First-principles study of hydrogen adsorption and permeation in reconstructed cubic erbium oxide surfaces

Understanding surface properties of Er₂O₃, especially in relation to adsorption and permeation of atomic hydrogen, is of considerable importance to the study of tritium permeation barriers. In this work, hydrogen diffusion pathways through the low-index (1 0 0), (1 1 0), and (1 1 1) surfaces of cubic Er₂O₃ have been calculated using density functional theory within the GGA (PBE) + U approach. The dependence of the effective U parameter on lattice constants, bulk moduli, and formation energies of Er₂O₃ has been investigated in detail. The energetics of hydrogen penetration from the surfaces to the solution site in bulk Er₂O₃ were defined using the optimum effective U value of 5.5 eV. For a low surface coverage of hydrogen (0.89 × 10¹⁴ H/cm²), a penetration energy of at least 1.7 eV was found for all the low-index erbium oxide surfaces considered. The results of the present study will provide useful guidance for future studies on modeling defects, such as grain boundaries and vacancies, in tritium permeation barriers.     

Compatibility of erbium oxide coating with liquid lithium–lead alloy and corrosion protection effect of iron layer

Li–Pb compatibility of Er₂O₃ and Er₂O₃-Fe two-layer coatings has been explored for an understanding of corrosion behaviors and effects of the protection layer. The coatings were peeled off after static Li–Pb immersion test at 600 °C due to the degradation of adhesion between the coating–substrate interface. A loss of Er and then subsequent corrosion of Er₂O₃ were shown after immersion at 500 °C for 500 and 1505 h. However, the outer Fe layer played a role to decrease corrosion rate of the coatings by comparing with the results of Er₂O₃ single layer coatings. Deuterium permeation measurements after corrosion tests at 500 °C showed that the Er₂O₃ coatings kept permeation reduction factors of 10²–10³ after 500 h immersion, but seriously degraded after 1505 h immersion. Corrosion mechanisms suggest that corrosion protection properties will be modified by an optimization of the outer Fe layer and a control of oxygen concentration in Li–Pb.     

On the study of catalytic membrane reactor for water detritiation: Membrane characterization

Tritium waste recycling is a real economic and ecological issue. Generally under the non-valuable Q₂O form (Q = H, D or T), waste can be converted into fuel Q₂ for a fusion machine (e.g. JET, ITER) by isotope exchange reaction Q₂O + H₂ = H₂O + Q₂. Such a reaction is carried out over Ni-based catalyst bed packed in a thin wall hydrogen permselective membrane tube. This catalytic membrane reactor can achieve higher conversion ratios than conventional fixed bed reactors by selective removal of reaction product Q₂ by the membrane according to Le Chatelier's Law.This paper presents some preliminary permeation tests performed on a catalytic membrane reactor. Permeabilities of pure hydrogen and deuterium as well as those of binary mixtures of hydrogen, deuterium and nitrogen have been estimated by measuring permeation fluxes at temperatures ranging from 573 to 673 K, and pressure differences up to 1.5 bar. Pure component global fluxes were linked to permeation coefficient by means of Sieverts’ law. The thin membrane (150 μm), made of Pd–Ag alloy (23 wt.%_Ag), showed good permeability and infinite selectivity toward protium and deuterium. Lower permeability values were obtained with mixtures containing non permeable gases highlighting the existence of gas phase resistance. The sensitivity of this concentration polarization phenomenon to the composition and the flow rate of the inlet was evaluated and fitted by a two-dimensional model.     

Deuterium permeation behavior of erbium oxide coating on austenitic,ferritic, and ferritic/martensitic steels

Tritium permeation barrier is required in fusion blanket for reduction of loss of fuel and health hazard. In this study, deuterium permeation experiments have been performed on four kinds of steels and erbium oxide coatings fabricated by a filtered arc deposition method. The permeation flux of uncoated samples shows diffusion-limited regime in the temperature range 573–723 K and the permeability is corresponding to literature data. The coated samples deposited at room temperature have been tested at 773 K. It is found that the coatings suppress the deuterium permeation to a close level in spite of different types of steel substrates. In addition, the exponent of the driving pressure slightly changes compared to the uncoated sample. However, the permeation regime is still near diffusion limited.     

