Study of helium diffusion in yttria: A multiscale approach based on the density functional theory and kinetic Monte Carlo, with transmission electron microscopy and thermo-desorption spectroscopy

Abstract

A known issue for future nuclear reactors is helium accumulation inside the steel structure materials, responsible for structural issues such as embrittlement and cracking. One possible solution is using new types of reinforced steel, such as oxide dispersion strengthened (ODS) steel. It consists of adding oxide nanoparticles to the Fe-based material, especially yttrium oxide (yttria, Y$_2$O$_3$), improving its properties. Therefore, one first step is understanding the helium diffusion inside this system. Very little is known about helium inside yttria, with most studies being theoretical ones. Based on this context, this work proposes a combined theoretical and experimental multiscale approach to investigate helium diffusion inside yttria. The theoretical approach starts with the density functional theory, used to model the atomic yttria cell and determine helium insertion sites. The transitions between the sites were described using the NEB method. Kinetic Monte Carlo was then employed to obtain the interstitial diffusion coefficient expression. It showed a limited diffusion at temperatures below 600 K, which may indicate a tendency for He to be blocked in the oxide. Then, the charged vacancies were explored. It showed that the vacancy further reduces helium diffusion. Finally, it was demonstrated that helium spreads across different vacancies and interstitial sites. The experimental part involved implanting helium ions at 50 keV in samples with nanometric or micrometric grains. Then, the specimens were characterised with transmission electron microscopy (TEM) and thermo-desorption spectroscopy (TDS) techniques. TEM did not evidence detectable bubbles even at the highest studied fluence (1\time 10^{16}$ cm$^{-2}$). The TDS highlighted different mechanisms for helium diffusion and the grain size's role, providing a model for diffusion coefficient calculation based on interstitial diffusion.