Detection Of Chlorpyrifos Based On Vis-NIR And Electrochemical Biosensor

Organophosphorus pesticides are widely used pesticides all around the world.In modern agricultural production,organophosphorus pesticides have played a vital role in preventing diseases and pests,improving the quality and yield of agricultural products and increasing farmers’ income.But at the same time,the nonstandard and excessive use of organophosphorus pesticides will lead to the accumulation of organophosphorus pesticides in farmland environment and ecosystem,which brings potential hidden dangers to human life,health and environmental safety.Effective detection methods of organophosphorus pesticides can deal with this potential risk.Therefore,it is of great significance to develop a rapid,convenient,accurate and cheap detection method of organophosphorus pesticides.Although traditional detection technologies such as gas chromatography,high performance liquid chromatography and chromatography-mass spectrometry can accurately detect organophosphorus pesticides,these methods usually require expensive instruments and equipment,professional operation skills,cumbersome operation steps and long detection time,which seriously limits the application scope and field of these detection methods,These traditional standard detection methods are difficult to be applied to the rapid or on-line detection of farmland environment.In recent years,visible near infrared spectroscopy and electrochemical biosensor detection technology have attracted extensive attention from scholars at home and abroad because of their fast,convenient,cheap and non-destructive characteristics.They show potential application value in the detection of organophosphorus pesticide residues.In this study,a representative organophosphorus pesticide chlorpyrifos(CP)was taken as the research object.The rapid detection of organophosphorus pesticide chlorpyrifos was studied by using visible near infrared spectroscopy and electrochemical biosensor technology with high sensitivity,and the hardware design of electrochemical biosensor was preliminarily studied and simulated.Around this research topic,this thesis mainly includes the following four parts:(1)Nondestructive qualitative analysis and detection of chlorpyrifos residue on cabbage leaves based on Vis-NIR.The cleaned 150 cabbage leaf samples were divided into 5 groups with 30 samples in each group.The cabbage leaves of each group were soaked in five chlorpyrifos pesticide solutions with different concentrations(1:1000,1:800,1:500,1:200,control group)to make the cabbage leaves residual chlorpyrifos pesticides with different concentrations.The visible near infrared reflectance data of cabbage leaves were collected by visible near infrared spectrometer.The full band spectral reflectance(R),reflectance first derivative(FD)and reflectance second derivative(SD)are preprocessed by grouping integration(summation)respectively,and the processed spectral data are modeled by BP neural network model to qualitatively identify the pesticide residue concentration of chlorpyrifos.From 30 samples in each group,24 samples are randomly selected,a total of 120 samples are used as the modeling set,and the remaining 6 samples in each group,a total of 30 samples are used as the prediction set.The experimental results show that when the average grouping number of spectral reflectance data is 25 groups and the first derivative(FD)of spectral reflectance is used for modeling,the modeling effect is the best.The recognition accuracy of the best modeling effect for the modeling set is97.50%,and the recognition accuracy of the prediction set can reach 96.67%.The modeling effect is good.This shows that based on the spectral data of the first derivative of the spectral reflectance of cabbage leaves,the average grouping integral preprocessing method of spectral data,combined with BP neural network model,can effectively and qualitatively detect the residue of organophosphorus pesticide chlorpyrifos on the surface of cabbage leaves.(2)Pretreatment algorithm based on spectral curve graph feature segmentation and its application in the detection of organophosphorus pesticide residues and tomato polyphenol oxidase.The visible near infrared spectral data curve is an approximate smooth curve,which shows that the spectral data is constrained on the smooth curve,and there is a lot of data redundancy between the spectral data.Based on the spectral curve graph feature segmentation algorithm,the space surrounded by the spectral curve and the coordinate axis is regarded as the spectral graph,the spectral graph is divided into small spectral graphs on average or non average,the area,perimeter,area perimeter ratio and other features of the spectral graph are extracted,and the segmented spectral graph features are used for modeling,The modeling effect is evaluated by the root mean square error(RMSEC,RMSEP)and recognition accuracy(PRC,PRP)of the modeling set and prediction set,and the optimal modeling effect in different segmentation numbers,different graphic features and different modeling