Preparation of Cr2O3 film by MOCVD as hydrogen permeation barrier

Cr₂O₃ film on structural material as hydrogen permeation barrier can be applied in many areas such as hydrogen storage devices, vacuum solar receivers and fusion reactors. In this study, the Cr₂O₃ film was prepared by MOCVD on 316L stainless steel using chromium(III) acetylacetonate as precursor. The film was characterized by X-ray diffraction (XRD), scanning electron microscope (SEM) and X-ray photoelectron spectroscopy (XPS). The hydrogen permeation inhibition performance of films was investigated by deuterium permeation experiment. The 366 nm thick Cr₂O₃ film on 316L could reduce the deuterium permeability by 24–117 times at 823–973 K, revealing efficient inhibition to hydrogen permeation. The Cr₂O₃ film is dense, crack-free and has a corundum structure which possesses a more stable structure than a metastable phase or an amorphous phase. Moreover, the crystalline Cr₂O₃ could be easily obtained by MOCVD at a low temperature, e.g. 773 K.     

An advance process of aluminum rich coating as tritium permeation barrier on 321 steel workpiece

We have proposed an advance three-step process, Al-electroplating in ionic liquid followed by heat treating and selectively oxidation, preparing aluminum rich coating as tritium permeation barrier (TPB). In present work, the advance process was applied to 321 steel workpieces. In the Al-electroplating, pieces were coated by galvanostatic electrodeposition at 20 mA/cm² in aluminum chloride (AlCl₃)–1-ethyl-3-methylimidazolium chloride (EMIC) ionic liquid. The Al coating on those pieces all displayed attractive brightness and well adhered to surface of pieces. Within the aluminizing time from 1 to 30 h, a series of experiments were carried out to aluminize 321 steel pieces with Al 20 μm coating at 700 °C. After heat treated for 8 h, a 30 μm thick aluminized coating on piece appeared homogeneous, free of porosity, and mainly consisted of (Fe, Cr, Ni)Al₂, and then was selectively oxidized in argon gas at 700 °C for 50 h to form Al₂O₃ scale. The finally fabricated aluminum rich coating, without any visible defects, had a double-layered structure consisting of an outer γ-Al₂O₃ layer with thickness of 0.2 μm and inner (Fe, Cr, Ni)Al/(Fe, Cr, Ni)₃Al layer of 50 μm thickness. The deuterium permeation reduction factor, PRF, of piece (Φ 80 × 2, L 150 mm) with such coating increased by 2 orders of magnitude at 600–727 °C. The reproducibility of the process was also showed.     

Compound cryopump for fusion reactors

We reconsider an old idea: a three-stage compound cryopump for use in fusion reactors such as DEMO. The helium “ash” is adsorbed on a 4.5 K charcoal-coated surface, while deuterium and tritium are adsorbed at 15–22 K on a second charcoal-coated surface. The helium is released by raising the first surface to ～30 K. In a separate regeneration step, deuterium and tritium are released at ～110 K. In this way, the helium can be pre-separated from other species. In the simplest design, all three stages are in the same vessel, with a single valve to close the pump off from the tokamak during regeneration. In an alternative design, the three stages are in separate vessels, connected by valves, allowing the stages to regenerate without interfering with each other. The inclusion of the intermediate stage would not affect the overall pumping speed significantly.The downstream exhaust processing system could be scaled down, as much of the deuterium and tritium could be returned directly to the reactor. This could reduce the required tritium reserve by almost 90%.We used a well-established free Direct Simulation Monte Carlo (DSMC) code, DS2V. At very high upstream densities (～10²⁰ molecules/m³ and above) the flow into the pump is choked. Enlarging the aperture is the only way to increase the pumping speed at high densities. Ninety percent of the deuterium and tritium is successfully trapped at 15 K (assuming that the sticking coefficient is 80–100% on the 15–22 K surface). On the other hand, the remaining 10% still exceeds the small amount of helium in the gas input.     

