Thermally activated defect tolerance of the oxygen vacancies in CeO₂ revealed by machine-learning molecular dynamics

Thermal instability and uncontrolled defect dynamics continue to be major obstacles to the reliable functioning of oxide-based energy materials under practical operating conditions. Due to its propensity for reversibly producing oxygen vacancies, cerium dioxide (CeO₂) is widely used in energy-related applications; nevertheless, the finite-temperature stability mechanisms leading to vacancy tolerance are still poorly understood. In this work, machine-learning molecular dynamics has been applied to thoroughly examine the dynamical and thermodynamic stability of oxygen vacancies in bulk CeO₂ at device-relevant temperatures (300-500 K). By directly linking vacancy formation energetics with finite-temperature lattice dynamics, a quantitative vacancy stability paradigm that encompasses both dynamic resilience and energetic accessibility has been presented. According to our findings, the oxygen vacancies in CeO₂ dynamically accommodate lattice distortions and intrinsically modest formation energies, which allow thermal disorders without long-range structural deterioration. Even at high temperatures, the root-mean-square displacement, mean-square displacement, and radial distribution analyses show reduced defect-driven diffusion and maintained crystallinity. We create a hierarchical vacancy stability phase map that outlines a wide defect-tolerant operating window for CeO₂ by integrating these descriptors. These findings establish the physical origin of defect tolerance in ceria and provide a material-level framework for the rational design and thermal optimization of defect-engineered oxide energy materials.