Surface helium effect on deuterium desorption from tungsten investigated via a novel 'sandwich' multilayer structure

Helium as intrinsic impurity species in a deuterium-tritium (D-T) thermonuclear fusion reactor will be retained in plasma-facing materials thereby affecting the retention and/or transport of hydrogen isotopes. However, laboratory studies introducing helium into the tungsten (W) matrix will inevitably create additional defects in the sample, which will complicate the conclusion drawn on the influence of helium on the uptake of hydrogen isotopes by W. In the present study, applying a novel ‘dumpling’ multilayer structure for temperature programmed desorption (TPD) measurements, we investigated the influence of surface helium on the thermal release behavior of deuterium (D) from D-W films. D was introduced into the W films by magnetron sputtering in D-admixed Ar atmosphere (co-sputtering). Helium was introduced into the system by either sputter deposition of a W layer in pure helium atmosphere or by low-energy helium plasma exposure of a pure-W layer on top of the D-W film. Being pre-characterized with RBS, four different samples including pure D-W layer, D-W layer covered with a co-sputtered He-W layer, with a He-implanted W layer and with a pure W layer were compared aiming at a systematic study on the effect of surface helium. The special feature of the results presented here is that the helium addition into the system does not introduce any additional defects into the D-W layer. The helium content is characterized with elastic recoil detection analysis (ERDA) using 16.4 MeV ¹⁷O and the D amount is measured using ³He nuclear reaction analysis (NRA). Both methods were performed for each sample before and after TPD measurement. Results showed that upon sample heating up to 800 K only part of the introduced helium was released independent of how helium was introduced while all D was released from each sample. The TPD spectra can be well separated for ⁴He and D₂ with the aid of the HD peak appearing always simultaneously with D₂ peak, which is very helpful in disentangling the surface helium effect on D effusion.