Researchers in Canada have discovered an alloy to improve the production of green hydrogen, thanks to a suggestion made by artificial intelligence (AI).
Race for best catalysts
A Toronto team, like the many scientists working on green hydrogen production elsewhere, are looking to generate hydrogen by passing electricity through metal electrodes in water.
The process currently consumes a lot of energy and usually involves expensive and rare metals, so the race is on to find catalysts that can make it cheaper and more efficient. One such lab found for example that chucking coffee granules into the mix helped to speed things up and made some of the ingredients recoverable for re-use.
Billions of alternatives narrowed down to 36,000
But instead of trial and error, the Canadian project, working alongside Carnegie Mellon University in Pennsylvania, turned to an AI program to help run through billions of alloy candidates and conduct simulated modelling to seek out the best one.
More than 36,000 metal oxide options were then run through the simulator before the most promising candidate underwent real world lab testing. Without the involvement of AI, it would have taken years for human researchers to narrow down the findings.
The lab tests involved the Canadian Light Source (CLS), a synchroton particle accelerator at the University of Saskatchewan, whose X-rays can help analyse how well any given catalyst is performing.
“What we needed to do is use that very bright light at the Canadian Light Source to shine it on our material and see how the atomic arrangements would change and respond to the amount of electricity that we put in,” team member Jehad Abed explained.
Ruthenium, chromium and titanium
The AI system’s suggestion that the best results would come from a specific combination of ruthenium, chromium, and titanium was borne out by the testing, with the alloy demonstrating 20 times better stability and durability than the team’s benchmark metal. Tests at the Advanced Photon Source at Argonne National Laboratory in Chicago confirmed the results, which were published in the Journal of the American Chemical Society.
While many fear the effects of AI in fields such as employment, intellectual property, armaments and security, Abed noted that “The computer was right about this alloy being more effective and stable. That was a breakthrough because it shows that this method for finding better catalysts is working,” he added. “What would take a person years to test, the computer can simulate in a matter of days.”
