Robust Pinched-Wave Design of a Size-Exclusion Simulated Moving-Bed Process for Insulin Purification

Abstract

A tandem simulated moving bed (SMB) has been developed previously to purify insulin from two impurities in two sequential steps. In this study, rate-model simulations are used to investigate how variations in operating and system parameters affect the yield and purity of the SMB. The parameters include the pump outputs, step time, bed voidage, size-exclusion factor, and extra-column dead volume. The results show that a robust pump configuration can minimize the effects of pump output deviations. The yield and purity of the tandem SMB are affected the most by the fluctuations of the zone II and IV pumps. In particular, a low-selectivity system is more sensitive to the variations in the operating parameters than a high-selectivity system. If all of the parameter deviations occur simultaneously, the standing-wave design has no sufficient margin to maintain the desired yield and purity. We have developed a new design method such that the standing waves are replaced by pinched waves. Rate-model simulations show that the design is robust. Even when the worst parameter deviations occur, the desired yield and purity can still be achieved. This design method is general and can be applied to other size-exclusion or linear isotherm systems.