Abstract:

Closed-loop feedback mechanisms are common in process and control engineering yet are rarely, if ever, applied to discovery chemistry. The successful incorporation of autonomous feedback mechanisms into chemical discovery processes has the potential to revolutionize and greatly accelerate the way in which blockbuster medicines and materials are found. This talk will establish the concept of molecular discovery engines: AI-driven automated chemical engines which use modular building blocks as fuel to generate molecules, map property space, and output molecular discoveries. Several case studies will be presented of molecular discovery engines built and used to solve frontier problems in the areas of molecular charge transport, global reaction condition optimization, photostable light-harvesting small molecules, and solid-state organic lasers.