Particle Bombardment and Energy Fluxes To the Vessel Walls in Fusion Devices
Based on an article by R. Behrisch (Atomic and Plasma-Material Interaction Data for Fusion, vol.1, p.7, 1991)
Magnetic Confinement of A Fusion Plasma
In the research of controlled thermonuclear fusion, the most promising scheme for the confinement of the hot hydrogen plasma is a strong toroidal magnetic field with closed nested magnetic surfaces, as in a tokamak or a stellarator. However, the confinement of a plasma in a magnetic field is limited. Besides freely moving and gyrating along the magnetic field lines, the plasma particles also diffuse and drift perpendicular to the magnetic surfaces. The plasma thus expands and comes in contact with the surrounding vessel walls which protect the plasma from the outer atmosphere. Neutral atoms in the plasma, neutrons and electromagnetic radiation are not confined by the magnetic field.
The first area of contact of the expanding plasma at the vessel walls is generally a limiter, being a solid which must be designed to withstand the large heat flux and particle bombardment from the plasma. The limiter determines the last closed magnetic surface (LCMS or separatrix). Plasma particles which diffuse and drift from the central plasma across the separatrix form the scrape-off layer (SOL) plasma in the volume outside the separatrix. Magnetic field lines from that volume finally intersect the sides of the limiters and/or the divertor plates. Because of the free movement of particles along the magnetic field lines, these areas are in direct contact with and bombarded by the SOL plasma.
The other areas of the vessel walls, which are at some distance from the LCMS, are mostly oriented nearly parallel to the magnetic field lines. These areas are hit by much smaller plasma fluxes which diffuse from the SOL plasma to the vessel walls instead of hitting a divertor plate or the sides of a limiter.
Finally the electromagnetic radiation, the bremsstrahlung and the recombination and line radiation, from the plasma is nearly uniformly distributed on the vessel walls. The vessel walls are further bombarded with energetic neutral atoms, which are created in charge exchange collisions between neutral atoms entering the plasma and ions in the plasma, and with 14 MeV neutrons for a D-T fusion plasma.
D + T → 4He (3.5 MeV) + n (14.1 MeV)
Particle and power exhaust
In a burning D-T fusion plasma the 3.5 MeV alpha particles from the fusion reaction must be confined for a sufficiently long time to transfer their energy to the plasma in order to balance the energy losses. However, the confinement of the alpha particles should not be too good and the upper limit for the alpha particle confinement time is given by the exhaust of the 4He particles (representing the ash of the D-T fusion process) necessary to prevent poisoning of the plasma by 4He. The D and T lost by fusion processes must be refuelled, for example by D-T gas puffing and/or by injection of D and T pellets.
The energy confinement time in a burning fusion plasma must be sufficient for the 4He energy and any additional power deposited in the plasma to sustain the plasma at the burn temperature and heat the new fuel. This is expressed by the Lawson criterion and/or the ignition criterion which represent a minimum and a maximum criteria for the product of the D-T density and the energy confinement time.
Vessel Wall Load and Plasma-Solid Transition
The particles and the energy leaving the plasma are deposited on the vessel wall structures. To remove the energy, very effective cooling, especially of the divertor tiles and the limiters, is needed. The particles will finally be thermalized and especially the helium has to be pumped away. The particle fluxes and energy deposition from the plasma on the different parts of the vessel wall depend on the parameters of the SOL plasma, the fluxes from the central plasma and the fluxes of particles recycling from the vessel walls.
In the transition region between the hot plasma and the solid, a local thermal equilibrium is generally not possible. Because of the higher velocity of the electrons, the solid is negatively charged with respect to the plasma and a sheath develops. For understanding and modelling vessel walls and the formation of the sheath, a detailed and reliable database is needed for the different atomic processes taking place when the hot plasma comes in contact with the surface of a solid.
Energy and Particle Fluxes
The values of the wall fluxes from a magnetically confined burning D-T fusion plasma to the different parts of the vessel walls have been obtained from the energy and particle exhaust conditions for a reactor. The particle fluxes to the vessel walls have a broad energy and angular distribution and they are different in the different wall areas. Thus, the data for the atomic interactions at the solid surfaces have to be known for a large range of parameters.
Energy fluxes of the vessel walls and divertor plates are determined by the fusion power and hence the neutron flux and He ions. The He ions must be confined for some time and transfer their energy to the plasma in order to balance the energy losses and the neutrons will immediately leave the plasma and penetrate into the vessel walls. Then they will be slowed down by collisions with the atoms of the wall structure material and deposit energy in an about 1 m thick layer called the blanket. Owing to scattering and nuclear reactions in the blanket, the actual neutron flux may finally be up to a factor of ten larger than the source flux. Hence the radiation damage and transmutations in the wall material due to the neutron bombardment and the necessary tritium breeding in reactions with lithium in the blanket structure are important and critical problems for fusion reactor design and for material selection.
When the surface layers of the divertor plates and vessel walls become saturated with hydrogen, neutral hydrogen molecules and atoms will be re-emitted at a rate which makes the reflected and re-emitted flux at a rate which makes the reflected and re-emitted flux about equal to the incident flux. Because of the large cross-section for charge exchange collisions, the flux of neutrals entering the plasma will cause a flux of energetic neutral atoms from the plasma back to the vessel walls. The energy distribution of the neutral flux leaving the plasma corresponds to the plasma temperature and thus the flux of energetic neutrals is routinely used to measure the plasma temperature.
