Desorption phase of thermal desorption spectroscopy (TDS) experiments performed on tungsten samples exposed to flux of hydrogen isotopes in fusion relevant conditions is analyzed using a reaction-diffusion model describing hydrogen retention in material bulk. Two regimes of hydrogen desorption are identified depending on whether hydrogen trapping rate is faster than hydrogen diffusion rate in material during TDS experiments. In both regimes, a majority of hydrogen released from material defects is immediately outgassed instead of diffusing deeply in material bulk when the evolution of hydrogen concentration in material is quasi-static, which is the case during TDS experiments performed with tungsten samples exposed to flux of hydrogen isotopes in fusion related conditions. In this context, analytical expressions of the hydrogen outgassing flux as a function of the material temperature are obtained with sufficient accuracy to describe main features of thermal desorption spectra (TDSP). These expressions are then used to highlight how characteristic temperatures of TDSP depend on hydrogen retention parameters, such as trap concentration or activation energy of detrapping processes. The use of Arrhenius plots to characterize retention processes is then revisited when hydrogen trapping takes place during TDS experiments. Retention processes are also characterized using the shape of desorption peaks in TDSP, and it is shown that diffusion of hydrogen in material during TDS experiment can induce long desorption tails visible aside desorption peaks at high temperature in TDSP. These desorption tails can be used to estimate activation energy of diffusion of hydrogen in material.

The mechanisms of growth and coalescence of helium nano-bubbles in tungsten are investigated using molecular dynamics simulations. It is shown that crystal symmetries and governed by them properties of dislocations, generated by the growing nano-bubbles, are responsible for main nano-bubble features revealed, including non-spherical shape and anisotropy of surrounding stress field. The transport of helium atoms in non-uniform stress field is simulated at different temperatures and the transport coefficients are determined. The implications of the considered dislocation and helium dynamics on nucleation and growth of bubbles in tungsten with implanted helium are discussed.

A toroidally symmetric tungsten ring inserted in the lower outer divertor of DIII-D was exposed to 25 repeated, attached L-mode shots in reverse- configuration. Radial profiles of the W gross erosion flux inferred in situ from spectroscopic measurements of the WI line (400.9 nm) during these experiments are well reproduced by ERO-D3D simulations of carbon and tungsten impurity erosion, transport and redeposition in the outer divertor region. Tungsten gross erosion is mainly induced by physical sputtering of tungsten by carbon impurities. The outward radial transport of carbon impurities in the outer divertor is shown to be mainly governed by drifts in the sheath region. In addition, the erosion and redeposition of carbon on tungsten, induced by the implantation of carbon into tungsten modeled with the homogeneous mixed material model, increases the effective flux of carbon impurities onto the tungsten ring (carbon recycling on tungsten). The dynamics of carbon implantation in tungsten is shown to be consistent with the plasma shot duration in DIII-D. Moreover, it is shown that the localized deposition of tungsten measured experimentally in the outboard region away from the tungsten ring is caused by the long-range radial transport of tungsten impurities in the outer divertor region induced by the interplay between poloidal and radial drifts. Such experimental measurements might provide direct quantitative estimations of tungsten net erosion. The modeling and analysis of carbon and tungsten erosion and redeposition presented in this paper demonstrates that various physical mechanisms and their synergistic effects need to be taken into account to accurately describe erosion, transport and redeposition of impurities in tokamak divertors.

Dedicated DIII-D experiments coupled with modeling reveal that the net erosion rate of high-Z materials, i.e. Mo and W, is strongly affected by carbon concentration in the plasma and the magnetic pre-sheath properties. Different methods such as electrical biasing and local gas injection have been investigated to control high-Z material erosion. The net erosion rate of high-Z materials is significantly reduced due to the high local re-deposition ratio. The ERO modeling shows that the local re-deposition ratio is mainly controlled by the electric field and plasma density within the magnetic pre-sheath. The net erosion can be significantly suppressed by reducing the sheath potential drop. A high carbon impurity concentration in the background plasma is also found to reduce the net erosion rate of high-Z materials. Both DIII-D experiments and modeling show that local 13 CH 4 injection can create a carbon coating on the metal surface. The profile of 13 C deposition provides quantitative information on radial transport due to E × B drift and the cross-field diffusion. The deuterium gas injection upstream of the W sample can reduce W net erosion rate by plasma perturbation. In H-mode plasmas, the measured inter-ELM W erosion rates at different radial locations are well reproduced by ERO modeling taking into account charge-state-resolved carbon ion flux in the background plasma calculated using the OEDGE code.

Thermal desorption spectra (TDSP) reported in the literature and resulting from thermal desorption experiments performed on tungsten samples exposed to deuterium at several fluences are analyzed using a reaction-diffusion model including up to ten different types of traps. The use of a large number of types of traps allows accurate fits of TDSP at all fluences using a unique broad spectrum of detrapping energy. Detrapping energies found in this work are in good agreement with detrapping energies predicted by density functional theory when several hydrogen atoms are trapped in a single tungsten vacancy. The comparison between the distribution of deuterium concentrations in material and distribution of trapped deuterium concentrations in TDSP suggests that traps characterized by a broad spectrum of detrapping energy, which may correspond to tungsten vacancies, are predominantly located near the material surface (≲10 nm).