Nuclear Fusion

Fusion is a physical process in which two light atomic nuclei combine to form a heavier nucleus. A small fraction of their mass is converted into energy according to Einstein’s equation, E = mc². Inside the Sun, hydrogen nuclei combine through several intermediate steps to form helium.

On Earth, the strong gravitational forces of the Sun are absent. Therefore, the positively charged fuel particles must be heated, compressed, or accelerated to extremely high temperatures or energies in suitable reactors so that they can overcome their electrical repulsion (the Coulomb force) and collide.

The most widely pursued approach on Earth is the fusion reaction between the hydrogen isotopes deuterium and tritium, which produces helium and a neutron. Around 80 percent of the energy released is carried by the neutron from the fusion zone into the surrounding components. In a future fusion power plant, this energy could then be used as heat. The two main approaches achieve this in different ways: in magnetic confinement fusion, a hot plasma is confined by magnetic fields. In laser fusion, a small fuel target is rapidly compressed and heated. Both approaches must reliably manage the resulting heat and neutrons, as well as the fuel and materials involved. [1]

Fusion is considered a potential component of a climate-friendly energy system because the fusion reaction itself does not release greenhouse gases such as CO₂. In the long term, fusion power plants could provide large amounts of low-carbon electricity and heat independently of weather conditions, complementing an energy system based largely on renewable sources. However, as with any form of energy generation, the construction, operation, and decommissioning of fusion facilities involve emissions and resource use. [2]

Another advantage is the very high energy density of the fuel: one kilogram of a deuterium-tritium mixture could theoretically release around 340 terajoules of fusion energy. [1] This is roughly equivalent to the energy content of 7,000 tonnes of natural gas (48 MJ/kg [3]). The amount of electricity that can actually be generated is lower because the energy must first be converted into electricity.

Fusion could also contribute to energy security. Deuterium can be obtained from water. Tritium occurs naturally only in trace amounts, but future fusion power plants are intended to produce it themselves using lithium. Water and lithium are more widely available worldwide than many fossil fuels. This could help reduce dependence on individual fuel suppliers and geopolitically concentrated import routes. [4]

The fuel combination currently being pursued most extensively consists of deuterium (D or ²H) and tritium (T or ³H). This reaction achieves high reaction rates under comparatively accessible conditions and is therefore considered the most technologically straightforward route toward a first fusion power plant.

Deuterium is a stable isotope of hydrogen and can be obtained from water. Tritium is radioactive, has a half-life of around 12.3 years, and occurs naturally only in very small quantities. The available global stocks are not sufficient to supply a large fleet of fusion power plants. Future facilities are therefore intended to produce tritium in a so-called breeding blanket. In this process, fusion neutrons interact with lithium-containing material, producing tritium, among other things. The tritium must then be extracted, purified, and fed back into the fuel cycle. ITER is intended to test different modules for this purpose, but it will not yet demonstrate complete tritium self-sufficiency for a power plant.

Tritium presents further challenges for fusion: because of its small atomic size, it can readily penetrate, diffuse through, and escape from materials. Suitable materials, barriers, closed process systems, and continuous monitoring are therefore required. [4]

Fusion and nuclear fission are based on the same physical principle: the mass defect. The difference lies in where the reaction occurs on the nuclear binding energy curve: light nuclei release energy when they fuse (fusion), while very heavy nuclei release energy when they split apart (fission). [1]

Nuclear fission primarily uses uranium-235 or plutonium-239. These materials are difficult to source geopolitically, require enrichment, and are highly radioactive. In fusion, the deuterium-tritium reaction is the most advanced approach (see Question 3).

In nuclear fission, released neutrons can trigger further fission reactions. This chain reaction must be technically controlled. Fusion does not involve a self-sustaining chain reaction: if plasma confinement, fuel supply, or energy input is interrupted, the fusion reaction stops. Nevertheless, fusion facilities must also manage risks associated with tritium, activated materials, stored magnetic or thermal energy, and residual heat.

A key difference concerns the radioactive substances produced and their half-lives. Nuclear fission produces radioactive fission products and transuranic elements. Some of these substances have half-lives ranging from tens of thousands to several million years and therefore need to be safely isolated from the environment for correspondingly long periods. In deuterium-tritium fusion, the immediate products are helium and high-energy neutrons. These are not radioactive. However, the neutrons can activate the materials used in the fusion chamber and adjacent components. Which radionuclides are produced and how long they remain radioactive depends strongly on the material composition, neutron exposure, and operating time.

