Single crystal fragmentation: Visualizing breakage model performance for pharmaceutical processes

The development of accurate crystal breakage kinetics is paramount to extending the capability of optimization and modelling methods within the pharmaceutical drug-product sector. However, there exists little experimental data on the performance on typical breakage models when applied to small-scale pharmaceutical crystals. Therefore, experimental investigations of crystal impacts on a steel target were carried out across a range of crystal sizes and impact velocities in order to determine critical breakage parameters for a given model. Direct observation of each impact was captured via a high-speed Shadowgraphy imaging technique, capable of 1μm crystal size resolution and up to a 125,000 frames-per-second capture rate. Crystals of Darunavir Ethanolate between a nominal 100→500μm size range were subject to impact velocities of 1→10m/s normal to the target surface. Furthermore, upon collision crystals were classified into one of three breakage modes: intact (Mode I), chipped (Mode II), or split/disintegrated (Mode III). The resulting data was then combined with a post-processing methodology utilizing a bi-variate histogram function and Kernel Density Estimation (KDE). This in turn enabled the extraction of experimentally observed crystal failure velocities, breakage modes, and selection functions detailing how crystal breakage varies with both crystal size and impact velocity. This data was then used to evaluate the performance of current literature in describing crystal breakage within pharmaceutical applications. Furthermore, the methodology implemented in this work can be applied to a range of particulate systems, yielding a flexible tool for pharmaceutical modelling capabilities.