https://phys.org/news/2018-10-discovery-superconducting-materials-informatics.html

This joint research team took advantage of the AtomWork database, which contains more than 100,000 pieces of data on inorganic crystal structures. The team first selected approximately 1,500 candidate material groups whose electronic states could be determined through calculation. The team then narrowed this list to 27 materials with desirable superconducting properties by actually performing electronic state calculations. From these 27, two materials SnBi2Se4 and PbBi2Te4 were ultimately chosen because they were relatively easy to synthesize.

The team synthesized these two materials and confirmed that they exhibit superconductivity under high pressures using an electrical resistivity measuring device. The team also found that the superconducting transition temperatures of these materials increase with increasing pressure. This data science-based approach, which is completely different from the conventional approaches, enabled identification and efficient and precise development of superconducting materials.