Scaling Down a Reactive Crystallization Using Various Mixing Models

Multiple commercially available mixing modeling tools were utilized to scale-down a reactive crystallization from the planned process performance qualification (PPQ) scale to the laboratory scale. Differences in the historical crystallization batch data with results in standard laboratory crystallization set-up drove the need to fully characterize the lab scale setup to ensure it was representative of the PPQ equipment train.
The mixing modeling tools were used to model mixing at both the PPQ and laboratory scales system and compared with experimental and batch data for physical properties (particle size and surface area) of the product. These tools encompassed both correlation-based and computational fluid dynamics based modeling (DynoChem, M-Star, and MixIT). In this presentation, the modeling results of numerous laboratory scale conditions and setups as well as large scale batch data were used to understand micro-mixing, meso-mixing, energy dissipation rate and strain rates. The modeling results were compare with experimental physical property data to give insight into the strengths and weaknesses of using these tools in identifying laboratory conditions and setup of the reactive crystallization.