Conceptual design of a water cooled breeder blanket for CFETR

China Fusion Engineering Test Reactor (CFETR) is an ITER-like superconducting tokamak reactor. Its major radius is 5.7 m, minor radius is 1.6 m and elongation ratio is 1.8. Its mission is to achieve 50–200 MW of fusion power, 30–50% of duty time factor, and tritium breeding ratio not less than 1.2 to ensure the self-sufficiency. As one of the breeding blanket candidates for CFETR, a water cooled breeder blanket with superheated steam is proposed and its conceptual design is being carried out. In this design, sub-cooling water at 265 °C under the pressure of 7 MPa is fed into cooling plates in breeding zone and is heated up to 285 °C with saturated steam generated, and then this steam is pre-superheated up to 310 °C in first wall (FW), final, the pre-superheated steam coming from several blankets is fed into the other one blanket to superheat again up to 517 °C. Due to low density of superheated steam, it has negligible impact on neutron absorption by coolant in FW so that the high energy neutrons entering into breeder zone moderated by water in cooling plate help enhance tritium breeding by ⁶Li(n,α)T reaction. Li₂TiO₃ pebbles and Be₁₂Ti pebbles are chosen as tritium breeder and neutron multiplier respectively, because Li₂TiO₃ and Be₁₂Ti are expected to have better chemical stability and compatibility with water in high temperature. However, Be₁₂Ti may lead to a reduction in tritium breeding ratio (TBR). Furthermore, a spot of sintered Be plate is used to improve neutron multiplying capacity in a multi-layer structure. As one alternative option, in spite of lower TBR, Pb is taken into account to replace Be plate in viewpoint of safety. In this contribution, study on neutronics and thermal design for a water cooled breeder blanket with superheated steam is reported.     

Preparation and investigation of aluminized coating and subsequent heat treatment on 9Cr–1Mo Grade 91 steel

Iron aluminide inner coating with alumina top layer is being considered as a potential solution for tritium permeation barrier and mitigating MHD pressure drop for liquid metal blanket concepts in the fusion reactor systems. Hot-dip aluminizing with subsequent heat treatment seems to offer a good possibility to produce aluminized coating with alumina top layer. 9Cr–1Mo Grade 91 steel samples were hot dipped in Al melt containing 2.25 wt% of Si at 750 °C for 3 min. Heat treatment was performed at 650, 750 and 950 °C for 5 h; samples were either air cooled or furnace cooled. Coatings have been evaluated by SEM, EDX, X-ray diffraction, microhardness, scratch adhesion and Raman spectroscopy. The thickness of the layers and phases formed were influenced by the heat treatment adopted. Fe₂Al₅ was the major phase present in the samples heat treated at 650/750 °C, whereas FeAl and α-Fe(Al) primarily made up the outer and inner layers respectively in the samples heat treated at 950 °C. Cooling method deployed affected the hardness. Air cooled samples had comparatively higher hardness than furnace cooled samples. The scratch test showed the adhesion for the samples heat treated at 950 °C was much better as compared to the samples heat treated at 650/750 °C. Raman spectroscopy analysis showed the presence of both α-Al₂O₃ and γ-Al₂O₃ on the surface of the samples heat treated at 950 °C, while Fe₃O₄ was present in the furnace cooled sample only.     

Feasibility of a multi-purpose demonstration neutron source based on a compact superconducting spherical tokamak

Tokamak neutron sources would allow near term applications of fusion such as fusion–fission hybrid reactors, elimination of nuclear wastes, production of radio-isotopes for nuclear medicine, material testing and tritium production. The generation of neutrons with fusion plasmas does not require energetic efficiency; thus, nowadays tokamak technologies would be sufficient for such purposes. This paper presents some key technical details of a compact (～1.8 m³ of plasma) superconducting spherical tokamak neutron source (STNS), which aims to demonstrate the capabilities of such a device for the different possible applications already mentioned. The T-11 transport model was implemented in ASTRA for 1.5 D simulations of heat and particle transport in the STNS core plasma. According to the model predictions, total neutron production rates of the order of ～10¹⁵ s^?1 and ～10¹³ s^?1 can be achieved with deuterium/tritium and deuterium/deuterium respectively, with 9 MW of heating power, 1.4 T of toroidal magnetic field and 1.5 MA of plasma current. Engineering estimates indicate that such scenario could be maintained during ～20 s and repeated every ～5 min. The viability of most of tokamak neutron source applications could be demonstrated with a few of these cycles and around ～100 cycles would be required in the worst cases.     

