Studies On The Fundamental Theories, Methodology And Application Of Capillary Electrophoresis

The capillary electrophoresis technique has been attracting people's more and more attention all over the world because it is brief and quick and has the advantages of high column effect, and low experimental consumption, and its applications have been further developed. The doctoral dissertation has made deep studies on the fundamental theories of CE, the quantitive methodology of CE, the separation of enantiomers of drugs using CE, the CE figureprints of traditional Chinese medicines and the application of the miniaturized total chemical analysis system( TAS) to drug analysis. The dissertation is composed of six chapters.In the first chapter, a review was provided on the development and the various models of CE. A brief introduction to the nonaqueous capillary electrophoresis (NACE) and the affinity capillary electrophoresis (ACE) was made. The various factors which influenced the velocity of the electroosmotic fluid was discussed, meanwhile, the related factor figure was set up. Both the CE-MS technique and the concise history of the miniaturized total chemical analysis system( TAS) were briefly reviewed.Chapter 2 firstly studied the energy changing rules of solutes in the capillary zone electrophoresis (CZE) by using the thermodynamic theory. A series of energy concepts about the solutes in CZE was set up, in which the separation free energy â–³ Gsep, mixed free energy â–³ Gmix, total kinetic energy variation â–³ T, total potential energy variation â–³ U, net increasing energy of solutes â–³ E, apparent working time teff, exhausted energy per peak W0 and the share of the net increasing energy were all included. The influences of the concentrations of the background electrolytes (BGE) and voltage applied to them were investigated by experiments. The conclusion is: Wo increased with the increasing of concentration of BGE or voltage, but T and teff decreased. â–³ T decreased as BGE concentration was increased, but increased as the voltage was applied highly. Under both conditions, â–³ Gsep and A U remained stable. Because â–³ T and â–³ U were much less thanE , they almost could be ignored. The separation free energy Gsep and the mixed free energy Gmix played important roles in the contribution to the net increasing energy of solutes. If we could have gained a better separation in CE, we should have selected both appropriate BGE concentration and appropriate voltage to perform the CE experiments. Secondly, the chapter was based on the Bronsted (proton) theory to regard the =Si-OH group on the interior wall of the fused silica capillary that immersed in the electrolyte solution as a weak acid, for it took part in the reversible proton transfer reaction. Since the equilibrium between the donating proton and accepting proton proceeded forever, the =Si-O- groups on the interior wall of fused silica capillary were definite under a constant acidity condition. We theoretically evaluated the approximate number of =Si-O- groups on the interior wall of fused silica capillary in a definite pH, which was benefited for understanding the concept of the electroosmotic flow (EOF). Thirdly, the chapter had studied the changing behavior of the solutes migration time in the capillary electrophoresis, the voltage applied varied from 2 kV to 30 kV and the ionic strength of the background electrolyte was increased from 10 mM to 100 mMo According to the experimental results, we put forward 4 empirical formulas as follows: lgtmig=a'-b'V , lgtmig= a" + b"I, lg/mig=b0- b1 V+ b2I l/tmig = b'0 -b'1V + b'2I, each of which was of good linearity. Meanwhile, we used the artificial neural network (ANN) modeling method to study the regularities that the migration time changed with the variations of both the voltage and the ionic strength. The 350 migration time obtained under different conditions for seven drugs were successfully predicted, in which both the voltage and the ionic strength were simultaneously included by the back-propagation (BP) ANN, moreover the algorithm of the ANN model was given. A three-layer net, namely a net conta...