Research On The Detection Method Of Pesticide Based On The Surface Plasmon Resonance Sensor And Spectral Absorption

Pesticide can effectively benefit the production of agricultural and animal husbandry very much,but its residue in agricultural products is a serious threat to people's health.To ensure that the pesticide does not pass through vegetables,fruits and other ways into the human body circulation,an effective detection of it is important.In this paper,we have studied the pesticide residue detection methods from the aspects of biosensor surface plasmon resonance(SPR)sensor and enzyme suppression spectral absorption characteristics.Firstly,we propose a matrix measurement scheme for the detection of temperature characteristic and refractive index(RI).And a double incident angle measurement method has been proposed to monitor the change of RI for the angle-scan SPR sensor.In addition,a dual-incident angle method and the dual-wavelength technique have been approved for the prism-based and fiber-optic SPR sensor,respectively.We also carry out the research of the quantitative detection model of pesticide concentration basing on the principle of enzyme inhibition and the method of single-sensor data fusion.The main work of this paper can be divided into five parts:The first part:mathematical model of SPR has been approved.According to the Maxwell equations and their boundary models propagating in the mediums,the process and resonance conditions of the photo-excited surface plasmon resonance are simulated.The reflectivity formula of Kretschmann-SPR has been deduced by combining the total reflection condition and the multilayer reflection.The second part:we design a simple angle scanning prism-SPR sensor system.The design of reflex parallel optical path simplifies the system structure,including the light source and intensity sensor CCD,parallel light path structure and temperature control module.With a motor scanning incident angle and temperature control,a matrix method has been proposed to measure the glucose mass concentration.A double incident angle measurement is applied with the angle-scan model SPR,which makes the angle-scan SPR monitoring the change of refractive index(molecular interaction)real time.Furthermore,with the detail analyses of the phenomenon caused by the improper selection of incidence angles,we propose the selection principle of the angles.The third part:combining SPR mathematical model,prism dispersion and metal Drude free electron model,the theoretical analysis model of temperature characteristic is constructed.A cross-sensitivity matrix method is proposed to simultaneously measure RI and temperature for the prism-based SPR sensor.We also analyze the error of the cross sensitivity matrix for the angle-interrogation mode and the wavelength-interrogation mode SPR respectively,and propose a piecewise linear correction method to improve the accuracy.A dual-incident measurement method is proposed for the prism SPR.The theoretical basis is analyzed in detail.And a proof-of-concept experiment is carried out to demonstrate the ability of the method.We have also theoretically analyzed the temperature effects on the performance of fiber-optic sensor based on SPR.Both of the thermo-optic effect in the fiber core and phonon-electron scattering along with electron-electron scattering in the metal layer are seriously considered in our theoretical model.A new dual-wavelength technique is described to simultaneously measure the changes in RI and temperature.As the technique taking full use of the signals at two fixed wavelengths without need of extra design of the sensor probe,it is practical and convenient.The fourth part:the optical detection platform of organic phosphorus pesticide residue spectral absorption based on enzyme inhibition principle is constructed for collecting the spectral data of the mixed solution of equal gradient concentration of O,O-dimethyl-O-(2,2-dichlorovinyl)phosphate(dichlorvos)and enzyme reagent,and the study of spectral absorption model is carried out.Monte Carlo Cross Verification(MCCV)is used to remove exception samples from the sample set.The sample set is divided into modeling set and validation set by concentration gradient method.Then,the single or combination of first derivative,standardized normal variate(SNV)and principal component analysis(PCA)are applied to process the modeling set and validation set.Finally,the radial basis function(RBF)neural network and partial least squares(PLS)model are established,respectively.The results show that spectral model based on the principle of enzyme inhibition can accurately predict the sample concentration,and the model built only with 400-430 nm band spectral data also can achieve good prediction results,indicating that there is a better correlation between the 400-430 nm band information and sample concentration.The fifth part:we introduce an algorithm derivation,device design and field experiment of a pesticide residue detection device based on single sensor data fusion.Group the data of single sensor and deduce the calculation method of data fusion.And the suppression rate calculation method without measuring the light source value is explained.In this way,the data redundancy is reduced,and the authenticity and reliability have been enhanced.