Scaling up temperature cycling-induced deracemization by suppressing nonstereoselective processes

The scale-up of Temperature Cycling-Induced Deracemization (TCID) of sodium bromate is feasible provided that two nonstereoselective processes are suppressed. Both nonstereoselective processes occur as a result of insufficient crystal breakage or attrition. In the absence of crystal breakage or attrition during the temperature cycles, large crystals emerge and the resulting small total crystal surface area is unable to sufficiently consume the supersaturation during cooling, resulting in nonstereoselective nucleation. This nonstereoselective process can be avoided by applying small temperature cycles involving small dissolving solid fractions. However, this leads to a slow deracemization rate. In addition, crystals undergo nonstereoselective agglomeration, which leads to the formation of large racemic agglomerates constructed of both chiral forms. To counteract their formation, secondary nucleation through crystal breakage was found to be a key requirement. At a large scale, a homogenizer was used to induce crystal breakage which, in combination with temperature cycles, led to the removal of racemic agglomerates as well as a significant increase in the deracemization rate. Overusing the homogenizer, however, caused the enantiomeric excess increase to stop. Our experiments show the importance of secondary nucleation in TCID of sodium bromate. However, secondary nucleation is currently not incorporated in the TCID process models. In the presence of a large amount of crystals which facilitates a sufficiently large crystal surface area at the highest temperature and careful use of the homogenizer, TCID leads to complete deracemization in volumes up to 1 L, demonstrating the potential to extend TCID to industrial applications.