UC San Diego

A fundamental obstacle to controlled fusion devices is the retention of hydrogenic fuel in Plasma Facing Material (PFM) that forms the first-wall and divertor target. A mixed plasma of primarily deuterium (D) and tritium (T) is magnetically confined and heated in the core of a confinement device. The resulting fusion reaction will create energetic ($>$ MeV kinetic energy) helium (He) and neutrons (n$^0$). The 14 MeV neutrons are unconfined and thus will induce damage throughout the PFM that can act to trap plasma ions that impinge on the surface, become neutral atoms, and then diffuse into the PFM. The ability to accurately quantify the T fuel loss to the PFM is needed to ensure that a reactor design can produce enough T to keep the reactor operating and to ensure radiological safety requirements. Tungsten (W) is a primary candidate for PFMs in current and future fusion devices. To avoid the use of radiologically activated samples, in this work neutron damage and T retention are simulated with heavy ion damage and D respectively. W samples were prepared, subjected to heavy ion displacement damage, and then exposed to a relatively low flux D\textsubscript{2} plasma to populate the induced defects with trapped D atoms. Nuclear Reaction Analysis (NRA) and Thermal Desorption Spectroscopy (TDS) were then used to measure the retention of D within the W samples.

The first-ever study of in-situ annealing of defects during damage production, referred to here as dynamic annealing, was carried out. Plasma-implanted D retention in polycrystalline W that had been previously subjected to copper (Cu) ion damage while holding the samples at a fixed elevated surface temperature was investigated. This approach allows us to determine if the annealing rate is fast enough relative to the rate of damage production to materially affect the defect density within the sample. Both NRA and TDS measure a significant reduction in D retention for samples damaged at elevated temperature. TDS quantitatively shows that the lowest energy trap remains largely unaffected while higher energy traps, induced by Cu ions, are annealed and approach intrinsic concentrations as the temperature during ion damage approaches 1243~K. Analysis of TDS data yields an activation energy of (0.10 $\pm$ 0.02)~eV for recovery of heavy-ion damage-induced traps at elevated temperature.

In order to accurately simulate the experimental implantation and thermal desorption of D in W, the capability of the Tritium Migration Analysis Program (TMAP7) needed to be expanded. TMAP7 can model systems with no more than three active traps per atomic species. To overcome this limitation, we developed a Pseudo Trap and Temperature Partition (PTTP) scheme allowing multiple inactively releasing traps to be accounted for by one pseudo trap, simplifying the system of equations to be solved. Without modifying the TMAP7 source code, the PTTP scheme is shown to successfully model the D retention using six traps. We demonstrate the full reconstruction from the plasma implantation phase through the controlled thermal desorption phase with detrapping energies near 0.9, 1.1, 1.4, 1.7, 1.9 and 2.1~eV for a W sample damaged at room temperature.

In the above experimental modeling, the spatial location and density of traps that define the total trap profile were not well constrained. Motivated by this issue, we devised a new Partial Thermal Desorption Spectroscopy (pTDS) technique to systematically and progressively depopulate trapped deuterium (D) from heavy ion damaged tungsten (W) trap sites to isolate and resolve both their spatial location and detrapping energies. Difference TDS profiles from samples with progressively higher pTDS peak temperature permits the isolation of traps that release between the two pTDS temperatures. Results indicate detrapping energies of 1.0, 1.2, 1.4, 1.5, 1.7, 1.9, and 2.5~eV with instantaneous surface recombination. NRA of these samples then shows three spatial zones of D populated defects: (I) the near surface at depths less than 0.1~$\mu$m, (II) the heavy ion displacement damage peaked near 1~$\mu$m, and (III) the remaining bulk with uniform intrinsic defects. The complete cycle of D$_2$ plasma loading, to pTDS, NRA, and finally full TDS could be accurately modeled with TMAP7 utilizing a Pseudo Trap and Temperature Partition (PTTP) scheme developed previously with these seven distinct release peaks distributed across the three spatial zones.