(88f) Process Modelling Tools for a Drown-out Crystallization Process Including Application to Scale-up

Computational fluid dynamics (CFD) is a powerful simulation tool that was used to investigate mixing (micro-, meso- and macromixing) effects, turbulence and shear on multi-scale batch crystallizers for a drown-out crystallization process. CFD was used to predict agitation conditions for transfer to a manufacturing scale. Mixing was found to be variable across kilo- and pilot-scale crystallizers but was not found to have an impact on the physical properties of the organic fine chemical (particle size, size distribution and crystal habit). A process modelling tool, DynoChem was used to estimate growth and nucleation kinetics from laboratory scale experimentation incorporating thermodynamic solubility information and Lasentec FBRM chord size distributions. The main factor influencing the nucleation rate constant was the level of initial solvent in the antisolvent at the start of the drown-out. The drown-out crystallization was found to be a nucleation dominated process during the early stages of addition as the equilibrium solubility was close to zero in the stirred tank and was independent of the mixing rate. As the equilibrium solubility increased gradually over the addition period, a threshold value of supersaturation corresponding to a specific solvent composition was reached whereby nucleation was gradually ?switched off', with the crystallization process becoming growth dominated thereafter.