Food Chemistry 181 (2015) 64–71
Contents lists available at ScienceDirect
Food Chemistry journal homepage: www.elsevier.com/locate/foodchem
Analytical Methods
Determination of pesticides in lettuce using solid–liquid extraction with low temperature partitioning Anna I.G. Costa a, Maria E.L.R. Queiroz a,⇑, Antônio A. Neves a, Flaviane A. de Sousa a, Laércio Zambolim b a b
Department of Chemistry, Universidade Federal de Viçosa, 36570-000 Viçosa, MG, Brazil Department of Phytopathology, Universidade Federal de Viçosa, 36570-000 Viçosa, MG, Brazil
a r t i c l e
i n f o
Article history: Received 29 July 2014 Received in revised form 30 January 2015 Accepted 14 February 2015 Available online 20 February 2015 Chemical compounds studied in this article: Chlorothalonil (PubChem CID: 15910) Methyl parathion (PubChem CID: 4130) Procymidone (PubChem CID: 36242) Endosulfan (PubChem CID: 3224) Iprodione (PubChem CID: 37517) k-Cyhalothrin (PubChem CID: 443046) Permethrin (PubChem CID: 40326) Cypermethrin (PubChem CID: 2912) Deltamethrin (PubChem CID: 40585)
a b s t r a c t This work describes the optimization and validation of a method employing solid–liquid extraction with low temperature partitioning (SLE/LTP) together with analysis by gas chromatography with electron capture detection (GC/ECD) for the determination of nine pesticides (chlorothalonil, methyl parathion, procymidone, endosulfan, iprodione, k-cyhalothrin, permethrin, cypermethrin, and deltamethrin) in lettuce. The method was found to be selective, accurate, and precise, with means recovery values in the range of 72.3–103.2%, coefficients of variation 612%, and detection limits in the range 0.4–37 lg kg1. The matrix components significantly influence the chromatographic response of the analytes (above 10%). The optimized and validated method was applied to determine the residual concentrations of the fungicides iprodione and procymidone that had been applied to field crops of lettuce. The maximum residual concentrations of the pesticides in the lettuce samples were 13.6 ± 0.4 mg kg1 (iprodione) and 1.00 ± 0.01 mg kg1 (procymidone), on the day after application of the products. Ó 2015 Published by Elsevier Ltd.
Keywords: Gas chromatography Pesticides Lettuce
1. Introduction Lettuce (Lactuca sativa L.), a low calorie salad vegetable belonging to the Asteraceae family (subfamily Cichoriaceae), is known worldwide and is a source of the vitamins A, B1, B2, and C, as well as essential elements such as calcium and iron (Altunkaya, Becker, Gökmen, & Skibsted, 2009; Rissato, Galhianea, Souzab, & Aponc, 2005). A limiting factor in the production of this vegetable is its susceptibility to diseases and pests that directly affect the marketable part of the plant and need to be carefully controlled. Chemical control has been most widely used, with the application of authorized pesticides including chlorothalonil (isophthalonitrile) and the dicarboximides procymidone and iprodione (ANVISA, 2014). Despite the benefits that can be obtained from the use of pesticides, their improper application can result in residual ⇑ Corresponding author. Tel.: +55 31 3899 1430. E-mail address: [email protected] (M.E.L.R. Queiroz). http://dx.doi.org/10.1016/j.foodchem.2015.02.070 0308-8146/Ó 2015 Published by Elsevier Ltd.
concentrations that exceed the maximum permissible limits of between 1 and 6 mg kg1 described in Brazilian legislation (ANVISA, 2014). In the European Union, the maximum permissible concentrations of authorized compounds are between 0.01 and 10 mg kg1 (European Union, 2013). There is also concern regarding the indiscriminate application of unauthorized active agents to crops. In reports issued annually by the Brazilian Program for Analysis of Pesticide Residues (PAPR), the most unsatisfactory results for lettuce crops have involved the presence of unauthorized chemicals (ANVISA, 2013). This situation demands the development and use of multiresidue methods of analysis that are inexpensive, reliable, and provide low limits of quantification in order to monitor the levels of contamination of lettuce by pesticides. The commonest methods used to extract pesticides from samples of lettuce involve an initial extraction into solvent followed by liquid–liquid partitioning, where the analytes are transferred to the organic phase, with undesirable co-extractives and some highly polar pesticides remaining in the aqueous phase (Fenoll, Hellin, Lopez, Gonzalez,
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
& Flores, 2007a, 2007b; Gelsomino, Petrovicová, Tiburtini, Magnani, & Felici, 1997; Oviedo, Toledo, & Vicente, 2003). These methods generally require large amounts of sample, sorbent and high quality organic solvents, as well as manipulation of the extracts. This has led to the development of faster and more efficient methods that are both inexpensive and environmentally friendly (Kristenson, Brinkman, & Ramos, 2006). In recent years, techniques for the analysis of pesticide residues in fruits and vegetables have evolved significantly in terms of simplification and improvements in sample extraction and purification. In particular, the technique of solid–liquid extraction with low temperature partitioning (SLE/LTP) has shown good results for the extraction of pesticides present in solid matrices such as potatoes (Dardengo et al., 2011), tomatoes (Pinho, Neves, Queiroz, & Silvério, 2010), butter (Marthe, Bittencourt, Queiroz, & Neves, 2010), and strawberries (Heleno, Queiroz, Neves, & Oliveira, 2014). The technique is based on the partition of analytes between the aqueous and organic phase resulted from temperature lowering (20 °C). The advantage of this method is that the sample components are frozen with the aqueous phase, whereas pesticides are extracted by the organic, without the need to clean-up one step. Other extraction techniques used for the analysis of pesticides include supercritical fluid extraction (SFE) (Lehotay, 1997; Rissato et al., 2005), microwave-assisted extraction (MAE) (Eskilsson & Bjorklund, 2000), pressurized liquid extraction (PLE) (Ramosa, Kristensonb, & Brinkmanb, 2002), solid phase extraction (SPE) (Balinova, Mladenova, & Shtereva, 2007; Faria, Maldaner, Santana, Jardim, & Collins, 2007), dispersion matrix solid phase extraction (DMSP) (Silva, Silva, & Navickiene, 2010), solid phase microextraction (SPME) (Hu et al., 2008), stir bar sorptive extraction (SBSE) (Barriada-Pereira, Serôdioc, González-Castro, & Nogueirac, 2010) and QuEChERS (quick, easy, cheap, effective, rugged, and safe), technique adopted in 2007 as official by the Association of Official Analytical Chemists (AOAC) that generates extracts that can be analyzed by liquid chromatography and/or gas chromatography coupled to mass spectrometry in series (GC– MS/MS and LC–MS/MS) (Koesukwiwata, Lehotay, Miao, & Leepipatpiboon, 2010). However, these techniques are more expensive, compared to SLE/LTP, and often require highly skilled technicians. The purpose of this study was to optimize and validate a methodology employing SLE/LTP with GC/ECD analysis for the simultaneous determination of the fungicides chlorothalonil, procymidone, and iprodione (compounds authorized by Brazilian legislation), and the insecticides methyl parathion, endosulfan, kcyhalothrin, permethrin, cypermethrin, and deltamethrin (compounds unauthorized by Brazilian legislation) (Table 1) in lettuce samples. The optimized and validated method was employed to determine the residual concentrations of pesticides in samples of lettuce to which authorized pesticides had been applied.