The targeted development of low-activation materials aims to enable a large proportion of replaced components to be reused, recycled, or disposed of with less effort after a cooling-off period of around 100 years. Tritium must also be handled safely. Fusion therefore also produces radioactive materials, but these are expected to be significantly less long-lived than those produced by nuclear fission – provided suitable materials and recycling concepts are successfully developed and implemented. [5]

To achieve a sufficient number of fusion reactions, temperature, particle density, and confinement time must collectively reach a certain threshold in order to overcome the Coulomb force. The two main approaches currently being pursued differ primarily in how they achieve this so-called triple product. [1]

In magnetic confinement fusion, a hot plasma is confined magnetically for an extended period. In laser fusion, a small fuel target is compressed and heated to extremely high levels, but only for a very brief moment. Both approaches will ultimately require similar basic systems, such as heat removal, a tritium fuel cycle, neutron shielding, and remote maintenance. However, the specific components and supply chains differ considerably. [5]

Magnetic Confinement Fusion

In magnetic confinement fusion, the D-T fuel mixture is first introduced into an evacuated reaction vessel and heated until it reaches the plasma state. Because a plasma consists of electrically charged particles, it can be confined by strong magnetic fields and kept away from the vessel walls. The main concepts are the tokamak and the stellarator. Magnetic fusion in particular requires technologies such as superconducting magnets, cryogenics, high-current systems, vacuum vessels, gyrotrons for plasma heating, divertors, blanket structures, cooling systems, and remote maintenance.

A stellarator generates the entire confining magnetic field using external, highly shaped coils. This means it does not require a plasma current for confinement and is fundamentally better suited to long or continuous operating periods. The challenges lie in the highly complex three-dimensional magnetic coils, demanding manufacturing processes, precise assembly, and unresolved questions about scaling laboratory results to power-plant scale.

Laser Fusion

In laser fusion, a fuel target measuring only a few cubic millimetres and containing deuterium and tritium is used. A large number of powerful laser beams compress and heat the target so rapidly that fusion conditions are created for a brief moment. The fuel is not confined by magnetic fields for an extended period; instead, it is held together by its own inertia during the extremely short compression phase.

In direct drive, the laser beams strike the fuel target directly. In indirect drive, the lasers first heat a cavity, generating X-rays that compress the target uniformly. Other concepts, such as fast ignition and shock ignition, are also being researched.

In 2022, the National Ignition Facility in the United States demonstrated an energy gain at the target level for the first time. In April 2025, 2.08 megajoules of laser energy were delivered to the target, producing 8.6 megajoules of fusion energy. However, the total energy consumption of the laser facility was significantly higher. [6] In addition to magnetic and laser fusion, other, partly hybrid concepts are also being researched.

A commercial fusion power plant does not yet exist. However, its basic design can be derived, for example, from existing tokamak and stellarator concepts.

At the centre is the fusion chamber containing the magnetically confined plasma. The first wall surrounds the plasma region and absorbs high heat loads. Behind it is the blanket, which slows down the high-energy neutrons, converts their kinetic energy into heat, protects components further out, and is intended to produce tritium using lithium. The divertor removes helium and impurities from the plasma and is exposed to particularly high heat loads.

A coolant transports the heat from the blanket and divertor to a downstream energy-conversion system. The basic principle is therefore similar to that of other thermal power plants: heat produces steam or drives another thermodynamic cycle, which in turn powers a turbine and generator to produce electricity. Additional systems are required for the magnets, vacuum, cooling, plasma heating, fuel processing, measurement, control, and remote maintenance.

The overall energy balance is crucial. A power plant will only deliver net electricity once the electrical energy it produces exceeds the total electricity consumption of the plant itself – including plasma heating or lasers, magnet cooling, pumps, the fuel cycle, and other auxiliary systems. [1] [5]

Fusion power plants would likely generate electricity in a similar way to other large thermal power plants, using turbines and generators. For grid integration, it is crucial whether they can operate continuously, flexibly, or in a pulsed mode, how quickly their output can be adjusted, and how much of the generated energy is required to run the plant itself.

