In this talk, I will focus on our recent computational work aimed at discovering new noncentrosymmetric (NCS) layered 214 Ruddlesden-Popper (RP) oxides. This was a particularly challenging problem, because NCS phases are seldom seen in the family 214-RP oxides – out of 105 compounds that have been experimentally synthesized, only two are known to be NCS phases. However, our estimated chemical space of “virtual” 214-RP oxides amount to ~3,300 compositions. Are there potential candidates in the virtual space that are NCS? We address this question by developing a novel approach that synergistically integrates group theory, informatics-based methods, and density functional theory. Our group theoretical analysis reveals that the condensation of two or more distortion modes, which describe suitable symmetry unique octahedral distortions or a combination of a single octahedral distortion with a “compositional” A-cation ordering mode would transform the aristotype into a NCS structure. With these symmetry guidelines, we formulated a data-driven informatics problem, founded on Bayes’ rules and information theoretic principles, to rationally search for combinations of A- and B-site elements that would induce the necessary octahedral distortions to stabilize NCS phases. Our data-driven models predict new andpreviously unexplored 214-RP oxides as potential NCS structures, which we evaluated with density functional theory and found promising results. I will conclude the talk highlighting the challenges we must overcome to achieve rational structure-driven design of complex materials.

Host: Turab Lookman, txl@lanl.gov