AI used to find new catalyst for CO2 conversion

Researchers have used artificial intelligence (AI) to find an improved method for transforming waste carbon into valuable products.

A team led by the University of Toronto enlisted the help of machine learning to speed up the process of identifying a catalyst that can efficiently convert carbon dioxide (CO2) into ethylene.

The result of the AI project is an electrocatalyst that has unprecedented efficiency. In addition, the new catalyst can be used with wind or solar power to store electricity.

“Using clean electricity to convert CO2 into ethylene, which has a $60 billion global market, can improve the economics of both carbon capture and clean energy storage,” said study senior author Professor Ted Sargent.

The research team has previously developed a number of world-leading catalysts to reduce the energy cost of the reaction that converts CO2 into ethylene and other carbon-based molecules. 

Better catalysts are always possible, but testing millions of potential material combinations is not realistic. This is where machine learning becomes very useful. Computer models, theoretical data, and algorithms can be used to identify the best candidates for catalysts without exhaustive tests. 

Professor Zachary Ulissi of Carnegie Mellon University specializes in the computer modeling of nanomaterials.

“With other chemical reactions, we have large and well-established datasets listing the potential catalyst materials and their properties,” said Professor Ulissi.

“With CO2-to-ethylene conversion, we don’t have that, so we can’t use brute force to model everything. Our group has spent a lot of time thinking about creative ways to find the most interesting materials.”

The algorithms created by Professor Ulissi and his team can predict what kinds of products a given catalyst is likely to produce, even without detailed modeling of the material itself.

The algorithms were applied to over 240 different materials, and four were predicted to have desirable properties for CO2 conversion. 

The top performer was found to be an alloy of copper and aluminum. After the two metals were bonded at a high temperature, some of the aluminum was etched away, leaving behind a nanoscale porous structure.

The new catalyst was tested in an electrolyzer, and its performance was the best ever recorded for this type of catalyst. 

According to Professor Sargent, the energy cost will need to be lowered even further so the system can produce ethylene that is cost-competitive. 

The strong performance of the new catalyst suggests that machine learning is a very strategic method to screen for potential catalysts.

“There are many ways that copper and aluminum can arrange themselves, but what the computations showed is that almost all of them were predicted to be beneficial in some way,” said Professor Sargent. “So instead of trying different materials when our first experiments didn’t work out, we persisted, because we knew there was something worth investing in.”

The study is published in the journal Nature.

—

By Chrissy Sexton, Earth.com Staff