Lila Sciences discovers high-performing palladium-based catalysts for green hydrogen through AI-driven screening

Lila Sciences/AI Science Factory

Lila Sciences' AI-directed laboratory screened 2,942 catalyst candidates across 53 material systems in three months, identifying six palladium-based families suitable for acidic oxygen evolution reaction in hydrogen production. The discovery addresses a longstanding supply-chain bottleneck: palladium is far more abundant than iridium or ruthenium, the current commercial catalysts.

• 2,942 catalysts screened across 53 systems and 26 elements in 3 months
• Six high-performing palladium-oxide families identified, including InMnPdOx and NiTaPdOx
• Lead candidate NiTaPdOx demonstrated stability over 1,000+ hours of operation
• AI Science Factory operated 17 times faster than conventional labs (240 samples per week)
• Research published on arXiv: 'AI-guided high-throughput discovery of iridium- and ruthenium-free palladium-oxide catalysts'