DataDKR · A Data-Driven Approach for the Optimization of Dynamic Kinetic Resolution Reactions

The optimization of asymmetric catalytic transformations is a complex multivariate problem where numerous interdependent variables must be tuned simultaneously. While data-science approaches have successfully begun to guide reaction optimization, their application has largely been confined to relatively simple systems. The present DataDKR project aims to extend these methods to the more challenging yet synthetically useful class of dynamic kinetic resolutions (DKRs). Experimental and computational workflows will be combined to generate high-quality datasets, which will underpin the construction of predictive, statistically robust models capable of capturing intrinsic aspects of DKR processes across different substrates and background racemization strategies. Two synthetically relevant case studies—the enantioselective synthesis of chiral-at-phosphorus and chiral-at-sulphur compounds via DKR—will serve as benchmarks for model development and generalization. By identifying and encoding the unique features that govern the reactivity and selectivity of DKRs, the project will deliver a broadly applicable, data-driven framework for the rational optimization of these complex systems, advancing synthetic organic chemistry and data science in tandem.