Strongly Correlated Electronic Properties of FeO Studied by the SCAN+U Functional

Abstract

The recently developed strongly constrained and appropriately normed (SCAN) meta-GGA exchange-correlation functional has been demonstrated to be able to give an accurate description of strongly correlated cuprates. However, it remains to be seen if SCAN calculations can predict the physics of compounds involving partially filled 3d orbitals. In this work, we present the first-principles electronic structure studies about the phase transition of FeO using the SCAN functional. Compared with the exchange-correlation functionals widely used in density functional theory, such as the local-density approximation (LDA) or the generalized gradient approximation (GGA), SCAN correctly predicts an antiferromagnetic insulating ground state. However, it still significantly underestimates the band gap for compounds involving partially filled 3d orbitals and, hence, predicts an incorrect phase transition pressure of FeO. Interestingly, the electronic structure and band gap properties of FeO from the SCAN calculations are similar to those from the GGA+U calculations with U0 = 0.7 eV. An on-site interelectronic interaction U is therefore still needed in order to correctly describe the phase transition physics of FeO for the SCAN calculations, predominantly as a simplified self-interaction-error reduction term that enhances the spatial compactness of 3d orbitals. The required value of U in SCAN+U is less than that required in GGA+U, and their difference is rightly equal to the value of U0. Most importantly, in comparison with GGA+U, the SCAN+U method no longer needs to use the prior-known knowledge to break orbital symmetry avoiding the metastability issue, benefiting from the spontaneous symmetry breaking it allows. The calculated properties of FeO from SCAN+U agree well with experiments.