پديدآورندگان :
Balsini Parvaneh Department of Chemistry, Sharif University of Technology, Tehran, Iran , Parastar Hadi h.parastar@sharif.edu Department of Chemistry, Sharif University of Technology, Tehran, Iran;
كليدواژه :
Pesticides , gas chromatography , chemometrrics , experimental design , multivariate calibration
چكيده فارسي :
Pesticides are materials that they used in agriculture for combat a variety of pests. Many persistent pesticides have been reported as cause of several diseases such as cancer. The U.S. Environmental Protection Agency (EPA) and European Commission (EC) used maximum residue limits (MRLs) to ensure food safety for consumers. Pesticides can be found in different samples such as environmental, vegetables, fruits and dairy products. Due to the complex sample matrices and low concentration of pesticides, extraction step has a very important role [1]. On the other side, various methods have been developed for the determination of pesticides in complex sample matrices such as UV-Vis spectrophotometry, high-performance liquid chromatography (HPLC) and gas chromatography (GC-MS and GC-FID) [2]. In this study, Quick, Easy, Cheap, Effective, Rugged and Safe (QuEChERS) combined with dispersive liquid–liquid micro-extraction (DLLME) has been used for extraction of pesticides in water samples. In this regard, simultaneous determination of 10 pesticides with QuEChERS-DLLME and GC-FID were designed, modelled and optimized using central composite design (CCD), backward multiple linear regression (MLR) and Nelder-Mead simplex optimization [3]. On this matter, 36 experiments for 5 extraction factors in 5 levels were generated. Also, multi-response optimization using Derringer desirability function was used to combine individual models of 10 pesticides and global optimum conditions were obtained which were 0.25 mL of acetonitrile, 39.34 µL of chloroform, 25 min of sonication time and 3.72% (w/v) salt in pH 6.7. After optimization of extraction procedure, multivariate calibration model for 10 pesticides was developed using Partial Least Squares (PLS) model in linear dynamic range (LDR) of 0.5–100 µgL-1. GC-FID profiles were baseline corrected and aligned before PLS modeling. Then, multivariate analytical figures of merit (AFOM) including sensitivity, selectivity and limit of detection (LOD) were calculated [4]. As an instance, the calculated LODs were below the MRLs of 10 pesticides which confirms the validity of the proposed method. Finally, the proposed analytical method was used for identification and quantification of 10 pesticides in real samples (i.e., tap water, milk).
