Symmetry breaking in density functional theory captures bulk peculiarities that were assumed to exclusively emerge from electron correlation*

Mean-field band theory, e.g., density functional theory (DFT), is often based on the symmetry-restricted, minimal-cell model, with the smallest number of possible magnetic and structural degrees of freedom. Such symmetry-restricted DFT fails to reproduce many experimental observations, such as local distortions, mass enhancement, and nematicity (in some Fe-based materials). It has been argued that the strong correlation approach is the exclusive theory needed for describing such peculiarities. We report that by removing the constraint of the minimal cell, the energy minimization in DFT finds in enlarged cells (large enough to allow for symmetry breakings) atomic and spin symmetry breakings, which can explain many observed anomalies and provide close agreements with experiments. We show how the energy-lowering symmetry breaking effects in mean-field DFT capture mass enhancement [1] in a range of perovskites e.g. d-electron SrVO3, SrTiO3, BaTiO3, LaMnO3 as well as p-electron CsPbI3 and SrBiO3, and local distortions and nematicity in iron-based compound FeSe [2]. Thus, symmetry breaking in DFT could describe effects that were previously attributed exclusively to complex correlated treatments.
[1] arXiv:2006.10099
[2] arXiv:1911.02670