Direct discovery · Chemistry · Materials science
AI-directed lab discovers durable palladium-oxide catalysts for acidic water electrolysis
Iridium- and ruthenium-free palladium-oxide catalyst families for acidic oxygen evolution, including a lead composition tested for more than 1,000 hours.
Summary
Lila Sciences reports that its automated laboratory synthesized and screened 2,942 oxide catalysts across 53 material systems and 26 elements, identifying six palladium-based families near the activity-stability frontier for acidic oxygen evolution. InMnPdOx retained an overpotential below 0.5 V for more than 1,000 hours in 1 M sulfuric acid, while the company reports 17-fold higher screening throughput than a standard laboratory and more than 90% less human time per sample.
AI role
A sequential-learning agent ranked compositions in a human-supervised closed loop combining synthesis, screening, characterization and multi-objective activity-stability optimization.
Narrative role
Combines a concrete, experimentally tested materials output with an unusually large closed-loop campaign and explicit throughput measurements, strengthening evidence that AI-directed laboratories can search counterintuitive chemical regions at scale.
Caveat
The evidence is a company-authored preprint without independent replication or peer review. Palladium is more available than iridium but is still a precious metal, and the lead material is not yet a commercial electrolyzer catalyst.