Determination Of Complex Mixed Solution Based On Fluorescence Spectroscopy Mechanism And Chemometrics

The detection of complex mixed solutions involves many fields such as medicine,agriculture,environment,food and so on.With the development of the economy,the increasing pollution of the water environment and the frequent occurrence of food safety incidents have caused great threats to people’s health and lives,and have become two hot issues of global concern.However,as the field of complex solution detection,water environment pollution and food safety,the analysis systems are complicated and often contain a mixture of multiple components,and even some unknown interferers exist in the system,making the determination difficult.Fluorescence spectroscopy,as an important photoelectric detection technology,can provide more complete spectral information,obtain rich physical and chemical information,and has the unique advantages of simplicity,speed,non-destructiveness,and high sensitivity.However,the current fluorescence spectroscopy analysis method has poor anti-interference ability to actual water environment and food complex solutions,and has weak ability to distinguish substances with serious overlapping fluorescence spectra.In addition,due to the recorded data is high-way and complex,it is difficult to qualitatively identify adulterated food.To this end,the three-way calibration algorithms were used to analyze the fluorescence spectrum data,and "mathematical separation" was used to qualitative and quantitative analysis of multiple pesticide residue in the water environment and food complex solutions.Then,chemical high-dimensional pattern recognition algorithms were established to analyze food-adulterated complex solutions,and the meaningful chemical composition were characterized and adulterated samples were identified.The main research work of this paper is as follows:(1)Based on quantum chemistry calculation methods,the mechanism analysis and calculation of the fluorescence spectra for carbendazim,chlorothalonil,carbaryl and tsumacide pesticide molecules were studied.The geometric structure parameters such as the bond length,bond angle and dihedral angle of each molecule was analyzed.The relationship between molecular fluorescence with the highest occupied orbital and the lowest unoccupied orbital in the front-line molecular orbitals were studied.The theoretical value of the fluorescence wavelength of each molecule transition from the excited state to the ground state was calculated.(2)The strategy that combines the excitation-emission matrix fluorescence spectrum with the parallel factor analysis and the unfold partial least squares coupling residual bilinearization was studied,and realized the simultaneous qualitative and quantitative analysis of carbendazim,chlorothalonil,carbaryl and tsumacide pesticide residues in the complex water environment.The Raman and Rayleigh scattering were eliminated using the new Delaunay triangulation method to achieve spectral preprocessing.In view of the highly overlapping fluorescence spectra of the four pesticides and the presence of unknown fluorescent interferences in the water environment,this research strategy used "mathematical separation" to achieve simultaneous quantitative analysis of multiple pesticide residues in complex mixed solution.The analytical figures of merits such as root mean squared error of prediction,sensitivity,limit of detection and limit of quantification were evaluated.This study provides a new idea for the online monitoring of trace organic pollutants in complex solution of the water environment.(3)The strategy that combines excitation-emission matrix fluorescence spectroscopy with a variety of iterative three-way calibration algorithms was studied to achieve simultaneous determination of bactericidal pesticide residues with overlapping spectra in peanut oil.From the perspective of practical application,the research compared the performance of several iterative three-way calibration algorithms such as parallel factor analysis,alternating trilinear decomposition,self-weighted alternating trilinear decomposition and alternating penalty trilinear decomposition in the selection of component number,convergence speed and quantitative analysis.The analytical figures of merits such as sensitivity,limit of detection and limit of quantification were evaluated.This strategy is of great significance for the detection of multiple hazardous substances in food solutions.(4)Combining the three-way calibration method alternating trilinear decomposition with chemical pattern recognition algorithms,the chemical high-dimensional pattern recognition methods such as alternating trilinear decomposition-linear discriminant analysis,alternating trilinear decomposition-partial least squares discriminant analysis,alternating trilinear decomposition-support vector machine and unfolded partial least squares discriminant analysis were established.They were used to analyze three-dimensional fluorescence data array of adulterated sesame oil.The meaningful chemical components in edible oil were characterized and the adulterated samples were classified.The classification accuracy,sensitivity,specificity and accuracy were evaluated.This analysis strategy is expected to become a powerful tool to prevent potential adulteration of food.