Doctoral Dissertations
Date of Award
12-1996
Degree Type
Dissertation
Degree Name
Doctor of Philosophy
Major
Chemical Engineering
Major Professor
Georges A. Guiochon
Committee Members
Paul R. Bienkowski, C. H. Byers, George C. Frazier
Abstract
Although there have been constant efforts to investigate simulated-moving bed technique over the past two decades, most of the works are based on truly moving bed, the countercurrent equivalency of simulated-moving bed. Only in a few works, the SMB was treated as a periodic process, but they were all based on numerical solution of different models. For a long time, the ideal model of linear chromatography, from which a set of simple solution for SMB can be possibly derived, has been ignored.
In this dissertation, the fundamental aspect of the Simulated Moving Bed chromatography (Sorbex type) has been investigated both theoretically and experimentally. The linear ideal and equilibrium-dispersive models of chromatography were solved and compared with the experimental results. The agreement is excellent. The results of ideal model is a set of algebraic equations. The equilibrium-dispersive model is demonstrated to be a solid tool for further design and optimization work.
It has been confirmed, in order to have a good separation under linear condition, the flow rates in each section and the switching time can be estimated provide (1) the retention of the two components are known; (2) the separation condition in each section satisfies the same safety margin; (3) the column efficiencies are moderate.
The performance of SMB depends not only on the chromatographic behavior of the two components but also on the configuration of SMB and the switching of the ports. Two different configurations were studied in this work. For the first time, we showed some oscillations observed in numerical simulation are caused by the periodic nature of SMB separation, not by numerical oscillation. A simple way to estimate the number of switching required to reach steady state is derived from ideal model solution, since the rate of SMB to reach steady state is independent on column efficiency, so this result can certainly be applied to column with any column efficiency.
Furthermore this theory is extended to systems which have different safety margins in different sections. The regions for any possible safety factor are discussed, there are no limit for safety factors in section I and IV, the options to have mobile phase recycling and stationary phase recycling are included. The region for all possible safety factors in section II and III is also discussed. Since most of the separation is done in these two sections, it gives the possibility to run SMB under different conditions to satisfy different production requirements. A simple optimization strategy under linear condition is presented, as long as the purity of the product is satisfied, the more close the values of safety factors to 1, the better because the condition allows more recycling, more feed input and the higher product concentration.
Recommended Citation
Yun, Tong, "A fundamental study of simulated moving bed chromatography under linear condition. " PhD diss., University of Tennessee, 1996.
https://trace.tennessee.edu/utk_graddiss/9892