2. Experimental 2.1. Reagents and solutions Standards of chlorothalonil (99.3%), procymidone (99.9%), iprodione (99.3%), deltamethrin (99.7%), and permethrin (92.2%) were acquired from Sigma–Aldrich (Seelze, Germany). Methyl parathion (99%), endosulfan (73.2%), and cypermethrin (92.4%) were obtained from Chem Service (West Chester, PA, USA). k-Cyhalothrin (86.5%) was obtained from Syngenta (São Paulo, Brazil). Bifenthrin (92.2%), used as an internal standard, was acquired from FMC (Brazil). The HPLC-grade acetonitrile and ethyl acetate (99.5%) solvents used in the extractions were purchased from Vetec (Rio de
65
Janeiro, Brazil). Anhydrous sodium sulfate (99%) used to remove water from the extracts was also obtained from Vetec. Stock 1000 mg L1 standard solutions of the pesticides were individually prepared in acetonitrile. Working solutions containing the nine pesticides at concentrations of 10 mg L1 (chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin) and 20 mg L1 (iprodione, permethrin, and deltamethrin) were prepared by diluting the stock standard solutions with the same solvent. A solution containing the internal standard (bifenthrin) at a concentration of 10 mg L1 was prepared in a similar way as the other standards. All solutions were stored in a freezer at a temperature of approximately 20 °C. 2.2. Equipment The analyses employed a gas chromatograph (Model GC-2014, Shimadzu) equipped with an electron capture detector (ECD) and an AOC-20i auto-injector. An ultrasonic bath (Unique, São Paulo, Brazil) operated at 150 W and 33 kHz was used during optimization of the extraction method. Other equipment included an analytical balance (Model BP 2215, Sartorius, Göttingen, Germany), a table shaker (Model TE 420, Tecnal, São Paulo, Brazil), a vortex mixer, a centrifuge (Excelsa II Model 206 MP, Fanem, São Paulo, Brazil), and a pH meter (Model DM 21, Digimed, São Paulo, Brazil). 2.3. Chromatographic conditions The chromatographic separation of the analytes was performed on an Agilent DB-5 capillary column (5% phenyl/95% dimethyl siloxane stationary phase, 30 m length, 0.25 mm internal diameter, and 0.1 lm film thickness). Nitrogen was used as the carrier gas at a flow rate of 1.2 mL min1. The column temperature program consisted of an initial hold at 150 °C for 2 min, followed by an increase at 40 °C min1 to 210 °C, a hold for 2 min, then an increase at 20 °C min1 to 240 °C, a hold for 5 min, and finally an increase at 15 °C min1 to 290 °C, with a hold for 5 min. The total analysis time was 20.3 min. The injector and detector temperatures were maintained at 280 and 300 °C, respectively. A 1 lL volume of sample was injected into the chromatograph, using a split ratio of 1:5. The pesticides were identified by comparing the retention times of the peaks obtained for the sample extracts with those of the standards. Quantification was achieved using the matrix superposition method, with fortification of pesticide-free lettuce samples using nine pesticide concentrations in the range 1.2–1340 lg kg1. Calibration graphs were constructed by plotting peak areas against concentrations. 2.4. Sample preparation for optimization studies Optimization and validation of the method employing solid–liquid extraction with low temperature partitioning (SLE/LTP) and analysis by gas chromatography with electron capture detection (GC/ECD) employed organic butterhead lettuces grown in the laboratory at the Plant Diseases Clinic of the Department of Phytopathology, Federal University of Viçosa. The lettuces were crushed in a mixer (Walita, São Paulo, Brazil) until homogenization was complete. A 4 g portion of the pulp was then placed in a 22 mL clear glass flask and fortified with 80 lL of a working standard solution containing chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin at concentrations of 10 mg L1 (giving concentrations of 0.2 lg g1 in the fortified sample), together with 80 lL of a working standard solution containing iprodione, permethrin, and deltamethrin at concentrations of 20 mg L1 (giving concentrations of 0.4 lg g1 in the fortified sample). The spiked samples were allowed to stand for 3 h in order to evaporate the solvent and promote greater
66
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
Table 1 IUPAC names and chemical formula of the pesticides studied. Compound
Formula
IUPAC name
Bifenthrin
2-Methyl-3-phenylbenzyl(1RS)-cis-3-(2-chloro-3,3,3-trifluoroprop1-enyl)-2,2-dimethylcyclopropanecarboxylate H3C
O CF
CH3
3
CH3
Cl
O
CH3
O
F3C
Cl
CH3
CH3
O
(Z)-(1S)-cis-
(Z)-(1R)-cis-
CN
Chlorothalonil
Tetrachloroisophthalonitrile
Cl
Cl
Cl
CN Cl
Methyl parathion
S OP(OCH3 )2
O2 N
Cl
Procymidone
O
O,O-dimethyl O-4-nitrophenyl phosphorothioate
N-(3,5-dichlorophenyl)-1,2-dimethylcyclopropane-1,2dicarboximide
N O
Cl Endosulfan
Cl
Cl
(1,4,5,6,7,7-Hexachloro-8,9,10-trinorborn-5-en-2,3-ylenebis methylene) sulfite
Cl
Cl Cl Cl Cl
Iprodione
O
O
3-(3,5-Dichlorophenyl)-N-isopropyl-2,4-dioxo imidazolidine-1carboxamide
N
N
O
Cl H
k-Cyhalothrin
O SO
CONHCH(CH3 )2
H
Cl
CO 2
F3 C
H
C
CN O
(S)-a-cyano-3-phenoxybenzyl (Z)-(1R,3R)-3-(2-chloro-3,3,3-trifluoro prop-1-enyl)-2,2-dimethylcyclopropanecarboxylate and (R)-acyano-3-phenoxybenzyl(Z)-(1S,3S)-3-(2-chloro-3,3,3-trifluoroprop1-enyl)-2,2 dimethylcyclopropane carboxylate
(R) (Z)-(1S)-CIS
Permethrin
Cl
C CH
Cl Cypermethrin
Cl Cl
Deltamethrin
C CH
Br Br
C CH
O C
O
O
CN CO 2CH
O C
O H
O
3-Phenoxybenzyl (1RS,3RS;1RS,3SR)-3-(2,2-dichlorovinyl)-2,2dimethylcyclo propanecarboxylate
3-Phenoxybenzyl (1RS,3RS;1RS,3SR)-3-(2,2-dichlorovinyl)-2,2 dimethylcyclo propanecarboxylate
(S)-a-cyano-3-phenoxybenzyl (1R,3R)-3-(2,2-dibromovinyl)-2,2 dimethylcyclopropanecarboxylate
CN O
interaction of the pesticides with the samples, prior to extraction. These samples were then used in optimization of the SLE/PLT technique. 2.5. Optimization procedure The parameters evaluated in optimization of the SLE/PLT technique applied to the lettuce samples were: addition of water (0, 0.5, 1, 2, and 4 mL); type of extraction solvent (8 mL of acetonitrile, or a mixture of 6.5 mL of acetonitrile and 1.5 mL of ethyl acetate); mode and time of agitation (mechanical agitation for 10 and 20 min, ultrasonication for 1 and 10 min, and vortex mixing for 0.5 and 1 min); addition of salt (the sample was added to a 2%
(w/v) solution of sodium chloride, instead of water); freezing duration (3, 6, 12, and 24 h); and centrifugation duration (1, 3, and 5 min at 1200g). The best conditions were identified using the average percentage recoveries of the pesticides after extraction assays performed in triplicate. 2.6. Optimized procedure for sample The SLE/LTP procedure consists of extracting 4 g of lettuce with the addition of 1 mL of water and 8 mL of the combined acetonitrile/ethyl acetate solvent (6.5:1.5, v/v). This mixture is stirred for 10 min on a shaker table at 25 °C and 200 rpm, centrifuged for
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
3 min at 1200g, and then allowed to freeze in the freezer at approximately 20 °C for 3 h. After this time, the extract is passed through a filter containing 1.5 g of anhydrous sodium sulfate. 1.5 mL of the recovered extract is transferred to a vial and stored in the freezer prior to GC/ECD analysis. 2.7. Validation of the method Validation of the SLE/PLT-GC/ECD methodology considered the following analytical parameters: selectivity, linearity, limits of detection (LOD) and quantification (LOQ), accuracy (recovery assays), and precision (repeatability and intermediate precision). In the recovery experiments, the pesticides were added to samples of lettuce at concentrations equal to one, two, and ten times the quantification limits. The procedures used to validate the proposed method were based on accepted national and international validation guidelines (ANVISA, 2003; ICH, 1995; INMETRO, 2010; MAPA, 2011; Ribani, Bottoli, Collins, Jardim, & Melo, 2004). 2.8. Evaluation of matrix effects The possible existence of matrix effects in the proposed method was investigated using two sets of analytical curves. The first curves were constructed using standard solutions of the nine pesticides in acetonitrile (pure solvent). The second curves were constructed using standards prepared in the organic extracts obtained using the SLE/PLT technique applied to pesticide-free lettuce samples. The concentrations of the standard solutions employed were in the ranges 10–360 lg L1 (chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, cypermethrin, and deltamethrin) and 20–370 lg L1 (iprodione and permethrin). The percentage matrix effect (ME) was determined from the slopes of the regression lines for the calibration curves prepared using the extract matrix and pure solvent, calculated using Eq. (1) (Pinho, Silvério, Neves, Queiroz, & Starling, 2010).
ME% ¼
aE 1 100 aS
ð1Þ
where aE = slope of the regression line for each pesticide prepared in the extract matrix; aS = slope of the regression line for each pesticide prepared in pure solvent.
