Mathematical Simulation Of Multiphase Reactive Chromatographic Process

Multiphase reactive chromatography is a chromatographic process with multi-mobile phases and chemical reactions. The difference between multiphase reactive chromatography and multiphase chromatography is the existence of the term of chemical reactions, which results in the difficulties in theoretic analysis and numerical simulation since the first-order partial differential equations of the process become inhomogeneous. In this thesis, the mathematical theory and method of simulation for multiphase reactive chromatography have been researched, and two progresses have been achieved. For one hand, the related non-linear theory on chromatography or reactive chromatography with multi-mobile phase was developed. And for another, the equations of multiphase multicomponent reactive chromatography were set up on the second thoughts of the coupling effects of chemical reactions and multiphase chromatographic process. As to the principles of the combined flooding and the movement rules of chemicals, the research advanced multiphase multicomponent chromatographic model and simulated the process of surfactant-polymer flooding. The second-order TVD scheme with flux limiter was applied to the analysis of a four-component reversible reaction, two-fluid-phase chromatographic process and a three-component cyclic reaction, two-fluid-phase chromatographic process. The analysis indicated that the flow sequence of multiphase reactive chromatography was codetermined by the adsorption on the stationary phase, the partition between the mobile phases and the saturation. When the coefficients of adsorption and partition were constant, the alternation of phase saturation would change the flow sequences of chromatography and enable the objective component to penetrate through the shocks of other components. Furthermore, the thesis also developed the moving boundary methods and proposed substituting the trial function with the orthogonal collocation method. The research results indicated that the method of orthogonal collocation was more effective in simulating the concentration distributions than the method of the trial function and its computation ran much faster than the second-order TVD scheme.