(302e) Development of a Thermodynamic Crystallization Using Data Rich Experimentation (DRE)

Until recently, an active pharmaceutical ingredient (API) had been isolated as a metastable form due to the complexity of the phase map, the propensity to maintain very high supersaturation levels, and the inability to nucleate the thermodynamic stable form. Data rich experimentation methods, integrating simple experimental techniques with state-of-the-art process analytical technologies (PAT), were utilized to develop a seeded thermodynamic crystallization process with good impurity rejection, high yield, and 55% improved volumetric productivity.

This study looked at the impact of the seed point on purity control, the crystallization path on filtration performance, and the drying conditions on residual solvent levels and generation of viable seed. A process fingerprint was developed and used to guide scale-up and ensure target product purity was achieved. The filtration process was assessed in the laboratory using a pocket filter experimental setup along with a DynoChem model. This allowed the crystallization path to be adjusted to optimize filtration performance. Using tools available in the Drying Process Development Lab (DPDL), a humid drying procedure was developed, and a DOE based model was used to define the operating space for meeting tight solvent specifications during scale-up. The ability to reconstitute dried solids to the desired form was demonstrated and key for the process to ensure formation of the thermodynamically preferred form. The process was scaled up from the laboratory through the pilot plant to the factory where multiple demonstration batches were completed successfully. Results of this investigation and the methods applied will be described in detail during this presentation.