First principles study of rare-earth oxides

TL;DR

This study uses a first-principles approach to investigate the electronic structure, valencies, and stability of rare-earth oxides, explaining their natural occurrence and predicting metastable phases.

Contribution

It applies a self-interaction-corrected local-spin-density method to determine valencies and stability of RE oxides, providing insights into their electronic structure and natural occurrence.

Findings

Ce, Pr, Tb in dioxides are tetravalent

Sesquioxides are predominantly trivalent

Predicted metastable NdO₂ phase

Abstract

The self-interaction-corrected local-spin-density approximation is used to describe the electronic structure of dioxides, REO$_2$, and sesquioxides, RE$_2$O$_3$, for the rare earths, RE=Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy and Ho. The valencies of the rare earth ions are determined from total energy minimization. We find Ce, Pr, Tb in their dioxides to have the tetravalent configuration, while for all the sesquioxides the trivalent groundstate configuration is found to be the most favourable. The calculated lattice constants for these valency configurations are in good agreement with experiment. Total energy considerations are exploited to show the link between oxidation and $f$-electron delocalization, and explain why, among the dioxides, only the CeO$_2$, PrO$_2$, and TbO$_2$ exist in nature. Tetravalent NdO$_2$ is predicted to exist as a metastable phase - unstable towards the formation of hexagonal Nd$_2$O$_3$.