Swift chemical sputtering and potential development for fusion reactor materials

One of the objectives of the International Thermonuclear Experimental Reactor (ITER) is to demonstrate prolonged fusion power production in deuterium-tritium plasma. The selection of plasma facing materials (PFMs) is a key issue for this objective, and multiple factors have to be taken into account. These include the lifetime of the materials (shortened by e.g. erosion and thermal fatigue), safety requirements (tritium retention and activation) and engineering aspects. Due to the ITER tokamak plasma design, the thermal load and particle flux are divided between different areas in the reactor. Consequently, the material requirements vary with location and the current choice for first wall material is beryllium and the divertor region is to be composed of carbon-fibre-composites (CFC) (strike point tiles) and tungsten (baffle and dome). When energetic atoms or ions escape from the hot plasma in a fusion reactor and hit a wall material, they can cause the material to erode. The erosion is well understood if the ion energy is high enough that the erosion is caused by physical sputtering, i.e. when the ion collides with a sample atom and directly kicks it out of the sample. Alternatively ions with thermal energies can also cause erosion if a chemical etching reaction can take place. In the particular case of hydrogen escaping from a fusion reactor plasma and hitting a carbon-based wall material, high carbon erosion has been observed for hydrogen ion energies which are so low (10-30 eV) that physical sputtering is impossible. On the other hand, no chemical etching reaction has been able to explain the erosion either. Using classical and quantum mechanical atomistic simulations of the ion-sample collision dynamics, we have shown that the observed erosion can be explained by a chemical sputtering mechanism, where the incoming ion attacks a chemical bond between two carbon atoms, and causes the bond to break. This mechanism requires an ion energy of only about 3 eV, and occurs on femtosecond time scales, whence we call it swift chemical sputtering. Sputtering and redeposition will unavoidably lead to formation of surfaces which are made up of a mixture of the three PFMs. The properties of these mixed materials can differ strongly from those of their constituents, hence, an understanding of their effects is considered to be crucial for the reactor operation. The understanding of the retention and recycling of hydrogen isotopes in Be is also of particular interest, since this will give insight into the undesired trapping of tritium and also into the plasma cooling effect by the release of the isotopes. When exploring the consequences of mixed materials and the influence of H isotopes on PFMs, realistic experiments with ITER relevant conditions are preferred, but not yet feasible. An excellent tool is, however, computer simulations and especially the sub-part of this field, the simulation technique based on Molecular Dynamics (MD), which allows for modelling of tens of millions atoms. The reliability of MD is proportional to the accuracy of the interatomic potentials used in the simulations, hence, great efforts must be made in developing proper potentials. We have developed interatomic potentials to be used in MD simulations to study the interplay between the fusion reactor materials beryllium, carbon and hydrogen. These potentials are so called analytical bond-order potentials (ABOP), which were initially developed by Tersoff (1988) to describe covalent solids and extended to metals by Brenner (1990). They are able to describe variations of the local chemical environment, such as bond-breaking, yet at the same time they are computationally efficient. In this contribution, we will present the swift chemical sputtering and show how this mechanism can explain many of the observed features of carbon erosion in fusion reactors. We will also present the development and performance of the recent Be-C-H potentials. (au)