(661f) Scale-up of Reaction and Crystallization Steps of a Commercial Manufacturing Process in a Multi-Purpose Plant Using Mechanistic Modeling

This work presents an approach on modeling a manufacturing process in a multi-purpose plant, that comprises multiple reaction and crystallization steps, from the laboratory to production scale. In a multi-purpose plant, more often than not, the equipment used might not be the most suited for a particular process (e.g., mass transfer limitations due to poor mixing capabilities of an impeller) and thus it becomes crucial to use modeling tools and first principle methodologies to increase process understanding and further optimize the process to highest performance possible.

The case-study to be shown comprises a manufacturing multi-step, multi-phase process already in commercial phase, in which the reaction and crystallization unit operations are both modeled in laboratory and large scales. It is assessed if the conditions used in today's manufacturing process are optimized and if there is room for improvement either in yield of final product (reduce impurity formation), or particle size distribution (narrower interval and more suited range for further particle engineering). The data source for the modeling is both offline (HPLC analysis) and online (FBRM) which can be used as-is in modeling tools such as Dynochem (no data processing needed) to adjust the kinetics and crystallization constants. Both models are derived from laboratory scale experiments and subsequently are used to predict the behavior of the large scale process. After having the models validated in large scale batches, they are used to propose new conditions for the large scale process, using narrower ranges inside the Design Space of the current manufacturing process, for either yield of final product or particle size distribution.

In parallel to the kinetics and crystallization modeling, potential mass transfer limitations are assessed with Dynochem and CFD (Computational Fluid Dynamics) tools, depending on the complexity of reactor and impeller designs (lab and large scale designs). The mixing efficiency and mass transfer coefficients are derived from this study and are added to the kinetics model to increase its robustness and improve the overall understanding of the process.