Understanding and modelling of crystallization mechanisms by in-situ analytical technologies monitoring
Résumé
The batch cooling crystallization of an industrial active pharmaceutical ingredient that exhibits a needle-like habit was investigated. A model of the time variations of key physical properties of particles (crystal size distribution CSD and specific area) was designed using an approach based on the simple population balance equation. The experiments were monitored thanks to three in-situ process analytical sensors (Mid-IR spectroscopy, in-situ microscopy and laser backscattering). Primary and secondary surface nucleation mechanisms as well as growth of the main crystal dimension (length) were described resulting in a six parameter model. The model roughly represents the effects of different cooling strategies on the supersaturation profile, the CSD and the specific area of the final particles. First, a very weak primary nucleation occurs leading to an initial population of a few large crystals. A second population of crystals is generated afterwards through secondary nucleation mechanisms. The particles issued from the second burst of nucleation largely dominate (in mass and in number) the final overall population. Second, two important features of the crystallization process are outlined, which are commonly observed during the industrial crystallization of APIs (Active Pharm. Ing.) in impure mother liquors.