Polished W discs exposed to pure He plasma in the PISCES-B linear-divertor-plasma simulator at 1120 and 1320 K are found to develop deeply nanostructured surface layers consisting of a conglomerate of amorphous ‘nanorods’. The growth of the thickness of the nanostructured layer is explored for exposure times spanning 300–(2.2 × 104) s in He plasmas of density ne ∼ 4 × 1018 m−3 and temperature Te ∼ 6–8 eV where the average He-ion surface-impact energy is ∼60 eV, below the threshold for physical sputtering. A nanostructured layer in excess of 5 µm thick is observed for the longest exposure time explored. The kinetics of the layer growth are found to follow Fick's law, characterized by an effective diffusive mechanism with coefficients of diffusion: D1120 K = 6.6 ± 0.4 × 10−12 cm2 s−1 and D1320 K = 2.0± 0.5 × 10−11 cm2 s−1. The diffusion of He atoms in W is considered too rapid to explain the observed growth of surface modification and points to the interplay of other mechanisms, such as the availability of thermal vacancies and/or the slower diffusion of He through the forming nanostructured layer.
Blister formation and D retention in W have been investigated for low energy (∼55 ± 15 eV), high flux (∼1022 m−2 s−1), high fluence (⩽4.5 × 1026 m−2) ion bombardment at moderate temperature (∼573 K) in mixed species D+He plasmas in the linear divertor plasma simulator PISCES-A. The amount of D retained in W is found to decrease significantly when compared with that in W exposed to pure D plasmas, as measured with high resolution thermal desorption spectroscopy. Scanning electron microscopy observations reveal the suppression of the blisters, a surface feature known to drive up retention, in the D + He mixture plasma exposed W samples. Reduced D retention is accompanied by the formation of nano-sized high density He bubbles in the near surface, observed with a transmission electron microscope (TEM). It is believed that the nano-bubbles act as a diffusion barrier to implanted D atoms and consequently reduce the amount of uptake in the W material. This newly observed effect implies that current predictions of D retention in W, in actual fusion devices, may be overestimated, since there will be He ash in fusion plasma. Toughness enhanced, fine-grained (grain size of ∼1 µm) W–TiC samples, exposed to pure D plasma conditions, also show little or no evidence of blistering. The measured D retention in the W–TiC samples was approximately 1 × 1019 D m−2 corresponding to about 2 × 10−7 of the implanted D fluence, and is very low compared with the retention in pure stress relieved W, which exhibited surface blisters and had a D retention of about 1 × 1021 D m−2.