(364b) A Predictive Material Sparing Method for Derisking Scale-up of Wet Stirred Media Milling for Nanocrystalline Suspensions Combining Microhydrodynamics and CFD-DEM

Delivering aqueous nanocrystalline suspensions of poorly soluble drugs is an important modality in the space of long acting injectables. The crystals on the order of a few hundred nanometers are commonly produced via wet stirred media milling, whereby agitated Yttrium stabilized zirconia or polymeric beads are used to crush the crystals through repeated bead-bead collisions. Among well-established methods of scale-up, the microhydrodynamic method is a commonly used approach in industry. The method works by obtaining the power density for a given mill, which when combined with bead properties and bead and drug loadings can be used to predict parameters controlling drug particle breakage, such as the collision frequency and the tendency for drug particles to be crushed. Nonetheless, this method for scale-up suffers from several deficiencies:

• The approach is not predictive, in that measurements at only one scale are not sufficient to guide scale up.
• Spatial distribution details such as bead distribution remain inaccessible
• Unable to predict and quantify effects of changes in excipients such as poloxamer content.
• No description of heat transfer and temperature control.

In this work, we present a complementary approach using CFD-DEM to predict the milling power density and heat transfer. This approach adds predictive power to the microhydrodynamic method by directly addressing the deficiencies in the model by solving detailed equations governing transport and collisions in the mill. We show that the approach is able to reduce material burden of process development by eliminating the need for extensive experimentation across scales. Additionally, we present observations on phenomenology in wet stirred media mills, which remains underexamined.