I´m no specialist, but could the reason be the automatic production of tritium in the lithium layer by the reactions

6Li + n → 4He + T + 4.78 MeV

7Li + n → 4He + T + n - 2.47 MeV

Is there a lithium layer ? I thought it was made of beryllium...isn't it?

I´ve learnt both. But Be is much more toxical. I don´t know the actual status, it´s not really my profession.

Carbon composite conducts bad the electricity, and the conductivity depends of temperature.

Li6 can be used as it is said by Hanno because it generates tritium and energy but Lithium reacts a lot with lots of compounds and it conducts very well the electricity.

If it would be used non magnetic steel the crio magnets would be preserved, but the magnetic field of the magnets would be partially shielded due eddy currents, such currents would heat the steel.

Steel can be mounted OUTSIDE of the magnets and after the lithium-6 wall. Outside the magnets there is very few B field and no eddy currents would affect to the steel.

It is needed steel arranged in sheets and cooled to avoid eddy currents protecting the magnets from neutrons and ion bombardment

First you have do distinguish the plasma facing materials which are the materials which are in direct contact with the plasma and the blanket where the n - Li reaction has to take place to win new fuelling material. Such a blanket does not exist in JET which is basically operated with pure Deuterium only. Only few experiments with additional Tritium have been performed.

The PFC has to withstand an ion bombardment (sputtering, heat load), but also neutron impact (14.1MeV) and very high UV radiaton which can reach regions far remote from the plasma. On the other hand one has to keep in mind the impact of the PFCs on the plasma. If material is eroded it can enter the plasma and radiate which cools the plasma and can become a show stopper. this favors ligh elements like Be, C which can be fully ionized in the plasma and have moderate Z. Such materials can reach concentrations in the low % range without stopping a reactor. high-Z materails like W are more critical. Here the concentrations have to be below 10^-5 to allow for a burning plasma.

In fact for a very long time C was the choice for the PFCs in fusion experiments. It has a moderate Z, cheap, easy to machine and is very tolerant versus heat overloads since there is ablation only but no melting which causes droplets which can cause further melt events and impurities for the plasma. But low Z materials have in general a higher erosion rate by ion impact than high Z materials and C is even chemically reactive to hydrogen forming hydrocarbons. This lead to high erosion yields and the formation of hydrogen rich carbon flaces which puile up in remote areas in the vessel. In the view of a fusion reactor this means rich of radioactive Tritium. therefore, a C machine would in very, very short time hit the Tritium inventory limit. That was the final C show stopper.

A current trend are high Z-materials (tungsten) due to the low erosion yields. one just has to control the tungsten influx into the confined plasma (under investigation, e.g. ASDEX Upgrade) or multi-material PFCs where the main chamber wall are made from low-Z materials, only the divertor from tungsten. Here the ansatz is that melt damages of the first wall have less impact on the plasma and therefore more acceptable.

Beside the pure material properties one has to keep in mind thet high mechanical stresses occur in a tokamak (heating cycle).All components have also to be produced. The JET first wall also is chosen in the view of future experiments where the wall elements have to be actively cooled. There arise questions like how to integrate the cooling pipes into your PFC material. How to avoid stresses between the materials during the heating cycle (thermal expansion coefficient).

So, I am not an expert in amorph glasses, but I am not really sure if such considerations have already been made. A solution for JET only is falling short.

Thank you very much, Hans, for your detailed explanation. Indeed the subject is very complex. I thought of a combination of both because carbon can conduct heat and electricity,  depending on its form ( graphite,  amorphous, or even nanotubes) and amorphous silica is currently used as a radioactive waste  because it can accommodate any kind of ion, and also can reform bonds, which would prevent cracks to some extent. The layers would pile in such a way that heat would flow in a preferred direction, and this could be optimized to improve the reactor performance.