Multiobjective Optimization of Unseeded and Seeded Batch Cooling Crystallization Processes

Abstract

Multiobjective optimization (MOO) of batch cooling crystallization is carried out for the development of optimal operating recipes for unseeded and seeded crystallization processes. Mean size and coefficient of variation (CV) are the two objectives considered for unseeded batch cooling crystallization of paracetamol, whereas mean size, CV, and nucleated mass are considered as objectives for the seeded batch cooling crystallization of potassium nitrate. In this work, along with finding the optimal temperature trajectories, the effect of choice of objectives on the final achievable Pareto front is analyzed using two different objective functions, namely number mean size and surface-weighted mean size. Further, the capability of MOO is also exploited to determine the feasible ranges of combinations of seed properties like initial seed mass and seed mean size for better crystal size distribution (CSD).