochlor Azoxystrobin Bentazone Boscalid Clomazone Cyproconazole_1 Cyproconazole_2 Deltamethrin 1 Deltamethrin 2 Desmedipham Dimethenamid-P Dimoxystrobin Fluazifop-P-butyl Flufenacet Linuron Metamitron Metconazole Metolachlor-S Pendimethalin Phenmedipham Picoxystrobin Prochloraz Propiconazole 1 Propiconazole 2 Prosulfocarb Quizalofop-P-ethyl Tebuconazole Thiacloprid
0.05
0.5
(k ¼ 2, α¼ 0.05)
Overall
GC
UPLC
GC
UPLC
GC
UPLC
GC
UPLC
GC
UPLC
122 (2) 107 (8) – 101 (4) 111 (6) 91 (8) – 106 (13) 82 (6) – 81 (10) 113 (3) 94 (4) 94 (18) – – – 109 (7) 119 (3) 119 (3) 101 (19) – 106 (5) 109 (8) – 86 (5) 90 (5) –
97 (18) 129 (7) – 102 (5) 128 (9) – 114 (7) – – 92 (8) – 116 (8) 107 (14) 123 (6) 112 (10) 130 (7) 105 (12) 126 (9) 75 (22) 100 (7) 110 (7) 103 (4) 112 (5) – 100 (8) 87 (20) 134 (8) 124 (10)
120 (3) 99 (8) – 84 (1) 122 (3) 91 (1) – 105 (8) 87 (5) 78 (4) 122 (2) 97 (5) 103 (4) 84 (4) – – 92 (10) 118 (1) 117 (4) 79 (6) 101 (4) 67 (6) 110 (6) 105 (13) – 94 (3) 90 (1) –
79 (7) 85 (3) 89 (9) 78 (3) 84 (2) 98 (7) 72 (1) – – 68 (5) 94 (11) 86 (2) 84 (3) 85 (2) 77 (4) 82 (4) 75 (4) 87 (2) 62 (13) 70 (4) 89 (4) 74 (7) 79 (3) – 76 (2) 79 (4) 85 (5) 86 (4)
119 (1) 100 (1) – 84 (1) 120 (1) 96 (3) – 96 (4) 81 (2) 81 (3) 120 (1) 103 (1) 102 (2) 85 (4) – – 96 (4) 115 (2) 117 (1) 80 (3) 101 (1) 69 (2) 94 (1) 92 (1) – 93 (1) 95 (1) –
98 (1) 94 (3) 60 (5) 91 (3) 94 (2) 109 (1) 94 (1) – – 84 (5) 92 (4) 98 (2) 92 (4) 98 (1) 89 (1) 93 (2) 89 (4) 96 (1) 79 (13) 84 (4) 99 (2) 86 (9) 92 (4) – 90 (5) 85 (6) 94 (3) 96 (3)
120 (2) 102 (7) – 90 (10) 118 (6) 92 (5) – 102 (9) 83 (5) 80 (4) 108 (19) 104 (7) 100 (5) 88 (11) – – 94 (7) 114 (5) 118 (3) 83 (9) 101 (10) 68 (4) 104 (8) 102 (11) – 91 (5) 92 (4) –
91 (9) 103 (4) 75 (9) 90 (4) 102 (4) 104 (7) 93 (3) – – 81 (6) 93 (8) 100 (4) 94 (7) 102 (3) 93 (5) 102 (4) 90 (7) 103 (4) 72 (16) 85 (5) 99 (4) 88 (7) 94 (4) – 89 (12) 84 (10) 104 (5) 102 (6)
13 14 – 20 15 11 – 19 14 14 39 15 10 24 – – 14 13 11 21 20 20 16 21 – 11 9 –
18 9 23 9 9 15 7 – – 16 16 8 14 6 11 9 15 8 36 16 9 15 9 – 12 22 11 11
acceptably accurate quantificative results in both GC–MS/MS and UPLC–MS/MS analyses. 3.4. Recovery study and estimation of measurement uncertainty For the purpose of evaluating the accuracy (trueness and precision) of the proposed method, a recovery study was carried out by spiking lupin, white mustard and sorghum samples with the target pesticides. The recovery and repeatability RSD values were determined by the analysis of six replicate spiked samples at each level (0.01, 0.05 and 0.5 mg kg 1). The results for calculation of recoveries were obtained by using matrix-matched standards prepared in lupin extracts. As can be seen in Table 2, most of the compounds presented satisfactory recoveries from lupin in the range between 70% and 120% with RSD values r20%, i.e. compliant with the EU criteria of SANCO/12571/2013 [20]. For a limited number of compounds, especially at the lowest spiking level of 0.01 mg kg 1, the recovery and RSD results were slightly outside the acceptance criteria. The most problematic compound was bentazone which generally did not present satisfactory results. But the overall recoveries of the target pesticides at the three spiking levels of 0.01, 0.05 and 0.5 mg kg 1 fell in the range between 68 and 120% (98% on average) and 72–104% (93% on average) with RSD values between 2 and 19% (7% on average) and 3–16% (6% on average) by GC–MS/ MS and UPLC–MS/MS techniques, respectively. Additional recovery evaluation was carried out on white mustard and sorghum (spiking level 0.05 mg kg 1, n¼6). In this experiment, the results for calculation of recoveries were also obtained by using matrix-matched standards prepared in lupin extracts. Generally, the results were very satisfactory and fulfilled the recovery and precision requirements of SANCO/12571/2013 (Table 3). Flufenacet and
Table 3 Average recoveries and RSDs in the GC–MS/MS and UPLC–MS/MS analysis of pesticides in white mustard and sorghum (spiking level 0.05 mg kg 1, n¼ 6). Pesticides
Recovery (%) White mustard
Acetochlor Azoxystrobin Bentazone Boscalid Clomazone Cyproconazole_1 Cyproconazole_2 Deltamethrin 1 Deltamethrin 2 Desmedipham Dimethenamid-P Dimoxystrobin Fluazifop-P-butyl Flufenacet Linuron Metamitron Metconazole Metolachlor-S Pendimethalin Phenmedipham Picoxystrobin Prochloraz Propiconazole 1 Propiconazole 2 Prosulfocarb Quizalofop-P-ethyl Tebuconazole Thiacloprid
Sorghum
GC
UPLC
GC
UPLC
70 (14) 78 (6) – 91 (7) 93 (9) 101 (10) – 97 (17) 87 (4) 106 (9) 86 (7) 109 (6) 96 (2) 22 (23) – – 110 (10) 79 (11) 100 (11) 99 (17) 90 (9) 45 (33) 102 (12) 95 (13) – 96 (5) 96 (3) –
93 (6) 85 (4) 25 (24) 77 (4) 91 (1) 96 (6) 75 (4) – – 78 (3) 91 (10) 88 (2) 91 (3) 84 (2) 106 (2) 105 (1) 79 (4) 92 (4) 87 (7) 79 (3) 100 (2) 89 (2) 84 (3) – 89 (1) 94 (3) 75 (1) 94 (2)
102 (11) 87 (7) – 100 (10) 107 (10) 105 (13) – – – 76 (8) 109 (6) 103 (14) 99 (11) 113 (8) – – 100 (20) 99 (6) 98 (18) 88 (12) 102 (12) 97 (5) 108 (10) 96 (17) – 101 (5) 104 (13) –
91 (3) 88 (3) 56 (17) 79 (5) 88 (2) 103 (4) 79 (6) – – 79 (2) 99 (11) 91 (3) 89 (3) 88 (4) 90 (4) 107 (2) 76 (3) 94 (2) 74 (7) 78 (1) 98 (1) 90 (5) 81 (2) – 84 (3) 92 (1) 74 (4) 95 (1)
202
S. Walorczyk et al. / Talanta 132 (2015) 197–204
prochloraz were the exceptions, which presented lower recoveries in white mustard in the case of GC–MS/MS analysis and so was bentazone which presented lower recoveries in white mustard and sorghum in the case of UPLC–MS/MS analysis. These phemomena were likely caused by the occurrence of strong matrix effects, as it was discussed previously in this section. Matrix-matching in the same matrix will be needed to obtain more accurate quantification for these problematic compounds. The limit of quantification (LOQ) was determined as the lowest spiking level at which the validation criteria were satisfied with average recovery 70–120% and RSD r20%, and it was 0.01 mg kg 1 (except for bentazone and metamitron). The data derived from the validation study (recovery and RSD) were used to estimate the measurement uncertainty associated with the analytical results when using the developed method to analyse real samples. Precision was identified as the main contribution to the uncertainty, but the uncertainty associated with the recovery, as calculated from rectangular distribution, was also included in the uncertainty budget to avoid underestimation of the total uncertainty [21]. Taking into account the overall recovery and RSD data (Table 2), the calculated measurement uncertainty values ranged from 9% to 39% (17% on average) and from 6% to 36% (13% on average), by using GC–MS/MS and UPLC–MS/MS techniques, respectively (coverage factor k ¼2, confidence level 95%). These values were lower than a default value of 750% recommended by the SANCO/12571/2013 document, demonstrating fitness for purpose of the developed and validated method.