Evaluating the hydrogen isotope absorption/diffusion coefficient of CVD-SiC at high temperature

Deuterium diffusion coefficient measurements of CVD-SiC were carried out using a solubility and diffusivity measurement apparatus to investigate the permeation mechanism of the hydrogen isotope through CVD-SiC. Experiments were conducted with thin-sheet-type samples with thicknesses of 0.1 mm, 1 mm, and 2 mm at 1073–1183 K. Total amount of occluded gas into or released gas from different thickness but same weight sample were expected to be the same, but unexpectedly differed by more than 50%. As the release rates after sufficient time had passed were almost the same, and the 1-mm-thick sample had twice the surface area of the 2-mm-thick sample, the measurements were probably affected by adsorbed gas on the surface. The value of D/L² (the diffusion coefficient divided by the square of the thickness), obtained by fitting to the theoretical formula but ignoring the early phase of discharge, was in good agreement for samples of different thickness at the same temperature, and was more than 5 orders of magnitude smaller than that obtained from the permeability measurement experiments. Therefore, we believe that the deuterium permeation through CVD-SiC is primarily dependent on the permeation rate through the grain boundaries.     

Uptake of deuterium in partially D-depleted residual codeposits left on DIII-D tiles after laboratory thermo-oxidation experiments

Recent evidence has shown that tokamak carbon-based codeposits may become partially or fully depleted of hydrogen through thermo-oxidation, as the hydrogen content of the codeposits is removed more rapidly than the carbon content. In this study we examine the ability of such partially-depleted residual DIII-D divertor codeposits to uptake deuterium upon subsequent exposure to deuterium gas or deuterium plasmas. The partially D-depleted specimens used here were obtained from a previous study where DIII-D codeposits were oxidized for 2 h at 623 K (350 °C) and 267 Pa (2 Torr) O₂ [J.W. Davis et al., Thermo-oxidation of DIII-D codeposits on open surfaces and in simulated tile gaps, J. Nucl. Mater. 415 (2011) S789–S792]. In the present study some of these specimens, having undergone prior oxidation, were exposed to D₂ glow discharge plasmas or D₂ gas at 20 kPa (150 Torr) at 300 or 523 K. In the case of plasma exposure, no uptake of D was observed, while an increase in D content was seen following D₂ gas exposures. When the gas exposure took place at 300 K, heating the specimens in vacuum to 623 K for 15 min led to the release of all of the increased D content. For the gas exposure at 523 K, the increase in D content was found to require longer (8 h) vacuum baking to remove. However, in a reference codeposit specimen (from a closeby location on the tile), which had not been previously oxidized, there was a similar increase in D content following D₂ exposure at 523 K, but it could not be released even following 8 h vacuum baking at 623 K.     

Tritium permeation in EUROFER in EXOTIC and LIBRETTO irradiation experiments

The reduced activation martensitic steel (RAFM) EUROFER is foreseen as a structural material in test breeder module (TBM) in ITER and breeder blanket in DEMO design. In a number of irradiation experiments conducted in high flux reactor (HFR) in Petten EUROFER was used as a containment wall of the breeder material, through which tritium permeation was monitored on line. Thus in EXOTIC-9/1 (EXtraction Of Tritium In Ceramics) experiment where Li₂TiO₃ pebbles were the breeder material, EUROFER was irradiated up to 1.3 dpa at 340–580 °C. In LIBRETTO experiments (LIBRETTO-4/1, -4/2 and -5) the breeder material was lead lithium eutectic which was in direct contact with the EUROFER containment wall. The neutron damage in steel achieved in the LIBRETTO experiments varied from 2 to 3.5 dpa. The irradiation temperature was 350 °C (LIBRETTO-4/1), 550 °C (LIBRETTO-4/2), and 300–500 °C (LIBRETTO-5).Tritium permeability was studied by varying the irradiation temperature and hydrogen concentration in the purge gas. From the analysis of the temperature transients performed in all four experiments yielded the tritium diffusion coefficients were derived, which appear to be factor ten lower than the literature data obtained in the gas driven permeation experiments.     

