Technoeconomic Mixed Integer Nonlinear Programming (MINLP) optimization for design of Liquid-Liquid Extraction (LLE) cascades in continuous pharmaceutical manufacturing of atropine

Continuous Pharmaceutical Manufacturing (CPM) offers operational and economic benefits over the currently dominant methods implemented by industry. The demonstrated continuous flow synthesis of atropine facilitates modeling and optimization of its upstream continuous manufacturing. This study implements MINLP optimization of a continuous Liquid-Liquid Extraction (LLE) flowsheet superstructure for total cost minimization of an upstream atropine CPM plant. The steady-state process model considers reactor design from regressed kinetic parameters including investigation of maximum allowable PFR reactor dimensions to maximize the benefits of API flow synthesis. Continuous LLE modeling considers solute partitioning between phases, UNIFAC-modeled liquid–liquid equilibria and mass transfer correlations for LLE design. Optimal LLE design configurations from the considered superstructure require fresh solvent added to the first vessel in co-current flow in addition to the countercurrent flow in the cascade. Optimization results indicate toluene as the best solvent choice, with total costs of 3.944 × 10⁶ GBP for a plant capacity of 10³ kg API year⁻¹.