(614c) Rapid Development of Population Balance Models Using an Automated Crystallization System Integrated with the User-Friendly Modeling Tool Crysiv

Crystallization is an essential process that has been extensively applied to separate and purify pharmaceuticals, agrochemicals, and other compounds for centuries in the chemical industry. The process has gained distinguished attention in the pharmaceutical industry since, by carefully setting the crystallization parameters, not only the downstream processes (e.g., drying, tableting) but also the quality of the product (e.g., bioavailability) can be improved. Despite primarily being a simple procedure, industrial crystallization processes are often highly complex systems (systems with different polymorphs, multiple crystallization mechanisms, etc.), significantly increasing the time and material required to develop essential products. Moreover, executing experiments for process development requires effort and individual expertise. When dealing with complex and vital crystallization processes, applying model-based design leveraging the advantages of simulation methods to optimize and better understand the underlying process is highly beneficial. One of the most widely applied modeling techniques for particulate processes is population balance modeling (PBM), which has been employed for process simulation and digital design to effectively decrease the required number of experiments, thus saving time and material. However, developing population balance models, designing, and performing suitable experiments, and successful parameter identification remain somewhat resource-intensive processes requiring considerable material, time, and often highly qualified personnel, which currently hinders the widespread application of this otherwise powerful framework. In this work, we demonstrate an automated crystallization platform [1], which is fully integrated with a user-friendly crystallization modeling software tool, CrySiV [2], as a holistic framework for the rapid development of crystallization digital twins. By enabling automated systems to execute crystallization experiments, the constant need for the presence of a professional can be mitigated, productivity can be enhanced, and the advantages of automated model development can be exploited. The systems consist of a computer-controlled MSMPR (mixed-suspension, mixed-product-removal) cascade system, which is developed to execute crystallization experiments autonomously according to a previously determined design. The automated crystallization unit can monitor the experiments using an integrated array of process analytical technology (PAT) tools, which are calibrated to monitor critical quality attributes (CQAs) needed for model development, including concentration, size, and particle count information. The system saves the results of the automated design of experiments (ADoE) in a format that is directly usable on the user-friendly crystallization modeling tool CrySiV, which allows a guided parameter identification and model-building process. This automated crystallization system can also be combined in a closed loop, enabling an iterative model development and experimental design procedure. The integration of an automated experimental platform with a user-friendly, efficient modeling tool can significantly reduce the time and effort needed for complex crystallization model development.

References:

1. Barhate, H. Kilari, W.L. Wu, Z.K. Nagy, Population balance model enabled digital design and uncertainty analysis framework for continuous crystallization of pharmaceuticals using an automated platform with full recycle and minimal material use, Chem. Eng. Sci, 287, 119688, 2024.
2. Szilagyi, W.-L. Wu, A. Eren, J. Mackey, S. Kshirsagar, E. Szilagyi, I. Ostergaard, K. Sinha, L. Mlinar, D. Pohlman, J. Chen, N. Nere, B. Moussa, M. Lovette, S. Black, A. Jawor-Baczynska, H. Li, B.-S. Yang, E. Irdam, D. Patience, R. McKeown, D. Juboor, M. Ketchum, D. Green, C. Polster, C. Burcham, D. Jarmer, J. Miles Merritt, L. Codan, J. Schoell, A. Cote, E. Sirota, Y.C. Liu, K. Girard, S. Kulkarni, Y. Yang, J. L. Quon, Z. K. Nagy, Cross-pharma collaboration for the development of a simulation tool for the model-based digital design of pharmaceutical crystallization processes (CrySiV), Cryst. Growth & Des, 21 (11), 6448-6464, 2021.