CGCNN-HD model architecture. Credit: Journal of the American Chemical Society (2024). DOI: 10.1021/jacs.4c01353
Researchers at the University of Toronto are using artificial intelligence to accelerate scientific progress in the quest for sustainable energy. Using the Canadian Light Source (CLS) at the University of Saskatchewan (USask), they have identified a new AI-generated catalyst “recipe” that offers a more efficient way to produce hydrogen fuel.
To produce green hydrogen, electricity generated from renewable sources is passed between two pieces of metal underwater, which releases oxygen and hydrogen gas. The problem with this process is that it currently requires a lot of electricity, and the metals used are rare and expensive.
Researchers are searching for the right alloy, or combination of metals, to act as a catalyst to make this reaction more efficient and less costly. Traditionally, this search has required trial and error in the lab, but this method is too time-consuming, like searching for a needle in a haystack.
“We're talking hundreds of millions, maybe billions, of possible alloys, and one of them could be the right answer,” said Jehad Abed, part of a team that developed a computer program to greatly speed up the search.
The findings, published in the Journal of the American Chemical Society, were published in 2017. At the time of the project, Abed was a doctoral student at the University of Toronto working under the supervision of Edward Sargent and with scientists at Carnegie Mellon University.
Credit: Canadian Light Source
The team's AI program took more than 36,000 combinations of metal oxides and ran virtual simulations to evaluate which combinations of ingredients would be most effective. Abed then tested the program's top candidates in the lab to see if its predictions were accurate.
The team used the CLS's ultra-bright x-rays to analyze the catalyst's performance during the reaction. “What we wanted to do was shine the extremely bright light from the Canadian Light Source on the material and see how the atomic arrangement changes and responds to the amount of electricity we put in,” Abed said. The team also used the Advanced Photon Source at Argonne National Laboratory in Chicago.
According to Abed, an alloy combining the metals ruthenium, chromium and titanium in specific ratios was the clear winner.
“The alloy recommended by the computer performed 20 times better than our benchmark metal in terms of stability and durability,” he said. “It lasted longer and performed more efficiently.”
While the AI program developed by Jehad and his colleagues shows great promise, the material itself still needs to undergo a lot of testing to ensure it will hold up under “real-world” conditions.
“The computer was right that this alloy was more effective and more stable. This was a breakthrough because it showed that this method of finding better catalysts is working,” Abed said. “What would take humans years to test, the computer can simulate in days.”
Researchers hope that AI can provide a faster route to finding the answers needed to make green energy practical and widely used.
Further information: Jehad Abed et al., “Pourbaix Machine Learning Framework Identifies Acidic Water Oxidation Catalysts with Suppressed Ruthenium Dissolution.” Journal of the American Chemical Society (2024). DOI: 10.1021/jacs.4c01353
Provided by: Canadian Light Source
Source: AI-powered team finds cheaper way to produce green hydrogen (August 29, 2024) Retrieved August 29, 2024 from https://phys.org/news/2024-08-team-ai-cheaper-green-hydrogen.html
This document is subject to copyright. It may not be reproduced without written permission, except for fair dealing for the purposes of personal study or research. The content is provided for informational purposes only.
