The erosion of solid surfaces by energetic particle bombardment — a primary mechanism of plasma-facing component wear that limits material lifetimes in fusion reactors.
Sputtering is the process by which atoms are ejected from a solid surface when it is struck by energetic particles — typically ions, neutral atoms, or neutrons. In fusion devices, plasma-facing components (PFCs) are continuously bombarded by ions and charge-exchange neutrals escaping the confined plasma, causing gradual erosion of the wall material. This erosion limits component lifetimes, contaminates the plasma with high-Z impurities, and redistributes material within the vacuum vessel in ways that can trap radioactive tritium.[1]
In physical sputtering, an incoming ion transfers kinetic energy to atoms in the target's surface layers through a cascade of binary collisions. If a surface atom receives enough energy to overcome its surface binding energy, it is ejected. The sputtering yield — the average number of atoms removed per incident ion — depends on the ion species, energy, angle of incidence, and the target material.[2]
Light ions such as deuterium and tritium produce lower sputtering yields on heavy target materials than heavier impurity ions. However, the flux of fuel ions to PFCs is enormous — on the order of 1023–1024 ions per square meter per second in the divertor — so even small yields translate into significant erosion over time.[1]
For carbon-based PFCs, an additional erosion mechanism operates: chemical sputtering. Incoming hydrogen isotopes react chemically with carbon atoms to form volatile hydrocarbons (primarily methane, CH4) that leave the surface regardless of the ion's kinetic energy. Chemical erosion yields peak around 600–800 K surface temperature and can exceed physical sputtering yields at low ion energies. This process, combined with carbon's propensity to co-deposit with tritium, led the fusion community to largely abandon carbon PFCs in favor of tungsten for next-generation devices.[3]
When sputtered wall atoms become ionized in the plasma edge and are accelerated back into the surface, they can cause self-sputtering. If the self-sputtering yield exceeds unity — meaning each returning atom ejects more than one new atom — a runaway erosion cascade can develop. This sets an upper limit on acceptable plasma-edge temperatures for a given wall material. For tungsten, self-sputtering yields approach unity at ion energies above roughly 100 eV for tungsten-on-tungsten impacts, making plasma detachment in the divertor a necessity for ITER and future reactors.[2]
Sputtering erosion constrains nearly every aspect of plasma-facing component design: material selection, component thickness and lifetime, divertor geometry, and the operational regime of the plasma edge. ITER's tungsten divertor is designed to survive roughly 5,000 full-power discharges before replacement, with the erosion budget carefully allocated across normal operation, transient events such as ELMs, and off-normal disruptions. Managing sputtering is inseparable from managing tritium retention, plasma contamination, and long-pulse reactor availability.[1]