An experimental effort to apply Flibe (a mixed molten salt of 2LiF+BeF2) to a self-cooled liquid blanket of a fusion reactor was carried out under a Japan–US collaboration called JUPITER-II. Maintaining Flibe under a reducing atmosphere is a key issue to transform TF to T2 with a faster reaction rate compared with the residence time in blanket. One of the tasks was to clarify whether or not the redox control of Flibe can be achieved with Be. The dissolution rate of a Be rod and the reaction rate of Be+2HF=BeF2+H2 in Flibe were experimentally determined. Sufficiently fast rates of the Be dissolution and the reduction reaction of HF to H2 were clarified by our redox control experiment. Close agreement was obtained between experiments and our simplified complete-mixing model. In particular, the reaction between Be and a F− ion immediately after their contact was found to be limited by diffusion of the F− ion.

The second Japan/US Program on Irradiation Tests for Fusion Research (JUPITER-II) began on April 1, 2001. Part of the collaborative research centers on studies of the molten salt 2LiF2–BeF2 (also known as Flibe) for fusion applications. Flibe has been proposed as a self-cooled breeder in both magnetic and inertial fusion power plant designs over the last 25 years. The key feasibility issues associated with the use of Flibe are the corrosion of structural material by the molten salt, tritium behavior and control in the molten salt blanket system, and safe handling practices and releases from Flibe during an accidental spill. These issues are all being addressed under the JUPITER-II program at the Idaho National Laboratory in the Safety and Tritium Applied Research (STAR) facility. In this paper, we review the program to date in the area of tritium/deuterium behavior, Flibe mobilization under accident conditions and testing of Be as a redox agent to control corrosion. Future activities planned through the end of the collaboration are also presented.

In order to investigate the viability of using Be as a redox agent in a molten Flibe (2LiF–BeF2) blanket, a series of kinetics experiments were performed in which HF was bubbled through Flibe with varying concentrations of dissolved Be. The feed gas consisted of 910–1800ppm HF and 0.1–0.2vol.% H2, with the balance comprised of He. A cylindrical rod of Be was contacted with the salt for periods of time ranging from 600 to 3600s, resulting in mole fractions in the salt ranging from 4.3×10−5 to 2.6×10−4. Initially, high HF conversion levels in excess of 90% were observed. As HF reacted with the Be, the conversion levels slowly dropped over a period of several hours to a few days. A simple kinetic model, which is first order in both HF and Be concentration has been coupled with a non-mixed reactor model to yield a good fit to the data. Application of this model indicates that Be should be suitable for keeping the TF concentration in the salt below 0.02ppb.

To understand the interactions between molten Flibe and Be, a metallic Be rod was immersed in molten Flibe at 803K under He atmosphere for 210h. The Be rod was significantly eroded during immersion in molten Flibe, and the Flibe changed from a clear crystal to a brownish-gray marble-like appearance. The concentration of Be0 in Flibe was evaluated by dissolving salt samples in acid solutions. This dissolution test is based on the reaction of Be0 with proton ions in the acid solutions to generate H2. Hydrogen gas was generated from Flibe contacted with Be under flowing He. The amounts of gas generated corresponded to mole fractions of [Be0]/[Flibe] ranging from 9.9×10−4 to 7.6×10−3. On the other hand, no H2 was generated from Flibe that was not exposed to Be or which was exposed to Be and then given significant H2–HF–He bubbling. These observations showed that Be does, indeed, dissolve in Flibe as Be0. The fact that no Be0 was detected after bubbling H2–HF–He into the salt indicates that Be0 is an effective redox agent for reacting HF.