Existing power plant sites could be attractive because they often already have strong grid connections, available land, industrial infrastructure, and, in some cases, cooling systems.

The FIRE roadmap of the German National Academy of Science and Engineering (acatech) therefore identifies brownfield sites – such as former coal-fired or nuclear power plant sites – as a potential option. [7] In addition to electricity, the heat generated could potentially also be used for industrial processes, district heating, or hydrogen production. Which application makes economic sense will depend on factors such as temperature, location, available off-takers, and plant availability.

Because the neutrons produced during fusion have no electric charge, they are not confined by the magnetic field. They penetrate the first wall and interact with the materials. This interaction can create lattice defects and gas inclusions in the material. As a result, components can swell, harden, become brittle, or lose thermal conductivity. [5]

In addition, atomic nuclei in the components can become activated and radioactive. Radioactive materials therefore arise primarily from the interaction of neutrons with the fusion chamber and its components, as well as from the handling of tritium.

The amount of these radioactive substances, as well as their activity and decay time, depends largely on the material, neutron dose, and operating time. The FIRE roadmap classifies them as predominantly low- to intermediate-level radioactive waste. Their half-lives are generally on the order of 100 years, making them more manageable than the radioactive products of nuclear fission. [7] Low-activation materials such as EUROFER, shielding, closed tritium systems, and remote maintenance are intended to reduce the resulting impacts. Processes for component replacement, reuse, recycling, and disposal still need to be further developed and tested at power-plant scale.

Fusion is not entirely risk-free; tritium, activated components, stored energy, and potential accident scenarios require a comprehensive safety concept. [5]

Fusion is still in the research and development phase; no fusion power plant is yet feeding electricity into a public grid. [9] However, important milestones have been achieved: magnetic-confinement facilities are increasingly able to control high-temperature plasmas for extended periods. At the National Ignition Facility in the United States, individual experiments have produced more fusion energy than the laser energy delivered to the target. A net energy gain for the facility as a whole has not yet been achieved. [6]

Governments and companies around the world are investing in different fusion approaches. [9] The largest international research project is ITER in France. [11] The facility is intended to demonstrate a plasma gain of Q = 10 – 500 megawatts of fusion power from 50 megawatts of external plasma heating. ITER will not, however, generate electricity. [10] [11]

In China, another major experimental facility, BEST, is under construction. According to current plans, it is expected to be completed by the end of 2027 and subsequently demonstrate a deuterium-tritium plasma as well as energy gain at the plasma level. At the same time, technologies for future electricity generation are intended to be tested. [12]

Private companies are also driving development forward. Commonwealth Fusion Systems plans to achieve a plasma gain of Q > 1 with SPARC in 2027. [13] This is intended to be followed by ARC as a first commercial facility, which, according to the company, is planned to begin supplying electricity to the grid in the early 2030s. These timelines are ambitious company targets, not independent forecasts. [13] [14]

The timelines vary depending on the concept, company, funding, and technological development. Many private companies are targeting the first demonstration facilities or initial electricity generation in the 2030s. These figures are company targets, not independent forecasts. [9] [13] [14]

Several milestones need to be distinguished. A scientific energy gain considers a defined part of the process, such as the plasma or the target. A technical net energy gain requires the entire facility to produce more usable energy than it consumes itself. After that, a fusion power plant must demonstrate that heat removal, the fuel cycle, maintenance, and electricity generation can operate reliably over extended periods. Only then does the question of economic viability, series production, and a broader contribution to the energy supply arise. [2] [15]

The German FIRE roadmap describes a first fusion power plant by 2045 as an ambitious development pathway that would require substantial acceleration and parallelisation. Key tasks include neutron testing, breeding blankets, materials qualification, and the development of industrial manufacturing capacity. [7]

For companies, it is important to note that the market for research equipment, prototypes, test facilities, and components will emerge well before the market for commercial fusion electricity. [16]

The materials used in a fusion power plant must withstand extraordinary and often opposing conditions at the same time. In magnetic-confinement systems, the plasma can reach temperatures well above 100 million degrees Celsius, while the superconducting magnets operate at around −269 °C inside a cryostat. Several protective and functional layers, together with an ultra-high vacuum, separate these regions.

