Study On Quantitative Analysis Methods Of Multiple Pesticides Based On THz-TDS Technology

Pesticides play a significant role in modern agriculture due to their ability to prevent pests and diseases for crops.However,with the large-scale application of pesticides,food safety issues induced by pesticide residues have become increasingly prominent.Two or more kinds of pesticides are mixed and used at different stages of crop growth,which leads to various pesticide residues.Therefore,there is an urgent demand to carry out quantitative analysis research on the mixtures of multiple pesticides and food matrix.As a new type of spectrum technology that has been continuously developed in recent years,terahertz(THz)spectroscopy can conveniently,non-destructively and effectively detect the physical and chemical information of materials.Thus,it has attracted comprehensive attention in the field of detection.In this paper,based on Terahertz time-domain spectroscopy(THz-TDS)technology,the quantitative analysis of imidacloprid and carbendazim in flour matrix were studied by utilizing chemometrics method,Voigt fitting combined with Partial Least Squares(PLS)method,respectively.The primary working contents are generalized as follows:1.Analysis of THz absorption spectrum and research of baseline correction method for pesticide absorption spectrum.The absorption spectra of imidacloprid and carbendazim were measured.Then the mixtures of these two pesticides and flour were analyzed and the absorption spectra of the mixture were obtained.Then,the baseline correction method of the absorption spectrum was investigated.In this paper,the Multiple Spectra Baseline Correction(MSBC)method which based on asymmetric least squares smooth was used to correct the baseline of absorption spectrum.The algorithm can effectively eliminate the scattering effect on the spectrum and improve the signal-to-noise ratio of the terahertz spectrum,thereby improving the accuracy of prediction model.2.The quantitative detection of imidacloprid,carbendazim and flour mixtures was researched based on THz-TDS technology combined with chemometrics method.Four regression methods,PLS,Principal Component Analysis(PCA),Support Vector Machine(SVM)and PCA-SVM,were employed to construct the relationship between absorption spectrum and sample concentration.Cross-validation was used to optimize the quantitative model.Root mean square error(RMSE)and correlation coefficient(R)were applied to evaluate the quantitative methods above.Finally,the quantitative results of four models were compared and analyzed.The results indicate that THz-TDS technology combined with chemometric methods can be used to simultaneously quantify multiple pesticide components in flour matrix.3.Simultaneous quantitative analysis of multiple pesticides in flour matrix was investigated based on Voigt fitting combined with PLS method.First,the baseline of the mixture spectrum was corrected using the MSBC method,and Voigt function was ustilized to analyze the corrected spectrum.Then,the samples were grouped into calibration sets and prediction sets.The calibration set was used to establish the quantitative model,and then the prediction set was predicted using the established model.In the case of different calibration sets,the calibration spectrum and the Voigt fitting spectrum were used as inputs of the PLS model respectively to predict the concentrations of imidacloprid and carbendazim in the mixture.Finally,the prediction results of the Voigt fitting combined with PLS method were compared with those of the PLS method alone.The results indicate that when the number of samples in calibration set is very small,the Voigt fitting combined with PLS method can give a better prediction result.The paper includes 42 pictures,7 tables,and 106 references.