Automation, in situ monitoring, and process control tools are implemented to understand and control the crystallization of an active pharmaceutical ingredient in development. As a first step in the study, the metastable zone is generated automatically by linking a laser backscattering probe to an automated laboratory reactor. Using the metastable zone as a guide, crystallization experiments with varying cooling rates and seeding protocols are conducted and monitored. The evolution of solution concentration and supersaturation is determined by transmitting data from an in situ total reflectance Fourier transform infrared (ATR-FTIR) spectroscope to the laboratory reactor. Supersaturation profiles coupled with data from the laser backscattering probe demonstrate the prevalence of primary and secondary crystal nucleation in the process. A cascaded proportional-integral controller is tuned and implemented to promote crystal growth over nucleation by maintaining supersaturation at low constant values. Nonlinear temperature profiles that result in crystals of larger size are thereby generated.

The utility of the growth mechanism to influence relative particle growth is investigated in a semi-batch reactor. A dynamic competitive growth model is developed and used to simulate the growth of two monodisperse polystyrene particle populations (bidisperse system) at 50 °C. Validation of the model with on-line density and on-line particle diameter measurements demonstrates that radical absorption into particles is more likely to occur by a collision mechanism than either by a propagation or diffusion mechanism. Experimental and simulation results show that relative narrowing of the bidisperse system increases as the weight fraction polymer inside the particles (W p) decreases. It is concluded that the ability to alter W p in a semi-batch reactor makes this mode of operation attractive for particle size distribution control.

No abstract

No abstract