Quantitative Structure Property Relationship And Its Application In Capillary Electrophoresis

The study of QSPR is to construct models between the structure of compounds and their biological activity or physical and chemical properties so as to develop new compounds and predict the properties of the compounds. It is a very active field in the application of computer in chemistry. Recently, capillary electrophoresis has been widely used in the field of life science, environmental science,food science,medicinal chemistry,forensic medicine and clinical diagnosis owing to its high efficient, rapidity and small amount of sample required. But compared with other analytical methods, there are more experimental parameters to be optimized in CE, which usually affect separation significantly. In many cases, it is difficult to find suitable experimental conditions for a given separation task. Prediction of separation conditions is not yet straightforward. So in this paper, QSPR and experimental design were applied in capillary electrophoresis to optimize the seperation conditions. The results indicated that the combination of experimental design and QSPR was found to be a powerful tool in predicting separationconditions in CE.There are seven parts in this dissertation, including a review on the background, research method and its application in analytical chemistry of quantitative structure property relationship (QSPR)and the application of experimental design in analytical chemistry; the application of experimental design and QSPR in the separation and determination the active components of Chinese traditional medicine and preparations by capillary electrophoresis (CE); determination of melamine in milk powder, liquid milk and fish feed by CE; and QSPR study on electrochemistry and optical properties of organic compounds by heuristic method (HM) and radial basis function neural network (RBFNN).In the first chapter, the background, research method and the application in analytical chemistry of QSPR and the application of experimental design in analytical chemistry were reviewed.In chapter 2, orthogonal design was used to the optimization of separation and determination of two active components aconitine and hypaconitine in traditional Chinese medicines by CE. The content of the aconitine and hypaconitine in Aconitum medicinal herbs were determined under the optimum separation conditions. In addition, a RBFNN with a"4-18-1"structure was developed based on the experimental results of orthogonal design and uniform design, and was applied to the prediction of peak resolution of the two active components under the optimum separation conditions given by orthogonal design. The predicted results were in good agreement with the experimental values, indicating that RBFNN is a potential way for the selection of separation conditions in CE.In chapter 3, a simple and fast capillary electrophoresis procedure with UV detection for the determination of matrine and oxymatrine in Kushen medicinal preparations was developed and optimized. Orthogonal design was used to optimize separation and determination conditions of the two active components. And the content of the active components were determined. In addition, multiple linear regression and a non-linear model using RBFNN approach were constructed for the prediction of the retention time of oxymatrine. The predicted results were in good agreement with the experimental values, indicating that RBFNN is a potential way for the prediction of separation time in CE.In chapter 4, a simple, rapid and low-cost capillary zone electrophoresis (CZE) method for analysis of melamine (MLE) in milk powder, liquid milk and fish feed has been proposed and compared with high performance liquid chromatography (HPLC). The method was successfully applied to the determination of melamine in milk powder, liquid milk and fish feed which has significant improvement than HPLC method prescribed in National Standards of People's Republic of China GB/T 22388-2008 and NY/T1372-2007. The performance of the CZE method evaluated in terms of precision, limits of detection, accuracy and quantification were comparable and in good agreement with those obtained with the HPLC method. Because of the shorter analysis time and low cost, CE is a more versatile and less expensive method for routine analysis of melamine than HPLC for a long term.In chapter 5, QSPR models were used to predict and explain binding constant to human serum albumin (HSA) (logK) of a variety of compounds determined by fluorescence quenching based on their structures alone. Stepwise multiple linear regression (MLR) and non-linear RBFNN were performed to build the models, respectively. The statistical parameter provided by the MLR model indicated satisfactory stability and predictive ability while the RBFNN predictive ability is somewhat superior. The proposed models were used to predict the binding constants of two bioactive components in traditional Chinese medicines (isoimperatorin and chrysophanol) whose experimental results were obtained in our laboratory and the predicted results are in good agreement with the experimental results. This QSPR approach can contribute to a better understanding of structural factors of the compounds responsible for drug-protein interactions, and can be useful in predicting the binding constants of other compounds.In chapter 6 and 7, a QSPR analysis has been conducted on the prediction of the half-wave potential (E1/2) and UV maximum absorption wavelength (λmax) of organic compounds by means of heuristic method (HM) and non-linear radial basis function neural network (RBFNN) modeling method. The statistical parameters provided by the two models indicated satisfactory stability and predictive ability. This QSPR approach can contribute to a better understanding of structural factors of the organic compounds influencing E1/2 andλmax , and can be useful in predicting E1/2 andλmax of other compounds.