Understanding hydrogen retention in damaged tungsten using experimentally-guided models of complex multispecies evolution

Abstract

Fuel retention in plasma facing tungsten components is a critical phenomenon affecting the mechanical integrity and radiological safety of fusion reactors. It is known that hydrogen can become trapped in small defect clusters, internal surfaces, dislocations, and/or impurities, and so it is common practice to seed W subsurfaces with irradiation defects in an attempt to precondition the system to absorb hydrogen. The amount of H can later be tallied by performing careful thermal desorption tests where released temperature peaks are mapped to specific binding energies of hydrogen to defect clusters and/or microstructural features of the material. While this provides useful information about the potential trapping processes, modeling can play an important role in elucidating the detailed microscopic mechanisms that lead to hydrogen retention in damaged tungsten. In this paper, we develop a detailed kinetic model of hydrogen penetration and trapping inspired by recent experiments combining ion irradiation, hydrogen plasma exposure, and thermal desorption. We use the stochastic cluster dynamics method to solve the system of coupled partial differential equations representing the mean field description of the multispecies system. The model resolves the spatial distribution of defects and hydrogen clusters during the three processes carried out experimentally and is parameterized with information from atomistic calculations. We find that the calculated thermal desorption spectra are broadly characterized by three H emission regions: (i) a low temperature one where dislocations are the main contributors to the release peaks; (ii) an intermediate one governed by hydrogen release from small overpressurized clusters with multiple overlapping peaks, and (iii) a high temperature one defined by clean isolated emission peaks from large underpressurized bubbles. These three temperature intervals are seen to largely correlate with the depth at which the clusters are found. The relevance of the ‘super abundant’ vacancy mechanism is assessed, finding that its main role is to transfer more clusters from the intermediate to the high temperature regions as its relevance increases. We find this picture to be in very good agreement with the experiments, adding confidence to the predictive potential of the models and their useto understand irradiation damage and plasma exposure effects in plasma facing components.

Perspectives on multiscale modelling and experiments to accelerate materials development for fusion

“…As an example, LiAlO 2 pellets used for tritium production in the Tritium Modernization Program exhibit remarkable dimensional and structural stability during irradiation, even to relatively high burnup ( ∼12% total Li) [342,343] . Characterization of irradiated pellets coupled with computational modelling reveal that the radiation damage is isolated to the interior of the grains, resulting from the tendency of Li to migrate to the grain boundaries (GBs) in the radiation-damaged lattice while Al and O are relatively immobile.…”

Coupling crystal plasticity and stochastic cluster dynamics models of irradiation damage in tungsten

“…where the set {g, s, k} represents the reaction rates of 0th (insertion), 1st (thermal dissociation, annihilation at sinks), and 2nd (binary reactions) order kinetic processes occurring inside Ω. Equation ( 18) can be straightforwardly extended to a PDE to capture diffusion due to defect concentration gradients [34]. Equation ( 18) is solved using the kinetic Monte Carlo (residence-time) algorithm by sampling, selecting and executing events from the set of rates {g, s, k} with the corresponding probability [32].…”

“…where D i and R −1 i are the diffusivity and the lifetime of mobile species i, R i = s + j kij N j . The reader is referred to past publications [32][33][34] for more in-depth details about the model. To set the stage for the coupling between CP and SCD that will be discussed in the next section, we simply note that here we consider defect absorption at sinks (1st order reaction), defined by the reaction rate:…”

“…While enjoying wide applicability in a multitude of scenarios, MFRT suffers from certain limitations that restrict its application to multispecies situations such as those found in nuclear materials subjected to a variable transmutation inventory. Techniques such as stochastic cluster dynamics (SCD) were developed with the aim to overcome these limitations due to intrinsic advantages in their mathematical formulation [32][33][34]. As well, much work has been done to modify CP and adapt it to irradiation scenarios, including sequential linkages of CP and MFRT [35][36][37].…”

Sputtering of the beryllium tungsten alloy Be2W by deuterium atoms: molecular dynamics simulations using machine learned forces

“…The problem of D/T retention and the erosion of the wall materials by the impacts of energetic particles including H isotopes and Be has been intensively studied by experiments together with computational modeling [17][18][19][20][21][22][23][24][25][26][27][28][29][30]. Temmerman and Doerner performed experiments on D retention in codeposited W layers.…”