Tin (Sn) is an attractive option for a liquid metal wall material for future fusion reactors. Control of tritium inventory is key for the successful operation of these reactors, but little data exists up until now on hydrogen isotope retention in Sn. Free surface Sn targets and Sn-based capillary porous structure targets were exposed to deuterium (D) plasma in nano-PSI and magnum-PSI respectively. The retained D inventory was determined using the methods of thermal desorption spectroscopy and nuclear reaction analysis. The retention dependence is somewhat complex due to the mixed composition of the exposed samples as well as their liquid nature. The D retained in both types of Sn targets was found to increase with increasing D plasma fluence. For free surface liquid Sn targets, both thermal desorption spectroscopy and nuclear reaction analysis measurements showed a negative relationship between D retention and sample temperature. For capillary porous structure Sn targets, D retained in the top layer measured by nuclear reaction analysis decreased with temperature while the total D retained measured by thermal desorption spectroscopy remained approximately constant. By extracting pure Sn pieces from the targets it was found that the amount of D retained in pure Sn was much lower than that in the whole Sn-based targets and was estimated to be about 10 −7 -10 −4 D/Sn. D retained at the Sn-wall interface was found to dominate the total amount of D retained in the whole sample and observed cavities between deposited Sn droplets and the wall are the leading candidates responsible for this. Cavity formation is proposed to be the main retention mechanism for D in liquid Sn targets, although enhanced solubility leading to supersaturation under a D plasma environment is mainly responsible for the observed higher D retention in pure Sn compared with normal solubility under D gas. When compared with tungsten, D in Sn samples is of the same order of magnitude at temperatures below 300 °C, but at higher temperatures at least one to two orders of magnitude higher, most likely due to D trapped in cavities.

In this work, Li-filled 3D-printed porous tungsten samples were exposed to deuterium (D) plasma in Magnum-PSI with a wide ion flux from 4 × 1022 to 1.5 × 1024 m−2 s−1 and with a corresponding wide temperature range from below Li melting point (180.5 °C) to above Li deuteride (LiD) melting point (∼690 °C). The formation, decomposition and melting of LiD have been directly observed in the experiment via infra-red thermometry and visually post-mortem while still in vacuo, and correlated to the D retained content. The LiD formation was characterized by a solid precipitate layer formed on the surface with high emissivity (0.6–0.9) characterized by a blue or dark blue color after exposure. The melting of Li–LiD layer was found to occur close to the temperature predicted by Li–LiD phase diagram. In situ nuclear reaction analysis (NRA) was applied to perform the measurement of D retained in Li samples immediately after exposure without breaking the vacuum. D depth profiles were determined by NRA, in which the highest D concentration (15–45 at.%) was found in the top several micrometers and decreases with depth to low levels (<5%) within 5–30 μm. No pure LiD layer was found on the sample surfaces, however a D concentration close to 50 at.% was observed on a Li-D co-deposited layer on the clamping ring in some cases. The experiments also indicate that the D retained increases with increasing temperature until ∼500 °C. At temperatures beyond ∼500 °C the dissociation of LiD starts to dominate and the deuterium retention started to decrease. Overall, D retained fraction for all cases was found to be below ∼2%, which is significantly different from literatures where full uptake has been suggested. A 1D reaction–diffusion (RD) model based on D diffusion and chemical reactions with Li has been built. D depth profiles from the RD modelling can roughly match that from NRA measurement and a low D retained fraction below ∼2% was also indicated by the model. The model can also help explain the relationship between D retained and the surface temperature and fluence. After D plasma exposure, either helium or H plasma was utilized to remove the retained D in Li and both were proved to be effective and the removal efficiency can be as high as 96% above 420 °C.

A physical model has been developed which includes high temperature liquid lithium evaporation, the expanding motion of the liquid lithium vapour cloud, the shielding effects of the vapour cloud on incident plasma particle bombardments, ejection suppressed analysis and a perpendicular field proposal, and photon radiation, heat flux and transport in the lithium vapour cloud plasma. The engineering outline design scheme and the relevant parameters for the liquid lithium surface divertor target plate configured by discrete tiny capillary arrays have been established. Splashing can be suppressed by utilizing discrete and electrical insulating capillary porous systems (CPSs), since the conductivity among the capillary cells has been cut off by adopting a special kind of ceramic composite material made of a non-conducting and unbreakable composite which is able to withstand high temperatures. The formula to describe the temperature-dependent evaporation power has been derived. The maximum temperature increases of the discrete plasma-facing liquid lithium surface divertor target plate have been compared under the high energy flux deposition of 10 MJ m −2 during a 1 ms time duration with or without evaporation power. The results show that a high surface heat load can be withstood by the designed discrete plasma-facing liquid lithium surface divertor target plate due to violent evaporation. The energy deposition of incident energetic particles and weakly relativistic electrons from the scrape-off layer have been calculated. A laboratory experimental facility to simulate liquid lithium surface interactions with plasma has been set up. Research on lithium evaporation, re-deposition and ejection suppressed experiments under high density linear plasma dumping is ongoing.

No abstract

Liquid metals have been proposed as potential divertor materials for future fusion reactors, and surface stability is a vital requirement for such liquid metal divertors (LMDs). Capillary porous structures (CPSs) have been applied to the design of liquid metal targets as they can avoid MHD instability by surface tension and provide a stable liquid surface. However, our previous work has found that liquid Sn surfaces can be very unstable in hydrogen plasma even in cases without magnetic fields. To increase our understanding of the interaction of liquid Sn surfaces with plasmas, in this work we systematically investigated the surface behaviors of liquid Sn in different plasma exposures in linear plasma devices, either in Nano-PSI at low flux and without magnetic field, or in Magnum-PSI with strong magnetic field strength. Surface instability leading to droplet ejection has been observed and recorded in the experiments. The ejection of droplets is not dependent on magnetic fields and plasma currents, and is found to be dependent on the plasma species and plasma flux and surface temperature. The CPS meshes applied in the experiments cannot completely avoid droplet ejection but can decrease droplet size and lower droplet production rate. In H plasma, droplets were observed once Sn melted even at low fluxes. For the case of N plasma, the appearance of droplets started at a temperature marginally higher than tin-nitride decomposition temperature. Only at high fluxes (~ 10 23 -24 m -2 s -1 ) and high temperatures (900 -1000 °C) were a few droplets observed in Ar or He plasma. For all cases, the ejection velocities of most droplets were around 1 -5 m/s. Bubble formation, growth and bursting in the plasma-species-supersaturated liquid Sn is proposed as the primary mechanism for the ejection of droplets. Plasma-enhanced solubility is responsible for the achievement of H/N-supersaturated liquid Sn, while high plasma flux implantation is responsible for Ar/He-supersaturated liquid Sn. Once the concentration of plasma species in liquid Sn reaches a certain supersaturation level, nucleation and growth of bubbles occur due to the desorption of dissolved plasma species from the liquid Sn. The formation and bursting of bubbles have been directly observed in the experiment. The sizes of most bubbles were estimated in the range of 40 -400 µm or even smaller. A bubble growth model based on Sievert's and Henry's laws is invoked to describe bubble growth in liquid Sn.