Multiresidue Analysis of Pesticides in Soil by Liquid-Solid Extraction Procedure

A multiresidue method for simultaneous determination of four pesticides (diazinon, acetochlor, aldrine and carbofuran) belonging to different pesticide groups, extracted from soil samples, is described. The method presented is based on liquid-solid extraction (LSE) and determination of pesticides, i.e. the pesticides were extracted by methanol-acetone mixture, purified on florisil column and eluted by ethyl acetate-acetone mixture. Optimization of the main parameters affecting the LSE procedure, such as the choice of purification sorbent, as well as the elution solvent and its volume, were investigated in details and optimized. Also, validation of the proposed method was done. Gas chromatography-mass spectrometry (GC-MS) was used for detection and quantification of the pesticides studied. Relative standard deviation (RSD) and recovery values for multiple analysis of soil samples fortified with 30 μg/kg of each pesticide were below 8% and higher than 89%, respectively. Limits of detection (LOD) for all the compounds studied were less than 4 μg/kg.


INTRODuCTION
The safety of food production depends on a number of factors, including the history of persistent toxic chemicals applied.As pesticides are the most widespread agrochemicals, their monitoring in plant material and soil in which it had grown is frequently performed.Considering the movement and transformation processes in which pesticides participate after being introduced into the environment (Đurović, 2011), as well as their potential persistence and effects they can produce, it seems that soil contamination over a long period of time is the biggest problem in terms of food safety.
Detailed investigation of pesticide residues began in mid-20 th century in reaction to early uses of organic pesticides.Numerous methods of analysis were developed, mostly using the conventional liquid-liquid extraction (LLE) or Soxhlet extraction.Although these extraction procedures may be considered conventional, no small number of publications dealing with their applications in pesticide residues determination in samples of different origin can be found in recent years (Kreuzig et al., 2000;Doong and Liao, 2001;Bavcon et al., 2003;Caballo-Lopez and Luque de Castro, 2003;Kodaka et al., 2003;Monkiedje et al., 2003;Navarro et al., 2003;Nawab et al., 2003;Stipicevic et al., 2003;Sassman et al., 2004;Barra et al., 2005;Wanga et al., 2008).On the other side, as these methods are part of a standard procedure, many laboratories currently use them for routine analysis.
Considering all of this and the fact that most of these methods are based on determination of pesticides belonging to the same chemical group, the aim of this study was to develop a multiresidue LSE method for simultaneous determination of pesticides that have distinct chemical structures and belong to different pesticide groups.The main parameters affecting LSE procedures, such as the choice of purification sorbent, as well as elution solvent and its volume, were investigated and optimized.Additionally, the most important analytical parameters, such as linearity, limit of detection (LOD), precision and confidence of the presented method were determined under the optimized LSE procedure.

Reagents and materials
The pesticides chosen for this study were: diazinon, acetochlor, aldrine and carbofuran (Dr Ehrenstorfer, Germany).The main physico-chemical properties of the pesticides studied are presented in table 1.Stock solutions (1 g/L) of each pesticide standard were prepared by dissolving the weighed amount in acetone (J.T. Baker, Deventer, Holland).Working standard mixed solutions (10 mg/L and 1 mg/L of each compound) were prepared weakly by diluting individual stock solutions with acetone.All solvents used in the study (hexane, acetone, methanol and ethyl acetate) were purchased from J. T. Baker (Deventer, Holland), as well as anhydrous sodium sulfate (99.0%purity).Florisil (60-100 mesh) was purchased from Serva (Germany), and silica gel (0.2-0.5 mm) and aluminum oxide (neutral) from Kemika (Croatia).Before use, sodium sulfate was dried for 24 h at 130 0 C, silica gel for 2 h at 110 0 C (SILICA), and florisil for 4 h at 600 0 C (FL, 600), 2 h at 300 0 C (FL, 300), and 18 h at 130 0 C (FL, 130).Aluminum oxide (Al 2 O 3 ) was used without previous treatment or purification.
An uncontaminated soil sample originating from Kikinda was used in the study.The main physico-chemical properties of the soil were: pH (H 2 O)=8.39;organic matter content=3.17%;sand content=73.96%;silt content =22.60%; clay content=3.44%.The soil was air dried and sieved (2 mm) before using.

Instrumentation
A gas chromatograph-mass spectrometer (GC/MS) (CP-3800/Saturn 2200, Varian, Australia) with 30 m x 0.25 mm x 0.25 µm VF-5ms column (Varian) was used.The GC was programmed as follows: initial temperature was 120 0 C, then increased to 170 0 C at 8 0 C/min and held for 4.5 minutes, increased to 280 0 C at 9 0 C/ min and held for 5.5 minutes.Helium was used as the carrier gas and its flow rate was 1.1 ml/min.
The ion trap mass spectrometer was operated in the electron impact/selected ion monitoring (EI/SIM) mode.The ion trap and transferline temperatures were set to 220 0 C and 250 0 C, respectively.One specific pesticide ion was selected for detection and quantification, while a second one was used for confirmation.The ions