Fig. 1. Dissipation patterns of pesticides after application on lupin.
It must be highlighted that most of the pesticides under this study could be analysed by either techniques, GC–MS/MS or UPLC– MS/MS, therefore the one to offer better analytical performance parameters for the particular pesticide and matrix combination was used when analysing real samples. 3.5. Application to real samples In this work, we focused on the study of dissipation rates of selected pesticides after application to lupin, white mustard, soya bean, sunflower and field bean. These plants were classified by the Polish Ministry of Agriculture and Rural Development as minor crops, for which there are limited options for protection against pests and pathogens. Our studies entailed the use of candidate pesticides having potential to be applied for the protection of these minor crops in field trials, followed by laboratory analysis for pesticide residues. Detailed data on tests that we have been conducted in experimental plots can be found in Supplementary information included with this article. The aim of our work was to increase the knowledge in the residue behaviour and dissipation study of the pesticides, which have recently been proposed for the use in comprehensive protection of selected minor crops by means of extension of authorization of pesticides which are already registered for use in Poland on other crops (i.e. major crops). The dissipation patterns of the pesticides were different probably due to the different chemical structures of the compounds, applied doses and preharvest times but the terminal pesticide residues, i.e. determined in the samples collected in the last sampling time, were always below the maximum residue levels (MRLs) [30]. Pesticide MRLs apply to edible plant parts but in this work unripe green plants were analysed in order to determine the dissipation patterns of the target pesticides and reassure preharvest intervals on the crops under study. Effectively, there was no need to analyse the final crops (seeds) because the target pesticides disappeared on the plants before harvest, and in most cases they fell below the method LOQ in the samples collected in the last sampling time. We can conclude that the proposed pesticide application schedules were correct and the data concerning the pesticide residue dissipation suggest that the proposed pesticides can be safely used on the studied minor crops, without the risk of the presence of problematic residues, provided that good agricultural practices have been followed. Examples of pesticide dissipation
Fig. 2. GC–MS/MS chromatogrrasams of a sample of lupin containing residues of (A) dimethenamid-P (0.016 mg kg 1), and (B) pendimethalin (0.028 mg kg 1). The sample was collected 32 days after spraying with Wing P 462.5 EC.
S. Walorczyk et al. / Talanta 132 (2015) 197–204
203
Fig. 3. UPLC–MS/MS chromatograms of (A) a sample of field bean containing thiacloprid (0.057 mg kg 1, 21 days after spraying with Proteus 110 OD), and (B) a sample of sunflower containing dimoxystrobin (0.236 mg kg 1, 14 days after spraying with Pictor 400 SC).
over time plots, showing concentration versus days after the last spraying, for different pesticides applied to lupin in a field experiment are shown in Fig. 1. Whereas, Fig. 2 shows GC–MS/MS chromatograms of a sample of lupin containing residues of dimethenamid-P (0.016 mg kg 1) and pendimethalin (0.028 mg kg 1) and Fig. 3 shows UPLC–MS/MS chromatograms of thiacloprid in field bean (0.057 mg kg 1) and dimoxystrobin in sunflower (0.236 mg kg 1). The results obtained in this work will aid in the establishment of safe and proper use of pesticides for extension of authorization on minor crops in Poland.
plants with the consideration of food safety, reduction of yield losses and threat to humans, farm animals and the environment” (contract No. HORkor.0660/IOR 2011-2015/2/2013), task No. 1.2. “Assessment of possible complex protection of minor crops”, and by Ministry of Science and Higher Education (Ministerstwo Nauki i Szkolnictwa Wyższego), project ID: POZ-07.
4. Conclusions
References
Matrices high in chlorophyll present a special difficulty in pesticide residue analysis owing to interferences which can negatively affect the method performance characteristics and create an analytical challenge. As a solution to these problems, we investigated the use of a new QuEChERS-based method of sample preparation and report it for the first time. We used a new sorbent, known as ChloroFiltr, which served for selective reduction of chlorophyll from green plant extracts, and was used in combination with PSA to keep the sample cleanup as straightforward as possible in the dispersive-SPE mode. The final analysis was carried out by using concurrent determinations by both GC–MS/MS and UPLC–MS/MS, and in the case of both techniques, the importance of using matrix-matched calibration was demonstrated as being of key importance for achieving accurate results. An important conclusion is also that relative consistency of matrix effects in both GC–MS/MS and UPLC–MS/MS analysis for a given pesticide between different matrices, makes it possible to quantify pesticide residues in different matrices using only one matrix-matched calibration curve. After validation, the developed method was successfully applied to study the dissipation patterns of pesticide residues after application on crops belonging the group of minor crops (lupin, white mustard, soya bean, sunflower and field bean).
[1] Regulation (EC) no. 1107/2009 of the European Parliament and of the Council of 21 October 2009 concerning the placing of plant protection products on the market and repealing Council Directives 79/117/EEC and 91/414/EEC. Available online: 〈http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:309:0001:0050: EN:PDF〉, (accessed 10.06.14). [2] Available online: 〈http://www.eppo.int/PPPRODUCTS/minor_uses/minor_uses. htm〉, (accessed 10.06.14). [3] Available online: 〈http://isap.sejm.gov.pl/DetailsServlet?id=WDU20130000659〉, (accessed 10.06.2014). [4] H.V. Botitsi, S.D. Garbis, A. Economou, D.F. Tsipi, Mass Spectrom. Rev. 30 (2011) 907–939. [5] European Food Safety Authority, The 2010 European Union report on pesticide residues in food, EFSA J. 11 (2013) 808 pp. Available online: 〈http://www.efsa. europa.eu/en/efsajournal/doc/3130.pdf〉, (accessed 10.06.14). [6] Y. Wang, H.Y. Jin, S.-C. Ma, J. Lu, R.-C. Lin, J. Chromatogr. A 1218 (2011) 334–342. [7] E. Muñoz, G. Muñoz, L. Pineda, E. Serrahima, F. Centrich, J. AOAC Int. 95 (2012) 1777–1796. [8] B. Łozowicka, M. Jankowska, E. Rutkowska, I. Hrynko, P. Kaczyński, J. Miciński, J. Nat. Med. 68 (2014) 95–111. [9] R.M. González-Rodríguez, R. Rial-Otero, B. Cancho-Grande, J. Simal-Gándara, J. Chromatogr. A 1196–1197 (2008) 100–109. [10] L. Li, W. Li, D. Qin, S. Jiang, F. Liu, J. AOAC Int. 92 (2009) 538–547. [11] L.-J. Qu, H. Zhang, J.-H. Zhu, G.-S. Young, H.Y. Aboul-Enein, Food Chem. 122 (2010) 327–332. [12] R. Xu, J. Wu, Y. Liu, R. Zhao, B. Chen, M. Yang, J. Chen, Chemosphere 84 (2011) 908–912. [13] M.Á. González-Curbelo, J. Harnández-Borges, L.M. Ravelo-Pérez, M.Á. Rodríguez-Delgado, Food Chem. 125 (2011) 1083–1090. [14] G. Chen, R. Cao, R. Liu, Food Chem. 125 (2011) 1406–1411. [15] M. Słowik-Borowiec, E. Szpyrka, S. Walorczyk, Bull. Environ. Contam. Toxicol. 89 (2012) 633–637. [16] H.G.J. Mol, A. Rooseboom, R. van Dam, M. Roding, K. Arondeus, S. Sunarto, Anal. Bioanal. Chem. 389 (2007) 1715–1754. [17] C.C. Leandro, P. Hancock, R.J. Fussell, B.J. Keely, J. Chromatogr. A 1166 (2007) 152–162. [18] S. Walorczyk, Rapid Commun. Mass Spectrom. 22 (2008) 3791–3801. [19] H. Kwon, S.J. Lehotay, L. Geis-Asteggiante, J. Chromatogr. A 1270 (2012) 235–245.
