Joachim Roth

Tritium inventory in ITER plasma-facing materials and tritium removal procedures

Interactions between the plasma and the vessel walls constitute a major engineering problem for next step fusion devices, such as ITER, determining the choice of the plasma-facing materials. A prominent issue in this choice is the tritium inventory build-up in the vessel, which must be limited for safety reasons. The initial material selection, i.e. beryllium (Be) on the main vessel walls, tungsten (W) on the divertor upper baffle and dome, and carbon fibre composite around the strike points on the divertor plates, results both from the attempt to reduce the tritium inventory and to optimize the lifetime of the plasma-facing components.In the framework of the EU Task Force on Plasma?Wall Interaction (PWI TF), the many physics aspects governing the tritium inventory are brought together. Together with supporting information from international experts represented by the ITPA SOL/DIV section, this paper describes the present status of knowledge of the in-vessel tritium inventory build-up. Firstly, the main results from present fusion devices in this field are briefly reviewed. Then, the processes involved are discussed: implantation, trapping and diffusion in plasma-facing materials are considered as well as surface erosion and co-deposition of tritium with eroded material. The intermixing of the different materials and its influence on hydrogen retention and co-deposition is a major source of uncertainty on present estimates and is also addressed.Based on the previous considerations, estimates for the tritium inventory build-up are given for the initial choice of ITER materials, as well as for alternative options. Present estimates indicate a build-up of the tritium inventory to the administrative limit within a few hundred nominal full power D?:?T discharges, co-deposition with carbon being the dominant process. Therefore, tritium removal methods are also an active area of research within the EU PWI TF, and are discussed. An integrated operational scheme to slow the rate of tritium accumulation is presented, which includes plasma operation as well as conditioning procedures.

Chemical erosion of carbon based materials in fusion devices

In the last two years the basic erosion mechanisms of carbon due to thermal and energetic hydrogen atoms have been resolved and the range of erosion yield data has been extended into new regimes of ion flux and energy. The present paper reviews the recent achievements in fundamental understanding. Model prediction of the dependence on ion energy and flux are compared with new experimental results from fusion experiments and plasma simulators. For graphites doped with B or Si the decrease of the erosion yield as function of energy and temperature indicates which elemental steps in the erosion process are influenced.

Revised formulae for sputtering data

Abstract A revision and improvement of analytic formulae for calculating sputtering yields is performed based on the large number of experimental and calculated sputtering yield data accumulated at IPP in the last three decades. The Bohdansky formula for calculating sputtering yields as a function of energy is revised by introducing a nuclear stopping cross-section based on an analytic fit to the KrC potential. New analytic expressions for the two fit parameters Q and E th of the Bohdansky formula are deduced. Yamamuras formula for the angular dependence of the sputtering yield is shown not to be valid for self-sputtering and for heavy projectiles at energies near the threshold.

Trapping, detrapping, and replacement of keV hydrogen implanted into graphite

Trapping and detrapping of low energy hydrogen isotopes in graphite is relevant to the collection of fuel particles in graphite probes for diagnostics of particle fluxes and energies in the plasma boundary layer, and to the recycling of fuel particles. The measurement of saturation concentrations by depth profiling of D may be influenced by detrapping of deuterium by the analyzing ion-beam. Detrapping yields have been measured for 1 and 8 keV deuterium implanted to saturation concentrations in papyex and pyrolytic graphite at room temperature by 790 keV H+, 400–2500 keV 3He+, and 400–1600 keV 14N+ ions. Initial detrapping yields of 1 keV implanted deuterium by 790 keV 3He+ and 14N+ are 4 and 56 D-atoms/ion, respectively. The detrapping yiled for 14N+-ions increases with increasing energy similar to the electronic stopping power of the ions indicating the influence of electronic excitation on the detrapping mechanism. Detrapping by 790 keV H+ was found to be negligible. Papyex and pyrolytic graphite show similar behavior. Careful determination of the deuterium saturation concentration from the D concentration profile by D(3He, α)H nuclear reaction method and taking detrapping into account yields a value of 0.4 D-atoms/C-atom. Furthermore, the replacement of hydrogen by deuterium at equal ranges (8keV H+, 7keV D+) has been measured in pyrolytic graphite with elastic recoil detection (ERD) by 2.6–2.8 MeV 4He ions. The H/D isotopic replacement behavior in saturated layers is similar to that found in metals at low temperatures and in low-Z coatings as TiC at room temperature.

