Oxygen vacancies beyond the dilute limit in doped CaMnO3 perovskites and implications for screening materials in thermochemical applications

TL;DR

This paper demonstrates that traditional screening methods for oxide perovskites in thermochemical energy storage are insufficient, proposing a new approach based on vacancy concentration and thermodynamic modeling to better predict material performance.

Contribution

It introduces a vacancy concentration-dependent formation energy framework and a thermodynamic model to improve screening of CaMnO3 perovskites for TCES applications.

Findings

Single OVFE is inadequate for screening due to inherent vacancies at operating conditions.

Vacancy formation energy curves aligned with experimental reduction enthalpies.

Doping mechanisms influence vacancy landscape through strain and redox environment modifications.

Abstract

Thermochemical energy storage (TCES) in oxide perovskites relies on reversible oxygen vacancy formation, and computational high-throughput screening of candidate materials has predominantly used the single oxygen vacancy formation energy (OVFE) as the key descriptor. We demonstrate that the OVFE is insufficient for screening cubic CaMnO3 perovskites, because the stoichiometric compound is not the minimum energy reference state; vacancies are inherently present at operating temperatures. Materials with negative single OVFEs are routinely excluded from screening datasets as unsuitable, but this reflects a mischoice of reference state rather than a genuine materials limitation, and risks discarding promising TCES candidates. We address this by computing OVFEs as a function of vacancy concentration using ab initio density functional theory, establishing the equilibrium vacancy concentration as the correct reference point. OVFE curves referenced to this minimum align with experimentally measured reduction enthalpies, providing a framework directly comparable to experiments. We further show that A-site and B-site doping modify the vacancy formation landscape through distinct mechanisms. A-site dopants act primarily through strain relaxation and symmetry breaking, while B-site dopants reshape the local redox environment and introduce strong configurational dependence. Finally, we develop a thermodynamic model incorporating configurational entropy that accurately predicts equilibrium oxygen stoichiometry as a function of temperature and oxygen partial pressure and reveals that selective reduction of Mn4+ versus B-site dopant ions can tune the onset temperature for vacancy formation. These results establish a screening framework for perovskite TCES materials and provide practical guidance for extending high-throughput workflows beyond the single-vacancy paradigm.