Unraveling Surface Hydroxylation Equilibrium and Proton Dynamics at the Water and Monoclinic ZrO ₂ Interface from Reactive Molecular Dynamics Simulations

Despite their importance in catalysis and electrochemical energy conversion, the equilibrium structure and proton dynamics of water/metal-oxide interfaces remain difficult to access at the molecular level. Herein, we investigate the water/monoclinic ZrO 2 (1̅11) interface using reactive molecular dynamics simulations enabled by a neural network potential trained on density functional theory calculations. Nanosecond-scale simulations with extended surface models reveal a strongly site-dependent surface hydroxylation equilibrium: 2-fold-coordinated oxygen atoms remain persistently protonated, whereas a fraction of 3-fold-coordinated sites undergo transient hydroxylation with an average lifetime of ∼200 ps. Proton transfer events occur frequently within the interfacial adlayer, and hydroxide ion migration via the Grotthuss-like mechanism is analyzed, yielding a diffusion coefficient of 19.3 Å 2 ns –1, indicating active ion transport on the surface. This work demonstrates the capability of machine-learning-potential-based molecular dynamics to resolve surface hydroxylation equilibrium and ion-transport phenomena at complex water/metal-oxide interfaces.