Acknowledgements This work was supported by Ministry of Agriculture and Rural Development (Ministerstwo Rolnictwa i Rozwoju Wsi) under LongTerm Programme of IPP–NRI for 2011–2015 “Protection of cultivated
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.talanta.2014.08.073.
204
S. Walorczyk et al. / Talanta 132 (2015) 197–204
[20] Document no. SANCO/12571/2013 (2014), Guidance document on analytical quality control and validation procedures for pesticide residues analysis in food and feed. Available online: 〈http://ec.europa.eu/food/plant/plant_protection_products/guidan ce_documents/docs/qualcontrol_en.pdf〉, (accessed 10.06.14). [21] S. Walorczyk, D. Drożdżyński, J. AOAC Int. 94 (2011) 1625–1642. [22] S.J. Lehotay, Methods Mol. Biol. 747 (2011) 65–91. [23] S.J. Lehotay, K. Ae Son, H. Kwon, U. Koesukwiwat, W. Fu, K. Mastovska, E. Hoh, N. Leepipatpoboon, J. Chromatogr. A 1217 (2010) 2548–2560. [24] O. Lacina, M. Zachariasova, J. Urbanova, M. Vaclavikova, T. Cajka, J. Hajslova, J. Chromatogr. A 1262 (2012) 8–18. [25] Available online: 〈https://www.unitedchem.com/catalog/sample-prep/solid-pha se-extraction-columns/enviro-clean/quechers-chlorofiltr〉, (accessed 19.07.14).
[26] 〈https://www.unitedchem.com/sites/default/files/docs/posters-and-papers/ SOFT_MarijuanaPoster_002_Lo-Res.pdf〉, (accessed 19.07.14). [27] S. Walorczyk, Talanta 120 (2014) 106–113. [28] C. Ferrer, A. Lozano, A. Agüera, A.J. Girón, A.R. Fernández-Alba, J. Chromatogr. A 1218 (2011) 7634–7639. [29] H. Stahnke, S. Kittlaus, G. Kempe, L. Alder, Anal. Chem. 84 (2012) 1474–1482. [30] Regulation (EC) no. 396/2005 of the European Parliament and of the Council of 23 February 2005 on maximum residue levels of pesticides in or on food and feed of plant and animal origin and amending Council Directive 91/414/EEC. Available online: 〈http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ: L:2005:070:0001:0016:en:PDF〉, (accessed 17.06.14).
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Determination of pesticide residues in samples of green minor crops by gas chromatography and ultra performance liquid chromatography coupled to tandem quadrupole mass spectrometry Stanisław Walorczyk n, Dariusz Drożdżyński, Roman Kierzek Institute of Plant Protection-National Research Institute, Władysława Węgorka 20, 60-318 Poznań, Poland
art ic l e i nf o
a b s t r a c t
Article history: Received 21 June 2014 Received in revised form 19 August 2014 Accepted 28 August 2014 Available online 8 September 2014
A method was developed for pesticide analysis in samples of high chlorophyll content belonging to the group of minor crops. A new type of sorbent, known as ChloroFiltr, was employed for dispersive-solid phase extraction cleanup (dispersive-SPE) to reduce the unwanted matrix background prior to concurrent analysis by gas chromatography and ultra-performance liquid chromatography coupled to tandem quadrupole mass spectrometry (GC–MS/MS and UPLC–MS/MS). Validation experiments were carried out on green, unripe plants of lupin, white mustard and sorghum. The overall recoveries at the three spiking levels of 0.01, 0.05 and 0.5 mg kg 1 fell in the range between 68 and 120% (98% on average) and 72–104% (93% on average) with relative standard deviation (RSD) values between 2 and 19% (7% on average) and 3–16% (6% on average) by GC–MS/MS and UPLC–MS/MS technique, respectively. Because of strong enhancement or suppression matrix effects (absolute values 420%) which were exhibited by about 80% of the pesticide and matrix combinations, acceptably accurate quantification was achieved by using matrix-matched standards. Up to now, the proposed method has been successfully used to study the dissipation patterns of pesticides after application on lupin, white mustard, soya bean, sunflower and field bean in experimental plot trials conducted in Poland. & 2014 Elsevier B.V. All rights reserved.
Keywords: Pesticide residue analysis Chromatography Tandem mass spectrometry Dispersive solid phase extraction Minor crops
1. Introduction A particular crop may be called “minor” if such small amounts are grown that it will provide a limited market for pesticides. If the crop is considered to be of low economic importance at a national level, the pesticide manufacturers have little interest to do expensive research and development work needed for the registration of pesticides for use on the crop, and as a consequence of such a situation, the crop will have limited options for protection against pests and pathogens. Therefore, there are little data, or effectively no data available, about efficacy and residue behaviour of candidate pesticides having potential to be applied for the protection of minor crops. For authorization of pesticides on minor crops, or for minor use, it is preferable to explore other possibilities for determining the efficacy and crop safety of pesticides than those based on the amount of data required for authorization on major crops [1,2]. Although minor crops are grown in relatively small amounts compared with major crops, they are of substantial economic
n
Corresponding author: Tel.: þ 48 618649181; fax: þ 48 618679180. E-mail addresses: [email protected], [email protected] (S. Walorczyk). http://dx.doi.org/10.1016/j.talanta.2014.08.073 0039-9140/& 2014 Elsevier B.V. All rights reserved.
importance in many countries. Which crops are minor largely depends on the specific country and region. In Poland, the list of the plants classified as minor crops is published by the Ministry of Agriculture and Rural Development, and it comprises various crop groups including vegetables, fruit and berry plants, industrial plants and cereals, herbaceous plants and forest nurseries plants [3]. In this work, we focus on the analysis of pesticide residues in green crops of various plants including lupin, white mustard, soya, sunflower, and field bean by using gas chromatography and ultra performance liquid chromatography coupled with tandem quadrupole mass spectrometry (GC–MS/MS and UPLC–MS/MS)-based methods. It must be highlighted, that the majority of up to now published pesticide residue methods were mainly focused on analysis of less complicated matrices such as fruit and vegetables [4]. The main reason might be the fact that fresh fruits and vegetables are consumed in larger amounts than other crops, so there is a risk for high intake of pesticide residues, especially when they are present above their legislative maximum residue levels (MRLs) [5]. However, another reason is that multiresidue pesticide analysis in matrices of high chlorophyll content is more difficult owing to matrix interferences and complicated extraction procedures [6]. In pesticide analysis, green matrices, high in chlorophyll, represent a particular challenge due to a massive load of co-extractives in the extract. The co-extracted
198
S. Walorczyk et al. / Talanta 132 (2015) 197–204
chlorophyll is one of the most problematic matrix interferences in pesticide residue analysis because of its non-volatile characteristics. It may hinder identification of the pesticides of interest by contamination of the chromatographic systems, causing downtime, and reducing the overall analytical performance [7,8]. The use of graphitized carbon black (GCB) as a sorbent has been shown to be effective for removal of co-extracted chlorophyll and other pigments from extract of plant crops [9–15]. On the other hand, GCB is well-known to adsorb pesticides with planar functionality leading to unsatisfactory recoveries of a number of pesticides susceptible to this adsorption, e.g. chlorothalonil, cyprodinil, fenazaquin, mepanipyrim, pirymethanil, prochloraz, quinoxifen, quintozene and thiabendazole [16–18]. The main objective of this work was to investigate the possibility of application of a new type of sorbent, known as ChloroFiltr, in order to reduce the content of chlorophyll from extracts of green plants belonging to the category of minor crops, and thereby reduce the unwanted background while not affecting the recovery of the target pesticides. For the final determination, concurrent analyses were carried out by using a programmable temperature vaporization injector (PTV) gas chromatography–tandem quadrupole mass spectrometry (GC–MS/MS) and ultra-performance liquid chromatography–tandem quadrupole mass spectrometry (UPLC–MS/MS) to achieve improved selectivity and high accuracy. Comprehensive method validation was carried out to evaluate fitness for the intended application, and it involved determination of recovery, precision, linearity, assessment of matrix effects and estimation of measurement uncertainty. The developed and validated method was applied to the study of dissipation patterns of pesticides in experimental plot trials after application to minor crops including lupin, white mustard, soya bean, field bean and sunflower.
