Standing wave design and optimization of a tandem size-exclusion simulated moving bed process for high-throughput recovery of neoagarohexaose from neoagarooligosaccharides

Abstract

A previous research confirmed the feasibility of using a size-exclusion tandem simulated-moving-bed (SMB) process in realizing the high-purity separation of neoagarohexaose from neoagarooligosaccharides (NAO) mixture. To promote the industrial application of such tandem SMB (named as "NAO-SMB" in this article) comprising two subordinate SMB units (Ring I and Ring II), we aimed at accomplishing its comprehensive optimization, which was carried out on the basis of a standing-wave- design method. It was found first that in a conventional frame where the same column length was set to be used in Ring I and Ring II under the column configuration of 2-2-2-2 in each ring, the separation sequence based on removing a larger-size impurity in Ring I and then smaller-size impurities in Ring II (S1 mode) led to higher throughput than the opposite separation sequence (S2 mode). By contrast, if the column lengths of the two rings were allowed to be set differently, the bed utilization of the S2 mode could increase more than that of the S1 mode, thereby causing the S2 mode to surpass the S1 mode in throughput. In addition, if the column configuration of each ring was allowed to be varied, the level of attainable throughput could be increased by allotting more columns in the enrichment zones containing the front and rear of product solute band in Ring I. Furthermore, it was confirmed that the simultaneous application of the aforementioned two methods (i.e., using different column lengths in two rings and placing more columns in the enrichment zones of Ring I) had an additional upward effect on each other. As a result, the optimized NAO-SMB under such frame, whose column length and configuration were 13.33 cm and 2-4-3-3 in Ring I and 18.67 cm and 2-3-5-2 in Ring II, could achieve 46% higher throughput than that under the conventional frame, and further lead to 133% higher throughput than the initial NAO-SMB reported in a previous study.