Rapid Detection Of Pesticide Residues Using Surface-enhanced Raman Spectroscopy

Pesticides play an important role in the growth of crops,which could effectively control pests and diseases and increase crops yield.However,the unreasonable application of pesticide could conduct some problems including excessive residues and water pollution,which would affect food security and human health.To ensure food security and human health,the detection of pesticide residues has become a great concern,and the related detection techniques have also become hotspot.Surface-enhanced Raman spectroscopy(SERS)has been applied to pesticide residues detection with its advantages including simple sample preparation and rapid speed.In this paper,SERS combined with chemometrics methods were used to explore pesticide residues in crop and irrigation water during the growth period.The main contents are as follows:(1)A rapid methodology for detecting three pesticide residues including phosemet,sulfoxaflor and fonofos based on SERS in water.First,water samples of pesticide residues containing different concentrations were centrifuged for SERS measurement.Then,spectral baseline were subtracted using polynomial fitting.Different classification models were adopted for spectral classification.The results showed that the accuracy(100%)of K-nearest neighbor model was the optimal for three pesticide spectra.Finally,different regression methods were established to realize quantitative analysis of spectra.The results indicated that PLSR method had the optimal predictive performance for sulfoxaflor and fonofos.RMSEP and R2pof this two PLSR models were 0.318,0.9991 and 0.515,0.9997,respectively.SVR model obtained the optimal predictive performance for phosemet with RMSEP of 0.207 and R2p of 0.9995.(2)A methodology for accurate detection of fenitrothion residues on maize was developed using SERS.Firstly,the residues of fenitrothion on maize was extracted by ethanol solvent.The extracting solution was used for SERS measurement.It could be known from results,the recovery detected by gas chromatography-mass spectrometer was higher than 92.5%;then,the detective limitation of pesticide residues determined by Principal component analysis was 0.5mg/kg.Finally,SVR combined with S-G convolution smoothing and wavelet transform(WT)was employed to establish models.The results indicated that SVR-WT regression model has the best quantitative analysis results with RMSEP of 0.1341 and R2P of 0.9996.(3)The SERS spectrum analysis software was designed to analyze pesticide residue spectra.According to interface message,the operator imported data,selected the corresponding algorithm and obtained the processed result with the type of graphic on interface,and then,these results could be observed clearly,directly and conveniently.Meanwhile,this sofftware contained parameter setting functions of samples and algorithms,and the optimal results could be obtained by changing these parameters.In addition,data management function of software could save output data for subsequent employment and search.This paper showed that SERS could be used to detect pesticide residues in water and maize.Meanwhile,chemometric methods(SVM and PLS)could were applied for qualitative and quantitative spectra analysis.The SERS analysis software for pesticide residue was designed to simplify the data processing for satisfying more requirement of researchers.