Model development of plasma implanted hydrogenic diffusion and trapping in ion beam damaged tungsten

A Cu ion beam is used to induce controlled levels of damage (10(-3), 10(-2), and 10(-1) dpa) in room temperature W samples. A single 5 MeV beam energy yielding a peaked damage profile 0.8 mu m into the material, or three beam energies (0.5, 2, and 5 MeV) producing a relatively uniform damage profile from the near surface up to 0.8 mu m were used. The W samples were then exposed to a D plasma ion fluence of 10(24) ions m(-2) at 380 K, and the resulting D retention was measured using the D(He-3, p)He-4 reaction analysis (NRA) and thermal desorption spectroscopy (TDS). We observe that within experimental error there is no significant difference in retention whether the damage profile is peaked or uniform. The increase in retention is observed to increase proportional to dpa(0.66) estimated from the dpa peak calculated from the SRIM program. A simplified retention model is proposed that provides concentration profiles that can be directly compared to NRA data and total retention measurements. Taking the trapping energies due to three defect types calculated from density functional theory (DFT), the only free-parameters are three defect densities of in-grain monovacancies, dislocations, and grain boundary vacancies, and we assume these defects to be the dominant trapping locations. The model can fit D retention data in a pristine W sample within the experimental error of the measurements, and in subsequent modeling these intrinsic defect densities are then fixed. We model the retention profile after ion damage by adding the SRIM predicted vacancy profile to the intrinsic monovacancy defect density. Since the increase in retention, and therefore the increase in vacancy production, does not increase linearly with dpa, a correction factor is multiplied to the predicted vacancy profile to fit the data. A new diffusion coefficient is calculated with the model that is a function of the concentration of trapped atoms. This calculation may resolve discrepancies of various diffusivity measurements and models in the literature.