AI System Rapidly Develops Promising Catalysts to Sustainably Convert CO2 into Methanol
-
Researchers developed an AI system and robot lab to rapidly create and test catalysts for synthesizing methanol from CO2.
-
In under 6 weeks, the system developed 150 promising, low-cost catalyst compositions with high conversion rates.
-
Methanol can be used as a sustainable fuel and chemical feedstock to replace fossil sources.
-
The AI uses Bayesian optimization to narrow down almost infinite catalyst possibilities to test the most promising.
-
By iterating between prediction, robotic synthesis and testing, the system can explore catalyst space and accelerate development.