Microstructure of Ni and stainless steel after multiple energy He and D implantation

Abstract Microstructural modifications of metal surfaces following multiple energy He+ and D+ implantations ( 250eV ≤ E ≤8000 eV ) at various temperatures T and doses φ were investigated by transmission electron microscopy (TEM). At these implantation conditions no surface blistering is observed. Room temperature He implantation into Ni at φ = 10 20 − to 1023 m−2 subsequently leads to lattice damage consisting of a high density of dislocation loops, dense ordered arrays of small He bubbles (diameter ∼2 nm), formation of elongated channels, and finally a steady-state microstructure consisting of an extensive network of channels and bubbles without detectable ordering. Based on experimental evidence, including thermal re-emission, a model for the evolution of this microstructure and for He release through channels is suggested. Stable bubble lattices in Ni are observed at T . At higher T random arrangements of facetted bubbles (diameter ≤ 40 nm) are found. Evidence for bubble formation after D+ implantations under similar experimental conditions could not be observed.

Hydrogen in tungsten as plasma-facing material

Materials facing plasmas in fusion experiments and future reactors are loaded with high fluxes (1020–1024 m−2 s−1) of H, D and T fuel particles at energies ranging from a few eV to keV. In this respect, the evolution of the radioactive T inventory in the first wall, the permeation of T through the armour into the coolant and the thermo-mechanical stability after long-term exposure are key parameters determining the applicability of a first wall material. Tungsten exhibits fast hydrogen diffusion, but an extremely low solubility limit. Due to the fast diffusion of hydrogen and the short ion range, most of the incident ions will quickly reach the surface and recycle into the plasma chamber. For steady-state operation the solute hydrogen for the typical fusion reactor geometry and wall conditions can reach an inventory of about 1 kg. However, in short-pulse operation typical of ITER, solute hydrogen will diffuse out after each pulse and the remaining inventory will consist of hydrogen trapped in lattice defects, such as dislocations, grain boundaries and irradiation-induced traps. In high-flux areas the hydrogen energies are too low to create displacement damage. However, under these conditions the solubility limit will be exceeded within the ion range and the formation of gas bubbles and stress-induced damage occurs. In addition, simultaneous neutron fluxes from the nuclear fusion reaction D(T,n)α will lead to damage in the materials and produce trapping sites for diffusing hydrogen atoms throughout the bulk. The formation and diffusive filling of these different traps will determine the evolution of the retained T inventory. This paper will concentrate on experimental evidence for the influence different trapping sites have on the hydrogen inventory in W as studied in ion beam experiments and low-temperature plasmas. Based on the extensive experimental data, models are validated and applied to estimate the contribution of different traps to the tritium inventory in future fusion reactors.