Evaluation of tritium diffusion through the Neutral Beam Injector calorimeter panel

The Neutral Beam Test Facility (NBTF) to be realized in Padoa will test the Neutral Beam Injection (NBI), one of the Heating and Current Drive Systems foreseen for ITER. The NBI is based on the acceleration of hydrogen or deuterium negative ions up to 1 MeV. This work has been aimed at assessing the tritium release from the NBTF in order to provide data for the safety analysis. In particular, the diffusion of the tritium through the neutral beam target material (the CuCrZr alloy calorimeter panels) has been assessed by using literature data of the diffusion coefficient. The tritium generated inside the calorimeter panels moves into both the vacuum and water side: the tritium diffusion flux has been evaluated during the beam-on (200 °C) and the beam-off (20 °C) phases of the NBTF experiments consisting of an interim campaign and a final test. The penetration depth of the tritium through the 2 mm thick CuCrZr alloy material has been also evaluated by using a Monte-Carlo code. As main result, the assessed diffusion flux of tritium during both the beam-on and the beam-off phases are modest. In fact, at the end of the interim campaign (100 days), about the 96% of the all generated tritium (626.5 MBq) exits the calorimeter while the residual tritium inventory (25 MBq) leaves the copper alloy with a diffusion time of about 1 month. At the end of the final test (14 days) about the 99% of the total generated tritium (1.023 × 10⁴ MBq) leaves the copper alloy and the remaining tritium inventory (152.2 MBq) is released by about 32 days. In both the interim campaign and the final test, more than the 99% of the total tritium is transferred into the vacuum side of the calorimeter panel while negligible tritium amounts enter the cooling water system thus showing a very low impact on the environment.     

Tritium release experiments on Li4SiO4 pebbles from TRINPC-I experiments: Effects of water adsorption and hydrogen gas

Out-of-pile tritium release experiments under different water uptake contents and purge gas chemistry were performed on Li₄SiO₄. Water measurement was performed on samples under different experimental procedures. It was found that water was adsorbed on the sample during its transferring and storage process. A strong dependence of tritium release behavior on water uptake was determined. By doping H₂ in the sweep gas, the formation of water in orthosilicate was observed in addition to the isotope exchange reaction with H₂ gas. Thermal desorption peaks of the water formation reaction and H₂ isotope exchange reaction appeared at 668 °C and 463 °C, respectively, at ramping rate of 5 °C/min.     

Assessment of tritium levels in rivers and precipitation in north-western Greece before the ITER operation

The project ITER aims to demonstrate that fusion is the energy source of the future. The prototype Tokamak machine is intended to start operation at about 2019 and tritium is one of the major contaminants that can be accidentally released in the environment. Nowadays environmental tritium levels are of natural origin except in the vicinity of nuclear facilities. The evaluation of background tritium levels is essential in the context of a future possibility of accidental tritium releases. For this purpose and also because of the lack of relevant information, an extended programme of river and rain water sampling was implemented in north-western Greece. Water samples from six major rivers in this area and rain water samples were analysed for tritium content. The rivers under investigation were Aliakmonas River, Pinios River, Arachthos River, Kalamas River, Aoos River and Louros River, which originate from the central Greek mountain range Pindos, and flow to Aegean and Ionian Sea.The tritium concentrations were determined by the Liquid Scintillator Analyser Tri-Carb 3170TR/SL. The statistical analysis of data revealed that there is a seasonal variation of tritium concentration in rain samples and a less pronounced seasonal variation in river samples. The weighted mean tritium concentration for rain samples was determined equal to 0.90 ± 0.08 Bq L^?1 (7.6 ± 0.7 TU) and the respective mean value for river samples was 0.94 ± 0.04 Bq L^?1 (7.9 ± 0.3 TU). Further analysis has demonstrated that river waters tend to show lower tritium concentrations than the concurrently measured tritium concentrations in rain samples, during spring and summer (at 47% and 71% of the sampling stations, respectively), while this observation is reversed during autumn and winter (at 44% and 35% of the sampling stations, respectively). This may be attributed to rain water remaining underground for a long period allowing tritium to decay and when it reappears as river water, the tritium concentration values are lower when compared to the rain water concentrations. Rough estimates of the residence time of underground waters in the study area provided values, which ranged from 0.5 to 11.7 years, with a mean value of 5.2 ± 0.9 years.     