2. Material and methods 2.1. Chemicals and reagents Acetonitrile and acetone (for residue analysis) were obtained from Witko (Łódź, Poland). Anhydrous magnesium sulphate (reagent grade) and Supel Que Citrate (EN) tubes containing 4 g magnesium sulphate, 1 g sodium chloride, 0.5 g sodium citrate dibasic sesquihydrate, 1 g sodium citrate tribasic dehydrate, water with 0.1% formic acid (LC–MS Chromasolv) and methanol with 0.1% ammonium acetate were obtained from Sigma-Aldrich Sp.z o. o. (Poznań, Poland). Enviro Clean extraction tubes containing 900 mg magnesium sulphate, 300 mg PSA and 150 mg ChloroFiltr were obtained from UCT (Bristol, PA, USA). Deionized water of resistivity 18.2 MΩ cm was prepared with a Milli-Q Plus system from Millipore (Billerica, USA). 2.2. Pesticide analytical standards Certified pesticide standards and internal standards triphenylphosphate (TPP) and simazine-d10 were obtained from Dr. Ehrenstorfer (Augsburg, Germany). Stock solutions (approximately 1000 μg mL 1) were prepared in acetone. Purities of the standards were accounted for when calculating the concentration of each stock solution. A blend stock solution of all pesticides at a concentration of 10 μg mL 1 was prepared in acetone. The working standards at 0.005, 0.01, 0.02, 0.05, 0.2, 0.5 and 1.0 μg mL 1 were prepared by diluting this stock solution with acetonitrile and 0.1% ammonium acetate in methanol/0.1% formic acid in water (1:1, v/v) for the GC–MS/MS and UPLC–MS/MS analysis, respectively. Matrix-matched standards were prepared differently for GC–MS/MS and UPLC–MS/MS analysis. For the GC–MS/MS
analysis, a volume (1.5 mL) of the standard at appropriate concentration was evaporated under nitrogen and reconstituted in acetonitrile sample extract at a concentration of 0.5 g mL 1. While for the UPLC–MS/MS analysis, a volume (1 mL) of the standard at appropriate concentration and 2 mL acetonitrile sample extract at a concentration of 0.5 g mL 1 were evaporated under nitrogen then reconstituted in 1 mL methanol with 0.1% ammonium acetate/water with 0.1% formic acid (1:1, v/v). 2.3. GC–MS/MS conditions The GC–MS/MS analysis of was carried out using a Varian CP3800 gas chromatograph coupled with a Varian 1200 triple quadrupole mass spectrometer (Varian Inc., Middelburg, The Netherlands). The analyte separation was obtained on a DB-5 30 m 0.25 mm 0.5 μm capillary column, protected by a deactivated guard column (2 m 0.53 mm). Helium (purity 99.9999%) at a flow rate of 1.2 mL min 1 was used as the carrier gas. The column oven temperature programme was as follows: 80 1C (held for 3 min), programmed at 30 1C min 1to 150 1C, then programmed at 10 1C min 1 to 300 1C (held for 10 min). Large volume injection (LVI) with programmed temperature vaporization (PTV) was used. The injection volume was 10 μL of sample extract in acetonitrile using a 100 μL syringe. The injector temperature programme was as follows: 70 1C (held for 0.5 min), programmed at 200 1C min 1to 300 1C (held for 15 min). The injector split ratio was initially set at 100:1, the splitless mode was enabled at 0.5 min, at 4 min the split ratio was set at 50:1, and it was reduced to 20:1 at 10 min. The mass spectrometer was operated in electron impact ionization mode (EI, 70 eV) with the filament current of 50 μA and electron multiplier voltage at 300 V above the voltage determined by automatic tuning with perfluorotribulylamine (PFTBA). The manifold ion source and transfer line temperatures were 40, 270 and 290 1C, respectively. The collision gas (argon, 99.9998% purity) was set at the collision cell pressure of 1.7 mTorr. Multiple reaction monitoring (MRM) transitions and other acquisition parameters can be found in supplementary data included with this article. Instrument control, data acquisition and evaluation was performed by using a Varian MS Workstation software, version 6.6. 2.4. UPLC–MS/MS conditions The UPLC–MS/MS analysis was carried out using a Waters ACQUITY UPLC ultra-performance liquid chromatography system (Milford, USA) coupled with a triple quadrupole mass spectrometer (Waters Inc., Micromass, Quattro Premier XE). The nebulizer and desolvation gas was obtained from a nitrogen generator model NM30-LA (Peak Scientific, Renfrewshire, Scotland, UK). The analyte separation was achieved using a BEH C18 100 mm 2.1 mm 1.7 μm UPLC column protected by a VanGuard PreColumn 5 mm 2.1 mm 1.7 μm. The temperature of the column was thermostated at 40 1C. The column was eluted with the mobile phase: water with 0.1% formic acid (A) and methanol with 0.1% ammonium acetate (B) at the flow rate of 0.3 mL min 1 using gradient mode. Gradient was programmed to increase the amount of B from an initial content of 10–100% in 6 min, held for 1 min and returned to the initial conditions (10% B) in 1 min (from 7 to 8 min), held for 1 min. The total run time was 9 min. Sample extract volumes of 5 μL were injected into the system. The temperature of the autosampler was thermostated at 25 1C. For the MS/MS data acquisition, the interface conditions were optimized for maximum intensity of the precursor ions: nebulizer and desolvation gas (nitrogen) flows were 100 L h 1 and 700 L h 1, respectively, source block and desolvation temperatures were 120 1C
S. Walorczyk et al. / Talanta 132 (2015) 197–204
and 350 1C, respectively. Argon (99.9998% purity) was used as the collision gas at the pressure of 6.9 10 3 mbar. The optimization of multiple reaction monitoring (MRM) transitions was performed individually for each analyte on the instrument used in this work. All the compounds were analysed using positive electrospray ionization mode (ESIþ ) except for bentazone which was analysed in the negative electrospray ionization mode (ESI ). Multiple reaction monitoring (MRM) transitions and other acquisition parameters can be found in supplementary data included with this article. The instrument was controlled using Waters MassLynx 4.0 software and data evaluation was carried out using Waters TargetLynx software. 2.5. Sample preparation procedure A 5 g homogenized sample was weighted into a polypropylene centrifuge tube (50 mL), internal standards (50 μL TPP at 150 μg mL 1 and 100 μL simazine-d10 at 87 μg mL 1) were added and the sample was and extracted with 5 mL deionized water (Milli-Q) and 10 mL acetonitrile on a laboratory shaker (Heidolph Promax 2020, Schwabach, Germany) for 10 min. Then 0.5 g disodium hydrogencitrate sesquehydrate, 1 g trisodium citrate dihydrate, 4 g anhydrous magnesium sulphate, and 1 g sodium chloride were added and immediately shaken for 1 min, and then centrifuged for 2.5 min at 4500 rpm. A 6 mL aliquot of the acetonintrile supernatant was transferred to a polypropylene centrifuge tube (15 mL) containing 0.9 g anhydrous magnesium sulphate, 0.3 g PSA, and 0.3 g ChloroFiltr. The contents were vortexed for 1 min and centrifuged for 2.5 min at 4500 rpm. For the GC–MS/MS analysis, an aliquot ( 0.4 mL) of the supernatant was transferred into an autosampler vial with a fixed insert. For the UPLC–MS/MS, 2 mL of the supernatant were evaporated to dryness under nitrogen, reconstituted with 1 mL 0.1% ammonium acetate in methanol/0.1% formic acid in water (1:1, v/v), and filtered using a 13 mm, 0.2 m GHP membrane into an autosampler vial. The final concentration of sample was 0.5 g mL 1 and 1 g mL 1 in the case of GC–MS/MS and UPLC–MS/MS analysis, respectively. 