He trapping and bubble formation in Ni, stainless steel 316 and amorphous alloys

Abstract He trapping and the formation of bubbles in surface layers of crystalline and amorphous metals following multiple energy He+ implantation (250 eV⩽E⩽8000 eV) at various irradiation temperatures Ti and fluences φ∗ were investigated by means of transmission electron microscopy, ion beam depth profiling and thermal desorption techniques. No surface blistering is observed at these implantation conditions. Materials under investigation were Ni, stainless steel (SS 316) and amorphous alloys of different composition (VITROVAC Ni78Si8B14, VITROVAC Ni40Fe40B20, METGLAS B-Ni2 Ni82.4(CrFeSiB) 17.6). After room temperature implantation at fluences φ ∗ ⩽φ c ∗ ≈ 2 × 10 21 He + m −2 He is effectively trapped in small, presumably overpressurized bubbles (radius r ≈ 1 nm in Ni and SS 316). In crystalline material the He bubbles arrange in ordered domains where they are preferentially aligned parallel to {111} and {220} matrix planes (“bubble lattice” with lattice parameters a1 ≈ 8 – 9 nm). The He containing surface layer of ≈ 150 nm thickness saturates at cHe = 20–30 at.% in Ni and SS 316. The experimental observations in SS 316 indicate that He re-emission through channels takes place at fluences φ ∗ > φ c ∗ in good agreement with a model for He re-emission [2] that accounts for the microstructural evolution at high implantation fluences in Ni. In addition, irradiation-induced precipitation of new phases is observed. In amorphous materials only random arrangements of small bubbles (r ≈ 1–4 nm) are formed. Considerably less He is trapped in the surface layer and depth profiles as well as thermal release spectra show distinct differences. He implantation into Ni and SS 316 at high Ti (300–500°C) leads to random arrangements of facetted bubbles (r ⩽ 25 nm). Evidence is found that coalescence of bubbles during continuous implantation results in the growth of bubbles that eventually intersect the surface and thus contribute to He release. At these temperatures the surface layer saturates at considerably lower He concentrations.

Erosion of Be as Plasma-Facing Material

Abstract The knowledge of the erosion of Be, BeO, Be 2 C, and Be 4 B due to ion bombardment is reviewed. The sputtering yield values obtained experimentally and by computer simulation are given for H, D, T, He, Be, C, O, Ne, and Ar bombardment. The dependence of the yield on ion energy and angle of incidence is discussed as well as the influence of the target temperature and surface roughness. For the case of C and O bombardment the result of surface composition changes due to accumulation of incident ions in the surface layer is discussed and effects due to simultaneous bombardment of hydrogen and C impurity ions are described. Finally, Be is compared to other potential plasma facing materials, such as graphite or tungsten, on the basis of the ratio of the tolerable plasma impurity concentration to the sputtering yield. The effects of a Maxwellian incident energy distribution of deuterium ions or the energy distribution typical for charge exchange neutral deuterium atoms are taken into account. Consequences for the use of Be in future fusion plasma experiments are regarded.

Status of knowledge of chemical erosion of carbon and critical issues for extrapolation to ITER

Chemical erosion of carbon based plasma-facing materials and the re-deposition of tritium containing carbon layers are critical processes for the material choice in ITER. At the occasion of the workshop on plasma-surface interaction related to fusion (PSIF) the present status of knowledge was defined, suggestions for improved measurements discussed and the importance of atomistic modelling of individual processes underlined. In the present paper, the status of knowledge for these processes is outlined and important unresolved issues are identified: the extrapolation of measured erosion yields at high-particle fluxes towards low energies as expected in detached divertor plasmas and the transport and deposition of hydrocarbon molecules to the divertor plate and walls.

Overview of EU CFCs development for plasma facing materials

To improve the properties of carbon materials, the tritium inventory should be reduced, chemical erosion and RES have to be suppressed to increase the resistance to water/oxygen at elevated temperatures. In addition, in the next generation devices, i.e., the International Thermonuclear Experimental Reactor (ITER), plasma disruption, slow transients, and ELMs, which can occur as off-normal events as the result of a transition from detached divertor operation to attached operation causes extremely high heat loading to carbon protection material. Therefore, Carbon fiber composites (CFCs) with high thermal conductivity (300 W m -1 K -1 at 20°C, 145 W m -1 K -1 at 800°C) are favourable. In framework of European Fusion Technology program, a great effort has been made to develop CFCs to meet all requirements. This paper presents an overview in progress of EU CFCs development. The characteristics of CFCs with respect to thermal-mechanical properties, erosion by plasma, tritium retention. H 2 O/O 2 reactions, and neutron irradiation effects were reported.