Chen Mao¹, Chi So¹, Lap Yin Leung¹, & Ariel R. Muliadi¹

¹ Small Molecule Pharmaceutical Sciences, Genentech, Inc., South San Francisco, CA, 94080, USA

Roller compaction is one of the most prevalent means of granulation in drug product development.  It is highly beneficial to predict the roll force (RF)-ribbon density (r) relationship to assist the material-sparing process development and scale up.  Although the classic Johanson model quantitatively describes such relationship, it has not been extensively employed due to the required knowledge of the effective angle of internal friction (d) and wall friction angle (f) of powders at the stress regimes commensurate to roller compaction.

In this study, we demonstrated that a simplified Johanson model, devoid of d and f, can be developed for the roller compaction of pharmaceutical powders.  Specifically, we provided a theoretical ground showing that the RF-maximum pressure (P_max) and RF-r relationship becomes independent of d and f, beyond a critical nip angle (a_c). We also showed that for most pharmaceutical roller compaction, a_c is lower than 17^o.  Because the nip angle increases with increasing f, one can maximize f in practical roller compaction runs by employing rolls with non-smooth surface, thereby leading to a roller compaction operation with the nip angle greater than 17^o.  Under this condition, the original Johanson model is drastically simplified to a single equation requiring only one material property (i.e. compressibility, K).

By performing pilot-scale roller compaction runs encompassing pharmaceutical powders with diverse K (microcrystalline cellulose and dicalcium phosphate), we demonstrated that the simplified Johanson model is capable of accurately predicting RF-r relationship.  A predictive tool in the form of graphical user interface (GUI) was developed based on the simplified model, and is being extensively used for in-house roller compaction formulation and scale-up development.