Model Predictive Control of Batch Enantiomer Crystallization Using Ternary Diagram Information
Additional Document Info
2018 AACC. Solutions of enantiomers that can be characterized as racemic compound forming systems in solvent have crystallization limitations based on solution thermodynamics that suggest that model predictive control (MPC) may provide an effective means for operating batch crystallization processes for such systems in a manner that meets thermodynamics-related operating constraints. While optimizing a performance measure tied to desired crystal properties, the MPC can seek to prevent the process state from entering regions in state-space associated with thermodynamically undesirable operating conditions (e.g., temperatures at which crystals formed will contain both enantiomers rather than a single pure enantiomer). The development of MPC also provides a framework for assessing appropriate final batch times for the batch operation through closed-loop simulations to assess potential trade-offs between longer batch operating times and optimization of the objective function when it depends on the final batch time. In this work, we demonstrate such principles through the development of an MPC for a racemic compound forming system (mandelic acid in water) that has received focus in the enantiomer crystallization literature.
Marcellos, C., Durand, H., Kwon, J., Barreto, A. G., da Cunha Lage, P. L., de Souza, M. B., Secchi, A. R., & Christofides, P. D.
complete list of authors
Marcellos, Caio Felippe Curitiba||Durand, Helen||Kwon, Joseph Sang-II||Barreto, Amaro Gomes||da Cunha Lage, Paulo Laranjeira||de Souza, Maurício Bezerra||Secchi, Argimiro Resende||Christofides, Panagiotis D