We have computationally investigated the use of the multisegment, multiaddition, plug-flow crystallizer (MSMA-PFC) for use in producing pharmaceutical crystals. A population balance framework was used to model the crystallization process. The dissolution of crystals can be modeled when solubility is below saturation. The evolved volume fraction distributions were optimized in a least-squares sense by manipulating a vector of decision variables in order to hit a target volume fraction distribution. The genetic algorithm was used for optimization. A reduced orthogonal array experimental design was used to examine the effect of several kinetic parameters and total crystallizer length. The results indicate that the parameters which govern nucleation are the most sensitive, followed by those for growth. Dissolution does not appreciably occur in any of the optimizations. The reason the optimization does not add any pure solvent is likely due to the addition of pure solvent causing a simultaneous decrease in concentration and decrease in residence time, which the optimization judges to be suboptimal.

• Purdue University, Department of Chemical Engineering

• Modeling

1. Growth
2. modeling
3. MSMA-PFC
4. Nucleation
5. population balance model
6. Research Article