2.6. Validation study The scope of the developed method comprises 25 pesticides (including 15 herbicides, nine fungicides and two herbicides). These are candidate pesticides having potential to be applied for control of the most important pests and pathogens affecting the production of the minor crops under study (i.e. field bean, lupin, sorghum, soya bean, sunflower and white mustard). Data on pesticides, matrices and conducted field experiments are detailed in Supplementary information included with this article. The developed method was subjected to validation study using green, unripe plants of lupin, white mustard and sorghum samples (previously checked to be free of the target pesticides) in order to determine linearity, matrix induced effects, recovery, precision, limit of quantification (LOQ) and measurement uncertainty. The calibration was carried out by internal standard method with reference to TPP and simazine-d10 which served as the internal standard for the GC–MS/MS and UPLC–MS/MS analytes, respectively. Linearity was studied by injecting standards at pesticide concentrations 0.005–1.0 μg mL 1 and 0.01–1.0 μg mL 1 in acetonitrile (GC–MS/MS) or 1 mL methanol with 0.1% ammonium acetate/water with 0.1% formic acid, 1:1, v/v (UPLC–MS/MS), and in matrix extracts of lupin, white mustard and sorghum. These concentration ranges corresponded to pesticide content in real samples between 0.01 and 2.0 mg kg 1 (GC–MS/MS) and 0.01 and 1.0 mg kg 1 (UPLC–MS/MS).
199
The percentage of matrix effect (%ME) was obtained for each analyte from the slopes of the calibration curves and was determined by comparing solvent and matrix-matched (lupin, white mustard and sorghum) slopes by using the formula: %ME¼100% (1slopesolvent/slopematrix) [19]. The recoveries were studied at the three spiking levels of 0.01, 0.05 and 0.5 mg kg 1 (n ¼6) on lupin and additional recovery tests at 0.05 mg kg 1 (n¼ 6) were carried out on white mustard and sorghum. Repeatability precision was expressed in terms of relative standard deviation (RSD) and calculated for each spiking level (n ¼6). The samples were spiked with the pesticides before proceeding with the extraction and the results from the recovery study were assessed for compliance with the European Union guidelines SANCO/12571/2013, according to which the average recovery should fall in the range of 70–120% with an associated RSD less or equal 20% [20]. The LOQ was defined as the lowest spiking level validated with satisfactory values of recovery (70–120%) and RSD (r20%). The measurement uncertainty was estimated based on the data obtained in the validation study. The major uncertainty sources included in the uncertainty budget were the repeatability of recoveries from spiked samples and uncertainty of the average recovery calculated from rectangular distribution. The relative expanded uncertainty was calculated by using the coverage factor k¼ 2 at the confidence level of 95% [21].
3. Results and discussion 3.1. GC–MS/MS and UPLC–MS/MS conditions MRM-based data acquisition methods with two or three MS/MS transitions for each analyte were developed on both instruments in order to carry out reliable quantification and identification of the pesticides in samples. Optimization of the conditions was started by using pesticide standards with the mass spectrometer operating in full scan mode to obtain the single stage MS spectrum. Then the fragmentation pattern of the selected precursor ions was studied in product ion scan mode at different collision energies (5–50 eV). In the case of UPLC–MS/MS, the pseudo-molecular ion [MþH] þ (and [M H] þ for bentazone) was present as the peak base of the MS spectrum for all the compounds except for desmedipham and phenmedipham. The [MþH] þ ion was used as the precursor for the MS/MS fragmentation. In the case of desmedipham and phenmedipham, due to lack of sensitivity of [Mþ H] þ ion, the ammonium adduct [MþNH4] þ was chosen as the precursor ion. For each target compound, the most selective and sensitive transitions were determined and used for the detection and quantification of the target analytes. To avoid chances of matrix interferences, two or three transitions were used except for internal standards (TPP and simazine-d10). The GC–MS/MS and UPLC–MS/MS acquisition conditions including precursor and product ions, collision energies and other parameters are detailed in Supplementary information included with this article. 3.2. Sample preparation and method validation Plant pigments including chlorophyll may contribute to high background in chromatograms and occurrence of matrix effects and thereby affect identification and quantification of the compounds of interest. The pigments can thermally degrade in the injection liner or accumulate in the columns which can lead to increased maintenance of the instrument. Thus, routine analysis of pesticide residues in chlorophyll containing matrices might be difficult and the proper choice of sorbent(s) is critical to remove
200
S. Walorczyk et al. / Talanta 132 (2015) 197–204
co-extracted matrix interferences, while assuring good recoveries of the target analytes. To prepare samples before chromatographic analysis, we employed the well-known and currently widely used quick, easy, cheap, effective, rugged and safe (QuEChERS) strategy to sample extraction and cleanup as this is a very flexible approach and offers many options for analysis depending on the range of pesticides and matrices being tested [22]. But we applied mechanical shaking for an extended time of 10 min (compared with 30 s of manual shaking in some other versions of the QuEChERS) [23]. Also, we lowered the sample amount (to 5 g) and added water (5 ml) before proceeding with the acetonitrile extraction to enable the extraction solvent penetrate better the plant tissues and ensure complete transfer of the analytes from naturally contaminated samples [24]. The QuEChERS technique inherently involves a dispersive-solid phase extraction (dispersive-SPE) step, for which the base sorbent is primary and secondary amine (PSA) which primarily removes fatty acids and other organic acids, sugars, and some pigments. The matrix constituents that might adversely affect the chromatographic analysis. In addition, a new type of sorbent, known as ChloroFiltr, was evaluated to help the dispersive-SPE performance and complement the PSA cleanup by reduction of the content of chlorophyll from green plant extracts [25,26]. Once the sample preparation method was selected, it was subjected to comprehensive validation study to prove its fitness for purpose through the assessment of analytical performance parameters including linearity, matrix effects, recovery, precision and estimation of measurement uncertainty. 