Accelerated materials design approaches based on structural classification: application to low enthalpy high pressure phases of SH3 and SeH3

Abstract We propose a methodology that efficiently asseses major characteristics in the energy landscape for a given space of configurations (crystal structures) under pressure. In this work we study SH3 and SeH3 , both of fundamental interest due to their superconducting properties. Starting from the crystal fingerprint, which defines configurational distances between crystalline structures, we introduce an optimal one dimensional metric space that is used to both classify and characterize the structures. Furthermore, this is correlated to the electronic structure. Our analysis highlights the uniqueness of the Im − 3m phase of H3S and H33Se for superconductivity. This approach is an useful tool for future material design applications.