Bond disproportionation, charge self-regulation and ligand holes in s-p and in d electron ABX3 perovskites by density functional theory

TL;DR

This paper uses density functional theory to explain the common phenomena of bond disproportionation, charge self-regulation, and ligand holes in various ABX3 perovskites, highlighting a universal self-regulating response mechanism.

Contribution

It demonstrates that DFT can naturally capture the SLE-to-DLE transition and associated phenomena across diverse ABX3 compounds without requiring strong correlation effects.

Findings

DFT describes SLE-to-DLE transition and ligand hole formation.

Bond length alternation and metallic to insulating transition explained.

No need for strong correlation effects beyond standard DFT.

Abstract

Some ABX3 perovskites exhibit different local environments (DLE) for the same B atoms in the lattice, an effect referred to as disproportionation, distinguishing such compounds from perovskites that have single local environments (SLE). The basic phenomenology of disproportionation involves the absence of B-atom charge ordering, the creation of different B-X bond length for different local environments, the appearance of metal (in SLE) to insulator (in DLE) transition, and the formation of ligand holes. We point out that this phenomenology is common to a broad range of chemical bonding patterns in ABX3 compounds, either with s-p electron B-metal cations (BaBiO3, CsTlF3), or noble metal cation (CsAuCl3), as well as d-electron cations (SmNiO3, CaFeO3). We show that underlying much of this phenomenology is the self-regulating response, whereby in strongly bonded metal-ligand systems with high lying ligand orbitals, the system protects itself from creating highly charged cations by transferring ligand electrons to the metal, thus preserving a nearly constant metal charge in DLE, while creating B-ligand bond alternation and ligand-like conduction band (ligand hole). We are asking what are the minimal theory ingredients needed to explain the main features of this SLE-to-DLE phenomenology, such as its energetic driving force, bond length changes, possible modifications in charge density and density of state changes. Using as a guide the lowering of the total energy in DLE relative to SLE, we show that density functional calculations describe this phenomenology across the whole chemical bonding range without resort to special strong correlation effects, beyond what DFT naturally contains. In particular, lower total energy configurations (DLE) naturally develop bond alternation, gaping of the metallic SLE state, and absence of charge ordering with ligand hole formation.