Efficient determination of pesticides in rice by QuEChERS and UV spectroscopy in combination with chemometrics

Abstract

AbstractBackground and ObjectivesTargeting the problems, like various residual pesticides, the high interference of sample matrix and the slow detection speed, this work studied the rapid screening method of multipesticide residues in rice by taking bensulfuron‐methyl, propanil and cypermethrin as examples.FindingsRice matrix solution was obtained by using the improved QuEChERS method, matrix standard solution, and solvent standard solution were prepared to obtain the linear range and matrix effect of each pesticide. QuEChERS method could rapidly pretreat samples, and the matrix effect was soft, with matrix effect values lower than 16%. Mixed solutions of the three pesticides in four combinations were prepared within the linear range. The type of pesticides in rice was determined by support vector machine (SVM), and the pesticide concentration was predicted by partial least square regression (PLSR). SVM provided a supervised identification method for pesticide residues, making the recognition accuracy higher than 90%. The error of PLSR in predicting pesticide concentration was lower than 5.36%.ConclusionsUV spectroscopy had fast test speed, so the combination of QuEChERS and UV spectroscopy could improve the efficiency of analysis. SVM combined with PLSR could preliminarily determine the type of pesticides in rice sample, and then accurately predict the concentration of pesticides.Significance and NoveltyThe QuEChERS sample pretreatment technology and UV spectroscopy in combination with chemometrics was expected to be a new technology for rapid detection of pesticide residues in samples.