Revealing the Non-Arrhenius Migration of Oxygen Vacancies at the CeO₂(111) Surface

Ceria (CeO₂) is renowned for its exceptional oxygen storage capacity, which makes it highly valuable in various applications. Central to its functionality is the migration of oxygen vacancies (V_O's). While previous studies have extensively examined the distribution of V_O's and Ce³⁺ polarons, their kinetic behaviors and interactions, especially in the presence of multiple vacancies, are not yet fully understood. In this study, we employ density functional theory (DFT), ab initio molecular dynamics (AIMD) simulations, and machine-learning methods to investigate these phenomena. Our findings reveal a nonmonotonic temperature dependence of the migration rate of V_O's at the CeO₂(111) surface. Our theoretical model further demonstrates that the migration of V_O's and the hopping of polarons are intricately coupled. Notably, frequent polaron hopping at high temperatures hinders V_O migration, indicating a non-Arrhenius mechanism. This finding is further validated through long-time molecular dynamics simulation enhanced with neural network potentials. Our results provide a microscopic understanding of the interplay between V_O's and Ce³⁺ polarons, offering crucial insights into the complex dynamics governing oxygen vacancy migration in ceria. This knowledge paves the way for improved material design and functionality.