Soil extraction optimization
The main parameters affecting the LSE procedure, such as the choice of purification sorbent, as well as elution solvent and its volume, were investigated and optimized.As our previous studies (Đurović et al., 2010a, 2010b) of multiresidue determination of pesticides by solid phase microextraction (SPME) showed that two successive extractions with methanol:acetone=1:1 mixture (MET-AC) were the best choice among different solvents tested (water, hexane, acetonitrile, acetone, methanol, methanol:acetone=1:1 and methanol:acetone:hexane=2:2:1), the same extraction mixture was chosen in this study, too.
LSE optimization was done using the following procedure: soil samples (10 g) were weighed in 50-mL centrifuge tubes and fortified with mixed pesticides at 30 µg/kg of each pesticide studied.The spiked samples were homogenized for 15 min using a rotary stirrer and left for 24 hours prior to further analysis.After addition of 2 g of Na 2 SO 4 (anh), the samples were homogenized twice with 15 mL of MET-AC for 30 min on the rotary stirrer and then centrifuged for 5 min at 4000 rpm.The extract was filtered and evaporated to dryness at 40 0 C using a rotary evaporator.The residues were redissolved in 5 mL of solvent (ethyl acetate-acetone mixture (4:1, v/v) (EA-AC) or acetonehexane mixture (1:4, v/v) (AC-HEX)) and 2 mL of the obtained solution was passed through a glass column previously filled with 1 g of Na 2 SO 4 (anh.)and 5 g of sorbent (florisil, silica gel or aluminum oxide).The sorbent was entered into the column with 30 mL of appropriate solvent (EA-AC or AC-HEX).Pesticides were eluted from the column by 30 mL of EA-AC or AC-HEX.Eluate was evaporated to dryness, the residues were redissolved in 2 mL of acetone, and 1 μL of the obtained solution was transferred into a vial for GC-MS analyses.

RESuLTS AND DISCuSSION
The main parameters affecting the LSE determination of selected pesticides in soil were investigated and optimized.Optimization was done by a well-structured step-by-step approach including the choice of a most suitable purification sorbent as well as elution solvent and its volume.

Selection of sorbent and elution solvent
As most pesticides included in this study have nonpolar or intermediate polar character (Table 1), it was suitable to use polar adsorbents for purification of the obtained soil extracts.Since silica gel, aluminum oxide and florisil (Mg deposited on silica gel) have an expressive polar character, these materials were included as sorbents in the study.Additionally, it is well known that the kind of sorbent activation could significantly affect purification efficiency (Zweig, 1978).Therefore, different ways of sorbent activation were included in our study, as presented in the section Materials and Methods/Reagents and materials.
On the other side, considering the characteristics of the sorbents used and the nonpolar or moderately polar nature of the studied pesticides, it was most appropriate to use nonpolar solvents or solvents of moderate polarity as elution solvent.In the presented study, two combinations of such solvents were tested (EA-AC and AC-HEX).
The results obtained for the combination of these three sets of experimental parameters (sorbent types, different ways of sorbent activation and different elution solvents used) are presented in figure 1.The recovery values of most of the selected pesticides and experimental parameters are shown to be higher than 70%, and therefore they may be considered as acceptable in terms of method confidence (Taverniers, et al., 2004).As shown in figure 1, somewhat lower recoveries were obtained for: carbofuran and the combination of FL, 130 as sorbent and AC-HEX as elution solvent, and SILICA and AC-HEX; diazinon and the combination of Al 2 O 3 and EA-AC; and aldrine with combined FL, 130 and AC-HEX, and SILICA and Al 2 O 3 with both elution solvents.An explanation of these results could be the exceptional complexity of the soil matrix and the fact that the samples were fortified with pesticides and left to rest for 24 h prior to analysis (intending to better simulate real-life conditions).
However, comparing the efficiency of sorbents, we concluded that florisil efficacy was highest for all pesticides studied, regardless of its activation mode or type of elution solvent used.And finally, comparing the results obtained for florisil as a sorbent, and both ways of its activation, and the type of elution solvents used, the best efficiency was achieved with the combination of FL, 130 and EA-AC.Therefore this combination of sorbent and solvent was chosen for further work.

Selection of optimum elution solvent volume
After selecting the most suitable purification sorbent and elution solvent, the next step was to determine the optimal elution solvent volume.In order to do that, after methanol-acetone extraction of the spiked soil samples and purification of the obtained extract on FL, 130 column, the volume of EA-AC as elution solvent was varied in the range of 20 to 35 mL.As figure 2 shows for all pesticides under study, the best method efficiency was achieved using 30 mL of EA-AC as elution solvent.Therefore, this volume was chosen as optimal for further work.

Method validation
The optimized LSE method, based on pesticides extraction by methanol-acetone mixture, purification on florisil column and elution by ethyl acetate-acetone mixture was validated.The most important analytical parameters, such as linearity, limit of detection (LOD), precision and confidence of the presented method were tested and determined.
Linearity was tested over the concentration range from 10 to 300 μg/kg for all pesticides studied.The obtained arrangements and correlation coefficients (R) are presented in table 2. The acquired correlation coefficients exceeded 0.99 for all compounds, indicating good linearity.Confidence and precision of the proposed LSE method were determined by performing four consecutive measurements of soil samples fortified with up to 30 µg/kg of all pesticides studied.The recovery and RSD values are presented in table 2. The recovery values exceeded 89% for all pesticides studied, indicating that the proposed method can be used for efficient determination of the selected pesticides from complex matrix samples such as soil.As RSD values for all compounds under study were below 8%, the proposed method is satisfactory in terms of precision, too.
The limit of detection (LOD) was determined to be 3.29 × s B (where s B is the blank standard deviation), according to IUPAC recommendations (Currie, 1999).As table 2 shows, the obtained LODs for all pesticides studied were less than 4 µg/kg, indicating that the sensitivity of the proposed method is good enough for soil sample analysis.

Figure 1 .
Figure 1.Dependence of method efficiency on sorbent types, different ways of sorbent activation and different elution solvents

Figure 2 .
Figure 2. Dependence of method efficiency on the volume of elution mixture of ethyl acetate-acetone (4:1, v/v)