Plasma-facing components such as the first wall and divertor must withstand high heat loads, intense neutron fluxes, and thermal cycling. Tungsten is of particular interest because of its very high melting point (3,422 °C) and its relatively low erosion under plasma exposure. Copper and copper alloys are used as heat sinks where energy needs to be transferred rapidly to a cooling system.

Low-activation steels such as EUROFER are being developed for load-bearing structures. Low activation means that the interaction with fusion neutrons produces as few radioactive substances as possible. The aim is to facilitate the later handling of irradiated components that need to be replaced. The challenge is to manufacture such steels in large quantities with as few activating impurities, such as nickel or cobalt, as possible, while maintaining the material properties required for reactor components.

Superconductors are required to generate strong magnetic fields. High-temperature superconductors can operate at higher temperatures and in stronger magnetic fields than conventional low-temperature superconductors. For industrial applications, however, further advances are needed in tape manufacturing, coil winding, joints, protection systems, cost reduction, and neutron resistance.

In laser fusion, the fusion chamber and its optics are exposed during each shot to neutrons, X-rays, high-energy particles, and target debris. Durable protective coatings and optics are required that can transmit a very large number of high-energy pulses with precision and reliability. The key challenge is not just the performance of individual materials, but the qualified interaction of surfaces, joints, cooling channels, and load-bearing structures. [5]

Fusion requires not only individual scientific breakthroughs, but also resilient supply chains spanning raw materials, qualified semi-finished products and components, testing, maintenance, take-back, and recycling. Tungsten illustrates how raw material supply and high-tech manufacturing interact. Global tungsten mining and processing are highly concentrated. In 2025, around 67,000 of the approximately 85,000 tonnes of global mine production came from China. [17] Other significant reserves are located in countries including Australia, Russia, and Vietnam. A broad expansion of fusion therefore requires diversified sources of supply, additional processing capacity, efficient use of materials, and recycling concepts. However, the precise demand will depend heavily on future plant designs and the pace at which the market develops. [18] [19]

Tungsten also presents technical challenges for the emerging fusion industry. For example, tungsten surfaces must be joined to thermally conductive copper or CuCrZr structures, cooling channels, and load-bearing steel structures. These composite structures must withstand high heat loads, thermal cycling, and neutron fluxes without layers delaminating or cracks forming. The manufacturing and qualification of such components could become a bottleneck. [5] [18]

Industrial opportunities therefore include semi-finished products, coatings, joining technologies, cooled modules, test facilities, quality assurance, and recycling. Comparable supply chains are also developing for magnets, lasers, optics, vacuum technology, power electronics, sensors, robotics, software, and fuel systems. Companies can already enter this emerging market with materials, prototypes, testing technologies, and manufacturing processes. [20]

Tritium must be handled in highly contained process systems and carefully accounted for. A future fuel cycle will include production from lithium, extraction from the blanket, purification and isotope separation, storage, metering, and return to the fusion chamber. [4]

Because tritium is a highly mobile hydrogen isotope, it can penetrate and, to some extent, permeate materials. This cannot be prevented entirely. Multiple containment barriers, suitable materials and coatings, tritium-compatible pumps, valves and storage systems, as well as extraction and purification systems, are intended to keep inventories and potential releases as low as possible. Processes are also needed to recover tritium from gases, water, and component surfaces. Sensors and accounting systems must be capable of detecting even very small quantities.

Many of these requirements create opportunities for industry in areas such as vacuum technology, sealing technology, coatings, process engineering, measurement technology, automation, and safety engineering. [5]

An illustrative example of a company from a traditional industry entering the fusion economy is the collaboration between Kyoto Fusioneering and Shimadzu to develop a prototype turbomolecular pump for use in tritium environments. [20]

For manufacturing companies, entry into the fusion industry does not begin with building a fusion reactor. Instead, it starts where an existing product or process can address a specific technical bottleneck in fusion. For many companies, the first business opportunity therefore lies in developing a prototype, test component, or subsystem. [16] [20]

Fusion facilities in Germany are generally subject to radiation protection law. Research facilities have already been licensed on this basis. At the same time, the legal framework is being further developed so that fusion facilities are explicitly covered and reliable, risk-based procedures can be established. In August 2026, the German federal government presented a draft amendment to the Radiation Protection Act. The draft explicitly identifies fusion facilities as installations for the generation of ionising radiation. The legislative process has not yet been completed. [21]

For future fusion power plants, regulatory issues will particularly concern tritium inventories, neutron radiation, potential releases, activated materials, occupational safety, site selection, maintenance, decommissioning, and waste management. In addition, conventional requirements under construction, environmental, water, and operational safety law will apply.

