Development of Hydrogen Retention Model Based on Plasma-Tungsten Interaction Analysis

Abstract

The hydrogen retention model was developed to explain temporal and spatial variation of hydrogen plasma properties (ne, Te) using hydrogen particle balance equation. The balance contains outflux of hydrogen ions from plasma to wall and influx of hydrogen neutrals from wall to plasma. The wall is defined as plasma-facing area where hydrogen retention occurs during operation. Tungsten was selected as wall material because it is a representative metal boundary as well as promising fusion plasma-facing material. Particle balance equations contain influx equations from wall to plasma as functions of set of desorption energies for various hydrogen retention reactions. Retention reactions can be formed and enhanced in volume of wall materials by intrinsic and extrinsic hydrogen retention sites depending on various plasma-wall interaction (PWI) conditions. The conditions consist of hydrogen solution, hydrogen oversaturation-induced vacancy trapping, implanted impurity-induced chemical trapping, physical damage-induced defect cluster trapping. The model was firstly constructed with the assumption that plasma properties can be changed by neutral gas influx from wall material because it changes boundary condition between plasma and wall. The influx is recycling flux, which is dependent on volume retention reactions of hydrogen in wall material because retention reactions decide amount of recycling flux and period of recycling. Thus, the purpose of hydrogen retention model is to expect the variation of plasma as functions of hydrogen retention reactions in volume of wall material. The volume retention reaction rate is governed by desorption energies (Edes) of specific set of retention reactions because different set of retention reactions are formed by different PWI conditions. Experiments was performed to obtain desorption energy data for various PWI conditions. The conditions consist of deuterium plasma exposure onto tungsten, deuterium plasma exposure onto carbon-implanted tungsten, deuterium plasma exposure onto defect-formed tungsten, gas-admixed (PHe or Ar ~10-20%) deuterium plasma exposure onto tungsten, deuterium plasma exposure onto recrystallized tungsten. Plasma was consistently exposed onto tungsten with electron cyclotron resonance (ECR) plasma system. The deuterium was used as hydrogen isotope which has higher measurement reliability of thermal desorption spectroscopy (TDS) than hydrogen. The set of desorption energies of specific retention reactions were obtained by using TDS. Because accurate measurement of desorption energy is the precondition for present work, the reliability of TDS was confirmed by international TDS round robin experiment (TDS-RRE). Hydrogen retention reactions in tungsten were figured out with corresponding desorption energies

hydrogen solution (Edes,0: 0.75-0.95 eV), hydrogen oversaturation-induced vacancy trapping (Edes,1: 1.84 eV), implanted carbon impurity-induced chemical trapping (Edes,2: 2.33 eV), physical damage-induced defect cluster trapping (Edes,3: 2.39 eV). In terms of variation effect in fusion-relevant condition, both He ash and Ar puffing gas effects were indirectly understood by using admixing condition

by surface modification and sputtering, both gases reduce retention amount (∆Nwall=50-90 %), which is over the density variation of admixed deuterium plasma (∆ne=10-20%). However, both gases did not change the desorption energy (∆Edes,i=0). For the same point of view, the effect of tungsten recrystallization was also analysed that can reduce hydrogen retention amount (∆Nwall=30-50%) due to reduced fabrication-defects without the change of desorption energy (∆Edes,i=0). The dimensions of volume retentions were extended from subsurface (nm ~ μm) to bulk (μm ~ mm) depending on incident energy of implanted plasma ions (100 eV/D2+), impurity ions (300 eV/C4+), and high energy ions (2.8 MeV/W2+). By using experimentally-obtained desorption energy data, the hydrogen retention model was constructed with long-term volume retention reactions and corresponding desorption energies. Based on the model with experimentally-obtained desorption energy data, validation was performed by expecting temporal plasma variation with wall recovery time. The wall recovery time explains settling time of plasma property as functions of retention reactions and corresponding desorption energies. The higher the desorption energy (0.75 ~ 2.39 eV), the longer the wall recovery time (0 ~ 14,400 sec) for equal wall temperature condition. Hence long-term volume retentions dominate the settling time of plasma property as a rate determining step. The spatial variation of plasma was observed with distance from wall to plasma, where the variation of plasma density occur by hydrogen recycling, is comparable to mean free path (MFP) between neutral particles (D2). The collisions within MFP result in that increased electron-impact ionization due to increased neutral gas (D2) density. The variation of plasma by volume retention reactions cannot be explained or expected by using conventional particle balance equation, which considers wall as fixed boundary condition such as constant recycling factor regardless of change under different PWI conditions. Therefore, this dissertation proved that analysis for expecting transition of plasma must consider volume retention reaction in wall as a transient boundary condition of plasma system.