3.3. Study of linearity and matrix effect Generally, the linearity parameters were highly satisfactory. The calibration curves were studied by using internal standardization in solvent solutions and matrix-matched in extracts of lupin, white mustard and sorghum, and gave the coefficients of determination (R2) of the least squared linear calibration curves 40.99. The exceptions occurred only for desmedipham, flufenacet, phenmedipham and prochloraz in the case of solvent-only calibration (GC–MS/MS), and bentazone in the case of solvent-only, lupin- and sorghum-matched calibration (UPLC–MS/MS). Detailed linearity data (equations and R2 values) obtained by using solvent-only and matrix-matched standards can be found in Supplementary information included with this article. The percentage of matrix effect (%ME) was calculated as the difference between the slope of the matrix-matched and solventonly calibration curves divided by the slope of solvent-only calibration curve. %ME in the range between 20 and 20% can be considered as insignificant because such variability is close to the repeatability RSD values [27,28]. But for most of the compounds under study, much higher values of %ME were present and both enhancement as well as suppression effects could be observed (Table 1). In the case of UPLC–MS/MS, the analyte responses produced by matrix-matched standards were nearly always substantially higher than the responses obtained for the equivalent standards in solvent, so the matrix enhancement effect occurred predominantly. Otherwise, in the case of GC–MS/MS analysis, the matrix effects occurred in a larger extent ( 116 to 82%) but no consistent pattern with respect to different analytes could be observed, i.e. enhancement, suppression, or negligible matrix effects o20% (absolute value) occurred for different compounds. The prevailing occurrence of enhancement effect in the case of UPLC–MS/MS technique was contrary to our expectation, because the ion suppression effect tend to be more common in positive electrospray mode (ESI þ) [29]. A possible explanation for the phenomenon we observed might be a specific suppression of the
Table 1 Matrix effects of the pesticides in the GC–MS/MS and UPLC–MS/MS analysis of lupin, white mustard and sorghum extracts, determined as 100% (1 slopesolvent/ slopematrix). Pesticides
Acetochlor Azoxystrobin Bentazone Boscalid Clomazone Cyproconazole_1 Cyproconazole_2 Deltamethrin 1 Deltamethrin 2 Desmedipham Dimethenamid-P Dimoxystrobin Fluazifop-P-butyl Flufenacet Linuron Metamitron Metconazole Metolachlor-S Pendimethalin Phenmedipham Picoxystrobin Prochloraz Propiconazole 1 Propiconazole 2 Prosulfocarb Quizalofop-P-ethyl Tebuconazole Thiacloprid
Lupine
White mustard
Sorghum
GC
UPLC
GC
UPLC
GC
UPLC
76 24 – 2 88 29 – 68 20 75 70 19 22 44 – – 31 45 39 79 26 11 40 45 – 4 23 –
43 42 62 46 38 38 43 – – 39 48 39 1 38 38 52 39 42 91 37 30 27 41 – 23 8 40 46
88 12 – 4 44 19 – 70 30 77 62 15 16 48 – – 25 48 39 82 20 116 25 26 – 4 23 –
53 51 57 53 48 43 51 – – 47 55 49 38 49 50 59 47 52 5 46 46 41 53 – 45 31 49 51
72 20 – 5 59 19 – 56 9 73 70 15 16 37 – – 28 48 33 77 20 2 30 36 – 4 14 –
43 43 64 43 32 37 40 – – 38 46 39 20 41 38 50 35 61 50 34 35 28 41 – 28 8 38 40
internal standard (simazine-d10) occurring due to co-eluting matrix compounds which resulted in analytes overestimation. However, this was not a major concern due to consistency of this effect among the matrices which led to sufficiently accurate quantification, provided that matrix-matched calibration would be used. By using either techniques (GC–MS/MS or UPLC–MS/MS), the matrix effects were relatively consistent among studied matrices for the majority of the pesticides, so application of matrixmatched standards in any matrix would give acceptably accurate results for quantification of pesticide residues in other matrices, as will be presented further in this section. The exceptions, in the case of GC–MS/MS technique, were flufenacet and prochloraz. Flufenacet exhibited positive effect in lupin and sorghum but negative in white mustard whereas prochloraz exhibited considerable stronger negative matrix effect in white mustard ( 116%) than in lupin ( 11%) and sorghum ( 2%). In the case of UPLC–MS/ MS technique, inconsistencies in the occurrence of matrix effects were observed for fluazifop-P-butyl, quizalofop-P-ethyl and pendimethalin. In the case of fluazifop-P-butyl and quizalofop-P-ethyl, stronger positive matrix effects was observed in white mustard than in lupin and sorghum. In contrast, in the case of pendimethalin, stronger negative matrix effects were observed in lupin ( 91%) and sorghum ( 50%) than in white mustard ( 5%). Whereas negligible matrix effects, with absolute values o20% in all matrices, were only observed in the case of boscalid, dimoxystrobin and quizalofop-P-ethyl when using the GC–MS/MS technique. Because of common occurrence of matrix effects in analysis of pesticide residues in chlorophyll containing matrices under this study, the use of matrix-matched calibration was necessary as a practical approach to compensate for matrix effects and achieve
S. Walorczyk et al. / Talanta 132 (2015) 197–204
201
Table 2 Average recoveries, RSDs and expanded uncertainties U, % (k ¼ 2, confidence level 95%) in the GC–MS/MS and UPLC-MS/MS analysis of pesticides in green lupin. Pesticides
Recovery, %
U, %
0.01
Acetochlor Azoxystrobin Bentazone Boscalid Clomazone Cyproconazole_1 Cyproconazole_2 Deltamethrin 1 Deltamethrin 2 Desmedipham Dimethenamid-P Dimoxystrobin Fluazifop-P-butyl Flufenacet Linuron Metamitron Metconazole Metolachlor-S Pendimethalin Phenmedipham Picoxystrobin Prochloraz Propiconazole 1 Propiconazole 2 Prosulfocarb Quizalofop-P-ethyl Tebuconazole Thiacloprid
0.05
0.5
(k ¼ 2, α¼ 0.05)
Overall
GC
UPLC
GC
UPLC
GC
UPLC
GC
UPLC
GC
UPLC
122 (2) 107 (8) – 101 (4) 111 (6) 91 (8) – 106 (13) 82 (6) – 81 (10) 113 (3) 94 (4) 94 (18) – – – 109 (7) 119 (3) 119 (3) 101 (19) – 106 (5) 109 (8) – 86 (5) 90 (5) –
97 (18) 129 (7) – 102 (5) 128 (9) – 114 (7) – – 92 (8) – 116 (8) 107 (14) 123 (6) 112 (10) 130 (7) 105 (12) 126 (9) 75 (22) 100 (7) 110 (7) 103 (4) 112 (5) – 100 (8) 87 (20) 134 (8) 124 (10)
120 (3) 99 (8) – 84 (1) 122 (3) 91 (1) – 105 (8) 87 (5) 78 (4) 122 (2) 97 (5) 103 (4) 84 (4) – – 92 (10) 118 (1) 117 (4) 79 (6) 101 (4) 67 (6) 110 (6) 105 (13) – 94 (3) 90 (1) –
79 (7) 85 (3) 89 (9) 78 (3) 84 (2) 98 (7) 72 (1) – – 68 (5) 94 (11) 86 (2) 84 (3) 85 (2) 77 (4) 82 (4) 75 (4) 87 (2) 62 (13) 70 (4) 89 (4) 74 (7) 79 (3) – 76 (2) 79 (4) 85 (5) 86 (4)
119 (1) 100 (1) – 84 (1) 120 (1) 96 (3) – 96 (4) 81 (2) 81 (3) 120 (1) 103 (1) 102 (2) 85 (4) – – 96 (4) 115 (2) 117 (1) 80 (3) 101 (1) 69 (2) 94 (1) 92 (1) – 93 (1) 95 (1) –
98 (1) 94 (3) 60 (5) 91 (3) 94 (2) 109 (1) 94 (1) – – 84 (5) 92 (4) 98 (2) 92 (4) 98 (1) 89 (1) 93 (2) 89 (4) 96 (1) 79 (13) 84 (4) 99 (2) 86 (9) 92 (4) – 90 (5) 85 (6) 94 (3) 96 (3)
120 (2) 102 (7) – 90 (10) 118 (6) 92 (5) – 102 (9) 83 (5) 80 (4) 108 (19) 104 (7) 100 (5) 88 (11) – – 94 (7) 114 (5) 118 (3) 83 (9) 101 (10) 68 (4) 104 (8) 102 (11) – 91 (5) 92 (4) –
91 (9) 103 (4) 75 (9) 90 (4) 102 (4) 104 (7) 93 (3) – – 81 (6) 93 (8) 100 (4) 94 (7) 102 (3) 93 (5) 102 (4) 90 (7) 103 (4) 72 (16) 85 (5) 99 (4) 88 (7) 94 (4) – 89 (12) 84 (10) 104 (5) 102 (6)
13 14 – 20 15 11 – 19 14 14 39 15 10 24 – – 14 13 11 21 20 20 16 21 – 11 9 –
18 9 23 9 9 15 7 – – 16 16 8 14 6 11 9 15 8 36 16 9 15 9 – 12 22 11 11
