2025 AIChE Annual Meeting

(363c) Hydrogenation Process Development in a Data-Rich Environment

Authors

Darren Whitaker, Takeda Pharmaceutical
Neda Nazemifard, University of Alberta
Charles Papageorgiou, Takeda Pharmaceuticals International Co.
Allan Myerson, Massachusetts Institute of Technology
Hydrogenation reactions play a critical role in the synthesis of active pharmaceutical ingredients (APIs) in the pharmaceutical industry. Traditionally, these reactions are developed in small-scale batch vessels, but the achievable pressure and temperature ranges in batch vessels are limited because of equipment constraints. Scaling up reaction conditions presents a challenge as mass transfer coefficients vary between batch vessels, influenced by differences in geometry and size. Continuous hydrogenation processes overcome these limitations, but are still constrained by the solubility of the starting material and often require large quantities of materials and catalysts for development. However, continuous flow systems offer enhanced safety profiles compared to batch processes due to precise dosing of hydrogen equivalents, reduced reaction volumes, and improved heat transfer. The tight control over critical process parameters often lead to improved quality and impurity profiles of the obtain product. The transferability between batch conditions and continuous flow conditions remains a significant challenge for the industry.

Herein, we present a workflow to accelerate process development for hydrogenation, scaling from milligram batch-scale to multi-gram continuous flow, by leveraging a data-rich environment and process analytical technology. To accelerate process development, we first investigate the desired chemistry (hydrogenation of a nitro group) through small-scale, high-throughput batch experiments to identify the reaction conditions for the preferred solvent, catalyst type, stability, and concentration. The chemistry is then transferred to continuous flow in a well characterized and fully automated packed bed continuous flow reactor. The process is optimized with an automated Design of Experiment approach for conditions that were initially limited by batch constraints. Dynamic flow ramps offer valuable insights into the mass transfer within the packed-bed reactor. The process stability was demonstrated through a continuous run lasting over 9 hours, producing more than 150g of the desired amino product.