Digital Design of the Crystallization of an Active Pharmaceutical Ingredient Using a Population Balance Model with a Novel Size Dependent Growth Rate Expression. From Development of a Digital Twin to In Silico Optimization and Experimental Validation

• DOI: 10.1021/acs.cgd.1c01108
• Publication Date: 2021-12-20T13:51:07Z
• Authors: Botond Szilágyi; Ayşe Eren; Justin L. Quon; Charles D. Papage...; Zoltán K. Nagy
• Keywords: Process Analytical Technology; Quality by Design; Critical quality attributes; Process modeling

Prediction and control of the product properties in crystallization processes are practical challenges pharmaceutical industry. Effective process design operation techniques needed to meet critical quality attributes (CQAs) minimize batch-to-batch variation. Mathematical modeling can enhance understanding save a considerable amount time, effort raw material when used development following guidelines Quality-by-Design (QbD) framework. When mathematical model is fitted validated with experimental data, it provides digital twin that enables execution silico experiments (DoEs), which particularly beneficial number factors increases or if expensive sparingly available, e.g., during early stage development. This work presents benefits by studying an active ingredient (API) from Takeda Pharmaceuticals International Co., referred hereafter as Compound A. A framework for construction, parameter estimation validation demonstrated through case study using population balance (PBM) approach. Secondary nucleation, size dependent growth (SDG), dissolution mechanisms considered. Size dependency introduced new formulation capturing considerably slower small crystals (D90 < 10 μm) while having larger crystal domain > 200 similar models literature. To make more industrially relevant, novel method developed recently authors applied use focused beam reflectance measurement (FBRM) data directly without further transformations. The kinetic parameters estimated minimizing difference between measured simulated concentrations, distributions (CSDs) maximizing correlation density FBRM counts. paper also illustrates SDG rate expression capture CSD dynamics better than standard models. DoE optimization, simulation results experimentally, demonstrating model-based