(133d) Optimization-Based Synthesis, Design and Evaluation of Process Concepts for Production of Pure Enantiomers

The two enantiomers of a substance are stereoisomers that differ only in their spatial structure. They have virtually identical physico-chemical properties, but often very different physiological effects. Thus, pure enantiomers are of vital economic relevance in the pharmaceutical industry. In corresponding productions, it is often more economic to synthesize the 1:1 mixture of both forms instead of the pure enantiomer. This makes necessary typically expensive subsequent separation step(s). Generally, several competing unit operations and process alternatives exist. Due to their relevance and their inherent complexity, focus is here on process concepts that utilize chromatographic operating modes like simulated moving bed (SMB) and steady-state recycling (SSR) chromatography.

In this work, we apply optimization-based process synthesis to address two questions. The first is to identify the most cost-effective production concept for a pure enantiomer based upon mathematical models of established unit operations and detailed cost functions. Nonlinear programming (NLP) is used to find optimal operating conditions for stand-alone SMB processes and known process combinations (i.e., hybrid separations). Furthermore, mixed-integer nonlinear programming (MINLP) problems are solved to optimize the structure of flow sheets that contain several recycles and additional unit operations like isomerization, crystallization, nanofiltration, and distillation. The second goal is to develop new advanced process concepts. An interesting example is a fully integrated SMB-reactor with an internal distribution of an isomerization reaction. The process increases the overall yield of enantiomer from 50 to 100% at superior performance in comparison to classical approaches. It was identified solely on the basis of structural optimization [1]. A practically feasible implementation using a ?reactive gradient? was proposed and successfully validated experimentally.

This work is supported by the European Commission under the collaborative research project ?IntEnant? (FP7-NMP2-SL-2008-214129).

[1] J. García Palacios, M. Kaspereit, A. Kienle, Chem Eng. Technol. 32 (2009) 1392