Improving productivity and process robustness of fluidized bed crystallization for the continuous separation of enantiomers

• Continuous chiral resolution at pilot-scale with 150 g/L/h and high purity > 99 %. • Novel resolution control removes chiral antipode without process interruption. • Process model used to find optimal trade-of between productivity and robustness. • Model-based optimization of operational parameters and crystallizer geometry. Preferential Crystallization (PC) is an efficient separation technique for chiral systems, as it enables the direct crystallization of one enantiomer from a racemic solution. The application of PC in a fluidized bed crystallizer (FBC) enables continuous operation with enhanced process robustness and integrated product classification. However, as typical for PC, the inevitable spontaneous primary nucleation of the undesired enantiomer remains a major challenge for continuous operation. Furthermore, the frequent absence of reliable knowledge regarding primary nucleation kinetics prevents successful model-based optimization of PC processes. This study demonstrates that productivity and process robustness of continuous PC can be improved even in the absence of knowing the rates of primary nucleation. A previously validated process model of the FBC is applied to identify the best trade-off between productivity and process robustness. Based on this, the operating and geometric parameters of the conically shaped FBC were optimized to maximize productivity. The optimization results were experimentally validated at pilot scale, where a productivity of 150 g/L/h were achieved in the continuous chiral separation of DL-asparagine monohydrate in water with product purities exceeding 99 % for both enantiomers. Furthermore, this study presents and experimentally demonstrates a strategy to remove during continuous operation possible product contamination caused by the nucleated chiral antipode, which allows maintaining operation over long times.