Search

Tungsten samples were heat-treated to achieve partial recrystallization and exposed to high ion flux deuterium plasma at different temperatures and fluences. Continuous stiffness nanoindentation measurements of near-surface hardness were performed in the grains of specific annealing states and of specific crystallographic orientation, determined by electron backscatter diffraction (EBSD); indentation pile-up was investigated using surface profilometry. Bulk hardness of unexposed tungsten does not strongly depend on grain orientation, but depends on the annealing state of the grain, with values between ∼4.3 GPa for recrystallized grains and ∼5.5 for non-recrystallized ones. Grains with surface normal orientation feature the least pile-up, while grains with orientation the most; pile-up also depends on the annealing state, being generally lower in recrystallized grains. Plasma exposure leads to the increase of hardness, most significantly near the surface. The width of plasma-affected zone increases with the increase of exposure temperature and fluence, as well in recrystallized grains, correlating with the increase of diffusion depth. Plasma exposure does not lead to the emergence of orientation-dependence of hardness. Both indentation pile-up and near-surface indentation pop-ins are generally suppressed by plasma exposure.

A collective Thomson scattering system has been developed for measuring ion temperature, plasma velocity and impurity concentration in the high density magnetized Magnum-PSI plasma beam, allowing for measurements at low temperature (4 × 1020 m−3, while avoiding laser plasma heating caused by inverse Bremsstrahlung. The collective Thomson scattering system is based on the fundamental mode of a seeded Nd:YAG laser and equipped with an LIVAR M506 camera (EBABS technology). The first collective Thomson scattering measurements are taken at the linear plasma generator Pilot-PSI, 40 mm downstream of the cascaded arc source. At this location, the ion temperature is about equal to the electron temperature in the bulk of the plasma beam.

Diamond is a promising candidate for enhancing the negative-ion surface production in the ion sources for neutral injection in fusion reactors; hence evaluation of its reactivity towards hydrogen plasma is of high importance. Single crystal and polycrystalline diamond samples were exposed in Pilot-PSI with the D+ flux of (4‒7)·1024 m−2s−1 and the impact energy of 7–9 eV per deuteron at different surface temperatures; under such conditions physical sputtering is negligible, however chemical sputtering is important. Net chemical sputtering yield Y = 9.7·10−3 at/ion at 800 °C was precisely measured ex-situ using a protective platinum mask (5 × 10 × 2 μm) deposited beforehand on a single crystal followed by the post-mortem analysis using Transmission Electron Microscopy (TEM). The structural properties of the exposed diamond surface were analyzed by Raman spectroscopy and X-ray Photoelectron Spectroscopy (XPS). Gross chemical sputtering yields were determined in-situ by means of optical emission spectroscopy of the molecular CH A-X band for several surface temperatures. A bell-shaped dependence of the erosion yield versus temperature between 400 °C and 1200 °C was observed, with a maximum yield of ∼1.5·10−2 at/ion attained at 900 °C. The yields obtained for diamond are relatively high (0.5–1.5)·10−2 at/ion, comparable with those of graphite. XPS analysis shows amorphization of diamond surface within 1 nm depth, in a good agreement with molecular dynamics (MD) simulation. MD was also applied to study the hydrogen impact energy threshold for erosion of [100] diamond surface at different temperatures.

International audience; This study demonstrates a fast one-step synthesis method for nanostructured carbon microparticles on graphite samples using high-flux plasma exposure. These structures are considered as potential candidates for energy applications such as Li-ion batteries and supercapacitors. The samples were exposed to plasmas in the linear plasma generator Pilot-PSI with an average hydrogen ion-flux of ~10 24 m-2 s-1. The parameter window was mapped by varying the ion energy and flux, and surface temperature. The particle growth depended mainly on the sample gross-erosion and the resulting hydrocarbon concentration in the plasma. A minimum concentration was necessary to initiate particle formation. The surface of the sample was covered with microparticles with an average growth rate of 0.2 μm/s, which is significantly faster than most chemical methods. The particles were initially volumetrically grown in in the gas-phase by a multi-phase process and after deposition on the sample their growth proceeded. Scanning and transmission electron microscopy reveal that the core of these microparticles can be made of an * Corresponding author. E-mail: d.aussems@differ.nl (D.U.B. Aussems) 2 agglomeration of nanoparticles, surrounded by crumpled layers of carbon nanowalls. Gas sorption analysis shows sufficient meso-and macropores for fast mass transport. In conclusion, this processing technique could be a novel synthesis route to nanostructure surfaces for electrochemical applications.

