Wide-range predictions of hydrogen-dependent vacancy diffusion in nickel from a near-DFT-accurate machine-learning potential

Vacancy-mediated atomic transport plays a central role in high-temperature creep and microstructural evolution. Pure Ni provides a representative model system because Ni is the principal matrix element of many Ni-based heat-resistant alloys, while its prototypical FCC structure enables the intrinsic effects of hydrogen to be isolated. However, quantitative understanding of how hydrogen modifies vacancy formation, migration, and the resulting self-diffusion in Ni over wide temperature and pressure ranges remains limited. We therefore develop a near-DFT-accurate machine-learning moment tensor potential for the Ni–vacancy–hydrogen system and map the vacancy diffusion coefficient over 300 – 1500 K and hydrogen gas pressures – . The equilibrium vacancy concentration is obtained from the vacancy formation free-energy, while the hydrogen-dependent jump frequency is evaluated from the activation free-energy using the excess hydrogen thermodynamic integration method. The maps show that hydrogen suppresses vacancy diffusion at low temperatures over the entire investigated range of , whereas it enhances vacancy diffusion at high temperatures. Incorporating the hydrogen-enhanced equilibrium vacancy concentration, the vacancy-mediated Ni self-diffusivity increases under hydrogen charging for all conditions, with the strongest acceleration at low temperature and high H 2 pressure.