acceptably accurate quantificative results in both GC–MS/MS and UPLC–MS/MS analyses. 3.4. Recovery study and estimation of measurement uncertainty For the purpose of evaluating the accuracy (trueness and precision) of the proposed method, a recovery study was carried out by spiking lupin, white mustard and sorghum samples with the target pesticides. The recovery and repeatability RSD values were determined by the analysis of six replicate spiked samples at each level (0.01, 0.05 and 0.5 mg kg 1). The results for calculation of recoveries were obtained by using matrix-matched standards prepared in lupin extracts. As can be seen in Table 2, most of the compounds presented satisfactory recoveries from lupin in the range between 70% and 120% with RSD values r20%, i.e. compliant with the EU criteria of SANCO/12571/2013 [20]. For a limited number of compounds, especially at the lowest spiking level of 0.01 mg kg 1, the recovery and RSD results were slightly outside the acceptance criteria. The most problematic compound was bentazone which generally did not present satisfactory results. But the overall recoveries of the target pesticides at the three spiking levels of 0.01, 0.05 and 0.5 mg kg 1 fell in the range between 68 and 120% (98% on average) and 72–104% (93% on average) with RSD values between 2 and 19% (7% on average) and 3–16% (6% on average) by GC–MS/ MS and UPLC–MS/MS techniques, respectively. Additional recovery evaluation was carried out on white mustard and sorghum (spiking level 0.05 mg kg 1, n¼6). In this experiment, the results for calculation of recoveries were also obtained by using matrix-matched standards prepared in lupin extracts. Generally, the results were very satisfactory and fulfilled the recovery and precision requirements of SANCO/12571/2013 (Table 3). Flufenacet and
Table 3 Average recoveries and RSDs in the GC–MS/MS and UPLC–MS/MS analysis of pesticides in white mustard and sorghum (spiking level 0.05 mg kg 1, n¼ 6). Pesticides
Recovery (%) White mustard
Acetochlor Azoxystrobin Bentazone Boscalid Clomazone Cyproconazole_1 Cyproconazole_2 Deltamethrin 1 Deltamethrin 2 Desmedipham Dimethenamid-P Dimoxystrobin Fluazifop-P-butyl Flufenacet Linuron Metamitron Metconazole Metolachlor-S Pendimethalin Phenmedipham Picoxystrobin Prochloraz Propiconazole 1 Propiconazole 2 Prosulfocarb Quizalofop-P-ethyl Tebuconazole Thiacloprid
Sorghum
GC
UPLC
GC
UPLC
70 (14) 78 (6) – 91 (7) 93 (9) 101 (10) – 97 (17) 87 (4) 106 (9) 86 (7) 109 (6) 96 (2) 22 (23) – – 110 (10) 79 (11) 100 (11) 99 (17) 90 (9) 45 (33) 102 (12) 95 (13) – 96 (5) 96 (3) –
93 (6) 85 (4) 25 (24) 77 (4) 91 (1) 96 (6) 75 (4) – – 78 (3) 91 (10) 88 (2) 91 (3) 84 (2) 106 (2) 105 (1) 79 (4) 92 (4) 87 (7) 79 (3) 100 (2) 89 (2) 84 (3) – 89 (1) 94 (3) 75 (1) 94 (2)
102 (11) 87 (7) – 100 (10) 107 (10) 105 (13) – – – 76 (8) 109 (6) 103 (14) 99 (11) 113 (8) – – 100 (20) 99 (6) 98 (18) 88 (12) 102 (12) 97 (5) 108 (10) 96 (17) – 101 (5) 104 (13) –
91 (3) 88 (3) 56 (17) 79 (5) 88 (2) 103 (4) 79 (6) – – 79 (2) 99 (11) 91 (3) 89 (3) 88 (4) 90 (4) 107 (2) 76 (3) 94 (2) 74 (7) 78 (1) 98 (1) 90 (5) 81 (2) – 84 (3) 92 (1) 74 (4) 95 (1)
202
S. Walorczyk et al. / Talanta 132 (2015) 197–204
prochloraz were the exceptions, which presented lower recoveries in white mustard in the case of GC–MS/MS analysis and so was bentazone which presented lower recoveries in white mustard and sorghum in the case of UPLC–MS/MS analysis. These phemomena were likely caused by the occurrence of strong matrix effects, as it was discussed previously in this section. Matrix-matching in the same matrix will be needed to obtain more accurate quantification for these problematic compounds. The limit of quantification (LOQ) was determined as the lowest spiking level at which the validation criteria were satisfied with average recovery 70–120% and RSD r20%, and it was 0.01 mg kg 1 (except for bentazone and metamitron). The data derived from the validation study (recovery and RSD) were used to estimate the measurement uncertainty associated with the analytical results when using the developed method to analyse real samples. Precision was identified as the main contribution to the uncertainty, but the uncertainty associated with the recovery, as calculated from rectangular distribution, was also included in the uncertainty budget to avoid underestimation of the total uncertainty [21]. Taking into account the overall recovery and RSD data (Table 2), the calculated measurement uncertainty values ranged from 9% to 39% (17% on average) and from 6% to 36% (13% on average), by using GC–MS/MS and UPLC–MS/MS techniques, respectively (coverage factor k ¼2, confidence level 95%). These values were lower than a default value of 750% recommended by the SANCO/12571/2013 document, demonstrating fitness for purpose of the developed and validated method.
Fig. 1. Dissipation patterns of pesticides after application on lupin.
It must be highlighted that most of the pesticides under this study could be analysed by either techniques, GC–MS/MS or UPLC– MS/MS, therefore the one to offer better analytical performance parameters for the particular pesticide and matrix combination was used when analysing real samples. 3.5. Application to real samples In this work, we focused on the study of dissipation rates of selected pesticides after application to lupin, white mustard, soya bean, sunflower and field bean. These plants were classified by the Polish Ministry of Agriculture and Rural Development as minor crops, for which there are limited options for protection against pests and pathogens. Our studies entailed the use of candidate pesticides having potential to be applied for the protection of these minor crops in field trials, followed by laboratory analysis for pesticide residues. Detailed data on tests that we have been conducted in experimental plots can be found in Supplementary information included with this article. The aim of our work was to increase the knowledge in the residue behaviour and dissipation study of the pesticides, which have recently been proposed for the use in comprehensive protection of selected minor crops by means of extension of authorization of pesticides which are already registered for use in Poland on other crops (i.e. major crops). The dissipation patterns of the pesticides were different probably due to the different chemical structures of the compounds, applied doses and preharvest times but the terminal pesticide residues, i.e. determined in the samples collected in the last sampling time, were always below the maximum residue levels (MRLs) [30]. Pesticide MRLs apply to edible plant parts but in this work unripe green plants were analysed in order to determine the dissipation patterns of the target pesticides and reassure preharvest intervals on the crops under study. Effectively, there was no need to analyse the final crops (seeds) because the target pesticides disappeared on the plants before harvest, and in most cases they fell below the method LOQ in the samples collected in the last sampling time. We can conclude that the proposed pesticide application schedules were correct and the data concerning the pesticide residue dissipation suggest that the proposed pesticides can be safely used on the studied minor crops, without the risk of the presence of problematic residues, provided that good agricultural practices have been followed. Examples of pesticide dissipation
Fig. 2. GC–MS/MS chromatogrrasams of a sample of lupin containing residues of (A) dimethenamid-P (0.016 mg kg 1), and (B) pendimethalin (0.028 mg kg 1). The sample was collected 32 days after spraying with Wing P 462.5 EC.
