Reactive precipitation of pharmaceuticals: Mixing, reaction, and nucleation kinetics

Crystallization process design in the pharmaceutical industry frequently targets a controlled, small-mean particle size material to meet formulation, clinical and regulatory needs. Pharmaceutical crystallization operations are typically carried out in batch or semi-batch mode, with little information about process kinetics. In order to produce particles of uniform small mean size (< 10 mum) directly, equipment configurations producing rapid mixing and uniformly high supersaturation, i.e. high rates of nucleation are needed. Reactive precipitation can achieve very high levels of supersaturation and are often employed to produce small particles. However, the interactions among liquid mixing, reaction and nucleation kinetics are not well understood in reactive precipitation processes.; Rapid mixing was achieved through an opposed jet Y-mixer which was developed and characterized using the 4th Bourne reaction. Results show a mixing time on the order of 5 ms for jet velocities of 20 m/s. An eddy growth model was applied to simulate physical mixing and to the 4th Bourne reaction conditions used to characterize mixing. With the model it was shown that a Damkohler number (defined as the ratio of mixing time to reaction time) of unity corresponded to a 4 th Bourne reaction selectivity of 0.05 which validated assumptions used in the previous literature.; The Y-mixer was applied to the reactive precipitations of benzoic acid and voriconazole over a range of supersaturation. Three techniques were developed for monitoring particle formation processes: turbidity monitoring, direct microscopic observation, and quenched particle size analysis. The quenched particle size distributions from a double-Y precipitator were used to monitor the time evolution of a particle size distribution. The results for benzoic acid and voriconazole agreed with the classical, combined homogeneous/heterogeneous nucleation rate for the well-mixed conditions. The eddy growth model was extended to simulate the effects of mixing on the reactive precipitations of benzoic acid and voriconazole.