For a fusion power reactor, the main characteristics of the materials close to the plasma are that they tolerate a high fluence from 14 MeV neutrons, yet produce a small amount of radioactive waste. Thus low activation materials are desirable, and must surely be developed if fusion is to provide a desirable energy source. ITER will typically produce damage of 3 dpa (displacements per atom) in the austenitic stainless steel of the first wall. With judicious use of existing low Nb and Co grades, this will permit most radioactive waste in ITER (except, essentially, from the blanket) to be cleared for unrestricted re-use a century from the end of operation.

For a commercial power reactor, and even in the DEMO device to immediately follow ITER, the damage that would be done to the first material walls if they were made of stainless steel would be above 300-500 dpa over a 30 year life. Even though these walls could be changed every few years, this is way beyond the capability of austenitic steels, which show significant swelling above damage levels of 30 dpa. Materials able to last longer, or which experience lower damage for a given neutron fluence, certainly need to be used. There are some promising candidates: low activation ferritic steels, and SiC composites, are able to withstand doses in excess of 150 dpa without swelling.

Because of its relatively low damage production rate, ITER cannot carry out the endurance qualification of materials for use in these subsequent devices (or it can but only at very high cost), and a specialised materials test facility (e.g. IFMIF - the international fusion materials irradiation facility) is needed, accumulating doses over smaller volumes at a much higher rate.

Despite these differences, many ITER conditions are highly relevant for reactor design choices, not least because ITER will be the first facility with a true fusion neutron spectrum, so that the materials tests even at the low dose available will give valuable information on fusion/fission correlations, which is important in interpreting the results of IFMIF:

• for diagnostics and their materials, which on today's experiments experience very little radiation, but which on ITER will be in areas of high dose and whose design can cause radiation streaming to deeper structures;
• for plasma-facing materials compatible with plasma purity and with heat removal from the plasma, their joint to underlying heat sinks and structures, and their coolant;
• for magnet structure which, on account of its large size, needs to be assembled using welds which introduce weak points liable to fatigue;
• for superconductor with regard to its behaviour in operation.

Among other materials, beryllium, tungsten and carbon fibre composite (CFC) are chosen for various regions of the first material wall facing the plasma. These materials are joined by various methods to copper-alloy heat sinks, which in turn are joined to stainless steel supporting structures. In the magnets area, niobium-tin alloy superconductor is used for the highest field components, and niobium-titanium elsewhere. In the diagnostic field, the key issue is to choose materials which maintain their optical, electrical or structural properties while experiencing high radiation dose.

Behind the choice of each material, and the particular grade to be used, is a large body of R&D to substantiate the properties, in the main carried out in the frame of the ITER EDA.