67
(Barbosa, 2010), in order to independently evaluate the residual concentrations of the two compounds. Analyses of the residual pesticides were performed using two heads of lettuce from each plot, together with a control lettuce. Collections were made on days 1, 3, 4, 10, and 15 after the application of iprodione, and on days 1, 3, and 4 after the application of procymidone. The samples were crushed in a mixer and the pulps were stored at 20 °C prior to extraction using the SLE/PLT technique and analysis of the extracts by GC/ECD. 3. Results and discussion 3.1. Chromatographic analysis Optimization of the chromatographic conditions used for the simultaneous analysis of the nine pesticides resulted in good separation of the compounds (Fig. 1). Quantification of endosulfan, k-cyhalothrin, permethrin, cypermethrin, and deltamethrin, which have more than one isomer, was performed by summing the peak areas of the different isomers. Fig. 1 shows a chromatogram obtained for a mixed standard solution (in acetonitrile) of chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin (at 100 lg L1), and iprodione, permethrin, and deltamethrin (at 200 lg L1). 3.2. Optimization of the technique The efficiency of solid–liquid extraction with low temperature partitioning (SLE/LTP) may be influenced by factors such as the volume of water added, choice of extraction solution, stirring time, centrifugation, freezing, and salt addition. Careful selection of the analytical parameters is therefore required in order to minimize analysis times, while ensuring good sensitivity and accuracy. Evaluation was made of the effect of the addition of water on the separation of the aqueous and organic phases after the SLE/ LTP freezing step. Different volumes of water (0, 0.5, 1, 2, and 4 mL) were added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred for 20 min on a shaker table at 25 °C and 200 rpm, and then allowed to freeze in the
2.9. Application of the proposed method The validated method was applied to determine the residual concentrations of procymidone and iprodione in butterhead lettuces grown at the field station of the Phytotechnology Department of the Federal University of Viçosa. Table 2 describes the pesticides used, the mode of application, the maximum residue limits (MRL) established by Brazilian legislation, the doses applied, and the withdrawal period for each pesticide (ANVISA, 2014). The fungicides were applied 20 days after transplanting lettuce seedlings. The experimental cultivation was divided into separate plots for each pesticide, applied at the recommended dose
Table 2 Pesticides applied to lettuce.
*
Pesticide
Mode of application
MRL (mg kg1)
Applied dose*
Withdrawal period (days)
Procymidone Iprodione
Leaf Leaf
5 1
1–1.5 kg ha1 150 g 100 mL1 of water
3 14
Application using commercial products.
Fig. 1. Chromatogram of a standard solution containing the pesticides in acetonitrile: (1) chlorothalonil (tR = 4.94); (2) methyl parathion (tR = 5.35); (3) procymidone (tR = 6.65); (4) a-endosulfan (tR = 6.92); (5) b-endosulfan (tR = 7.61); (6) iprodione (tR = 7.79); (7) bifenthrin (internal standard) (tR = 9.51); (8) k-cyhalothrin (tR = 10.66 and 10.99); (9) permethrin (tR = 12.30 and 12.54); (10) cypermethrin (tR = 13.76 and 14.00); (11) deltamethrin (tR = 15.51 and 15.74).
68
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
freezer at approximately 20 °C for 12 h. Average percentage recoveries of 56.5 ± 4.1% were obtained using 1 mL of water, and the results were statistically similar for most of the compounds (at 95% confidence, using ANOVA with the post hoc Tukey test). It was therefore decided to use the addition of 1 mL of water to the sample in order to keep the extraction volume as small as possible. Another important parameter that must be evaluated in the optimization of an extraction technique is the type of solvent. Acetonitrile, acetone, and ethyl acetate are the three solvents most commonly used in multiresidue methods for the determination of pesticides in different matrices (Mastovska & Lehotay, 2004). Here, the influence of the polarity of the extraction solvent on pesticide recovery was investigated using either acetonitrile or a mixture of acetonitrile/ethyl acetate (6.5:1.5, v/v). 1 mL of water was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent (acetonitrile or a mixture of acetonitrile/ethyl acetate). This mixture was stirred for 20 min on a shaker table at 25 °C and 200 rpm, and then allowed to freeze in the freezer at approximately 20 °C for 12 h. The addition of ethyl acetate decreases the polarity of the medium, favoring the extraction of less polar pesticides. However, the maximum volume of ethyl acetate that could be added to the mixture was 1.5 mL, in order to avoid phase separation prior to freezing (Vieira, Neves, & Queiroz, 2007). Recovery percentages above 65.2 ± 0.9% were obtained for all the pesticides using the solvent mixture, which was therefore employed in the subsequent experiments. Similar findings were reported by Pinho, Neves, et al. (2010), who observed that less polar mixtures gave the best recoveries for the same group of pesticides present in tomato samples (Pinho, Neves, et al., 2010). Agitation results in greater interaction between the sample and the solvent, and can have an important influence on extraction efficiency. In this case, 1 mL of water was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred (different conditions), and then allowed to freeze in the freezer at approximately 20 °C for 12 h. The best results were obtained here using vortexing (for 1 min) (recovery percentages P70%) and mechanical stirring (for 10 min), with most pesticides showing no significant differences between these two methods (at the 95% confidence level, using ANOVA with the Tukey post hoc test). Mechanical stirring for 10 min was therefore selected, which provided good recovery percentages (P79.3%) with low standard deviations (66%) and greater ease of operation. The effect of salt addition on the extraction was evaluated by replacing the volume of water (1 mL) with the same volume of a 2% (w/v) solution of NaCl. This volume was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred for 10 min on a shaker table at 25 °C and 200 rpm, and then allowed to freeze in the freezer at approximately 20 °C for 12 h. It was found that for most of the compounds studied, the addition of salt did not significantly alter the extraction efficiency (P76.0%) (at a 95% confidence level, ANOVA with the Tukey post hoc test). This could be explained by the low solubility of the compounds in water. It was therefore decided not to add salt to the extraction solution. The influence of centrifugation time on the volume of recovered extract was also evaluated. This step was introduced to improve sedimentation of the solid matrix and recovery of the organic phase. In these assays, 1 mL of water was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred for 10 min on a shaker table at 25 °C and 200 rpm, centrifuged (different conditions) at 1200g and then allowed to freeze in the freezer at approximately 20 °C for 12 h. A time of 3 min
Fig. 2. Chromatograms of the extracts obtained for (A) a pesticide-free lettuce sample, and (B) lettuce containing the analytes, where the peaks are labeled as follows: (1) chlorothalonil (tR = 4.94); (2) methyl parathion (tR = 5.35); (3) procymidone (tR = 6.65); (4) a-endosulfan (tR = 6.92); (5) b-endosulfan (tR = 7.61); (6) iprodione (tR = 7.79); (7) bifenthrin (internal standard) (tR = 9.51); (8) k-cyhalothrin (tR = 10.66 and 10.99); (9) permethrin (tR = 12.30 and 12.54); (10) cypermethrin (tR = 13.76 and 14.00); (11) deltamethrin (tR = 15.51 and 15.74).
was found to be sufficient to provide satisfactory separation (volume recovered extract, 7.7 mL) and was therefore selected for use in the procedure. Freezing is one of the most important stages of the technique, as it enables phase separation and clean-up of the extract. Under the conditions identified in the preceding tests, it was found that freezing of the aqueous phase occurred after 2–3 h at a temperature of about 20 °C. A freezing time of 3 h was therefore selected, which was sufficient for extract clean-up and resulted in interference-free chromatograms as well as recoveries exceeding 78.7 ± 0.1% for most of the pesticides. Considering the results of the optimization procedure and practical aspects of the technique, the final SLE/LTP procedure required only a small extraction mixture volume and avoided the need for evaporation and solvent exchange steps. This reduced the risks of contamination and sample loss, enabling the achievement of recoveries ranging from 72.3% to 103.2%. 3.3. Validation of the analytical method 3.3.1. Selectivity The chromatograms showed satisfactory peak resolution, and the selectivity of the method for the pesticides was confirmed by comparing the chromatograms obtained for the pesticide-free extract and the extract fortified with chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin (at 0.1 mg L1) and with iprodione, permethrin, and deltamethrin (at 0.2 mg L1) (Fig. 2). 3.3.2. Linearity and limits of detection and quantification of the method In order to evaluate the linearity of the method, lettuce samples spiked with concentrations of between 0.0012 and 1.34 mg kg1 of the pesticides were extracted and analyzed using the optimized SLE/LTP-GC/ECD procedure. The analytical curves constructed for the compounds were then evaluated using least squares linear regression. The limits of detection (LOD) for the proposed method were determined based on three times the baseline noise signal obtained for the pesticide-free lettuce extract (the blank). The
69
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71 Table 3 Linearity statistics, limits of detection (LOD), and limits of quantification (LOQ) for pesticides extracted from samples of lettuce, and maximum residue limit (MRL) values. Pesticides
Linearity 1
Linear range (mg kg Chlorothalonil Methyl parathion Procymidone Endosulfana Iprodione k-Cyhalothrin Permethrin Cypermethrin Deltamethrin a
0.0012–1.01 0.0990–1.31 0.0264–1.03 0.0016–1.01 0.0330–1.04 0.0396–1.04 0.1320–1.34 0.0330–1.04 0.0330–1.04
)
LOD (mg kg1)
LOQ (mg kg1)
0.0004 0.0280 0.0070 0.0005 0.0090 0.0110 0.0370 0.0090 0.0090
0.0012 0.0924 0.0231 0.0015 0.0297 0.0363 0.1221 0.0297 0.0297
r 0.9936 0.9981 0.9982 0.9992 0.9978 0.9972 0.9967 0.9970 0.9960
MRL values (mg kg1) ANVISA (2014)
Codex Alimentarius (2013)
European Union (2013)
10.00
0.01 0.01 0.01 0.05 10.00 0.50 0.05 2.00 0.50
6.00 5.00 1.00
2.00 0.70 2.00
a-Endosulfan + b-endosulfan.
