(205e) Digital Twin Design of an Agrochemical Crystallization Process Using a 2D Population Balance Model Approach

In the agrochemical industry, crystallization is a key separation process that is used in the to separate the agrochemical actives from reaction mixtures and impure solutions. A poorly designed crystallization step can lead to poor yield, low purity, and long filtration time. To minimize process cycle time, a series of design of experiments led by the Quality-by-Design (QbD) approach can be used by practice to optimize the operating profile of a crystallization process (i.e. temperature profile, solvent/antisolvent addition profile, pH profile).¹ However, an exhaustive list of experiments covering all factors is needed to explore the whole design space experimentally. To avoid excessive open-loop experiments and personnel exposure to toxic chemicals, recently introduced Quality-by-Control (QbC) methodology controls critical quality attributes (CQAs) through implementation of feedback control algorithm to determine the operating profile of the process.² For rapid process design, direct design or model free (mfQbC) approaches can be used to generate a robust operating profile leading the system to the desired CQAs. However, mfQbC results are often suboptimal with new process changes (scale-up, mixing conditions, and solvent addition), which requires model-based QbC (mbQbC) to further optimize the mfQbC results.^3,4

In this work, a digital twin of an industrial agrochemical crystallization process was developed to generate a temperature cycling operating profile with the objective of maximizing crystal size to reduce filtration time. A 2D population balance model (PBM) solved via semi-discrete finite volume method was developed. Parameter estimation experiments to determine the 2D kinetics was performed using a well-mixed 2D PBM and laboratory scale experiments. Scale-up and antisolvent addition effects were incorporated in the digital twin by using a multicompartmental PBM approach with parameters refined from large-scale experiments. Various compartments were required to capture wall-effects from large jacket to vessel temperature differences and high shear environment near the impeller area.

References:

(1) Bondi, R. W.; Drennen, J. K. Quality by Design and the Importance of PAT in QbD; Academic Press, 2011; Vol. 10. https://doi.org/10.1016/B978-0-12-375680-0.00005-X.

(2) Su, Q.; Ganesh, S.; Moreno, M.; Bommireddy, Y.; Gonzalez, M.; Reklaitis, G. V.; Nagy, Z. K. A Perspective on Quality-by-Control (QbC) in Pharmaceutical Continuous Manufacturing. Comput. Chem. Eng. 2019, 125, 216–231. https://doi.org/10.1016/j.compchemeng.2019.03.001.

(3) Szilagyi, B.; Eren, A.; Quon, J. L.; Papageorgiou, C. D.; Nagy, Z. K. Application of Model-Free and Model-Based Quality-by-Control (QbC) for the Efficient Design of Pharmaceutical Crystallization Processes. Cryst. Growth Des. 2020, 20 (6), 3979–3996. https://doi.org/10.1021/acs.cgd.0c00295.

(4) Öner, M.; Bach, C.; Tajsoleiman, T.; Molla, G. S.; Freitag, M. F.; Stocks, S. M.; Abildskov, J.; Krühne, U.; Sin, G. Scale-up Modeling of a Pharmaceutical Crystallization Process via Compartmentalization Approach. In Computer Aided Chemical Engineering; Elsevier B.V., 2018; Vol. 44, pp 181–186. https://doi.org/10.1016/B978-0-444-64241-7.50025-2.