Controlling the microstructure of ZnO nanoparticles embedded in Sapphire by Zn ion implantation and subsequent annealing

(0 0 0 1) α-Al₂O₃ single crystals (sapphire) were implanted with Zn ions of 60 keV at a fluence of 1 × 10¹⁷ ions/cm². Transmission electron microscopy and optical absorption spectroscopy studies show the formation of ZnO nanoparticles in the sapphire substrate after the implanted sample was annealed at 700 °C in oxygen ambient. The photoluminescence spectrum of the annealed sample indicates the formation of ZnO nanoparticles with perfect lattice structure. The selected-area electron diffraction pattern proves that the ZnO nanoparticles have the (0 0 0 2) orientation which follows the orientation of Al₂O₃ substrate. The result shows that the crystallographic orientation of nanoparticles obtained through ion implantation is defined by the substrate.     

Tritium analysis in hydraulic oil waste by oxidation technique

The aim of the present study is to investigate a method to evaluate the tritium activity in hydraulic oil waste generated during the operation of Romanian Cernavoda Nuclear Power Plant.The method is based on a combustion technique using the 307 PerkinElmer^® Sample Oxidizer model.The hydraulic oil samples must be processed prior to counting to avoid color quenching (the largest source of inaccuracy) because these samples absorb in the region of 200–500 nm, where scintillation phosphors emit.Prior to combustion of the hydraulic oil waste, tritium recovery degree and tritium retention degree in the circuits of combustion system were evaluated as higher than 98% and less than 0.08%, respectively.After combustion, tritium activity was measured by a 2100 Tri-Carb^® Packard model liquid scintillation analyzer.The blank counts were 16.25 ± 0.50 counts/min, measured for 60 min. The significant activity level value was 6.53 counts/min, at a preselected confidence level of 95%. The Minimum Detectable Activity of a 0.2 mL hydraulic oil sample was calculated to 1.09 Bq/mL. Therefore, the developed method is sensitive enough for the tritium evaluation in the ordinary hydraulic oil waste samples.     

Inactive commissioning of a micro channel catalytic reactor for highly tritiated water production in the CAPER facility of TLK

In future DT fusion machines, several events will generate highly tritiated water (HTW). Among potential techniques for HTW processing, isotopic swamping in a catalytic membrane reactor (PERMCAT) appears promising. The experimental demonstration of PERMCAT for HTW processing has started in the CAPER facility at the Tritium Laboratory of Karlsruhe (TLK). Without any HTW source, such water has to be produced on purpose.Catalytic HT oxidation would ensure clean operation but could be critical for operation due to possible occurrence of explosive mixture. A tritium compatible micro-channel catalytic reactor (μCCR) has been designed and manufactured to produce up to 10 mL min^?1 of HTW with very high specific tritium activity (stoichiometric DTO: 5.2 × 10¹⁶ Bq kg^?1). Prior to its integration in CAPER for tritium operation, this reactor has been commissioned at different feed flow rates, gas composition (air or Helium), and temperature. The results demonstrate the good performances of the μCCR in producing water.The combination of the μCCR with the O₂ sensor represents a reliable system able to produce HTW in a safe way and without radioactive waste. Accordingly, the CAPER facility can be upgrade in order to continue the R&D activity on HTW processing with PERMCAT.     

Effect of ion-irradiation on the microstructure and microhardness of the W-2Y2O3 composite materials fabricated by sintering and hot forging

A W-2Y₂O₃ material was developed in collaboration with the Plansee Company (Austria). An ingot of the material having approximate dimension of 95 mm × 20 mm was fabricated by mixing the elemental powders followed by pressing, sintering and hot forging. The microstructure of the W-2Y₂O₃ composite was investigated using transmission electron microscopy (TEM). The microhardness was studied using nano-indentation technique. We observed that the W-grains having a mean size of about 1 μm already formed and these grains contain very low density of dislocations. The size of the yttria particles was between 300 nm and 1 μm and the Berkovich hardness was about 4.8 GPa. The specimens were irradiated/implanted with Fe and He ions at JANNuS facility located at Orsay/Saclay, France. The TEM disks kept were irradiated/implanted at 300 and 700 °C using Fe and He ions with an energy of 24 and 2 MeV, respectively. The calculated radiation dose was about 5 dpa produced by Fe ions and total He content is 75 appm at both 300 and 700 °C. From the TEM investigation of irradiated samples, few radiation loops are present on the W grains, whereas on yttria particles, the radiation induced damages appear as voids. Berkovich hardness of the irradiated sample is higher than that of the non-irradiated sample. Results on the microstructure and microhardness of the ion-irradiated W-2Y₂O₃ composites are presented in detail.