Modelling of Rheological Properties of Pharmaceutical Powder Mixtures for Direct Compression: A Statistical Approach

The rheological properties of pharmaceutical mixtures are exceedingly difficult to understand and control. Despite their impact on final product quality, no generally valid approach to define and predict mixture rheology has been discovered so far. The cause is a complexity of factors affecting the particulate mixture behaviour. This study aimed at addressing part of this problem by suggesting an approach to evaluate and subsequently predict flow properties of binary particulate mixtures from parameters of individual components to better guide the formulation development. This study investigated several parameters of powder flowability under conditions with different stress levels using the FT4 powder rheometer. Studied mixtures contained ibuprofen as a model active pharmaceutical ingredient with a variety of common fillers. Results revealed that mixture behaviour varies according to given flow conditions and cannot be characterised by a single parameter and parameters appropriate to unit operations of the intended process should be used. Outputs confirmed the complexity of rheological properties of mixtures that are not always the linear combination of rheological properties of starting materials. On the other hand, developed multilinear regression models could predict studied mixture behaviour quite well by accounting for interdependencies between critical parameters, composition, and particle characteristics capturing the complexity of particle interaction interplay beyond the simple additivity of pure material properties. The prediction is valid for model ibuprofen and the other active APIs with similar characteristics. Thus, models could streamline selection of suitable mixtures for direct compression or highlight potentially problematic rheological aspects of a mixture to be addressed.