(287a) Model Based Optimization of Cooling Crystallization of Active Pharmaceutical Ingredients Undergoing Thermal Degradation
Crystallization is one of the most important separation processes in the pharmaceutical, food and chemical industries. In the pharmaceutical industry, it serves as both a separation unit operation to produce high-purity crystalline products and as a particulate design process. Cooling crystallization is the most widely used crystallization strategy to purify a new drug molecule under development. However, many drugs degrade at high process temperatures which leads to reduced product yield. Reduction of the initial process temperature decreases the rate of degradation, but simultaneously decreases the process yield. In this work a model-based optimization problem is formulated for unseeded cooling crystallization of Aspirin, an industrially relevant API. Two different optimization objectives are considered in this study viz. maximization of product crystal size and minimization of degradation product concentration, which give widely different optimal cooling profiles. Investigation of the fundamental mechanisms viz. particulate nucleation and growth reveals major differences between the two optimal scenarios. Experiments are also carried out in the laboratory using online Process Analytical Technology tools viz. Focused Beam Reflectance Measurement, PVM and online Ultra-High Performance Liquid Chromatography to validate the model simulations. It is found that there is substantial plant-model mismatch for one of the optimal scenarios, which necessitated model reformulation to include the adsorption of degradation impurity onto the product crystals.