(164c) The Use of Cfd and Population Balance for Anti-Solvent Crystallization Scale-up

Anti-solvent crystallization, widely used in the pharmaceutical industry, is one of the preferred ways of crystallizing thermo labile materials from solution. In this process, the crystallizer performance depends not only on the intrinsic crystallization kinetics but also on the physical processes such as inter-phase mass flow within a multiphase crystallizer. Mixing in conjunction with anti-solvent addition is expected to have significant effect on the supersaturation distribution, nucleation process and final crystal particles size distribution. For batch processes the effects of hydrodynamics and mixing are known to have a major impact on process scale-up.

CFD simulations combined with population balance equation capture the effect of mixing on anti-solvent crystallization of organic compounds and provide a way to determine optimum process conditions and rules for scale-up from laboratory to industrial scale. A DOE with CFD simulations was used to determine the effect of (1) agitation rate, (2) position of anti-solvent feed, and (3) anti-solvent feeding rate on the distribution of the anti-solvent concentration in the crystallizer. Based on these results, anti-solvent feeding strategy was developed to optimize the process conditions, and to avoid unfavorable high local supersaturation as a result of poor mixing. The results of the simulation scale-up with (NdD²)/q, where q is the volumetric flow rate of anti-solvent addition, N the agitation rate, d the diameter of the impeller and D the diameter of the crystallizer. A second DOE with CFD simulations and population balance has been conducted to formulate scale-up rules when crystallization kinetics are introduced. A second scale-up parameter emerging from this DOE is the Damkohler number, Dacry = tmix / tcry, the ratio of mixing to crystallization time, with the latter defined as the ratio of average crystal size to crystal growth rate.