Researchers from the National Institute of Standards and Technology (NIST) have integrated density functional theory (DFT) and artificial intelligence (AI) to discover new hydride superconductors. This approach uses high-throughput DFT calculations and a graph neural network (GNN) model to predict superconducting transition temperatures under various pressures. The team found over 120 promising hydride structures. This method enhances the efficiency and accuracy of screening potential superconductors, aiming to facilitate the discovery of materials that work at higher temperatures and more practical conditions.
For more details, visit Phys org: