The Research Of The Mechanism Of Multifrequency Large Amplitude Pulse Voltammetry On Electronic Tongue

Multifrequency large amplitude pulse voltammetry, developed by our group, is a new electrochemical analysis method for excitation and collection signals of electronic tongue. Based on it, we developed a novel electronic tongue, so called multifrequency large amplitude voltammetric electronic tongue. This work focus on how to find some equations to explain the response of multifrequency large amplitude pulse voltammetry. The single potential pulse response of multifrequency large amplitude voltammetry is divided into three segments, and three transient electrochemical methods, i.e. solution equivalent circuit analysis, diffusion control analysis and mixed-controll analysis, are used to explain the three divided segments respectively. Also, the response of multifrequency large amplitude voltammetry is integrated by time to meet the principle of chronocoulometry. The results show that the whole responses of multifrequency large amplitude voltammetry can be explained by these four transient electrochemical methods, i.e. solution equivalent circuit analysis, diffusion control analysis, mixed-controll analysis and chronocoulometry. Then a new electrochemical model, so called electrochemical combination model, is build by the equations of those four transient electrochemical methods and the division methods of response of multifrequency large amplitude voltammetry. The new model can be used to calculate the electrochemical parameters of the analysis solutions. Three dairy products are used to evaluate the effective of the new electrochemical model. The response of multifrequency large amplitude voltammetry from the three dairy products, i.e. CC Nutri-Express, BB pure milk and Yedao coconut milk, with different shelf time, are fitted by the equations of electrochemical combination model. The parameters from electrochemical combination model are used to distinguish the quality of dairy products with different shelf time. The original eigenvalues of multifrequency large amplitude voltammetry are used as reference. The works get the following conclutions.(1) The research of solution equivalent circuit shows that the equations of solution equivalent circuit model could applicate in calculation of Rs, Rct and Ï„ of multifrequency large amplitude voltammetry. The sampling rate should be20Kpoints/s. The multifrequency large amplitude pulse voltammetry from all six sensors of sensors array of smartongue can be used for calculation of Rs and the best one is from platinum electrode of1Hz segment. The iteration of Itâ†'∞of multifrequency large amplitude pulse voltammetry is effective for calculation of Rct and Ï„. The optimized sensor and frequency is silver1Hz and10Hz segments.(2) The research of diffusion control shows that the data from the last5%of response of potential pulse can meet the equations of diffusion control model. But all the data should be preprocessed. Three data preprocessing methods, i.e., Savitzky-Golay fitting method, polynomial fitting method and a notch filter, are tried in the experiments. The polynomial fitting exhibits the best results and the data proprecessed by polynomial fitting can get the good linear relationship from I and t-1/2. The research gets another result that the data model will be deteriorated by the frequency increases. No sensors of smartongue can meet the diffusion control model of0.01mol/L glucose and100%alcohol by the10Hz and100Hz of multifrequency large amplitude pulse voltammetry. The optimized data for diffusion control model is from platinum electrode and1Hz segment.(3) The research of hybride control shows that the data of multifrequency large amplitude pulse voltammetry from the1/2Imax to3/4Imax can meet hybride control model well and also the data need to be preprocessed before fitting the model. Three data preprocessing methods, i.e., Savitzky-Golay fitting method, polynomial fitting method and a notch filter, are tried and polynomial fitting method gets the best results as same as diffusion control model fitting. In this research, the difference of reversible reaction and non-reversible in the hybride control model is discussed. The two equations from the reversible reaction and non-reversible reaction respectively are merged into one equation. The calculation results show that1Hz of multifrequency large amplitude pulse voltammetry from all the sensors except tungsten electrode can meet hybride control model. But the data model also will be deteriorated by the frequency increases. No sensors of smartongue can meet the hybride control model of all five characteristic solutions by the10Hz and100Hz of multifrequency large amplitude pulse voltammetry.(4) The research of chronocoulometry shows that the discret integral curve from multifrequency large amplitude pulse voltammetry can be used for chronocoulometry model fitting. The response from pulse potential and removal potential of multifrequency large amplitude pulse voltammetry fit two equations respectively. The results show that it is difficult for fitting the chronocoulometry model by the signal response from week electrolyte solutions,0.01mol/L glucose and100%alcohol. No sensors of smartongue are suit for analysis of0.01mol/L glucose and100%alcohol based on chronocoulometry model. But for strong electrolyte solutions, the response from titanium electrode and1Hz of multifrequency large amplitude pulse voltammetry is best choice.(5) The research builds the electrochemical combination model by the equations from four transient electrochemical methods, i.e. solution equivalent circuit, diffusion control, mixed-control and chronocoulometry. The electrochemical combination model is used to explain the response of multifrequency large amplitude pulse voltammetry. Three dairy products, i.e. CC Nutri-Express, BB pure milk and Yedao coconut are used to evaluate the effective of the model. The parameters of resoponse of multifrequency large amplitude pulse voltammetry of three dairy products with different shelf time are calculated by the equations of the electrochemical model. The original eigenvalues of multifrequency large amplitude pulse voltammetry is used as reference to be campared. The results show that the electrochemical model can explain the quality change of products well. The score plots of three dairy products based on principal component analysis either can exhibit the change directions of the samples following their shelf time, or can predict the shelf time of unknown samples. The relative predicts errors are21.12%,28.18%and25.34%for CC Nutri-Express, BB pure milk and Yedao coconut milk, respectively. Although the relative errors are more than5%, the results can meet the requirements of food industry, since the shelf time is a comprehensive indicator and the samples has no preprocessing before analysis.The new constructed electrochemical model can be used as basic theoretics for future research of multifrequency large amplitude pulse voltammetry.