The Safety and Tritium Applied Research (STAR) Facility, a US DOE National User Facility at the Idaho National Engineering and Environmental Laboratory (INEEL), comprises capabilities and infrastructure to support both tritium and non-tritium research activities important to the development of safe and environmentally friendly fusion energy. Research thrusts include (1) interactions of tritium and deuterium with plasma-facing-component (PFC) materials, (2) fusion safety issues [PFC material chemical reactivity and dust/debris generation, activation product mobilization, tritium behavior in fusion systems], and (3) molten salts and fusion liquids for tritium breeder and coolant applications. This paper updates the status of STAR and the capabilities for ongoing research activities, with an emphasis on the development, testing and integration of the infrastructure to support tritium research activities. Key elements of this infrastructure include a tritium storage and assay system, a tritium cleanup system to process glovebox and experiment tritiated effluent gases, and facility tritium monitoring systems.

The Safety and Tritium Applied Research (STAR) facility at the Idaho National Engineering and Environmental Laboratory (INEEL) is currently being commissioned to investigate tritium-related safety questions for fusion and other technologies. The tritium inventory for the STAR facility will be maintained below 1.5 g to avoid the need for STAR to be classified as a Category 3 nuclear facility. A key capability in successful operation of the STAR facility is the ability to receive, inventory, and dispense tritium to the various experiments underway there. The system central to that function is the Tritium Storage and Assay System (SAS).The SAS has four major functions: (1) receiving and holding tritium, (2) assaying, (3) dispensing, and (4) purifying hydrogen isotopes from non-hydrogen species.This paper describes the design and operation of the STAR SAS and the procedures used for tritium accountancy in the STAR facility.

Diffusion coefficients of hydrogen isotopes in Flibe were correlated with making reference to previous relating data of F- ion self-diffusivity and Flibe viscosity and so on. Rates of tritium permeation through structural materials in a fusion reactor system with Flibe blanket were estimated comparatively under conditions with or without a Flibe permeation barrier. A way to lower the tritium leak rate below a level regulated by law was proposed, and its effectiveness was discussed.

Carbon fiber composites (CFCs) are often suggested as armor material for the first wall of a fusion plasma chamber because of carbon's low atomic number, high thermal conductivity, and high melting point. However, carbon is chemically reactive in air and readily absorbs tritium. Accordingly, it is believed that during a loss-of-vacuum accident (LOVA), the CFC armor will react with the air ingress and release its absorbed tritium. The mobilization of this tritium and the carbon monoxide produced by the CFC-air chemical reaction are both safety concerns. This paper discusses the MELCOR thermal-hydraulic analysis of a simulated LOVA for the SOMBRERO fusion design. The MELCOR analysis is important because it included data from recent oxidation experiments that studied the advanced CFC NB31. A previous MELCOR analysis of a simulated SOMBRERO LOVA event suggested that the ingress of air would aggressively oxidize the CFC. While the current analysis revealed initial first-wall temperatures that exceed those of the prior analyses, the trend reversed 161 h after the onset of the LOVA. The calculated wall temperatures at the back of the blanket for the current analysis were consistently lower than those previously calculated using the older data. Accordingly, the conclusion is that a LOVA event for a fusion design similar to SOMBRERO may not be as grave as once predicted.

Experimental studies to investigate the behavior of deuterium and tritium in the molten salt Flibe (2LiF · BeF2), have been conducted as part of the Japan–US joint research program (JUPITER-II). Measurements of deuterium transport were made in a cylindrically symmetric, dual permeation probe assembly containing 400 cc of Flibe. An exact analytical transport solution in cylindrical coordinates was fit to the measured permeation data, and this analysis derived deuterium diffusion and solubility coefficients of 8.0 × 10−10 m2/s and 3.1 × 10−4 mol/m3 Pa at 600 °C, respectively, and 3.0 × 10−9 m2/s and 1.0 × 10−4 mol/m3 Pa at 650 °C. The diffusion coefficients were about a factor of two less than previous results derived from capillary-reservoir diffusion measurements with tritium. Solubility results were significantly greater than previously measured for D2 but they were comparable to those for DF in Flibe. The results suggest that the dominant deuterium transport species in Flibe was D+F− for these experiments.