Cavity-ring-down spectroscopy (CRDS) was implemented to measure the re-deposition of liquid tin under a high flux plasma beam in the linear plasma device Pilot-PSI. A capillary porous system (CPS) consisting of a molybdenum cup and tungsten meshes (pores diameters of 0.2 mm and 0.44 mm) was filled with tin and exposed to argon plasma. The absorption of a UV laser-beam at 286.331 nm was used to determine a number of sputtered neutral tin atoms. The incoming flux of argon ions of ∼50 eV was 1.6-2.7 × 1023 m-2 s-1, and the sample temperature measured by pyrometry varied from 850 °C to 1200 °C during exposures. The use of CRDS for measuring absolute number of particles under such plasma exposure was demonstrated for the first time. The number of sputtered tin particles in the cavity region assuming no losses would be expected to be 5.5 × 1011-1.2 × 1012 while CRDS measurements showed only 5.7-9.9 × 108. About 98-99.8% of sputtered particles were therefore found to not reach the CRDS observation volume. Spectroscopic ratios of Sn I to Sn II ions, as well as equilibrium considerations, indicate that fast ionization as well as plasma entrainment of neutrals is responsible for the discrepancy. This would lead to high re-deposition rates, implying a lowered contamination rate of core plasma and lower required replenishment rates at high-flux conditions than would otherwise be expected.

Liquid metal infused capillary porous structures (CPSs) are considered as a potential divertor solution for DEMO due to their potential power handling capability and resilience to long term damage. In this work the power handling and performance of such Sn-based CPS systems is assessed both experimentally and via modelling. A Sn-CPS target was exposed to heat fluxes of up to 18.1 MW m−2 in He plasma in the Pilot-PSI linear device. Post-mortem the target showed no damage to nor any surface exposure of the underlying W-CPS felt. The small pore size (∼40µm) employed resulted in no droplet formation from the target in agreement with calculated Rayleigh-Taylor and Kelvin-Helmoholtz instability thresholds. The temperature response of the Sn-target was used to determine the thermal conductivity of the mixed Sn-CPS material using COMSOL modelling. These values were then used via further finite element analysis to extrapolate to DEMO relevant monoblock designs and estimate the maximum power handling achievable based on estimated temperature windows for all component elements of the design. For an optimized design a heat-load of up to 20 MW m−2 may be received while the use of CPS also offers other potential design advantages such as the removal of interlayer requirements. Keywords: Liquid-metal, DEMO, CPS, Tin, Power-exhaust, Plasma-wall interaction

Samples made of tungsten (W) doped either with titanium carbide (W-1.1TiC) or tantalum carbide (W-3.3TaC) were exposed to a low-energy (40 eV/D), high-flux (1.8–5 × 10 23 D/m 2 s) deuterium (D) plasma at temperatures of about800 K, 1050 K, and 1250 K to a fluence of about1 × 10 27 D/m 2 . The deuterium (D) inventory in the samples was examined by nuclear reaction analysis and thermal desorption spectroscopy. At 800 K the D bulk concentrations and total D inventories in W-1.1TiC and W-3.3TaC were more than one order of magnitude higher compared to that in pure polycrystalline W. At 1050 K and 1250 K the D concentrations in all types of samples were very low (≤10 −5 at. fr.); however the D inventories in W-1.1TiC were significantly higher compared to those in W-3.3TaC and pure W. It is suggested that D trapping inside the carbide precipitates and at their boundaries is essential at high temperatures and high incident fluxes, especially in W-1.1TiC.

Samples with tungsten (W) and tungsten-yttrium (Y) coatings on molybdenum were exposed to plasmas simulating the divertor conditions of ITER. The exposed surfaces were studied using in-situ laser induced breakdown spectroscopy diagnostics (LIBS) and the results were compared to those obtained by other surface characterization methods. Our results show that LIBS is a reliable technique for in situ monitoring of erosion, deposition, and fuel retention processes under reactor-relevant conditions. In the regions of the highest gross erosion the thickness of the remaining tungsten layer is thinnest and at the central part of samples both mixing and deposition of components took place. LIBS also allowed reliable recording of D retention. D accumulation is influenced by the surface temperature as well as by the surface irregularities.