algorithm combinations is traversed and explored.The preprocessing algorithm based on spectral curve graph feature segmentation proposed in this paper is applied to the spectral reflectance data of cabbage leaves and tomato absorbance spectral data respectively.The preprocessing algorithm has good application effect.For the spectral reflectance data of cabbage leaves,the non average segmentation algorithm Nasr / mlr-21 based on the area perimeter ratio of spectral curve graph has the best modeling effect.The root mean square errors of modeling set and prediction set are0.198 and 0.278 respectively.The recognition accuracy of modeling set and prediction set are 77.50% and70% respectively.The running time of computer program of modeling algorithm is 3.47 seconds,Compared with the traditional partial least squares modeling algorithm,the root mean square error of modeling set and prediction set is 0.412 and 0.457,the recognition accuracy of modeling set and prediction set is 65.83% and 60%,and the running time is 54 seconds.Nasr / mlr-21 modeling method not only greatly improves the modeling effect,but also improves the operation efficiency of the algorithm and shortens the modeling time.For tomato spectral absorbance data,the feature average segmentation algorithm ASR / mlr-21 based on the area perimeter ratio of spectral curve graph has the best modeling effect.The root mean square errors of modeling set and prediction set are 1.74 and 1.99 respectively,and the correlation coefficients of modeling set and prediction set are 0.98 and 0.97 respectively.Compared with the direct application of partial least squares(PLS),the root mean square errors of modeling set and prediction set are 2.94 and 3.32 respectively,and the correlation coefficients of modeling set and prediction set are 0.92 and 0.90 respectively.ASR / mlr-21 algorithm greatly reduces RMSEC and RMSEP,effectively improves RC and RP,and the modeling algorithm improves the recognition accuracy of tomato polyphenol oxidase.(3)Detection of organophosphorus pesticide chlorpyrifos by screen printed electrode electrochemical biosensor based on bound hemoglobin and gold nanoparticles / molybdenum disulfide nano sheet chitosan modified.Gold nanoparticles(Au NPs)and molybdenum disulfide(Mo S2)nanosheets chitosan(chitosan)were used to modify the screen printed electrode(SPE)to improve the conductivity of the screen printed electrode and increase its specific surface area.Hemoglobin was fixed to the electrode surface by using the biocompatibility and adsorption of chitosan to make an electrochemical biosensor Au NPs / Mo S2 CS / BHb/ SPE.Because chlorpyrifos,an organophosphorus pesticide,can specifically bind to hemoglobin and form a thin film with poor conductivity on the surface of the biological electrode,which will hinder the electron transfer channel on the electrode surface and inhibit the redox peak current on the electrode surface.Using this mechanism,the linear relationship between chlorpyrifos,an organophosphorus pesticide,and the peak current inhibition ratio is established,It can realize the quantitative detection of organophosphorus pesticide chlorpyrifos.The experimental results show that the biosensor has a good linear range of 0.004 ~28.52 for the detection of chlorpyrifos μ M.The limit of detection(LOD)was 5.6 nm.Reliable experiments verify the repeatability,reproducibility,stability and anti-interference ability of the biosensor.The detection effect of chlorpyrifos pesticide residues in real cabbage leaves and leek leaves is good,and the actual detection recovery is 87% ~ 109%.The experimental results show that the biosensor has the advantages of high sensitivity,good stability,simple fabrication and low cost.The biosensor has potential application value and has important reference value for the electrochemical detection of organophosphorus pesticides.(4)The hardware circuit design of the proposed electrochemical biosensor is preliminarily studied and explored.The basic circuit of the electrochemical biosensor based on AT89C51 single chip microcomputer is designed by using Proteus development software,and the circuit simulation is carried out.The ultimate goal of the design and development of electrochemical biosensor is to design a fast,accurate,simple and portable test instrument.The electrochemical biosensor Au NPs / mos2-cs / BHb / SPE designed in this study mainly uses the square wave voltammetry(SWV)electrochemical test technology.In order to apply the SWV technology to the portable miniaturized test instrument,the SWV test hardware circuit is designed and realized by using AT89C51 micro single chip microcomputer,digital to analog converter,amplification circuit,analog-to-digital converter and display circuit,The circuit system can send out stepped square wave,namely SWV,and can adjust the parameters of SWV electrochemical test technology through key circuit,so as to collect current signal.The circuit system is simulated by Proteus Software,and the circuit system works well,which lays a foundation for the further development of electrochemical test micro system.