Optimal design of simulated moving bed chromatography for chiral separation
Abstract
Simulated moving bed (SMB) chromatography has received attention for chiral separations since early 1990s and is, today, considered as a cost-effective preparativescale purification technology. In this study, a design method for chiral SMB separation was developed and verified by both simulations and experiments. The verified design method was further used for optimization and for understanding the effects of various parameters on SMB performance. An improved standing wave design (SWD) method was developed for nonlinear isotherm systems with significant mass transfer effects and an operating pressure limit. The design method was verified with rate model simulations and then tested for enantioseparation of phenylpropanolamine. Both high purity (>99%) and high yield (>99%) were achieved experimentally using an SMB with a pressure limit of 2.4 MPa. Standing wave analysis was used to systematically examine the effects of purity and yield requirements, system parameters, and material properties on productivity and desorbent requirement, with and without a pressure limit. The system parameters studied include particle size, column length, interparticle void fraction, total number of columns, column configuration, extra-column dead volume, and feed concentration. The material properties studied include saturation capacity, adsorption equilibrium constants, intraparticle void fraction, and intraparticle diffusivity. The standing wave analysis shows explicitly the effects of each parameter. An efficient optimization tool was developed based on SWD using a grid search method. Both system parameters (particle size, column length, column diameter, total number of columns, column configuration, and feed concentration) and operating parameters (zone flow rates and switching time) were optimized to achieve the maximum productivity or the minimum separation cost. High-pressure SMB systems (5.2 MPa) can have higher productivity but low- and medium-pressure SMB systems (1.0 and 2.4 MPa, respectively) are more economical. A new optimization technique based on a genetic algorithm was incorporated with SWD. The new optimization method can reduce computation time by at least two orders of magnitude compared to the grid search method, for the simultaneous optimization of six system parameters and five operating parameters. The new optimization method was used for multi-objective optimization to generate Pareto optimal solutions.
Degree
Ph.D.
Advisors
Wang, Purdue University.
Subject Area
Chemical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.