The materials for the vessel walls, limiters and divertor files must be able to withstand the plasma load, i.e. the particle and power fluxes and the heat pulses during plasma discharges, without major erosion and destruction. There will be erosion at the vessel walls by sputtering, arcing and sublimation due to the load from the plasma, and the eroded atoms may partly enter plasma. In the plasma, they represent impurity atoms which cause a reduction of the fusion reactions that is due to the dilution of the D-T ions for a given plasma pressure and an extra energy loss by radiation. The tolerable concentrations of impurities in a burning fusion plasma depend on their atomic number Z, and they decrease proportionally to Zn, with 2 < n < 6. Therefore low-Z materials such as Be, B, Borides, carbides and in particular carbon are favored for the plasma facing components.
Another important criterion for selecting the plasma facing material is its property to withstand the high thermal load from the plasma without melting and excessive cracking. Regarding this criterion, carbon has favorable properties because it does not melt, except at extremely high pressure. In present plasma experiments the best plasma parameters have been achieved with the plasma facing components made of carbon and/or coated with carbon, boron and beryllium.
If the plasma in front of the vessel walls can be kept at a very low temperatures, i.e. well below the threshold energy for sputtering which is in the range of 5-20 eV, also a refractory metal such as tungsten or a compound may be considered.
Plasma-Solid Interaction, Atomic Processes and Required Data
The plasma-solid interaction is necessary for the exhaust of 4He and the alpha energy. For better understanding, control and prediction of the plasma-solid interaction and for the selection of plasma facing materials, a detailed and reliable atomic data base is needed in several topics:
The electrical coupling between the plasma and the solid
The build up of electric field between the plasma and the solid leads to a modification of the energy deposition as well as the angular and energy distribution of particle fluxes from the plasma to the vessel walls. Electric arcs may ignite between the plasma and the vessel wall. The sheath potential between the plasma and a solid surface, the emission of electrons from the solid surface due to the impact of ions, electrons and photons must be considered. Thus, detailed data are needed for ion, electron and photon induced electron emission yields, especially for the materials and surface conditions of limiters and divertor plates as well as for the wall fluxes. The ion fluxes to the vessel wall will be predominantly D, T, and He, but also C and O ions and all ions from the wall material, probably with higher charge states, resulting in larger acceleration in the sheath giving high impact energies.
The energy deposition
The energy deposition from the plasma by charged particles, energetic neutrals and electromagnetic radiation can cause modifications of the wall material. In designing a first fusion reactor, the major problem is to find a material with sufficiently high thermal conductivity to keep the plasma facing surface at a tolerably low temperature to prevent surface melting and massive evaporation. In addition, the material must be able to withstand the thermal shocks occurring during plasma buildup and shutdown, plasma instabilities and disruptions without major surface melting, sublimation and surface cracking.
Plasma particle balance by Ion reflection, implantation, trapping and re-emission
The ions leaving the plasma and impinging on the different part of the vessel walls are lost from the plasma. A fraction of them may be recycled if they are reflected at the vessel walls, the other ions are implanted and may be permanently trapped in the surface layers of the vessel walls. During ion bombardment, previously implanted ions may be desorbed from the surface layers of the vessel walls, while the implanted ions may be trapped. The plasma has to be replenished with the ions permanently trapped in the surface layers of the vessel walls, for example by gas puffing or pellet injection. For controlling and predicting the plasma density and composition, detailed atomic data for ion reflection (including the energy, angular and charge state distributions of the reflected atoms) and for implantation, diffusion, trapping, saturation, and bombardment induced and thermal release are needed for the materials considered for plasma facing components. The data should be provided for targets with different surface topographies and different temperatures, for the bombardment conditions at the different parts of the plasma facing components, with respect to ion masses, ion energies, angles of incidence and current densities. More information is needed on temporary and permanent trapping or 'matrix trapping' of hydrogen and helium in the vessel walls by implantation and codeposition (i.e. bombardment of a solid surface by a flux of carbon and/or metal ions in addition to the flux of hydrogen and helium ions).
Impurity release at the vessel walls
One of the most critical problems in fusion reactor design is the erosion at divertor plates, limiters and vessel walls due to the impact of energetic plasma particles and power deposition. This erosion causes a loss of material, resulting in an undesired material transport along the vessel walls, impurities. The impurities in the plasma, their release from the vessel walls are caused by the plasma exposure, their ionization in the boundary plasma, and their diffusion into the plasma and back onto the vessel walls. The major erosion processes at the vessel walls are physical sputtering, chemical sputtering, thermal sublimation and radiation enhanced sublimation due to ion and neutral particle bombardment and surface heating. A large amount of data for total erosion have been obtained mostly at normal incidence on several relevant wall materials, however, for many important quantities are still only few measured, for example for the dependence of the sputtering yield on the angle of ion incidence with respect to the target normal, the possible dependence of the erosion yield on the ion current density, and the energy distribution of the sputtered particles for different ion bombardment parameters. At the divertor surfaces, where the magnetic field lines intersect at nearly grazing incidence, large angles of incidence will prevail and the dependence of the sputtering yield and the distribution of the sputtering material on the surface topography need to be predicted.