Table 4 Repeatability, intermediate precision, and recovery values achieved with the proposed SLE/LTP-GC/ECD procedure. Pesticide
Chlorothalonil Methyl parathion Procymidone Endosulfana Iprodione k-Cyhalothrin Permethrin Cypermethrin Deltamethrin a
a-Endosulfan + b-endosulfan.
*
LOQ: limit of quantification.
Coefficient of variation (%)
Recovery (%)
Repeatability (n = 7) 2 LOQ*
Intermediate precision (n = 21) 2 LOQ
LOQ
2 LOQ
10 LOQ
5.2 3.3 3.3 6.5 3.9 3.4 3.9 1.0 5.6
3.7 6.3 3.4 10.6 7.6 6.4 5.0 6.1 6.5
101.8 ± 5.3 76.9 ± 2.4 73.1 ± 2.5 85.7 ± 0.4 80.1 ± 5.4 73.2 ± 2.5 73.9 ± 2.7 82.4 ± 1.1 86.7 ± 4.7
103.2 ± 0.9 77.6 ± 2.9 75.9 ± 1.9 86.5 ± 3.9 78.2 ± 3.9 79.9 ± 6.5 76.3 ± 3.7 84.6 ± 8.6 89.4 ± 10.6
97.1 ± 0.3 77.1 ± 1.6 75.9 ± 0.6 84.4 ± 0.9 81.0 ± 1.9 74.2 ± 3.4 72.3 ± 2.5 81.8 ± 2.2 87.2 ± 1.6
limits of quantification (LOQ) were determined considering a signal 10 times greater than the baseline noise (Ribani et al., 2004). The linearity statistics, detection limits, and quantification limits for the different pesticides are shown in Table 3, together with national and international MRL values (ANVISA, 2014; Codex Alimentarius, 2013; European Union, 2013). The correlation coefficients (r) obtained for all the compounds were greater than 0.99, indicating the good linearity of the method (ANVISA, 2003), and for most of the pesticides the LOQ values were below the maximum residue limits set by ANVISA, the Codex Alimentarius, and the European Union.
3.3.3. Accuracy and precision of the method The accuracy of the technique was determined by recovery assays in which samples of pesticide-free lettuce were fortified at three concentration levels (LOQ, 2 LOQ, and 10 LOQ), in triplicate. The percentage recovery values ranged from 72.3 ± 2.5% to 103.2 ± 0.9% (Table 4). The repeatability of the method was evaluated according to the coefficients of variation (CV%) obtained for lettuce samples fortified with pesticides at concentrations of 2 LOQ, using seven replicates. The coefficients of variation obtained (Table 4) were indicative of good repeatability, since values up to 20% are considered acceptable for complex samples (Ribani et al., 2004). The intermediate precision of the method was determined using the coefficients of variation obtained for seven lettuce samples fortified with each compound at a concentration level equal to 2 LOQ and analyzed on three separate days (n = 21). Analyses on days 1, 3, and 5 were performed by the same analyst, using the same equipment. The CV values obtained varied from 3.4% to 10.6% (Table 4). Analytical procedures used to analyze pesticide residues should be able to achieve recoveries of 70–120%, on average, at each forti-
fication level, and show CV values 620% (MAPA, 2011). The values obtained here conformed to these requirements, so it could be concluded that the SLE/LTP-GC/ECD procedure was suitable for pesticide analysis. The performance of the proposed method was compared with other reported extraction and analytical techniques used for the determination of pesticides in lettuce (Table 5). The detection limits obtained here were either comparable to or better than those reported previously, and the proposed method was slightly more sensitive for some of the pesticides. Similar results in terms of recovery and CV values were obtained in other studies using SLE/ LTP-ECD, SPE-GC/ECD, DMSP-GC/MS, and SESB/LD-GC/MS/LVI techniques (Balinova et al., 2007; Barriada-Pereira et al., 2010; Rissato et al., 2005; Silva et al., 2010).
3.3.4. Evaluation of matrix effects Matrix interference in the chromatographic response was evaluated using analytical curves obtained for the pesticides prepared in pure solvent and in the SLE/LTP extract. The matrix effect was determined from the slopes of the regression lines calculated for the two analytical curves, using Eq. (1) (Pinho, Silvério, et al., 2010). The matrix effects (in terms of percentage) calculated for each pesticide are shown in Table 6. The results showed that the chromatographic responses for all the pesticides suffered from significant matrix effects (above 10%) (Hajslova & Zrostlíková, 2003). The compounds chlorothalonil, methyl parathion, k-cyhalothrin, cypermethrin, and deltamethrin showed positive interferences, with higher responses obtained for the compounds prepared in the extract, compared to those prepared in pure solvent (at the same concentrations). A possible explanation could be occupation of the active sites of the injector insert by co-extractives present in the matrix, so that only some of the analyte was adsorbed, resulting in a higher
70
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
Table 5 Comparison of the proposed method with other methods for the determination of pesticides in lettuce. Present work Pesticide Extraction technique Instrumentation Mass of sample (g) Volume and type of extraction solvent Linear range (mg kg1) LODt (mg kg1) CVu (%) Recovery (%) a b c d e f g h i j k l m n o p q r s t u
a
b
c
Rissato et al. (2005) d
e
IS , OP , OC , DC , PY SLE/LTPh GC/ECDm 4 8.0 mL (ACNp:EACq – 6.5:1.5) 0.0012–1.34 0.0004–0.037 1.0–10.6 72.3–103.2
c
f
b
OC , ON , OP , PY SLEi GC/ECDm 25 80.0 mL (ACTr) – 0.001–0.008 4.2–8.8 70.0–83.0
e
Balinova et al. (2007) g
d
b
CB , DC , OP , PY SPEj GC/ECDm 10 20.0 mL (ACTr) – 0.0005–0.001 1.0–21.0 74.0–118.0
e
Silva et al. (2010) g
c
b
CB , OC , OP , DC DMSPk GC/MSn 4 30.0 mL (ACNp) 0.0125–1.25 0.01–0.02 0.6–8.0 50.0–120.0
d
Barriada-Pereira et al. (2010) OCc SBSE/LDl GC/MS/LVIo 0.1 4.0 mL (METs) 0.0001–5.0 0.0004–0.04 0.0–17.7 3.2–68.3
IS – isophthalonitrile pesticides. OP – organophosphorus pesticides. OC – organochlorine pesticides. DC – dicarboximide pesticides. PY – pyrethroid pesticides. ON – organonitrogen pesticides. CB – carbamate. SLE/LTP – solid–liquid extraction with low temperature partitioning. SLE – solid–liquid extraction. SPE – solid phase extraction. DMSP – dispersion matrix solid phase. SBSE/LD – stir bar sorptive extraction liquid desorption. GC/ECD – gas chromatography/electron capture detector. GC/MS – gas chromatography/mass spectrometry. GC/MS/LVI – gas chromatography/mass spectrometry with large volume injection. ACN – acetonitrile. EAC – ethyl acetate. ACT – acetone. MET – methanol. LOD – limit of detection. CV – coefficient of variation.
Table 6 Percentage matrix effects for the pesticides.
Table 7 Concentrations of pesticides in lettuces collected on different days after application.
Pesticide
ME (%)
Chlorothalonil Methyl parathion Procymidone Endosulfan Iprodione k-Cyhalothrin Permethrin Cypermethrin Deltamethrin
28.6 23.6 34.6 15.3 56.2 34.4 13.4 47.8 83.6
chromatographic response (Sousa et al., 2012). In contrast, the compounds permethrin, endosulfan, iprodione, and procymidone showed matrix-induced chromatographic response reduction (Mastovska & Lehotay, 2004), with lower responses obtained using the extraction matrix, compared to the pure solvent (at the same concentrations). The negative interferences could have been due to problems associated with the injection of analytes in the extract matrix, with gradual accumulation of non-volatile matrix components in the GC system resulting in the formation of new active sites, hence reducing the analyte responses. Similar results regarding the effect of matrix were obtained by Koesukwiwat and collaborators, analyzing pesticides in fruits and vegetables using QuEChERS and low-pressure gas chromatography–time-of-flight mass spectrometry (Koesukwiwata et al., 2010). A solution to errors induced by matrix effects is to construct calibration curves using the pesticides prepared in the extract matrix, instead of in pure solvent (Ribani et al., 2004).