Methane was injected into a high density low temperature hydrogen plasma beam ( n e ⩽ 10 21 m −3 , T e ∼ 1 eV) in Pilot-PSI to study carbon transport and provide a benchmark case for modeling. Gas injection was performed through molybdenum samples that acted as collectors of redeposits. The CH A–X band emission plumes are localized near the injection hole with a typical width of ∼3 mm and a decay length of ∼1.5 mm. The geometry of the emission profile is not influenced by the electron temperature. The redeposition efficiency and codeposit structure depend on the surface temperature. Surface temperatures below 500 K promote growth of a-C:H films across the surface and lead to redeposition of up to 38% of injected carbon. Above 800 K the redeposition efficiency decreases strongly and carbon microparticles are found near the injection hole.

A steady-state high-flux H or He plasma beam was balanced against the pressure of a Sn vapor cloud for the first time, resulting in a self-regulated heat flux intensity near the liquid surface. A temperature response of the liquid surface characterized by a decoupling from the received heating power and significant cooling of the plasma in the neutral Sn cloud were observed. The plasma heat flux impinging on the target was found to be mitigated, as heat was partially dissipated by volumetric processes in the vapor cloud rather than wholly by surface effects. These results motivate further exploration of liquid metal solutions to the critical challenge of heat and particle flux handling in fusion power plants.

The first study of the interaction of metallic dust (tungsten, aluminum) adhered on tungsten substrates with transient plasma heat loads is presented. Experiments were carried out in the Pilot-PSI linear device with transient heat fluxes up to 550 MW m−2 and in the DIII-D divertor tokamak. The central role of the dust-substrate contact area in heat conduction is highlighted and confirmed by heat transfer simulations. The experiments provide evidence of the occurrence of wetting-induced coagulation, a novel growth mechanism where cluster melting accompanied by droplet wetting leads to the formation of larger grains. The physical processes behind this mechanism are elucidated. The remobilization activity of the newly formed dust and the survivability of tungsten dust on hot surfaces are documented and discussed in the light of implications for ITER.

For tungsten exposed to low-energy hydrogen-plasmas, it has been thought that grains with 〈 1 1 1 〉 surface normal are most susceptible to blistering while those with 〈 0 0 1 〉 surface normal are virtually impervious to it. Here, we report results showing that non-uniformity of blister distribution depends on the state of the surface due to polishing. In electrochemically polished material blisters appear on the grains with all orientations, while in mechanically polished material blister-free areas associated with particular orientations emerge. On the other hand, blistering is shown to have a strong dependence on the level of deformation within particular grains in partially recrystallized material.

The concentration of helium in tungsten fuzz-like nanostructures has been measured by means of thermal desorption spectroscopy. Fuzz was formed on the W surface under intensive plasma irradiation at 1500 K. The helium content was measured first in the as-irradiated sample, and then in a similar sample with the fuzz mechanically scraped from the sample surface. The difference gave the He content in the fuzz, which was estimated to be He/ W = (13 ± 4)%.

The effect of severe plastic deformation on the deuterium retention in tungsten exposed to high-flux low-energy plasma (flux ∼ 1024 D/m2/s, energy ∼ 50 eV, and fluence up to 3 × 1026 D/m2) at the plasma generator Pilot-PSI was studied by thermal desorption spectroscopy and scanning electron microscopy. The desorption spectra in both reference and plastically deformed samples were deconvolved into three contributions attributed to the detrapping from dislocations, deuterium-vacancy clusters, and pores, respectively. The plastically induced deformation, resulting in high dislocation density, does not change the positions of the three peaks, but alters their amplitudes as compared to the reference material. The appearance of blisters detected by scanning electron microscopy and the desorption peak attributed to the release from pores (i.e., deuterium bubbles) were suppressed in the plastically deformed samples but only up to a certain fluence. Beyond 5 × 1025 D/m2, the release from the bubbles in the deformed material is essentially higher than in the reference material. Based on the presented results, we suggest that a dense dislocation network increases the incubation dose needed for the appearance of blisters, associated with deuterium bubbles, by offering numerous nucleation sites for deuterium clusters eventually transforming into deuterium-vacancy clusters by punching out jogs on dislocation lines.

Taking the example of tungsten, we demonstrate that high-flux plasma exposure of recrystallized and plastically deformed samples leads to principal differences in the gas trapping and associated surface modification. Surface of the exposed pre-deformed samples exhibits ruptured µ m-sized blisters, a signature of bubbles nucleated close to the surface on the plastically induced dislocation network. Contrary to the recrystallized samples, no stage attributable to gas bubbles appeared in the desorption spectrum of the deformed samples demonstrating the strong impact of dislocations on hydrogen retention.

