Desenvolvimento e validação de método multirresíduo para determinação de pesticidas em grãos de soja por QUECHERS e DLLME

Abstract

A method was developed for the determination of chlorpyrifos, α - endosulfan, β - endosulfan, endosulfan sulfate, malathion, metolachlor, parathion-methyl, permethrin and trifluralin in soybean using as sample preparation the QuEChERS method combined with Dispersive Liquid-Liquid Microextraction (DLLME). In the development of the method a clean-up step using the adsorbents Activated Neutral Alumina (ANAt), Florisil, Celite, C-18 and silica gel was evaluated. The effect of type of extractor solvent, disperser solvent volume and addition of salt (ionic strength) were evaluated in the DLLME step applying a 2n factorial design. After developed the proposed method, the QuEChERS sample preparation procedure consisted of adding 10.0 mL CH3CN and 10.0 mL of deionized water and 1.00 g of the sample ground soybeans, followed by stirring and adding of 2.00 g of anhydrous MgSO4 and 1.00 g NaCl. After agitation and centrifugation, the clean-up of 3.00 mL of the extract with 0.250 g of MgSO4 and ANAt, followed by stirring and centrifugation was done. The procedure was repeated once. In step DLLME 1.00 mL of the extract was transferred to a tube containing 5.00 mL of aqueous 5% NaCl (w/v), followed by addition of 100 μL of n-hexane, stirring and centrifugation, and 40 μL of the collected organic phase for determination of analytes by GC-MS. The type of extractor solvent disperser solvent volume and addition of salt (ionic strength) have a significant effect (p = 0.05) in the DLLME step for most analytes evaluated. The matrix superposition method was suitable for calibration since the matrix effect was observed for most analytes. The proposed method provided recoveries between 81-111% with coefficients of variation (CV) lower than 17%, except to permethrin whose CV reached 45%, for fortification levels between 5.00 and 500 mg kg-1. The limits of detection and quantification of this method ranged from 2.60 to 46.6 μg kg-1 and 5.00 to 100 μg kg-1, respectively. These results meet the maximum residue limits established by ANVISA, Codex Alimentarius and EU, with the exception of malathion and chlorpyrifos (meet the MRL established by the Codex Alimentarius and the US, but not the ANVISA) and parathion-methyl (meets only the MRL established by ANVISA) .The method was applied to three samples of soybean and none of the evaluated pesticides was detected. The results indicate that the proposed method can be an alternative for the determination of pesticides evaluated in soybean. The combination of DLLME and QuEChERS methods promotes the reduction of the limit of quantification, taking into consideration the amount of sample used. The proposed method is rapid, easy to perform, and avoids the use of chlorinated solvents and sophisticated equipment for the sample preparation.