Robust tandem simulated moving bed for insulin purification

Sungyong Mun, Purdue University

Abstract

Large-scale size exclusion chromatography (SEC) is used to purify insulin from high molecule weight proteins and zinc chloride before crystallization in an existing process of Eli Lilly and Co. A system of two four-zone simulated moving bed (SMB) units is invented to replace the existing batch SEC in order to increase yield and throughput and reduce eluent consumption. A systematic approach based on the standing wave design, rate model simulations, and experiments was used to develop the tandem SMB process. Compared to the current batch SEC process, the tandem SMB has 10% higher yield, 400% higher throughput, and 72% lower eluent consumption. Efficient startup and shutdown strategies were developed to reduce the unsteady state period that has undesirable effects on the SMB performance. The startup and shutdown strategies are proved with results from computer simulations and validated with experimental data. An efficient optimization tool based on the standing wave design was developed to find the optimal tandem SMB for insulin purification. Both system and operating parameters were optimized to achieve the lowest purification cost. Variations in parameters can affect the yield and purity of the SMB. To maintain high yield and high purity against a given set of parameter deviations, a new design method was developed. In this design, the four concentration waves are pinched waves under normal conditions and become standing waves when the worst parameter deviations occur. For a pharmaceutical like insulin, the product should be clearly linked to its source. In addition, long residence time of a protein in SMB can result in aggregation or denaturation. To shorten the residence time of insulin and maintain complete batch integrity, either a partial feeding strategy or non-standing wave operating conditions can be used in combination with short zone lengths and a minimal eluent gap between feed batches. All the methods developed in this study are general and can be applied to other size-exclusion or linear isotherm systems.

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
.

Share

COinS