Crystallization Research for Controlling Particle Size Distribution (PSD)

Orel Mizrahi, Chemical Engineering, Biotechnology and Materials, Ariel University, Teva Innovative R&D, Netania, Israel

Teva Pharmaceuticals LTD, innovative R&D.

All over the world in the pharmaceutical industry, tons of active pharmaceutical ingredients (API) are disqualify for use due to particle size distribution (PSD) that does not fit specifications. Today process development technology provides solutions for this problem. The most advanced approach in this field is using of simulation software. In this research predictive model for final crystallization of Rasagiline mesylate was built. The crystallization process was developed and characterized in the aim of control of particle size distribution (PSD). Such model provides the opportunity for fast process optimization and prediction of process parameters effects. During the development solubility curves were built for the Rasagiline mesylate/isopropanol (IPA)/water mixture. A solvent mixture was chosen for the model building experiments to be IPA/water 2%. Preliminary experiments were performed in order to characterize the process and to find the effects of water percentage, cooling profile and seeding ratio on product PSD. An SMOM model was built and found to be inaccurate, giving much lower PSD span than found in the experiments. It was established that other mechanism is acting during crystallization than nucleation or growth. A QMOM model was built and size dependent growth was assumed to be this mechanism. Good fit between experimental and calculated results was achieved for the PSD and for the solute concentration during the crystallization. The model predicts accurately the PSD span and its statistical parameters. A series of additional experiments performed to test the sensitivity of the model for changes of the main parameters (supersaturation, seed ratio and cooling profile). It was found that the model accurately predicts the PSD under different experimental conditions.