Scientists from the Department of “Physical Problems of Materials Science” at the Institute of Nuclear Physics and Technology of the National Research Nuclear University MEPhI have developed a technology for producing a composite material for future thermonuclear reactors. The scientists have come up with a way to combine a tungsten mesh and the steel base of the reactor structure into a single element.

Rosatom/Global Look Press
Imagine the wall of a thermonuclear reactor — a facility in which scientists are trying to replicate the reactions occurring in the interior of the Sun. Streams of hot plasma are hurled at this wall. That is why choosing the optimal material facing the plasma is one of the main challenges for designers of future power plants. Today, the most popular material for such surfaces is tungsten: it has a record‑high melting point (3422°C) and excellent thermal conductivity.
However, solid tungsten has a weak point — plasma instabilities, during which enormous energy is briefly delivered to certain sections of the wall. This is where the idea of “liquid‑metal” protection comes in handy: if you cover a section of the wall with a thin layer of liquid lithium, it evaporates and creates a cloud that shields the surface from the thermal impact.
To make lithium rise to the surface of the wall on its own, capillary‑porous structures (CPS) are used — essentially, a tungsten sponge made of compressed thin wires. The liquid metal rises through the capillaries from the reservoir upward, toward the plasma, and is renewed naturally, without pumps.
The problem is that such a “sponge” needs to be securely attached to the steel base — the reactor vessel. Spot welding and clamps, as experiments have shown, do not provide good thermal contact: the structure overheats, melts locally, and breaks down. The ideal solution is to make the base and the porous structure a single whole by partially impregnating the tungsten sponge with a metal melt compatible with the steel base of the tokamak.
This is where the difficulties begin. Steel, the traditional structural material for tokamaks, forms a brittle intermetallic compound Fe₇W₆ when it comes into contact with tungsten. This layer forms at the interface between the fiber and the matrix, and the cracks filled with steel spread further due to the difference in thermal expansion coefficients. The team of authors from the Department of Physical Problems of Materials Science at MEPhI — N.S. Popov, A.N. Suchkov, M.Yu. Zharkov, and their colleagues — tested two methods for filling a porous tungsten structure with molten material: casting into a copper mold and immersion infiltration. It turned out that in both cases, the steel matrix contains intermetallic compounds, and the tungsten fiber partially dissolves. The problem of protecting the fiber from the steel melt was not solved: the coatings inevitably dissolve, and consequently, the intermetallic compounds are released again.
The solution was suggested by existing alloys TNI (tungsten‑nickel‑iron). Previous studies had shown that if, instead of steel, one uses an alloy with a nickel‑to‑iron ratio close to 2:1 (66 atomic percent nickel and 34 atomic percent iron), tungsten does not form brittle intermetallic compounds but dissolves in a ductile solid solution. It was precisely this composite — WNi34Fe — that was obtained using the casting method. Its structure contains neither brittle intermetallic layers nor cracks at the phase boundary, despite the noticeable difference in thermal expansion coefficients. At the same time, the tungsten content in such a composite reaches 42 vol.% — significantly higher than in steel counterparts (about 25 %).
“Ni34Fe alloy is the optimal matrix material for the composite. Its structure contains no brittle layers or cracks, and the content of tungsten, which is resistant to high temperatures, is significantly higher than in steel counterparts. Essentially, we have come up with an intermediate layer between the steel reactor vessel and the tungsten structure. It turned out to be something like a sandwich that won’t overheat and break down inside the reactor,” says Nikita Popov, an engineer at the Department of Physical Problems of Materials Science at MEPhI.
The work, supported by a grant from the Russian Science Foundation, demonstrated the main point: it is fundamentally possible to create a single monolithic block consisting of a “tungsten porous structure — metal matrix,” which can be directly attached to the steel reactor vessel with cooling channels. And the best way to do this is by immersion infiltration, which allows you to control the melt temperature, sample heating, and interaction time. Next up is optimizing the process parameters to minimize the dissolution of the tungsten fiber and the amount of undesirable phases. However, the direction of the search is already clear: where steel fails, a nickel‑iron alloy comes to the rescue. Thermonuclear energy, as is well known, is still in the distant future, but each such step in materials science brings closer the day when a small sun will be lit on Earth.
A number of media outlets, including Moskovsky Komsomolets, wrote about the research conducted by scientists at MEPhI.





