Game-changing method makes green hydrogen production cheaper

Researchers are using AI to advance breakthroughs in producing green hydrogen.

Normally, to produce green hydrogen, electricity generated from renewable sources is passed between two metal electrodes in water, resulting in the release of oxygen and hydrogen gases. However, this process is currently energy-intensive and relies on metals that are both rare and costly.

However, the researchers confirmed with the Canadian Light Source at the University of Saskatchewan that an AI-generated catalyst recipe offers a more efficient method for producing hydrogen fuel.

They have been seeking the optimal alloy or metal combination to act as a catalyst, aiming to make the reaction more efficient and affordable. Traditionally, this search involved time-consuming trial and error in the lab, which is inefficient for such a complex task.

New AI program accelerates alloy search

According to Jehad Abed, who was part of a team that developed a computer program to significantly speed up the process, the search involves hundreds of millions or billions of alloy candidates, any of which could be the right answer.

The AI program developed by the team analyzed over 36,000 metal oxide combinations through virtual simulations to determine the most promising options. Abed then tested the top candidate in the lab to verify the program’s predictions.

Next, the team utilized the CLS’s ultra-bright X-rays to assess the catalyst’s performance during the reaction.

“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,” Abed explained.

Alloy of ruthenium, chromium, and titanium emerges as top choice

The researchers also utilized the Advanced Photon Source at Argonne National Laboratory in Chicago. As Abed pointed out, the alloy—combining ruthenium, chromium, and titanium in specific proportions—proved to be the clear winner.

The computer’s recommended alloy demonstrated 20 times better stability and durability than the benchmark metal, proving to be both long-lasting and efficient.

While the AI program shows great promise, the material still requires extensive testing to ensure it performs well under real-world conditions, Abed claims.

The computer accurately identified the alloy as more effective and stable, marking a breakthrough in the method for finding better catalysts. This process, which would take years for a person to test, can be simulated by the computer in just days.

Thus, the researchers remain hopeful that AI will provide a faster route to making green energy practical for widespread use.