Pesticide
Procymidone Iprodione
Concentration (mg kg1) Day 1
Day 3
Day 4
Day 10
Day 15
1.00 ± 0.01 13.6 ± 0.4
0.50 ± 0.01 1.80 ± 0.70
0.10 ± 0.01 0.30 ± 0.10
0.03 ± 0.01
Contents lists available at ScienceDirect
Food Chemistry journal homepage: www.elsevier.com/locate/foodchem
Analytical Methods
Determination of pesticides in lettuce using solid–liquid extraction with low temperature partitioning Anna I.G. Costa a, Maria E.L.R. Queiroz a,⇑, Antônio A. Neves a, Flaviane A. de Sousa a, Laércio Zambolim b a b
Department of Chemistry, Universidade Federal de Viçosa, 36570-000 Viçosa, MG, Brazil Department of Phytopathology, Universidade Federal de Viçosa, 36570-000 Viçosa, MG, Brazil
a r t i c l e
i n f o
Article history: Received 29 July 2014 Received in revised form 30 January 2015 Accepted 14 February 2015 Available online 20 February 2015 Chemical compounds studied in this article: Chlorothalonil (PubChem CID: 15910) Methyl parathion (PubChem CID: 4130) Procymidone (PubChem CID: 36242) Endosulfan (PubChem CID: 3224) Iprodione (PubChem CID: 37517) k-Cyhalothrin (PubChem CID: 443046) Permethrin (PubChem CID: 40326) Cypermethrin (PubChem CID: 2912) Deltamethrin (PubChem CID: 40585)
a b s t r a c t This work describes the optimization and validation of a method employing solid–liquid extraction with low temperature partitioning (SLE/LTP) together with analysis by gas chromatography with electron capture detection (GC/ECD) for the determination of nine pesticides (chlorothalonil, methyl parathion, procymidone, endosulfan, iprodione, k-cyhalothrin, permethrin, cypermethrin, and deltamethrin) in lettuce. The method was found to be selective, accurate, and precise, with means recovery values in the range of 72.3–103.2%, coefficients of variation 612%, and detection limits in the range 0.4–37 lg kg1. The matrix components significantly influence the chromatographic response of the analytes (above 10%). The optimized and validated method was applied to determine the residual concentrations of the fungicides iprodione and procymidone that had been applied to field crops of lettuce. The maximum residual concentrations of the pesticides in the lettuce samples were 13.6 ± 0.4 mg kg1 (iprodione) and 1.00 ± 0.01 mg kg1 (procymidone), on the day after application of the products. Ó 2015 Published by Elsevier Ltd.
Keywords: Gas chromatography Pesticides Lettuce
1. Introduction Lettuce (Lactuca sativa L.), a low calorie salad vegetable belonging to the Asteraceae family (subfamily Cichoriaceae), is known worldwide and is a source of the vitamins A, B1, B2, and C, as well as essential elements such as calcium and iron (Altunkaya, Becker, Gökmen, & Skibsted, 2009; Rissato, Galhianea, Souzab, & Aponc, 2005). A limiting factor in the production of this vegetable is its susceptibility to diseases and pests that directly affect the marketable part of the plant and need to be carefully controlled. Chemical control has been most widely used, with the application of authorized pesticides including chlorothalonil (isophthalonitrile) and the dicarboximides procymidone and iprodione (ANVISA, 2014). Despite the benefits that can be obtained from the use of pesticides, their improper application can result in residual ⇑ Corresponding author. Tel.: +55 31 3899 1430. E-mail address: [email protected] (M.E.L.R. Queiroz). http://dx.doi.org/10.1016/j.foodchem.2015.02.070 0308-8146/Ó 2015 Published by Elsevier Ltd.
concentrations that exceed the maximum permissible limits of between 1 and 6 mg kg1 described in Brazilian legislation (ANVISA, 2014). In the European Union, the maximum permissible concentrations of authorized compounds are between 0.01 and 10 mg kg1 (European Union, 2013). There is also concern regarding the indiscriminate application of unauthorized active agents to crops. In reports issued annually by the Brazilian Program for Analysis of Pesticide Residues (PAPR), the most unsatisfactory results for lettuce crops have involved the presence of unauthorized chemicals (ANVISA, 2013). This situation demands the development and use of multiresidue methods of analysis that are inexpensive, reliable, and provide low limits of quantification in order to monitor the levels of contamination of lettuce by pesticides. The commonest methods used to extract pesticides from samples of lettuce involve an initial extraction into solvent followed by liquid–liquid partitioning, where the analytes are transferred to the organic phase, with undesirable co-extractives and some highly polar pesticides remaining in the aqueous phase (Fenoll, Hellin, Lopez, Gonzalez,
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
& Flores, 2007a, 2007b; Gelsomino, Petrovicová, Tiburtini, Magnani, & Felici, 1997; Oviedo, Toledo, & Vicente, 2003). These methods generally require large amounts of sample, sorbent and high quality organic solvents, as well as manipulation of the extracts. This has led to the development of faster and more efficient methods that are both inexpensive and environmentally friendly (Kristenson, Brinkman, & Ramos, 2006). In recent years, techniques for the analysis of pesticide residues in fruits and vegetables have evolved significantly in terms of simplification and improvements in sample extraction and purification. In particular, the technique of solid–liquid extraction with low temperature partitioning (SLE/LTP) has shown good results for the extraction of pesticides present in solid matrices such as potatoes (Dardengo et al., 2011), tomatoes (Pinho, Neves, Queiroz, & Silvério, 2010), butter (Marthe, Bittencourt, Queiroz, & Neves, 2010), and strawberries (Heleno, Queiroz, Neves, & Oliveira, 2014). The technique is based on the partition of analytes between the aqueous and organic phase resulted from temperature lowering (20 °C). The advantage of this method is that the sample components are frozen with the aqueous phase, whereas pesticides are extracted by the organic, without the need to clean-up one step. Other extraction techniques used for the analysis of pesticides include supercritical fluid extraction (SFE) (Lehotay, 1997; Rissato et al., 2005), microwave-assisted extraction (MAE) (Eskilsson & Bjorklund, 2000), pressurized liquid extraction (PLE) (Ramosa, Kristensonb, & Brinkmanb, 2002), solid phase extraction (SPE) (Balinova, Mladenova, & Shtereva, 2007; Faria, Maldaner, Santana, Jardim, & Collins, 2007), dispersion matrix solid phase extraction (DMSP) (Silva, Silva, & Navickiene, 2010), solid phase microextraction (SPME) (Hu et al., 2008), stir bar sorptive extraction (SBSE) (Barriada-Pereira, Serôdioc, González-Castro, & Nogueirac, 2010) and QuEChERS (quick, easy, cheap, effective, rugged, and safe), technique adopted in 2007 as official by the Association of Official Analytical Chemists (AOAC) that generates extracts that can be analyzed by liquid chromatography and/or gas chromatography coupled to mass spectrometry in series (GC– MS/MS and LC–MS/MS) (Koesukwiwata, Lehotay, Miao, & Leepipatpiboon, 2010). However, these techniques are more expensive, compared to SLE/LTP, and often require highly skilled technicians. The purpose of this study was to optimize and validate a methodology employing SLE/LTP with GC/ECD analysis for the simultaneous determination of the fungicides chlorothalonil, procymidone, and iprodione (compounds authorized by Brazilian legislation), and the insecticides methyl parathion, endosulfan, kcyhalothrin, permethrin, cypermethrin, and deltamethrin (compounds unauthorized by Brazilian legislation) (Table 1) in lettuce samples. The optimized and validated method was employed to determine the residual concentrations of pesticides in samples of lettuce to which authorized pesticides had been applied.
