Research On Rapid Detection Of Pesticide Residues On Tomato Based On Flexible Materials And SERS

With the rapid development of economic and the continuous improvement of people’s requirements for life,the demand for fruits and vegetables is also increasing.In order to reduce the yield reduction of fruits and vegetables caused by pests and diseases,the abuse of pesticides has become more and more serious,which has caused a series of food safety problems.The purpose of this research was to study the method of rapid detection of pesticide residues on tomato surface,using flexible tape as a support,combining with AuNPs and coagulant to make a flexible SERS substrate.Through "adsorption-separation-dripping",the collection and detection of thiabendazole,carbendazim and chlorpyrifos on tomato surface were realized.The main research contents and conclusions of were as follows:1.The effects of different commercial tapes on the SERS signal of pesticides were studied,and the sampling efficiency of the tapes was analyzed.The results showed that:(1)Among the four commercial tapes(scotch tape,paper tape,3M 9080 super tape,3M post-it),the background noise of scotch tape and 3M 9080 super tape were weak and will not interfere with the characteristic peaks of the pesticide.(2)The sampling rate of the tape was studied with the analysis of variance,and the type of tape had a significant impact on the sampling rate.Whether it is on the glass surface or the tomato surface,the sampling performance of 3M9080 super glue was significantly better than that of transparent glue.The average sampling rates of thiabendazole,carbendazim,and chlorpyrifos on tomato with 3M 9080 super tape were 64.9%,65.3%,and 60.3%,respectively,and the value of Relative Sandard Deviation were less than 10%,indicating that this method meet the requirements of practical application and had the potential for quantitative detection of pesticides.2.The parameters such as the concentration of AuNPs,the type of coagulant,the amount of AuNPs and coagulant which make the best SERS signal enhancement of the three pesticides were determined.The results showed that:(1)When the AuNPs was concentrated 2.5 times,the SERS signals of thiabendazole,carbendazim and chlorpyrifos were the strongest.(2)Using NaCl as the coagulant,the enhancement of pesticide SERS signal was significantly better than that of KI and NaBr.(3)When the amount of AuNPs was 80 pL,the SERS signal intensity enhancement of thiabendazole pesticide was the best.The SERS signals of carbendazim and chlorpyrifos pesticides were the strongest,when the amount of AuNPs was 120 μL.(4)When the ratio of AuNPs to NaCl was 8:1,the SERS signal enhancement was the best.For thiabendazole,carbendazim and chlorpyrifos,the optimum amount of NaCl was 10μL,15 μL and 15 μL,respectively.(5)The optimized flexible SERS substrate had good sensitivity.The detection limits for the three pesticide standard solutions of thiabendazole,carbendazim and chlorpyrifos were 8 ng/cm2,20 ng/cm2 and 40 ng/cm2 respectively.There was a good linear correlation between SERS signal intensity and the concentration of thiabendazole,carbendazim and chlorpyrifos,the Correlation coefficient(R2)were 0.992,0.973 and 0.989 respectively.The characteristic peak intensity of 20 randomly selected detection sites on the flexible SERS substrate was counted.The RSD of the characteristic peak intensity were all less than 9%,which met the requirements of rapid detection in practical applications.3.The rapid and non-destructive detection of pesticide residues on tomato based on portable Raman spectrometer and flexible SERS substrate were realized.The results showed that:(1)The quantitative detection limits for thiabendazole,carbendazim and chlorpyrifos on tomato with flexible SERS substrate were 20 ng/cm2,36 ng/cm2 and 80 ng/cm2 respectively.(2)The Partial Least Squares(PLS)and Least Squares Support Vector Machine(LSSVM)models based on the full spectrum,the characteristic variables selected by Competitive Adaptive Reweighted Sampling(CARS)and Successive Projections Algorithm(SPA),and the band of characteristic peak were established respectively.Among them,the LSSVM model based on the characteristic peak band had the best accuracy and strong generalization ability.The Correlation Coefficient of Prediction(Rp2)for thiabendazole,carbendazim and chlorpyrifos were 0.960,0.933 and 0.864 respectively,and the Residual Predictive Deviation(RPD)were 5.044,3.504 and 2.507 respectively.In general,it was feasible to use portable Raman combined with a flexible SERS substrate to detect pesticide residues on the surface of tomatoes quickly and non-destructively.