Time-modulated helium (He) ion energy (e.g. VBias = −50 + 25·sin(2πfRF · t), fRF = 13.56 MHz) is demonstrated to strongly affect the development of tungsten (W) surface morphology that results from He plasma irradiation in the DIONISOS linear plasma experiment. Nano-tendril bundles (NTBs), which appear as isolated 'islands' of nano-tendrils, can rapidly grow on an otherwise smooth W surface. This is in contrast to previously seen full-surface coverage of nano-tendril growth known as 'fuzz'. When tall NTBs form, less than 15% of the surface contains nano-tendrils. The NTB surface coverage changes with growth conditions and the total volume of nano-tendrils in the NTBs is observed to be up to a factor of 16 larger than when fuzz is grown. This indicates that long-range W surface transport underlies nano-tendril formation.Surface temperature 870–1220 K, the DC bias potential −30 to −70 V, and the ion flux density 4.4 × 1021–1.1 × 1022 He · m−2 · s−1 are varied in the experiments. NTBs form at similar conditions as fuzz with the critical difference being the RF modulation of the ion energy bombarding the W, another indication of the importance of W surface transport. Mass loss measurements indicate net erosion with a yield of 1–8 × 10−4 W/He when NTBs form; erosion that is not attributable to chemical or physical sputtering by He or impurities in the plasma. The erosion is correlated to the NTB growth, based on post-exposure inspection by electron microscopy indicating that NTBs are prone to loss from the surface. NTB growth is compared to the empirical growth-erosion model of fuzz, showing NTBs grow up to a factor of 100 times taller than the expected fuzz layer depth under DC bias conditions. Insights into nano-tendril growth provided by this new growth regime are discussed. Strategies to mitigate W fuzz growth may inadvertently result in rapid localized nano-tendril bundle growth with a higher probability of dust production.
Simulation of fuzz formation on tungsten surface and its impact on sputtering and hydrogen retention by MD-MC hybrid method