2. Experimental 2.1. Reagents and solutions Standards of chlorothalonil (99.3%), procymidone (99.9%), iprodione (99.3%), deltamethrin (99.7%), and permethrin (92.2%) were acquired from Sigma–Aldrich (Seelze, Germany). Methyl parathion (99%), endosulfan (73.2%), and cypermethrin (92.4%) were obtained from Chem Service (West Chester, PA, USA). k-Cyhalothrin (86.5%) was obtained from Syngenta (São Paulo, Brazil). Bifenthrin (92.2%), used as an internal standard, was acquired from FMC (Brazil). The HPLC-grade acetonitrile and ethyl acetate (99.5%) solvents used in the extractions were purchased from Vetec (Rio de
65
Janeiro, Brazil). Anhydrous sodium sulfate (99%) used to remove water from the extracts was also obtained from Vetec. Stock 1000 mg L1 standard solutions of the pesticides were individually prepared in acetonitrile. Working solutions containing the nine pesticides at concentrations of 10 mg L1 (chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin) and 20 mg L1 (iprodione, permethrin, and deltamethrin) were prepared by diluting the stock standard solutions with the same solvent. A solution containing the internal standard (bifenthrin) at a concentration of 10 mg L1 was prepared in a similar way as the other standards. All solutions were stored in a freezer at a temperature of approximately 20 °C. 2.2. Equipment The analyses employed a gas chromatograph (Model GC-2014, Shimadzu) equipped with an electron capture detector (ECD) and an AOC-20i auto-injector. An ultrasonic bath (Unique, São Paulo, Brazil) operated at 150 W and 33 kHz was used during optimization of the extraction method. Other equipment included an analytical balance (Model BP 2215, Sartorius, Göttingen, Germany), a table shaker (Model TE 420, Tecnal, São Paulo, Brazil), a vortex mixer, a centrifuge (Excelsa II Model 206 MP, Fanem, São Paulo, Brazil), and a pH meter (Model DM 21, Digimed, São Paulo, Brazil). 2.3. Chromatographic conditions The chromatographic separation of the analytes was performed on an Agilent DB-5 capillary column (5% phenyl/95% dimethyl siloxane stationary phase, 30 m length, 0.25 mm internal diameter, and 0.1 lm film thickness). Nitrogen was used as the carrier gas at a flow rate of 1.2 mL min1. The column temperature program consisted of an initial hold at 150 °C for 2 min, followed by an increase at 40 °C min1 to 210 °C, a hold for 2 min, then an increase at 20 °C min1 to 240 °C, a hold for 5 min, and finally an increase at 15 °C min1 to 290 °C, with a hold for 5 min. The total analysis time was 20.3 min. The injector and detector temperatures were maintained at 280 and 300 °C, respectively. A 1 lL volume of sample was injected into the chromatograph, using a split ratio of 1:5. The pesticides were identified by comparing the retention times of the peaks obtained for the sample extracts with those of the standards. Quantification was achieved using the matrix superposition method, with fortification of pesticide-free lettuce samples using nine pesticide concentrations in the range 1.2–1340 lg kg1. Calibration graphs were constructed by plotting peak areas against concentrations. 2.4. Sample preparation for optimization studies Optimization and validation of the method employing solid–liquid extraction with low temperature partitioning (SLE/LTP) and analysis by gas chromatography with electron capture detection (GC/ECD) employed organic butterhead lettuces grown in the laboratory at the Plant Diseases Clinic of the Department of Phytopathology, Federal University of Viçosa. The lettuces were crushed in a mixer (Walita, São Paulo, Brazil) until homogenization was complete. A 4 g portion of the pulp was then placed in a 22 mL clear glass flask and fortified with 80 lL of a working standard solution containing chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin at concentrations of 10 mg L1 (giving concentrations of 0.2 lg g1 in the fortified sample), together with 80 lL of a working standard solution containing iprodione, permethrin, and deltamethrin at concentrations of 20 mg L1 (giving concentrations of 0.4 lg g1 in the fortified sample). The spiked samples were allowed to stand for 3 h in order to evaporate the solvent and promote greater
66
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
Table 1 IUPAC names and chemical formula of the pesticides studied. Compound
Formula
IUPAC name
Bifenthrin
2-Methyl-3-phenylbenzyl(1RS)-cis-3-(2-chloro-3,3,3-trifluoroprop1-enyl)-2,2-dimethylcyclopropanecarboxylate H3C
O CF
CH3
3
CH3
Cl
O
CH3
O
F3C
Cl
CH3
CH3
O
(Z)-(1S)-cis-
(Z)-(1R)-cis-
CN
Chlorothalonil
Tetrachloroisophthalonitrile
Cl
Cl
Cl
CN Cl
Methyl parathion
S OP(OCH3 )2
O2 N
Cl
Procymidone
O
O,O-dimethyl O-4-nitrophenyl phosphorothioate
N-(3,5-dichlorophenyl)-1,2-dimethylcyclopropane-1,2dicarboximide
N O
Cl Endosulfan
Cl
Cl
(1,4,5,6,7,7-Hexachloro-8,9,10-trinorborn-5-en-2,3-ylenebis methylene) sulfite
Cl
Cl Cl Cl Cl
Iprodione
O
O
3-(3,5-Dichlorophenyl)-N-isopropyl-2,4-dioxo imidazolidine-1carboxamide
N
N
O
Cl H
k-Cyhalothrin
O SO
CONHCH(CH3 )2
H
Cl
CO 2
F3 C
H
C
CN O
(S)-a-cyano-3-phenoxybenzyl (Z)-(1R,3R)-3-(2-chloro-3,3,3-trifluoro prop-1-enyl)-2,2-dimethylcyclopropanecarboxylate and (R)-acyano-3-phenoxybenzyl(Z)-(1S,3S)-3-(2-chloro-3,3,3-trifluoroprop1-enyl)-2,2 dimethylcyclopropane carboxylate
(R) (Z)-(1S)-CIS
Permethrin
Cl
C CH
Cl Cypermethrin
Cl Cl
Deltamethrin
C CH
Br Br
C CH
O C
O
O
CN CO 2CH
O C
O H
O
3-Phenoxybenzyl (1RS,3RS;1RS,3SR)-3-(2,2-dichlorovinyl)-2,2dimethylcyclo propanecarboxylate
3-Phenoxybenzyl (1RS,3RS;1RS,3SR)-3-(2,2-dichlorovinyl)-2,2 dimethylcyclo propanecarboxylate
(S)-a-cyano-3-phenoxybenzyl (1R,3R)-3-(2,2-dibromovinyl)-2,2 dimethylcyclopropanecarboxylate
CN O
interaction of the pesticides with the samples, prior to extraction. These samples were then used in optimization of the SLE/PLT technique. 2.5. Optimization procedure The parameters evaluated in optimization of the SLE/PLT technique applied to the lettuce samples were: addition of water (0, 0.5, 1, 2, and 4 mL); type of extraction solvent (8 mL of acetonitrile, or a mixture of 6.5 mL of acetonitrile and 1.5 mL of ethyl acetate); mode and time of agitation (mechanical agitation for 10 and 20 min, ultrasonication for 1 and 10 min, and vortex mixing for 0.5 and 1 min); addition of salt (the sample was added to a 2%
(w/v) solution of sodium chloride, instead of water); freezing duration (3, 6, 12, and 24 h); and centrifugation duration (1, 3, and 5 min at 1200g). The best conditions were identified using the average percentage recoveries of the pesticides after extraction assays performed in triplicate. 2.6. Optimized procedure for sample The SLE/LTP procedure consists of extracting 4 g of lettuce with the addition of 1 mL of water and 8 mL of the combined acetonitrile/ethyl acetate solvent (6.5:1.5, v/v). This mixture is stirred for 10 min on a shaker table at 25 °C and 200 rpm, centrifuged for
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
3 min at 1200g, and then allowed to freeze in the freezer at approximately 20 °C for 3 h. After this time, the extract is passed through a filter containing 1.5 g of anhydrous sodium sulfate. 1.5 mL of the recovered extract is transferred to a vial and stored in the freezer prior to GC/ECD analysis. 2.7. Validation of the method Validation of the SLE/PLT-GC/ECD methodology considered the following analytical parameters: selectivity, linearity, limits of detection (LOD) and quantification (LOQ), accuracy (recovery assays), and precision (repeatability and intermediate precision). In the recovery experiments, the pesticides were added to samples of lettuce at concentrations equal to one, two, and ten times the quantification limits. The procedures used to validate the proposed method were based on accepted national and international validation guidelines (ANVISA, 2003; ICH, 1995; INMETRO, 2010; MAPA, 2011; Ribani, Bottoli, Collins, Jardim, & Melo, 2004). 2.8. Evaluation of matrix effects The possible existence of matrix effects in the proposed method was investigated using two sets of analytical curves. The first curves were constructed using standard solutions of the nine pesticides in acetonitrile (pure solvent). The second curves were constructed using standards prepared in the organic extracts obtained using the SLE/PLT technique applied to pesticide-free lettuce samples. The concentrations of the standard solutions employed were in the ranges 10–360 lg L1 (chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, cypermethrin, and deltamethrin) and 20–370 lg L1 (iprodione and permethrin). The percentage matrix effect (ME) was determined from the slopes of the regression lines for the calibration curves prepared using the extract matrix and pure solvent, calculated using Eq. (1) (Pinho, Silvério, Neves, Queiroz, & Starling, 2010).
ME% ¼
aE 1 100 aS
ð1Þ
where aE = slope of the regression line for each pesticide prepared in the extract matrix; aS = slope of the regression line for each pesticide prepared in pure solvent.
