Interfacial Properties of Water on Highly Oriented Pyrolytic Graphite Trained with Machine-Learned Force Fields

This work investigates the structural and dynamical behavior of confined water at highly oriented pyrolytic graphite (HOPG) surfaces containing atomic vacancies and heteroatom dopants (N, O, and S) using ab initio molecular dynamics simulations enhanced with machine-learned force fields (MLFF). Both structural defects and chemical dopants significantly modify interfacial water structure, hydrogen bonding, and molecular mobility. Vacancies increase surface-energy heterogeneity and promote stronger water adsorption, while dopants introduce polar sites that enhance hydrogen bonding and water-surface coupling. Oxygen and nitrogen doping markedly strengthen interfacial interactions, whereas sulfur doping produces moderate effects. Velocity autocorrelation function and vibrational density of state analyses reveal enhanced damping and red-shifted low-frequency modes, consistent with stronger hydrogen bonding and restricted molecular motion. Mean square displacement and diffusion analyses confirm increased perturbation of water dynamics near vacancy sites at the surface compared to water farther from the surface, along with reduced mobility of interfacial water near doped surfaces relative to pristine HOPG. Overall, surface heterogeneity enhances interfacial ordering while suppressing dynamics, offering molecular-level insight for tailoring the hydrophilicity and reactivity of graphitic materials in catalytic, electrochemical, and energy applications.