Carbon Fiber Composites (CFCs) are often suggested as armor material for the first wall of a fusion plasma chamber due to carbon's low atomic number, high thermal conductivity, and high melting point. However, carbon is chemically reactive in air and will react with ingress air during a Loss of Vacuum Accident and release tritium fuel that has been retained in the carbon. Tritium mobilization and carbon monoxide generation via CFC oxidation are both safety concerns. This paper discusses chemical reactivity experiments that were performed using the state-of-the-art 3-dimensional NB31 CFC produced by SNECMA and a laminar reaction gas of Ar–21 vol% O2. Oxidation reaction rates were measured for CFC temperatures of 525, 600, 700, 800, 900, and 1000 °C and a 100 standard cubic centimeters per minute (sccm) Ar–O2 flow rate. Experiments were also performed at CFC temperatures of 700 and 1000 °C and a 1000 sccm Ar–O2 flow rate. Mass spectral analyses of the exhaust reaction gas suggested that carbon monoxide was the primary reaction at the CFC surface and carbon dioxide was readily produced in the exiting reaction gas. The measured reaction rates compare well with the literature and were used to produce a CFC oxidation curve that is recommended for use in fusion safety analyses.

Abstract Experiment of D2 permeation through Ni facing with purified Flibe is being carried out under the Japan-US joint research project (JUPITER-II). The experiment is proceeding in the following phases; (i) fabrication and assembly of a dual-probe permeation apparatus, (ii) a single-probe Ni/D2 permeation experiment without Flibe, (iii) a dual-probe Ni/D2 permeation experiment without Flibe, (iv) Flibe chemical purification by HF/H2 gas bubbling, (v) physical purification by Flibe transport through a porous Ni filter, (vi) Ni/Flibe/D2 permeation experiment, and (vii) Ni/Flibe/HT permeation experiment. The present paper describes results of the single and dual Ni/D2 permeation experiments in detail.

The second Japan/US Program on Irradiation Tests for Fusion Research (JUPITER-II) began on April 1, 2001. Part of the collaborative research centers on studies of the molten salt 2LiF-BeF2 (also known as Flibe) for fusion applications. Flibe has been proposed as a self-cooled breeder in both magnetic and inertial fusion power plant designs over the last twenty years. The key feasibility issues associated with the use of Flibe are the corrosion of structural material by the molten salt, tritium control in the molten salt blanket system, and safe handling practices and releases from Flibe during an accidental spill. An overview of the experimental program to address the key feasibility issues is presented.

Steam chemical reactivity experiments were conducted for several ITER-like tokamak plasma-facing-component (PFC) materials: NB31 and NS31 carbon fiber composites (CFCs), W–1%La, DShG-200 Be, and Be (S65C and Kawecki PO Ductile Be) specimens irradiated to fast neutron fluences ranging from 5×1019 to 1×1021 n/cm2. Experiments were performed at 800–1100 °C for CFC, 550–1000 °C for W–1%La, 500–900 °C for unirradiated Be, and 600–800 °C for irradiated Be. Average hydrogen generation rates are presented for these new measurements as a function of temperature, and the results are compared to previous studies on similar materials. In general, the new results extend the temperature and material range from previous studies, are consistent with previous work, and provide Arrhenius-type hydrogen generation expressions useful for safety assessment calculations. Little irradiation influence on oxidation behavior was observed for Be irradiated to 1×1021 n/cm2, whereas significant influence was observed in previous tests on Be irradiated to 5×1022 n/cm2.

This paper reports the results of mass-spectrometer experiments on the vaporization properties of pure Li and two different Sn–25at.%Li alloys heated in a Knudsen effusion source at temperatures from 800 to 1200 K. A Li vaporization energy of 172 kJ/mol was obtained from the analysis of three combined Sn–25at.%Li experiments. Li vapor pressures for Sn–25at.%Li were derived from the data by comparing the IT (Li+ ion current×temperature) products from Sn–25at.%Li experiments with the measured IT product from a pure Li experiment at 816 K for which the vapor pressure is 1.8 Pa. The derived vapor pressures are more than three orders of magnitude below the corresponding vapor pressures for pure Li, and they are generally in good agreement with previous data obtained from a variety of different measurement techniques.

