Using complex layer melt crystallization models for the optimization of hybrid distillation/melt crystallization processes

Layer melt crystallization is a highly selective method for the separation of narrow boiling mixtures which are difficult to separate with conventional separation techniques like distillation due to low driving forces. Contrawise, layer melt crystallization has the drawback of limited capacity due to the direct connection between crystal product and required cooled surface. Here, the combination of the high throughput distillation and highly selective layer melt crystallization into an integrated hybrid process can lead to enormous benefits. Since the separation efficiency of the crystallization is not predictable, it has to be described with empirical correlation. Here, studies from literature use strongly simplified correlations by, e.g. assuming complete separation. This bears the serious risk of overestimating the efficiency of the hybrid process. Further, the effective post purification step sweating was not implemented into hybrid processes in studies from literature. This study fills this gap in literature. A distilliation/melt crystallization hybrid process is optimized by realistically describing crystallization separation efficiency and by implementing sweating. The required crystallization models are presented and experimentally validated. The optimization of the hybrid process is done with different modelling depths and the results underline impressively the importance of the adequate description of the crystallization separation efficiency. (C) 2014 Elsevier B.V. All rights reserved.