Plasma cleaning of diagnostic first mirrors for the nuclear fusion machine ITER

First Mirrors (FMs) will play a crucial role in optical diagnostic systems of future fusion devices like the International Thermonuclear Experimental Reactor (ITER). Unlike today's tokamaks, forthcoming reactors are expected to produce a high level of radiations and neutrons, preventing the use of common optical components (windows, fibers). Instead, an alternative solution based on an optical labyrinth embedded in the neutron shielding and employing metallic mirrors was proposed. Being the first element of the optical path which allows light to cross the neutron shielding, FMs will be placed close to the thermonuclear plasma and will, therefore, be subject to intense thermal and radiations loads, bombardment by plasma particles (mainly charge-exchange neutrals, CXNs) and deposition of material eroded from the plasma-facing components. Especially net deposition of particles eroded from the First Wall (FW), i.e. mainly beryllium (Be) and tungsten (W), can degrade the reflectivity of FMs severely compromising the reliability of the optical diagnostics. Although passive mitigation techniques are predicted to reduce the amount of Be and W on FMs, optical degradation cannot be fully suppressed; in situ mirror cleaning techniques are indispensable. Discharge plasma cleaning is currently considered as the most promising method to tackle this issue.
The main goal of this thesis is to provide an exhaustive study of plasma cleaning techniques of FMs in conditions as close as possible to future fusion reactors. To obtain suitable results for the fusion community and in particular for ITER, experiments were not only conducted in the University of Basel but in numerous international facility as the Swiss Plasma Center in Lausanne, the Paul Scherrer Insitute in Villingen, the National Institute for Laser, Plasma and Radiation Physics (INFLPR) in Bucharest or in two tokamaks, the Joint European Torus (JET) and the Experimental Advanced Superconducting Tokamak (EAST), respectively located in England and China.
Significant outcomes have been obtained from the experimental investigations. Several types
of plasma generation processes were considered and one of them exhibited very promising results, namely capacitively coupled radio-frequency (RF) discharge. Different types of deposits, trying to mimic the one that could be found in ITER were obtained by magnetron sputtering technique and intensively studied. Cleaning tests done on typical ITER contaminant material (Be or W) as well as on a Be proxy (Be is toxic), aluminium (Al), either in pure or mixed thin films, obtained from existing tokamak or deposited in laboratory showed very promising results.
After that, cleaning experiments using larger mirrors up to 200x300 mm2, approaching the final design of ITER's edge Thomson scattering (ETS) FM, were performed in Basel and
provided encouraging results. Despite the type of dielectric properties of the mirror's surface
(conducting, insulating or half-insulating/half-conducting) the cleaning process systematically
exhibited homogeneous etching over the surface. Furthermore, as a strong magnetic field (up to 3.5 T on the mirror locations) will be present in ITER, a thorough experimental investigation on the effect of external B-field on the etching proprieties was conducted. The strength (up to 3.5 T) and orientation of the B-field to the normal of the mirror surface (from 0 to 90°) were shown to be parameters strongly affecting etching rate and homogeneity of the cleaning process. With the B-field being always more tangential to the surface, the etching was getting more and more inhomogeneous. When the B-field was at 90°, e.g. parallel to the surface, the sputtering was impossible on half the mirror. This observation led to a major review of some ITER diagnostics designs with all first mirrors being now repositioned to have a maximum angle of 85° between their normal and the B-field. The first ever cleaning trials in an existing tokamak (EAST) performed on the ETS mock-up will be introduced.
Finally, in the optic of in situ RF plasma cleaning integration in ITER, extensive work was started on the so-called driven grounded electrode technique, where the idea is to ground (e.g. short-circuit) the FMs to ease the implementation in ITER.