67
(Barbosa, 2010), in order to independently evaluate the residual concentrations of the two compounds. Analyses of the residual pesticides were performed using two heads of lettuce from each plot, together with a control lettuce. Collections were made on days 1, 3, 4, 10, and 15 after the application of iprodione, and on days 1, 3, and 4 after the application of procymidone. The samples were crushed in a mixer and the pulps were stored at 20 °C prior to extraction using the SLE/PLT technique and analysis of the extracts by GC/ECD. 3. Results and discussion 3.1. Chromatographic analysis Optimization of the chromatographic conditions used for the simultaneous analysis of the nine pesticides resulted in good separation of the compounds (Fig. 1). Quantification of endosulfan, k-cyhalothrin, permethrin, cypermethrin, and deltamethrin, which have more than one isomer, was performed by summing the peak areas of the different isomers. Fig. 1 shows a chromatogram obtained for a mixed standard solution (in acetonitrile) of chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin (at 100 lg L1), and iprodione, permethrin, and deltamethrin (at 200 lg L1). 3.2. Optimization of the technique The efficiency of solid–liquid extraction with low temperature partitioning (SLE/LTP) may be influenced by factors such as the volume of water added, choice of extraction solution, stirring time, centrifugation, freezing, and salt addition. Careful selection of the analytical parameters is therefore required in order to minimize analysis times, while ensuring good sensitivity and accuracy. Evaluation was made of the effect of the addition of water on the separation of the aqueous and organic phases after the SLE/ LTP freezing step. Different volumes of water (0, 0.5, 1, 2, and 4 mL) were added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred for 20 min on a shaker table at 25 °C and 200 rpm, and then allowed to freeze in the
2.9. Application of the proposed method The validated method was applied to determine the residual concentrations of procymidone and iprodione in butterhead lettuces grown at the field station of the Phytotechnology Department of the Federal University of Viçosa. Table 2 describes the pesticides used, the mode of application, the maximum residue limits (MRL) established by Brazilian legislation, the doses applied, and the withdrawal period for each pesticide (ANVISA, 2014). The fungicides were applied 20 days after transplanting lettuce seedlings. The experimental cultivation was divided into separate plots for each pesticide, applied at the recommended dose
Table 2 Pesticides applied to lettuce.
*
Pesticide
Mode of application
MRL (mg kg1)
Applied dose*
Withdrawal period (days)
Procymidone Iprodione
Leaf Leaf
5 1
1–1.5 kg ha1 150 g 100 mL1 of water
3 14
Application using commercial products.
Fig. 1. Chromatogram of a standard solution containing the pesticides in acetonitrile: (1) chlorothalonil (tR = 4.94); (2) methyl parathion (tR = 5.35); (3) procymidone (tR = 6.65); (4) a-endosulfan (tR = 6.92); (5) b-endosulfan (tR = 7.61); (6) iprodione (tR = 7.79); (7) bifenthrin (internal standard) (tR = 9.51); (8) k-cyhalothrin (tR = 10.66 and 10.99); (9) permethrin (tR = 12.30 and 12.54); (10) cypermethrin (tR = 13.76 and 14.00); (11) deltamethrin (tR = 15.51 and 15.74).
68
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
freezer at approximately 20 °C for 12 h. Average percentage recoveries of 56.5 ± 4.1% were obtained using 1 mL of water, and the results were statistically similar for most of the compounds (at 95% confidence, using ANOVA with the post hoc Tukey test). It was therefore decided to use the addition of 1 mL of water to the sample in order to keep the extraction volume as small as possible. Another important parameter that must be evaluated in the optimization of an extraction technique is the type of solvent. Acetonitrile, acetone, and ethyl acetate are the three solvents most commonly used in multiresidue methods for the determination of pesticides in different matrices (Mastovska & Lehotay, 2004). Here, the influence of the polarity of the extraction solvent on pesticide recovery was investigated using either acetonitrile or a mixture of acetonitrile/ethyl acetate (6.5:1.5, v/v). 1 mL of water was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent (acetonitrile or a mixture of acetonitrile/ethyl acetate). This mixture was stirred for 20 min on a shaker table at 25 °C and 200 rpm, and then allowed to freeze in the freezer at approximately 20 °C for 12 h. The addition of ethyl acetate decreases the polarity of the medium, favoring the extraction of less polar pesticides. However, the maximum volume of ethyl acetate that could be added to the mixture was 1.5 mL, in order to avoid phase separation prior to freezing (Vieira, Neves, & Queiroz, 2007). Recovery percentages above 65.2 ± 0.9% were obtained for all the pesticides using the solvent mixture, which was therefore employed in the subsequent experiments. Similar findings were reported by Pinho, Neves, et al. (2010), who observed that less polar mixtures gave the best recoveries for the same group of pesticides present in tomato samples (Pinho, Neves, et al., 2010). Agitation results in greater interaction between the sample and the solvent, and can have an important influence on extraction efficiency. In this case, 1 mL of water was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred (different conditions), and then allowed to freeze in the freezer at approximately 20 °C for 12 h. The best results were obtained here using vortexing (for 1 min) (recovery percentages P70%) and mechanical stirring (for 10 min), with most pesticides showing no significant differences between these two methods (at the 95% confidence level, using ANOVA with the Tukey post hoc test). Mechanical stirring for 10 min was therefore selected, which provided good recovery percentages (P79.3%) with low standard deviations (66%) and greater ease of operation. The effect of salt addition on the extraction was evaluated by replacing the volume of water (1 mL) with the same volume of a 2% (w/v) solution of NaCl. This volume was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred for 10 min on a shaker table at 25 °C and 200 rpm, and then allowed to freeze in the freezer at approximately 20 °C for 12 h. It was found that for most of the compounds studied, the addition of salt did not significantly alter the extraction efficiency (P76.0%) (at a 95% confidence level, ANOVA with the Tukey post hoc test). This could be explained by the low solubility of the compounds in water. It was therefore decided not to add salt to the extraction solution. The influence of centrifugation time on the volume of recovered extract was also evaluated. This step was introduced to improve sedimentation of the solid matrix and recovery of the organic phase. In these assays, 1 mL of water was added to 4 g of spiked sample prepared with 8 mL of the extraction solvent mixture (6.5 mL of acetonitrile and 1.5 mL of ethyl acetate). This mixture was stirred for 10 min on a shaker table at 25 °C and 200 rpm, centrifuged (different conditions) at 1200g and then allowed to freeze in the freezer at approximately 20 °C for 12 h. A time of 3 min
Fig. 2. Chromatograms of the extracts obtained for (A) a pesticide-free lettuce sample, and (B) lettuce containing the analytes, where the peaks are labeled as follows: (1) chlorothalonil (tR = 4.94); (2) methyl parathion (tR = 5.35); (3) procymidone (tR = 6.65); (4) a-endosulfan (tR = 6.92); (5) b-endosulfan (tR = 7.61); (6) iprodione (tR = 7.79); (7) bifenthrin (internal standard) (tR = 9.51); (8) k-cyhalothrin (tR = 10.66 and 10.99); (9) permethrin (tR = 12.30 and 12.54); (10) cypermethrin (tR = 13.76 and 14.00); (11) deltamethrin (tR = 15.51 and 15.74).
was found to be sufficient to provide satisfactory separation (volume recovered extract, 7.7 mL) and was therefore selected for use in the procedure. Freezing is one of the most important stages of the technique, as it enables phase separation and clean-up of the extract. Under the conditions identified in the preceding tests, it was found that freezing of the aqueous phase occurred after 2–3 h at a temperature of about 20 °C. A freezing time of 3 h was therefore selected, which was sufficient for extract clean-up and resulted in interference-free chromatograms as well as recoveries exceeding 78.7 ± 0.1% for most of the pesticides. Considering the results of the optimization procedure and practical aspects of the technique, the final SLE/LTP procedure required only a small extraction mixture volume and avoided the need for evaporation and solvent exchange steps. This reduced the risks of contamination and sample loss, enabling the achievement of recoveries ranging from 72.3% to 103.2%. 3.3. Validation of the analytical method 3.3.1. Selectivity The chromatograms showed satisfactory peak resolution, and the selectivity of the method for the pesticides was confirmed by comparing the chromatograms obtained for the pesticide-free extract and the extract fortified with chlorothalonil, methyl parathion, procymidone, endosulfan, k-cyhalothrin, and cypermethrin (at 0.1 mg L1) and with iprodione, permethrin, and deltamethrin (at 0.2 mg L1) (Fig. 2). 3.3.2. Linearity and limits of detection and quantification of the method In order to evaluate the linearity of the method, lettuce samples spiked with concentrations of between 0.0012 and 1.34 mg kg1 of the pesticides were extracted and analyzed using the optimized SLE/LTP-GC/ECD procedure. The analytical curves constructed for the compounds were then evaluated using least squares linear regression. The limits of detection (LOD) for the proposed method were determined based on three times the baseline noise signal obtained for the pesticide-free lettuce extract (the blank). The
69
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71 Table 3 Linearity statistics, limits of detection (LOD), and limits of quantification (LOQ) for pesticides extracted from samples of lettuce, and maximum residue limit (MRL) values. Pesticides
Linearity 1
Linear range (mg kg Chlorothalonil Methyl parathion Procymidone Endosulfana Iprodione k-Cyhalothrin Permethrin Cypermethrin Deltamethrin a
0.0012–1.01 0.0990–1.31 0.0264–1.03 0.0016–1.01 0.0330–1.04 0.0396–1.04 0.1320–1.34 0.0330–1.04 0.0330–1.04
)
LOD (mg kg1)
LOQ (mg kg1)
0.0004 0.0280 0.0070 0.0005 0.0090 0.0110 0.0370 0.0090 0.0090
0.0012 0.0924 0.0231 0.0015 0.0297 0.0363 0.1221 0.0297 0.0297
r 0.9936 0.9981 0.9982 0.9992 0.9978 0.9972 0.9967 0.9970 0.9960
MRL values (mg kg1) ANVISA (2014)
Codex Alimentarius (2013)
European Union (2013)
10.00
0.01 0.01 0.01 0.05 10.00 0.50 0.05 2.00 0.50
6.00 5.00 1.00
2.00 0.70 2.00
a-Endosulfan + b-endosulfan.
