Researchers at North Carolina State University (NCSU; Raleigh; www.ncsu.edu) and the University of North Carolina (Chapel Hill; www.unc.edu), with the support of Eastman Chemical Co. (www.eastman.com), have developed a closed-loop autonomous catalysis platform for identifying high-performing catalyst formulations and for investigating how reaction conditions affect selectivity.
The new platform, called Flex-Cat, “couples parallel miniaturized batch reactors for pressurized gas-liquid chemistry with a hierarchical, plate-constrained Bayesian optimization framework for mixed discrete (ligand identity) and continuous (process) variables,” the researchers say. Combining robotics, automated product analysis and artificial intelligence, Flex-Cat discovers both high-performing catalysts and catalysts whose behavior can be “tuned” to make different products under different reaction conditions. NCSU professor Milad Abolhasani says: “It is not enough to identify the right catalytic material. You also have to find the right temperature, pressure and concentration for that catalyst to work efficiently and selectively. That creates a vast experimental search space.”
Flex-Cat can prepare catalysts, run reactions, analyze products and use the results to decide which experiments to run next. In a recent study published in Nature Communications, the researchers used Flex-Cat to study rhodium-catalyzed hydroformylation of propylene. A key challenge in hydroformylation is controlling which aldehyde isomer is produced. Researchers can direct Flex-Cat to automatically optimize the catalyst and conditions to maximize the yield of a single aldehyde isomer product. The system performed 680 experiments using 16 chemically diverse phosphorus-based ligands. Flex-Cat identified catalyst-condition combinations that improved catalyst activity by more than 2.5-fold, expanded the accessible range of product selectivity and uncovered ligands that could be programmed to favor different products under different conditions.