Development and Characterization of a Single Stage Mixed-Suspension, Mixed-Product-Removal Crystallization Process with a Novel Transfer Unit

Abstract

A continuously operated single stage mixed-suspension, mixed-product-removal (MSMPR) crystallizer using intermittent withdrawal via a dip pipe with combined pressure/vacuum was successfully developed for the manufacture of active pharmaceutical ingredients. Approximately 5.8% of the total operating volume was intermittently removed at a high velocity using vacuum. The transfer line was also periodically purged with nitrogen to ensure complete removal of residual solids. In situ process analytical technologies (focused beam reflective measurement (FBRM) and process video microscopy (PVM)) were successfully applied to monitor and characterize the MSMPR crystallization process. In this study, a cooling crystallization of paracetamol from an aqueous isopropyl alcohol solution was investigated. Experimental results indicate that the crystallization system was able to operate without any clogging issues for over 10 residence times, before which the system had approached steady state. Three different start-up strategies for continuous crystallization were investigated, and the results indicate that the chord length distributions at steady state were the same for all cases. Also, starting the continuous operation from a saturated solution that was seeded with product from a previous MSMPR run offered the quickest route to steady state. To better control and scale up the crystallization process, the nucleation and crystal growth kinetics of the model compound were also determined through use of the newly developed process. The growth rates were found to be size dependent, and an exponential three-parameter model was employed to characterize the size-dependent growth. It was seen that the crystal growth rate was extremely low and increased linearly with particle size when the particle size was below 10 μm. However, the growth rate increased dramatically with particle size when the particle size was between 10 and 1000 μm. The nucleation kinetics was correlated by the semiempirical equation BTOT = 1.11 × 1015MT0.98Gavg1.12. The orders of the total nucleation rate with respect to the magma density and average growth rate were 0.98 and 1.12, respectively. Therefore, the effect of supersaturation (or residence time) and magma density on the steady state crystal size was investigated.