Magnetic ordering effects on vacancy migration in austenitic Fe

Austenitic steel is an attractive construction material for high-temperature applications, with service lifetimes being inversely proportional to metal vacancy diffusion rates. Despite the ubiquity of austenitic steel, there remains significant disagreement in how to calculate Fe vacancy formation and vacancy migration energies due to ambiguity in how to model the Fe paramagnetic (PM) ordering. Specifically, density functional theory (DFT) can efficiently adopt either nonmagnetic (NM) and spin-polarized (SP) approaches, but not directly access the disordered state of PM ordering. This study uses a combination of DFT, Ising models, and kinetic Monte Carlo (KMC) simulations to elucidate the proper magnetic structure that should be used in simulating austenitic Fe and the resulting diffusion rates. Ising models based on DFT spin-flip energetics show Fe vacancies stabilize local ferromagnetic (FM) domains, and that these domains exist up to the melting temperature even without the presence of a vacancy. Thus, SP calculations which collapse into FM ordering should be used in DFT calculations. DFT calculations demonstrate that FM lowers the vacancy hopping barrier Em v compared to NM calculations. Further, KMC results demonstrate that >99.9% of vacancy hops occur in local FM domains. Overall, this work demonstrates that vacancy formation and hopping are best modeled by FM-initialized SP calculations in fcc Fe, and by extension in austenitic stainless steel. It also lays the foundation for appropriate vacancy migration modeling in other materials operated within their PM temperature ranges.