(324d) Population Balance Modeling of Kinetic-Controlled Incorporation of Structurally Related Impurities for Crystallization Systems

Crystallization is a key separation and purification step in drug substance manufacturing. Critical quality attributes such as crystal size distribution (CSD), morphology, polymorphic form and chemical purity often dictate the performance of downstream processes as well as the bioavailability, and shelf life of the final products. Ideally, pure crystalline products can be obtained through crystallization due to mismatches in molecular volume or unfavourable steric interactions between the desired active pharmaceutical ingredients (API) and impurity compounds.¹ In practice, some impurities such as degradants, unreacted reagents, side products, and solvents from the synthesis step can compromise crystal purity.

The presence of impurities in a crystallization medium may alter the crystallization kinetics and lead to a reduction in productivity.² Furthermore, impurities in pharmaceutical products may be biologically detrimental above certain thresholds. The identification and control of impurity concentrations to acceptable limits is therefore critical in the pharmaceutical industry.³ Impurities may be incorporated into growing crystals through agglomeration, liquid inclusions, adsorption of mother liquor on crystal surfaces, co-crystal formation and solid solutions.⁴ Of these mechanisms, solid solutions and co-crystal formation are thermodynamically controlled. In general, kinetically controlled inclusion mechanisms such as agglomeration, surface deposition and lattice incorporation are commonly encountered during crystallization process development. Impurity identification and rejection is often achieved through extensive experimentation based on traditional design of experiments (DoE) approaches which may be infeasible in early-stage process development due to the limited availability of the API. The development of mathematical models of crystallization processes can allow in silico process simulation and lead to reduced numbers of wet experiments.^5–7

The objective of this work is to develop a population balance model for the discrimination of impurity incorporation mechanisms in the presence of multiple structurally related impurities on the crystallization of active pharmaceutical ingredients. A 2D population balance model is developed to capture the sensitivity of crystal aspect ratios in the presence of low impurity concentrations. The population balance model (PBM) is solved using a high-resolution finite volume method. The crystallization kinetics describing nucleation, growth, and agglomeration of the model compound paracetamol (acetaminophen) are obtained in the presence of p-acetaminobenzoic acid and acetanilide. The kinetic parameters are obtained by minimizing an objective function that considers solute concentration, crystal aspect ratios and particle counts from batch cooling crystallization experiments of paracetamol in ethanol. Experimental results show that at low cooling rates, larger crystals with higher aspect ratios are obtained in the presence of p-acetaminobenzoic acid when compared with cases involving pure paracetamol and the crystallization of paracetamol in the presence of acetanilide. The larger crystal size and lower particle counts obtained in the presence of p-acetaminobenzoic acid suggests that p-acetaminobenzoic acid alters the crystallization kinetics of paracetamol to a growth-dominated process. Acetanilide on the other hand has been found to reduce the crystal size and aspect ratios when compared to the cases with pure paracetamol. Experiments conducted at higher cooling rates in the presence of both impurities show lower agglomeration and larger crystal sizes when compared with the crystallization of pure paracetamol at the same cooling rate. In the future, the population balance model will be extended to the impurity incorporation mechanisms for a different commercial API/impurity system, including those that are kinetically controlled and the generation of design spaces within which the product quality attributes are met.

References

(1) Chakrabarti, R.; Vekilov, P. G. Dual Mode of Action of Organic Crystal Growth Inhibitors. Cryst. Growth Des. 2021, 21 (12), 7053–7064.

(2) Beckmann, W. Crystallization: Basic Concepts and Industrial Applications; 2013.

(3) Capellades, G.; Bonsu, J. O.; Myerson, A. S. Impurity Incorporation in Solution Crystallization: Diagnosis, Prevention, and Control. CrystEngComm 2022, 24 (11), 1989–2001.

(4) Urwin, S. J.; Levilain, G.; Marziano, I.; Merritt, J. M.; Houson, I.; Horst, J. H. Ter. A Structured Approach to Cope with Impurities during Industrial Crystallization Development. ACS Publ. 2020, 24 (8), 1443–1456.

(5) Abramov, Y. A.; Zelellow, A.; Chen, C. Y.; Wang, J.; Sekharan, S. Novel Computational Approach to Guide Impurities Rejection by Crystallization: A Case Study of MRTX849 Impurities. Cryst. Growth Des. 2022, 22 (12), 6844–6848.

(6) Borsos, A.; Majumder, A.; Nagy, Z. K. Multi-Impurity Adsorption Model for Modeling Crystal Purity and Shape Evolution during Crystallization Processes in Impure Media. Cryst. Growth Des. 2016, 16 (2), 555–568.

(7) Nordstrom, F. L.; Linehan, B.; Teerakapibal, R.; Li, H. Solubility-Limited Impurity Purge in Crystallization. Cryst. Growth Des. 2019, 19 (2), 1336–1346.