Table 4 Repeatability, intermediate precision, and recovery values achieved with the proposed SLE/LTP-GC/ECD procedure. Pesticide
Chlorothalonil Methyl parathion Procymidone Endosulfana Iprodione k-Cyhalothrin Permethrin Cypermethrin Deltamethrin a
a-Endosulfan + b-endosulfan.
*
LOQ: limit of quantification.
Coefficient of variation (%)
Recovery (%)
Repeatability (n = 7) 2 LOQ*
Intermediate precision (n = 21) 2 LOQ
LOQ
2 LOQ
10 LOQ
5.2 3.3 3.3 6.5 3.9 3.4 3.9 1.0 5.6
3.7 6.3 3.4 10.6 7.6 6.4 5.0 6.1 6.5
101.8 ± 5.3 76.9 ± 2.4 73.1 ± 2.5 85.7 ± 0.4 80.1 ± 5.4 73.2 ± 2.5 73.9 ± 2.7 82.4 ± 1.1 86.7 ± 4.7
103.2 ± 0.9 77.6 ± 2.9 75.9 ± 1.9 86.5 ± 3.9 78.2 ± 3.9 79.9 ± 6.5 76.3 ± 3.7 84.6 ± 8.6 89.4 ± 10.6
97.1 ± 0.3 77.1 ± 1.6 75.9 ± 0.6 84.4 ± 0.9 81.0 ± 1.9 74.2 ± 3.4 72.3 ± 2.5 81.8 ± 2.2 87.2 ± 1.6
limits of quantification (LOQ) were determined considering a signal 10 times greater than the baseline noise (Ribani et al., 2004). The linearity statistics, detection limits, and quantification limits for the different pesticides are shown in Table 3, together with national and international MRL values (ANVISA, 2014; Codex Alimentarius, 2013; European Union, 2013). The correlation coefficients (r) obtained for all the compounds were greater than 0.99, indicating the good linearity of the method (ANVISA, 2003), and for most of the pesticides the LOQ values were below the maximum residue limits set by ANVISA, the Codex Alimentarius, and the European Union.
3.3.3. Accuracy and precision of the method The accuracy of the technique was determined by recovery assays in which samples of pesticide-free lettuce were fortified at three concentration levels (LOQ, 2 LOQ, and 10 LOQ), in triplicate. The percentage recovery values ranged from 72.3 ± 2.5% to 103.2 ± 0.9% (Table 4). The repeatability of the method was evaluated according to the coefficients of variation (CV%) obtained for lettuce samples fortified with pesticides at concentrations of 2 LOQ, using seven replicates. The coefficients of variation obtained (Table 4) were indicative of good repeatability, since values up to 20% are considered acceptable for complex samples (Ribani et al., 2004). The intermediate precision of the method was determined using the coefficients of variation obtained for seven lettuce samples fortified with each compound at a concentration level equal to 2 LOQ and analyzed on three separate days (n = 21). Analyses on days 1, 3, and 5 were performed by the same analyst, using the same equipment. The CV values obtained varied from 3.4% to 10.6% (Table 4). Analytical procedures used to analyze pesticide residues should be able to achieve recoveries of 70–120%, on average, at each forti-
fication level, and show CV values 620% (MAPA, 2011). The values obtained here conformed to these requirements, so it could be concluded that the SLE/LTP-GC/ECD procedure was suitable for pesticide analysis. The performance of the proposed method was compared with other reported extraction and analytical techniques used for the determination of pesticides in lettuce (Table 5). The detection limits obtained here were either comparable to or better than those reported previously, and the proposed method was slightly more sensitive for some of the pesticides. Similar results in terms of recovery and CV values were obtained in other studies using SLE/ LTP-ECD, SPE-GC/ECD, DMSP-GC/MS, and SESB/LD-GC/MS/LVI techniques (Balinova et al., 2007; Barriada-Pereira et al., 2010; Rissato et al., 2005; Silva et al., 2010).
3.3.4. Evaluation of matrix effects Matrix interference in the chromatographic response was evaluated using analytical curves obtained for the pesticides prepared in pure solvent and in the SLE/LTP extract. The matrix effect was determined from the slopes of the regression lines calculated for the two analytical curves, using Eq. (1) (Pinho, Silvério, et al., 2010). The matrix effects (in terms of percentage) calculated for each pesticide are shown in Table 6. The results showed that the chromatographic responses for all the pesticides suffered from significant matrix effects (above 10%) (Hajslova & Zrostlíková, 2003). The compounds chlorothalonil, methyl parathion, k-cyhalothrin, cypermethrin, and deltamethrin showed positive interferences, with higher responses obtained for the compounds prepared in the extract, compared to those prepared in pure solvent (at the same concentrations). A possible explanation could be occupation of the active sites of the injector insert by co-extractives present in the matrix, so that only some of the analyte was adsorbed, resulting in a higher
70
A.I.G. Costa et al. / Food Chemistry 181 (2015) 64–71
Table 5 Comparison of the proposed method with other methods for the determination of pesticides in lettuce. Present work Pesticide Extraction technique Instrumentation Mass of sample (g) Volume and type of extraction solvent Linear range (mg kg1) LODt (mg kg1) CVu (%) Recovery (%) a b c d e f g h i j k l m n o p q r s t u
a
b
c
Rissato et al. (2005) d
e
IS , OP , OC , DC , PY SLE/LTPh GC/ECDm 4 8.0 mL (ACNp:EACq – 6.5:1.5) 0.0012–1.34 0.0004–0.037 1.0–10.6 72.3–103.2
c
f
b
OC , ON , OP , PY SLEi GC/ECDm 25 80.0 mL (ACTr) – 0.001–0.008 4.2–8.8 70.0–83.0
e
Balinova et al. (2007) g
d
b
CB , DC , OP , PY SPEj GC/ECDm 10 20.0 mL (ACTr) – 0.0005–0.001 1.0–21.0 74.0–118.0
e
Silva et al. (2010) g
c
b
CB , OC , OP , DC DMSPk GC/MSn 4 30.0 mL (ACNp) 0.0125–1.25 0.01–0.02 0.6–8.0 50.0–120.0
d
Barriada-Pereira et al. (2010) OCc SBSE/LDl GC/MS/LVIo 0.1 4.0 mL (METs) 0.0001–5.0 0.0004–0.04 0.0–17.7 3.2–68.3
IS – isophthalonitrile pesticides. OP – organophosphorus pesticides. OC – organochlorine pesticides. DC – dicarboximide pesticides. PY – pyrethroid pesticides. ON – organonitrogen pesticides. CB – carbamate. SLE/LTP – solid–liquid extraction with low temperature partitioning. SLE – solid–liquid extraction. SPE – solid phase extraction. DMSP – dispersion matrix solid phase. SBSE/LD – stir bar sorptive extraction liquid desorption. GC/ECD – gas chromatography/electron capture detector. GC/MS – gas chromatography/mass spectrometry. GC/MS/LVI – gas chromatography/mass spectrometry with large volume injection. ACN – acetonitrile. EAC – ethyl acetate. ACT – acetone. MET – methanol. LOD – limit of detection. CV – coefficient of variation.
Table 6 Percentage matrix effects for the pesticides.
Table 7 Concentrations of pesticides in lettuces collected on different days after application.
Pesticide
ME (%)
Chlorothalonil Methyl parathion Procymidone Endosulfan Iprodione k-Cyhalothrin Permethrin Cypermethrin Deltamethrin
28.6 23.6 34.6 15.3 56.2 34.4 13.4 47.8 83.6
chromatographic response (Sousa et al., 2012). In contrast, the compounds permethrin, endosulfan, iprodione, and procymidone showed matrix-induced chromatographic response reduction (Mastovska & Lehotay, 2004), with lower responses obtained using the extraction matrix, compared to the pure solvent (at the same concentrations). The negative interferences could have been due to problems associated with the injection of analytes in the extract matrix, with gradual accumulation of non-volatile matrix components in the GC system resulting in the formation of new active sites, hence reducing the analyte responses. Similar results regarding the effect of matrix were obtained by Koesukwiwat and collaborators, analyzing pesticides in fruits and vegetables using QuEChERS and low-pressure gas chromatography–time-of-flight mass spectrometry (Koesukwiwata et al., 2010). A solution to errors induced by matrix effects is to construct calibration curves using the pesticides prepared in the extract matrix, instead of in pure solvent (Ribani et al., 2004).
Pesticide
Procymidone Iprodione
Concentration (mg kg1) Day 1
Day 3
Day 4
Day 10
Day 15
1.00 ± 0.01 13.6 ± 0.4
0.50 ± 0.01 1.80 ± 0.70
0.10 ± 0.01 0.30 ± 0.10
0.03 ± 0.01