Flibe–tritium experiment in the Japan–US joint project (JUPITER-II) was initiated in 2001. H/D isotopic exchange experiments were conducted to select a Flibe-facing material. Because of hydrogen isotope interactions with carbon, Ni crucibles were selected for Flibe/tritium behavior experiments. A Flibe–tritium pot with two Ni (or Cu) permeable probes was designed. The rate of the overall tritium permeation through the Flibe-facing Ni or Cu was estimated by numerical simulation using TMAP4 code. Diffusion in bulk Flibe was found to be the rate-determining step for purified Flibe.

A preliminary design is presented for the FLIQURE or Fusion LIQUid Release Experiment. This experimental system is designed to measure the mobilization of tritium, fluorine, and other constituents in molten Flibe following trace neutron irradiation at fusion-relevant temperatures (500-1000°C). Trace neutron irradiation is achieved using a spontaneous fission 252Cf source that produces a relatively uniform concentration of tritium in the Flibe. The experimental goals aim to better understand the mechanisms, mobilization rates, and physiochemical forms of tritium mobilized from the Flibe along with other potentially detectable radioactive isotopes and toxic-material species under inert-gas, air, and steam-ingress conditions. System design details are discussed which include neutronic studies to optimize tritium production, thermal design to maintain and isolate molten Flibe, and instrumentation to meet experimental goals.

This paper reports the results of a systematic investigation into retention of deuterium implanted into various forms of W including: reduction-rolled, powder-metallurgy foil; discs of W1%La alloy; W and W2C prepared by chemical vapor deposition (CVD) and annealed C-coated W discs. Deuterium was implanted at energies of 0.5 keV/D with fluxes of ∼3×1019D/m2 s and fluences of 3×1023D/m2, for samples at temperatures from 23°C to 400°C. Retained deuterium quantities were measured using thermal desorption spectroscopy (TDS). Retention in annealed CVD-W and W1%La, for implantation temperatures less than 200°C, is below that in annealed W foil, indicating that trapping may be affected by the different material defect structures. For implantation temperatures less than 300°C, retention in CVD-W2C is somewhat higher than that in CVD-W, indicating trapping could be enhanced by trace carbon impurities, differences in W2C material structure and recoil carbon-induced material damage. Implantation into C-coated W resulted in orders of magnitude more retention than in uncoated material, because of trapping in the carbon coating.

Tungsten brush accommodates thermal stresses and high heat flux in fusion reactor components such as plasma facing surfaces or armor. However, inherently higher surface areas are introduced with the brush design. We have tested a specific design of tungsten brush in steam between 500 and 1100°C. Hydrogen generation and tungsten volatilization rates were determined to address fusion safety issues. The brush design prepared from 3.2-mm diameter welding rods had a packing density of 85%. We found that both hydrogen generation and tungsten volatilization from brush, fixtured to represent a unit within a larger component, were less than projections based upon the integrated total surface area (TSA). Steam access and the escape of hydrogen and volatile oxide from void spaces within the brush are restricted compared with specimens with more direct diffusion pathways to the test environment. Hydrogen generation rates from restrained specimens based on normal surface area (NSA) remain about five times higher than historic rates based on total surface area. Volatilization rates from restrained specimens based upon normal surface area (NSA) were only 50% higher than our historic cumulative maximum flux plot (CMFP) for tungsten. This study has shown that hydrogen generation and tungsten volatilization from brush do not scale according to predictions with earlier determined rates. Volatilization rates from brush with higher packing density could, in fact, approach those from flat surfaces.