# Magnetic Superexchange and Mott Insulator Mechanisms in Cubic Perovskites: From First-Principles to Canonical Models

## TL;DR

This paper explores magnetic and insulating properties of perovskite materials using simulations and finds that traditional models miss key factors in stabilizing these states.

## Contribution

The study reveals that bonding orbitals and electronic backdonation, not accounted for in traditional models, are crucial for stabilizing insulating states in perovskites.

## Key findings

- Antiferromagnetic ordering is predicted but stabilized by bonding orbitals not included in canonical models.
- Electronic backdonation plays a key role in stabilizing insulating states via different mechanisms in KNiF3 and KVF3.

## Abstract

The ground state
of many insulating, open-shell transition-metal
perovskites with a 180° metal–ligand–metal bridge
is antiferromagnetic (AFM), as predicted by Anderson’s superexchange
interaction or Hubbard’s model. These well-established, standard
models show how these systems are insulators due to the minimization
of the interactions between electrons, at the cost of localizing the
electrons on the metal ions. In this work, we carry out first-principles
simulations on the cubic perovskites KNiF3 and KVF3, analyzing electron densities, energies and bond indices.
Although our calculations predict an antiferromagnetic ordering (AFM),
in agreement with canonical superexchange models, we show through
various indicators that the stabilization of this phase is not mainly
associated with the antibonding magnetic orbitals but rather with
bonding orbitals not included in the models. In particular, these
traditional descriptions of superexchange do not adequately describe
the ligand-to-metal electronic backdonation, which is an important
element for stabilizing the insulating state of the two studied perovskite
fluorides, albeit by diametrically different mechanisms: (1) reducing
electron–electron repulsion in KNiF3, as proposed
by Hubbard, whereas (2) enhancing electron–nuclear attraction
in KVF3. Our findings highlight some of the limitations
of these foundational models and offer a novel perspective on the
understanding of magnetism.