Researchers have developed DigMethPy, an AI platform that accelerates the discovery of catalysts for methane pyrolysis. This process breaks down methane into hydrogen gas and solid carbon, avoiding the CO2 emissions of traditional steam methane reforming. The platform analyzes over 40,000 data points from scientific literature to predict which molten metal alloys will work best as catalysts. The AI approach replaces slow trial and error in the lab with digital screening, focusing on properties like atomic charge and hydrogen adsorption. The team already used it to identify nickel-iron based alloys with high catalytic activity. This could lower the cost and time to develop clean hydrogen production methods that don't rely on carbon capture.
