CHAPTER 16. Digital Design and Operation of Continuous Crystallization Processes via Mechanistic Modelling Tools

This chapter will outline the application of digital design tools, specifically mechanistic models in the pharmaceutical industry, to enable more efficient development, design and control of API crystallization processes. Calibrated mechanistic models for crystallization can facilitate a more efficient workflow for studying and designing new continuous crystallization process, by characterising the process kinetics using batch crystallization data. The calibrated crystallization kinetics can then be utilised to inform the configuration of the continuous crystallization process, such as the number of stages to employ and the set-points for each stage. This analysis can be performed in silico via simulation prior to performing any continuous crystallization runs. It was shown in the first case study that the mechanistic model of the continuous crystallization can subsequently be employed to optimise the process to desired objective functions and provide a means to understand the various processing trade-offs. Finally, it was also shown how these mechanistic models can be further applied to facilitate the more efficient development of online MPCs. In comparison to the data-driven approach, the mechanistic models have the potential to significantly reduce the experimental effort, as well as material usage for MPC development.