A new algorithm opens the door for using artificial intelligence and machine learning to study the interactions that happen on the surface of materials.
Scientists and engineers study the atomic interactions that happen on the surface of materials to develop more energy efficient batteries, capacitors, and other devices. But accurately simulating these fundamental interactions requires immense computing power to fully capture the geometrical and chemical intricacies involved, and current methods are just scratching the surface.
"Currently it's prohibitive and there's no supercomputer in the world that can do an analysis like that," says Siddharth Deshpande, an assistant professor in the University of Rochester's Department of Chemical Engineering. "We need clever ways to manage that large data set, use intuition to understand the most important interactions on the surface, and apply data-driven methods to reduce the sample space."
By assessing the structural similarity of different atomic structures, Deshpande and his students found that they could get an accurate picture of the chemical processes involved and draw the relevant conclusions by analyzing just two percent or fewer of the unique configurations of surface interactions. They developed an algorithm reflecting this insight, which they described in a study published in Chemical Science.
In the study, the authors used the algorithm to, for the first time, analyze the intricacies of a defective metal surface and how it affects the carbon monoxide oxidation reaction, which can, in turn, aid in understanding the energy losses in an alcohol fuel cell. Deshpande says the algorithm they developed supercharges density functional theory, a computational quantum mechanical modeling method that he calls the "workhorse" for the past several decades for studying the structure of materials.
"This new method becomes the building ground to incorporate machine learning and artificial intelligence," says Deshpande. "We want to take this to more difficult and challenging applications, like understanding the electrode-electrolyte interference in batteries, the solvent-surface interactions for catalysis, and multi-component materials such as alloys."
