Confocal Micro-Raman Spectrometry Determination of Multi-Pesticide Formulations Using Least-Squares Support Vector Regression

The feasibility of confocal micro-Raman spectroscopy for determination of two pesticide formulations was investigated, including chlorpyrifos and malathion. Confocal micro-Raman spectra of different concentrations of pesticide solutions were collected at excitation wavelength of 780 nnn. Calibration models of the concentrations of pesticide solutions versus the Raman intensity were established by curve fitting and least-squares support vector regression (LS-SVR) with four different preprocessing methods. Compared to curve fitting, the performance of LS-SVR was slightly better, with higher correlation coefficients (r) of 0.984 and 0.963, and lower root mean square error (RMSE) of 0.221 and 0.260 for chlorpyrifos and malathion. Based on the results, it was concluded that confocal micro-Raman spectroscopy is a potential tool to determine the pesticide formulation. The method can eliminate the reagent consumption and waste generation, and avoid the contact of the operator with toxic products for food quality and safety control.