False metals and real insulators*

Very often, the prediction of metallic electronic structure in mean-field theory for compounds known to be insulators (listed below) is attributed to a lack of correlation. We show that the intricate coupling of electronic, spin and lattice degrees of freedom afforded by symmetry broken mean-field theories such as DFT (rather than strong or dynamic correlation) correctly predicts the appropriate insulating or metallic state in a range of compounds. The symmetry-breaking modes that lower the total energy while opening a gap include:
(i) spin ordering as demonstrated in CuBi₂O₄, NiO, and YNiO₃;
(ii) formation of a distribution of local spin (polymorphous) networks as shown in paramagnetic LaTiO₃, and YTiO₃;
(iii) structural symmetry breaking such as (a) bond disproportionation in Li-doped TiO₂, SrBiO₃, and rare earth nickelates, (b) pseudo-Jahn-Teller distortions in LaMnO₃, and (c) spontaneous formation of vacancy arrays in Ba₄As₃;
(iv) allowing spin-orbit coupling in CaIrO₃ and Sr₂IrO₄.
Importantly, the accounting of the above symmetry-breaking modes allows distinguishing real metals (e.g., SrVO₃ and Ca₆Al₇O₁₆) from the above example of false metals.
O.I. Malyi, A. Zunger Applied Physics Reviews, doi: 10.1063/5.0015322.