For internationally active companies, different licensing approaches, technical rules, and standards in the respective target markets are also relevant. Harmonised terminology, testing procedures, and safety requirements could help facilitate market access and avoid the need for repeated qualification efforts.

As of: September 1, 2026.
The sources were used to provide technical and contextual background for the information presented.

[1] Grundlagen der Fusion. Modul 1, Sander Korteweg; herausgegeben von FuseNet, o. J.

[2] Fusion Frequently Asked Questions International Atomic Energy Agency (IAEA), o. J.

[3] 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 2: Energy, Intergovernmental Panel on Climate Change (IPCC), 2006.

[4] ITER Organization: Tritium Breeding

[5] Fusion Materials. Module 4 Simone Mingozzi und Dario Cruz; herausgegeben von FuseNet, o. J.

[6] Lawrence Livermore National Laboratory: Target Breakthrough Enabled Fusion Record at NIF

[7] FIRE – Forschungs- und Innovationsroadmap Fusionsenergie: Zwischenergebnisse. Fusionsenergie in Deutschland – Entwicklungspfade zum ersten
     Fusionskraftwerk
     acatech – Deutsche Akademie der Technikwissenschaften e. V.; Projektleitung: Claudia Eckert und Robert
     Schlögl, 2026. https://www.acatech.de/projekt/forschungs-und-innovationsroadmap-fusionsenergie-fire/

[8] Radiotoxizität von Kernspaltung und Fusion J. Jacquinot, Nuclear Fusion 50, 2010, Artikelnummer laut Ausgangstext: 0144001.

[9] The Global Fusion Industry Report 2026: Fusion Companies Survey by the Fusion Industry As-sociation Fusion Industry Association (FIA), 2026.
      https://www.fusionindustryassociation.org/news/from-the-fia/#industry-reports

[10] ITER Organization: ITER Members

[11] ITER Organization: In a Few Lines - Ziele von ITER

[12] Chinese Academy of Sciences: BEST Research Plan und geplanter Abschluss 2027

[13] Commonwealth Fusion Systems: SPARC

[14] Commonwealth Fusion Systems: ARC

[15] Weiterführend: acatech Impuls - Kernfusion, 2024

[16] Auf dem Weg zum ersten Fusionskraftwerk in Deutschland: Handlungsempfehlungen der deutschen Fusionsindustrie für die politische
       und technologische Agenda
       Industrieverband Pro-Fusion e. V., Februar 2026.
       https://www.pro-fusion.org/news-stories/pro-fusion-publishes-actionable-plan-for-path-to-germanys-first-fusionpower-plant

[17] Tungsten. Mineral Commodity Summaries 2026 Souleymane H. Saloum; U.S. Geological Survey (USGS), 2026.

[18] Supply and demand of tungsten in a fleet of fusion power plants E. Day-San, G. C. Blackett, M. Dorn-hofer, A.
       K. Manduku, M. D. Anderton, L. Tanure und T. P. Davis, 2025.

[19] Safeguarding the sustainable tungsten supply: Advances in green beneficiation and metallurgical technologies 
       Chenquan Ni, Feng Tian, Degang Liu, Yong Liang und Chang Liu, 2026.

[20] Industriekooperation in der Fusion: Hightech für Fusionsanlagen – Wie profitieren Industrieunternehmen von Aufträgen für die Fusionsforschung?
        Max-Planck-Institut für Plasmaphysik, Karlsruher Institut für Technologie und Forschungszentrum Jülich, aktualisierte und erweiterte Ausgabe
        2022.
        https://www.ipp.mpg.de/5288054/industriekooperationen

[21] Bundesrat, Drucksache 461/26: Entwurf zur Modernisierung des Strahlenschutzrechts

Note: Timelines for the first fusion power plants are development targets, not guaranteed forecasts.
Statements regarding materials, waste, and costs depend on the specific plant design.