An autonomous lab for data-driven homogeneous catalysis.

Area of Science:

• Catalysis
• Chemical Engineering
• Materials Science

Background:

• Homogeneous catalyst discovery and optimization are critical for efficient chemical transformations but remain challenging.
• Developing selective catalysts requires exploring vast chemical spaces of ligands and reaction conditions.

Purpose of the Study:

• To develop an autonomous platform, Flex-Cat, for accelerated homogeneous catalyst discovery and optimization.
• To couple miniaturized reactors with Bayesian optimization for efficient exploration of catalyst design space.

Main Methods:

• Flex-Cat utilizes parallel miniaturized batch reactors for pressurized gas-liquid chemistry.
• A hierarchical, plate-constrained Bayesian optimization framework handles mixed discrete (ligand identity) and continuous (process) variables.
• Rhodium-catalyzed hydroformylation of propylene with diverse phosphorus-based ligands was used as a model system.

Main Results:

• Flex-Cat conducted 680 experiments across three multi-objective optimization campaigns.
• Identified ligand-reaction condition regions achieved over 2.5-fold improvements in turnover frequency.
• Discovered ligands with condition-programmable selectivity inversion, expanding regioselectivity range.

Conclusions:

• Flex-Cat provides a scalable and generalizable approach for autonomous catalyst development.
• The platform effectively links discovery-scale experiments to process-relevant outcomes.
• Autonomous optimization significantly enhances catalyst performance and expands accessible chemical space.