Using AI to screen vast numbers of element combinations and identify the most promising candidates for detailed quantum analysis, an international research consortium has demonstrated that machine learning can dramatically accelerate the discovery of superconducting materials.
The SuperC consortium, led by Professor Paivi Torma of Aalto University, used this approach to identify two previously unknown superconductors: YRu3B2 and LuRu3B2. Both have properties derived from electrons forming flat bands within the kagome lattice structure. The machine learning algorithm first screens a huge number of possible material combinations and performs quantum calculations on the most powerful candidates. Collaborators at Rice University, led by Professor Emilia Morosan, then synthesized and experimentally tested both materials. The results of this study were published in the journal Physical Review Research.
The key lies in the scale this method allows. Of the more than 7,000 known superconductors, researchers have been able to predict the theoretical feasibility of only about 20 due to the computational complexity involved. Tolma said the AI-driven approach could potentially increase the number of materials that can be screened into the billions.
The consortium’s broader goal is to discover room-temperature superconductors by 2033. Such materials have the potential to radically reduce global energy consumption, Tolma argued, especially in computing and data center infrastructure, where heat generation is a significantly increasing cost. SuperC was founded in 2023 and is funded by sources including the Kavli Foundation and the Jane and Aatos Erkko Foundation.
sauce