We have investigated the formation of various carbon nanostructures using extreme plasma fluxes up to four orders of magnitude larger than in conventional plasma-enhanced chemical vapor deposition processing. Carbon nanowalls, multi-wall nanotubes, spherical nanoparticles and nanotips are among the structures detected with electron microscopy methods. Precursor injection or surface pretreatment were not required for the synthesis of the nanostructures. Preliminary experiments with varied plasma composition, sample bias and surface temperature have demonstrated the potential for optimizing the growth of the nanostructures in the current experimental set-up.

We report measurements of chemical sputtering yields of graphite exposed to low temperature nitrogen plasmas. The influence of surface temperature and incoming ion energy on the sputtering yields has been investigated in two distinct ion flux density regimes. Sputtering yields grow consistently with increasing temperatures in experiments with low flux density (Gamma(i) approximate to 10(20) m(-2)s(-1) - 10(21) m(-2)s(-1)) and high flux density (Gamma(i) approximate to 10(23) m(-2)s(-1)). Moreover, empirical fitting of the data suggests that the temperature of 670 degrees C is optimal for chemical sputtering at high flux density. Negative biasing of the samples was used to vary the ion energy in the low flux density regime. The sputtering yield in this case increases from 0.07 atoms/ion for E-i = 1.5eV to 0.19 atoms/ion for E-i = 35 eV. After taking into account the dependence of the yields on temperature and ion energy, we evidenced a flux dependence of sputtering, similar to that found for chemical sputtering of carbon by hydrogen.

Fine-grain graphite samples were exposed to high density low temperature ( ⁠ n e ∼ 10 20 m − 3 , T e ∼ 1 eV ⁠ ) hydrogen plasmas in the Pilot-PSI linear plasma generator. Redeposition of eroded carbon is so strong that no external precursor gas injection is necessary for deposits to form on the exposed surface during the bombardment. In fact, up to 90% of carbon is redeposited, most noticeably in the region of the highest particle flux. The redeposits appear in the form of carbon microparticles of various sizes and structures. Discharge parameters influence the efficiency of the redeposition processes and the particle growth rate. Under favorable conditions, the growth rate reaches 0.15 μ m/s. The authors used high resolution scanning electron microscopy and transmission electron microscopy to study the particle growth mode. The columnar structure of some of the large particles points toward surface growth, while observation of the spherical carbon nanoparticles indicates growth in the plasma phase. Multiple nanoparticles can agglomerate and form bigger particles. The spherical shape of the agglomerates suggests that nanoparticles coalesce in the gas phase. The erosion and redeposition patterns on the samples are likely determined by the gradients in plasma flux density and surface temperature across the surface.

The formation of metallic nanostructures by exposure of molybdenum and tungsten surfaces to high fluxes of low energy helium ions is studied as a function of the ion energy, plasma exposure time, and surface temperature. Helium plasma exposure leads to the formation of nanoscopic filaments on the surface of both metals. The size of the helium-induced nanostructure increases with increasing surface temperature while the thickness of the modified layer increases with time. In addition, the growth rate of the nanostructured layer also depends on the surface temperature. The size of the nanostructure appears linked with the size of the near-surface voids induced by the low energy ions. The results presented here thus demonstrate that surface processing by low-energy helium ions provides an efficient route for the formation of porous metallic nanostructures. (C) 2012 American Vacuum Society.

Samples of EUFROFER, a reduced activation ferritic martensitic steel, were exposed in the linear plasma device Pilot-PSI to a deuterium (D) plasma with incident ion energy of similar to 40 eV and incident D flux of 2-6 x10(23) D/m(2) s to fluences up to 10 27 D/m(2) at surface temperatures ranging from 400 K to 950 K. The main focus of the study lays on the surface morphology changes dependent on the surface temperature and the surface composition evolution, e.g., the enrichment in tungsten; but also the erosion and the D retention are studied. The created surface morphology varies strongly with surface temperature from needle-like to corral-like structures. The visible lateral length scale of the formed structures is in the range of tens of nanometres to above 1 mu m and exhibits two thermal activated regimes below and above similar to 770 K with activation energies of 0.2 eV and 1.3 eV, respectively. The lateral variation of the enrichment of heavy elements on the surface is correlated to this surface morphology at least in the high temperature regime, independent of the origin of the enrichment (intrinsic from the sample or deposited by the plasma). Also the erosion exhibits temperature dependence at least above similar to 770 K as well as a fluence dependence. The amount of deuterium retained in the top 500 nm is almost independent of the exposure temperature and is of the order of 10(18) D/m(2), which would correspond to a sub-monolayer D coverage on the surface. The retained D in the volume summing up over the complete samples exceeds the D retained close to the surface by one order of magnitude. (C) 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license.