The influence of electric fields and neutral particles on the plasma sheath at ITER divertor conditions

Abstract

The purpose of this thesis is to support the optimization of the ‘exhaust-pipe’, or so-called ‘divertor’, of the nuclear fusion experiment ITER, a large international fusion reactor now under construction in the south of France. We focus particularly on two ‘tools’ for optimization of the plasma conditions in the divertor: electric fields and neutral particles. We look at how these ‘tools’ affect the plasma conditions at divertor surfaces. These conditions determine the type and rates of plasma-surface interaction processes and ultimately the lifetime of these surface materials. A plasma boundary phenomenon that can be changed by the presence of electric fields is the so-called ‘Debye sheath’. This is a voltage drop in the transition between plasma and surface. Extremely localized, it will extend only a few micrometers from the ITER divertor plates into the plasma. However, its voltage is a crucial parameter for the interaction of plasma with these plates, since it determines the impinging ion energy. The change in sheath voltage may be particularly large in ITER where conditions are conducive to the development of large electric fields. This is partly due to the low electron temperature, such that electrical resistivity is high. It is also partly due to the high ion fluxes, which allow large currents to flow since electric currents through the divertor plates are limited by the ion flux. We will see that neutral particles will also influence the boundary conditions in the ITER divertor. Their influence is important, because in contrast to existing tokamak divertors, the ion flux will be so high that the plasma will not be transparent for neutral atoms. Atoms will exchange energy and momentum with plasma particles. Clearly, experiments are required to study the consequences of neutral particles and also electric fields on divertor boundary conditions in the plasma conditions foreseen for ITER. To perform these experiments systematically, we created the projected ITER divertor plasma conditions as closely as possible in a linear laboratory experiment, Pilot-PSI. Not only is this linear experiment unique in its production of the particle and heat fluxes expected in the ITER divertor, it is also able to produce parameters corresponding to the whole range of present day tokamak divertors. As a linear machine, it has large advantages over tokamak divertor experiments. The diagnostic accessibility is significantly improved and plasma parameters can be controlled much more directly. We began the project with the development of two non-intrusive diagnostic techniques for the study of electric fields and atomic neutral density in the Pilot-PSI beam. The first diagnostic developed uses optical emission spectroscopy to probe radial electric fields via the E X B ion rotation drift. Although these rotating ions do not emit radiation, we could estimate their drift velocity by observing radiation from excited neutral atoms. These atoms are coupled to both the hot, rotating ions as well as to cold non-rotating neutrals. A procedure was developed to obtain the ion rotation velocity as well as the ion temperature from measured spectra. Radial electric field strengths could then be deduced. We measured electric field strengths in Pilot-PSI of up to 16 kV/m. Secondly, we needed measurements of the atomic neutral density. Laser induced fluorescence (LIF) is generally well-suited for this purpose, since it probes ground state atomic densities directly and with high spatial resolution. However we found, at the high electron density conditions in Pilot-PSI, the fluorescence signal to be severely limited in strength and the background emission signal to be large. Sensitive LIF measurements were not possible. Absorption spectroscopy provided a good alternative. With this diagnostic we determined an upper limit on the atomic density in the centre of the beam, from which we could calculate the ionization degree (> 85% near the plasma source).We also found that as the electron density in Pilot-PSI was increased to ITER relevant values, there was a strong rise in the neutral atomic density in the beam and also in the proportion of molecules in the vessel that were strongly rovibrationally excited. Electric fields and ion temperatures could also be determined, and were in line with values from optical emission spectroscopy. Finally, we also obtained estimations for the dissociation degree in the vessel (~ 7%) and the proportion of rovibrationally excited molecules entering the plasma beam (~ 30%). The next step was to learn to understand and manipulate the radial electric fields in the beam of Pilot-PSI. We found that the radial electric field at the plasma source exit increased with nozzle diameter of the source and with magnetic field strength. The electric fields (and associated electric current) were found to penetrate into the beam outside of the plasma source with a characteristic length increasing with magnetization of the beam. We could then imitate the situation in a tokamak where electric fields in the plasma interface with electrically conducting divertor surfaces. We experimentally verified that electric current will flow through these conducting surfaces. Furthermore, we confirmed that the local sheath voltage can increase substantially from its typical value without biasing, 2.5kTe/e up to the total voltage difference applied. The sheath voltage increases at positions for which the current into the target is positive. Since the sheath voltage determines ion energies at the target, this may have negative consequences for the lifetime of divertor materials. Especially when there are heavy impurities present in the divertor, the threshold energy for physical sputtering may be surpassed. Experiments confirmed that sheath voltage increase at a floating target (for which the electrical potential is free to change) is avoided if an insulator inhibits surface current. We conclude that material damage reduction can be obtained by placing insulating inserts between electrically floating divertor plates. Finally, we addressed the issue of heating by neutral particles that are reflected from the divertor plates back into the plasma, carrying energy from the sheath. This heating effect will be important in ITER because of the strong ion-neutral coupling projected. We studied its effect in Pilot-PSI, where we amplified its impact by increasing the sheath voltage with target biasing. The result was an increase in the electron temperature measured close to the target. Also, the electron density was observed to decrease while ion flux to the target remained constant. Since electron and ion densities are equal in a quasineutral plasma, this implies rarefication caused by plasma acceleration. We attribute this acceleration to the Bohm criterion, which states that the plasma must accelerate to at least the sound velocity at the sheath edge. Since an increased temperature corresponds to an increase in the sound velocity, extra acceleration close to the target must result. These results are significant because they show that neutral atoms reflected from divertor plates in ITER will have a significant influence on the plasma boundary conditions. This will affect the rates of a range of processes at the plasma-wall interface. One important example is the redeposition rate of eroded divertor plate material. The observed effects are particularly striking when sheath voltages are enhanced either by electric fields in the plasma or by negative plate biasing, but will also play a role when divertor plates are floating or grounded. In conclusion, this thesis presents an experimental study of the influence of electric fields and neutral particles on the plasma conditions close to tokamak divertor plates. Since diagnostic access to tokamak divertors is limited and measurements of densities, temperatures, velocities and ion energies are minimal, good care should be taken in predicting values for these parameters. The predicted effects will be particularly strong at ITER divertor relevant conditions, where electric fields can be large and ion-neutral coupling strong.