(104b) Multi-Scale Characterization of Powder Compaction Spanning Single Particle, Single Tablet Fabrication, and Industrial Rotary Tablet Press

In a continuous tableting line, hundreds of tablets are produced every minute with the overall plant throughput being multiple kilograms per hour. Tablets need to have certain Critical Quality Attributes (CQAs) at the end of the production line; hence a Distributed Control System (DCS) plays the crucial role of controlling various unit operations by using a network of different sensors. The DCS system needs predictive, robust and computable models to inform changes in upstream and downstream process parameters. These models must be able to capture the dynamics of the system induced by variations in the blend composition, material flow properties, production rate, etc. Therefore, it is important to characterize the powder and the compaction process thoroughly.

We perform this process characterization at three distinct length scales, namely the particle scale, the single tablet scale, and the bulk production scale. Specifically, we characterize individual pharmaceutical powder particles under diametrical compression using a Shimadzu MCT‐W 510 micro-compression tester, at small deformation rates. We characterize the fabrication of single tablets using a Gamlen D Series compaction simulator. Even though the compactor simulator is much more highly instrumented than most industrial rotary tablet presses, the compaction rate is smaller than that of tablet presses. We also use a Sotax AT4 tester to assess in-line the CQAs of a few selected tablets made in a continuous tableting line using a Natoli BLP-16 press.

It is worth noting that we collect data at different length and time scales and thus special care needs to be taken in characterizing the compaction process and in developing predictive models for the DCS system. In this talk, we will discuss the similarities and differences between the three components of this multi-scale characterization approach. We will also show that models at various length scales account for different physical mechanisms and all are needed to robustly capture the dynamics of the system induced by changes and perturbations in the continuous tableting line, such as variations in the blend composition, material flow properties, and production rate.