Green Chemistry-Oriented Multiobjective Optimization with Semiautomated Process Monitoring for Apixaban Intermediate Synthesis

Optimizing synthetic processes for active pharmaceutical ingredients is essential for improving the economic viability, safety, and environmental performance. This study presents a dynamic closed-loop framework integrating Bayesian optimization (BO), real-time spectroscopic monitoring, and automated execution to optimize the synthesis of apixaban intermediate 3-morpholino-1-(4-nitrophenyl)-5,6-dihydropyridin-2(1H)-one. Reaction parameters were systematically screened and refined using multiobjective BO, targeting high conversion, low reagent usage, and minimized environmental burden. The ECP100 automation system enabled precise control of temperature-sensitive operations such as exothermic chlorination, improving the dosing efficiency and safety. Across key steps, reagent cost was reduced by 27%, electricity consumption was reduced by 31%, and process mass intensity was reduced by 54%, while side-product content was reduced by 98%. Green chemistry metrics improved by 51% overall. Moreover, equipment availability increased by 71%, enabling high-value task substitution and yielding a 406% increase in potential value output based on an opportunity cost estimate. These results validate a scalable, cost-effective, and environmentally aligned strategy for green process development in pharmaceutical manufacturing.