(610d) Polymorph Selection in Continuous Crystallizers in the Presence of Wet Milling

Vetter, T., University of Manchester
Li, Y., Tianjin University
The design of controlled processes that ensure the consistent production of a desired polymorph is an important goal in the context of pharmaceutical manufacturing.1 The most stable crystalline form can be obtained from a batch crystallization process either through seeding or by allowing enough time for metastable forms to convert to the stable form. On the other hand, producing metastable forms in a batch process relies on delicate process control strategies (temperature, solute concentration and batch time control). As shown in the work by Farmer et al.4 which polymorph is obtained from a continuous crystallizer at steady state can be controlled by selecting appropriate combinations of residence time, operating temperature and inlet concentration. For cases where the crystallization process is dominated by nucleation and growth, they have shown that the polymorphic outcome in such a continuous process can be accurately predicted once two modified Damköhler numbers are known. In essence, these dimensionless groups represent ratios of the characteristic time for nucleation and growth, the process time (residence time of the crystallizer) and the feed concentration.

In this contribution, we are expanding on this previous works and consider the case where wet milling is used to tune the particle size distribution and potentially alter the polymorphic outcome. We thus present a mathematical model of continuous crystallization with wet milling for polymorphic systems. The model relies on one-dimensional population balance equations including breakage terms. Making the process model dimensionless, we identified a surprisingly small number of additional dimensionless groups that describe the process outcome in terms of polymorphic form when breakage is present. The resulting model is solved using a finite volume scheme (FVS)5. Operating regions where the metastable/stable polymorph can reliably be obtained can again be identified. In comparison to the work by Farmer et al.4, we show that the operating region where the metastable/stable polymorph is obtained can be enlarged by tuning key dimensionless numbers in the breakage kinetics. We combine this type of analysis with an investigation of the productivity of such systems and report regions of attainable particle sizes4,5

We show that introduction of a wet suspension mill therefore represents an attractive way to obtain the desired polymorph even if the underlying kinetics of nucleation and growth are disadvantageous. Apart from this finding, we show that introducing breakage increases process productivity and allows to tune particle size more extensively.

[1] Hermanto, M. W.; Chiu, M. S.; Woo, X. Y.; Braatz, R. D. Robust optimal control of polymorphic transformation in batch crystallization. AIChE J. 2007,53,2643–2650.
[2] Farmer, T. C.; Carpenter, C. L.; Doherty, M. F. Polymorph selection by continuous crystallization. AIChE J.2016, 62, 3505−3514.
[3] Kumar, R.; Kumar, J. Numerical simulation and convergence analysis of a finite volume scheme for solving general breakage population balance equations. Appl Math Comput. 2013, 219, 5140-5151.
[4] Vetter, T.; Burcham, C.L.; Doherty, M.F. Regions of attainable particle sizes in continuous and batch crystallization processes, Chem. Eng. Sci., 2014, 106, 167-180.
[5] Vetter, T.; Burcham, C.L; Doherty, M.F. Designing robust crystallization processes in the presence of parameter uncertainty using attainable regions, Ind. Eng. Chem. Res., 2015, 54, 10350-10363.