Simulation of fuzz formation on tungsten surface and its impact on sputtering and hydrogen retention by MD-MC hybrid method

Tungsten (W) is considered the primary material for the divertor and first wall in future fusion reactors. Under the irradiation of helium (He) plasma, a porous layer named ‘fuzz’ forms on W surface, accompanied by changes in its physical properties. In this paper, the early stage of fuzz is constructed using a molecular dynamics (MD) and Monte Carlo (MC) hybrid method. It provides a more physically realistic structure by allowing the nanostructure to emerge naturally from the interplay of atomic-scale processes. The generated fuzz structures are subsequently employed to evaluate the changes in surface properties, including both sputtering and hydrogen (H) permeation/retention. Simulation results show that the fuzz layer ceases to grow once it reaches a certain thickness. Further analysis indicates that the formation of fuzz structure can suppress the diffusion and accumulation of He from the surface to bulk material. The simulation results also indicate that higher ion energy results in greater He retention, which consequently leads to an increase in the porosity of the fuzz layer. Simulations of He atoms cascades show that, although the energy directly transferred from He atoms (<100 eV) to W atoms is lower than the sputtering threshold energy (∼8.9 eV), W surface modifications and sputtering can still occur via adatoms formation. Compared with flat W surface, fuzzy surface tends to recapture the sputtering W atoms, exhibiting lower net sputtering yield under the same He irradiation conditions. The bombardment by He atoms can also promote the curling of fuzz structures, which may subsequently influence W surface diffusion. Finally, the He–H hybrid simulations indicate that the formation of fuzz structure significantly suppresses the H permeation and retention in the W bulk area and enhances H retention on W surface. © 2026 The Author(s). Published by IOP Publishing Ltd on behalf of the IAEA.