S. Walorczyk et al. / Talanta 132 (2015) 197–204
203
Fig. 3. UPLC–MS/MS chromatograms of (A) a sample of field bean containing thiacloprid (0.057 mg kg 1, 21 days after spraying with Proteus 110 OD), and (B) a sample of sunflower containing dimoxystrobin (0.236 mg kg 1, 14 days after spraying with Pictor 400 SC).
over time plots, showing concentration versus days after the last spraying, for different pesticides applied to lupin in a field experiment are shown in Fig. 1. Whereas, Fig. 2 shows GC–MS/MS chromatograms of a sample of lupin containing residues of dimethenamid-P (0.016 mg kg 1) and pendimethalin (0.028 mg kg 1) and Fig. 3 shows UPLC–MS/MS chromatograms of thiacloprid in field bean (0.057 mg kg 1) and dimoxystrobin in sunflower (0.236 mg kg 1). The results obtained in this work will aid in the establishment of safe and proper use of pesticides for extension of authorization on minor crops in Poland.
plants with the consideration of food safety, reduction of yield losses and threat to humans, farm animals and the environment” (contract No. HORkor.0660/IOR 2011-2015/2/2013), task No. 1.2. “Assessment of possible complex protection of minor crops”, and by Ministry of Science and Higher Education (Ministerstwo Nauki i Szkolnictwa Wyższego), project ID: POZ-07.
4. Conclusions
References
Matrices high in chlorophyll present a special difficulty in pesticide residue analysis owing to interferences which can negatively affect the method performance characteristics and create an analytical challenge. As a solution to these problems, we investigated the use of a new QuEChERS-based method of sample preparation and report it for the first time. We used a new sorbent, known as ChloroFiltr, which served for selective reduction of chlorophyll from green plant extracts, and was used in combination with PSA to keep the sample cleanup as straightforward as possible in the dispersive-SPE mode. The final analysis was carried out by using concurrent determinations by both GC–MS/MS and UPLC–MS/MS, and in the case of both techniques, the importance of using matrix-matched calibration was demonstrated as being of key importance for achieving accurate results. An important conclusion is also that relative consistency of matrix effects in both GC–MS/MS and UPLC–MS/MS analysis for a given pesticide between different matrices, makes it possible to quantify pesticide residues in different matrices using only one matrix-matched calibration curve. After validation, the developed method was successfully applied to study the dissipation patterns of pesticide residues after application on crops belonging the group of minor crops (lupin, white mustard, soya bean, sunflower and field bean).
[1] Regulation (EC) no. 1107/2009 of the European Parliament and of the Council of 21 October 2009 concerning the placing of plant protection products on the market and repealing Council Directives 79/117/EEC and 91/414/EEC. Available online: 〈http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:309:0001:0050: EN:PDF〉, (accessed 10.06.14). [2] Available online: 〈http://www.eppo.int/PPPRODUCTS/minor_uses/minor_uses. htm〉, (accessed 10.06.14). [3] Available online: 〈http://isap.sejm.gov.pl/DetailsServlet?id=WDU20130000659〉, (accessed 10.06.2014). [4] H.V. Botitsi, S.D. Garbis, A. Economou, D.F. Tsipi, Mass Spectrom. Rev. 30 (2011) 907–939. [5] European Food Safety Authority, The 2010 European Union report on pesticide residues in food, EFSA J. 11 (2013) 808 pp. Available online: 〈http://www.efsa. europa.eu/en/efsajournal/doc/3130.pdf〉, (accessed 10.06.14). [6] Y. Wang, H.Y. Jin, S.-C. Ma, J. Lu, R.-C. Lin, J. Chromatogr. A 1218 (2011) 334–342. [7] E. Muñoz, G. Muñoz, L. Pineda, E. Serrahima, F. Centrich, J. AOAC Int. 95 (2012) 1777–1796. [8] B. Łozowicka, M. Jankowska, E. Rutkowska, I. Hrynko, P. Kaczyński, J. Miciński, J. Nat. Med. 68 (2014) 95–111. [9] R.M. González-Rodríguez, R. Rial-Otero, B. Cancho-Grande, J. Simal-Gándara, J. Chromatogr. A 1196–1197 (2008) 100–109. [10] L. Li, W. Li, D. Qin, S. Jiang, F. Liu, J. AOAC Int. 92 (2009) 538–547. [11] L.-J. Qu, H. Zhang, J.-H. Zhu, G.-S. Young, H.Y. Aboul-Enein, Food Chem. 122 (2010) 327–332. [12] R. Xu, J. Wu, Y. Liu, R. Zhao, B. Chen, M. Yang, J. Chen, Chemosphere 84 (2011) 908–912. [13] M.Á. González-Curbelo, J. Harnández-Borges, L.M. Ravelo-Pérez, M.Á. Rodríguez-Delgado, Food Chem. 125 (2011) 1083–1090. [14] G. Chen, R. Cao, R. Liu, Food Chem. 125 (2011) 1406–1411. [15] M. Słowik-Borowiec, E. Szpyrka, S. Walorczyk, Bull. Environ. Contam. Toxicol. 89 (2012) 633–637. [16] H.G.J. Mol, A. Rooseboom, R. van Dam, M. Roding, K. Arondeus, S. Sunarto, Anal. Bioanal. Chem. 389 (2007) 1715–1754. [17] C.C. Leandro, P. Hancock, R.J. Fussell, B.J. Keely, J. Chromatogr. A 1166 (2007) 152–162. [18] S. Walorczyk, Rapid Commun. Mass Spectrom. 22 (2008) 3791–3801. [19] H. Kwon, S.J. Lehotay, L. Geis-Asteggiante, J. Chromatogr. A 1270 (2012) 235–245.
Acknowledgements This work was supported by Ministry of Agriculture and Rural Development (Ministerstwo Rolnictwa i Rozwoju Wsi) under LongTerm Programme of IPP–NRI for 2011–2015 “Protection of cultivated
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.talanta.2014.08.073.
204
S. Walorczyk et al. / Talanta 132 (2015) 197–204
[20] Document no. SANCO/12571/2013 (2014), Guidance document on analytical quality control and validation procedures for pesticide residues analysis in food and feed. Available online: 〈http://ec.europa.eu/food/plant/plant_protection_products/guidan ce_documents/docs/qualcontrol_en.pdf〉, (accessed 10.06.14). [21] S. Walorczyk, D. Drożdżyński, J. AOAC Int. 94 (2011) 1625–1642. [22] S.J. Lehotay, Methods Mol. Biol. 747 (2011) 65–91. [23] S.J. Lehotay, K. Ae Son, H. Kwon, U. Koesukwiwat, W. Fu, K. Mastovska, E. Hoh, N. Leepipatpoboon, J. Chromatogr. A 1217 (2010) 2548–2560. [24] O. Lacina, M. Zachariasova, J. Urbanova, M. Vaclavikova, T. Cajka, J. Hajslova, J. Chromatogr. A 1262 (2012) 8–18. [25] Available online: 〈https://www.unitedchem.com/catalog/sample-prep/solid-pha se-extraction-columns/enviro-clean/quechers-chlorofiltr〉, (accessed 19.07.14).
[26] 〈https://www.unitedchem.com/sites/default/files/docs/posters-and-papers/ SOFT_MarijuanaPoster_002_Lo-Res.pdf〉, (accessed 19.07.14). [27] S. Walorczyk, Talanta 120 (2014) 106–113. [28] C. Ferrer, A. Lozano, A. Agüera, A.J. Girón, A.R. Fernández-Alba, J. Chromatogr. A 1218 (2011) 7634–7639. [29] H. Stahnke, S. Kittlaus, G. Kempe, L. Alder, Anal. Chem. 84 (2012) 1474–1482. [30] Regulation (EC) no. 396/2005 of the European Parliament and of the Council of 23 February 2005 on maximum residue levels of pesticides in or on food and feed of plant and animal origin and amending Council Directive 91/414/EEC. Available online: 〈http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ: L:2005:070:0001:0016:en:PDF〉, (accessed 17.06.14).
