Model-Based Solvent Selection during Conceptual Process Design of a New Drug Manufacturing Process

• Published in: Organic process research & development Vol. 13; no. 4; pp. 690 - 697
• Main Authors: Hsieh, Daniel, Marchut, Alexander J, Wei, Chenkou, Zheng, Bin, Wang, Steve S. Y, Kiang, San
• Format: Journal Article
• Language: English
• Published: American Chemical Society 17.07.2009
• ISSN: 1083-6160; 1520-586X
• DOI: 10.1021/op900058e

Using models, we have demonstrated an efficient approach to identify optimal solvent compositions during conceptual design of an active pharmaceutical ingredient (API) process. A ternary solvent system was considered for a reaction, extraction, distillation, and crystallization sequence. Two thermodynamic models, NRTL-SAC and NRTL, as well as Aspen modeling tools, were employed to predict the liquid−liquid, vapor−liquid, and solid−liquid phase behaviors. We used these modeling tools to identify a solvent composition space for the reaction that allows for reasonable reaction volume while continuously removing a byproduct into a second aqueous phase. This composition also reduces API loss during subsequent aqueous extractions. Furthermore, the composition of the organic phase allows for an efficient azeotropic distillation during solvent exchange, resulting in a shorter cycle time needed to achieve the desired composition for final crystallization. Overall solvent usage for the process is also significantly reduced. This approach was applied retrospectively to a late-stage API process under experimental development and was validated with the production of API of excellent quality at the pilot scale with solvent compositions of the process in agreement with those predicted by the models.