How Does Nuclear Fusion Work?

What nuclear fusion is

Nuclear fusion is a process in which light atomic nuclei combine to form a heavier nucleus. The reaction can release energy because the products have slightly less mass than the original nuclei; the mass difference is converted into energy according to the relationship described by Einstein’s equation, E=mc². Fusion powers the Sun and other stars. On Earth, research commonly focuses on combining isotopes of hydrogen, especially deuterium and tritium, to produce helium and a high-energy neutron. The reaction’s energy and the difficulty of controlling the fuel make fusion both scientifically demanding and potentially useful as an energy source.

An atomic nucleus contains positively charged protons and electrically neutral neutrons. Because positive charges repel each other, nuclei must approach closely enough for the strong nuclear force to bind them. In a star, gravity confines hot matter over vast distances and times. A laboratory experiment cannot reproduce a star’s size or gravitational pressure, so it uses different methods to heat and confine fuel. The fuel becomes a plasma: an electrically charged state of matter in which electrons are separated from nuclei. Fusion research must create conditions that allow enough nuclei to collide and fuse while keeping the plasma controlled long enough to produce useful energy.

Fusion is distinct from nuclear fission. Fission splits heavy nuclei, such as certain uranium or plutonium isotopes, into smaller nuclei and releases energy. Existing nuclear power plants generally use controlled fission. Fusion combines light nuclei and is not the process used by those plants. The two processes have different fuels, reaction products, engineering requirements, and waste profiles. Calling both “nuclear” describes the involvement of atomic nuclei, not an identical technology.

Why achieving fusion is difficult

The fuel for a fusion reaction must be extremely hot so that nuclei move fast enough to overcome their electrical repulsion. Temperature alone is not sufficient. Researchers also need enough particles in a suitable volume and must maintain the conditions for long enough. This combination is often summarized through plasma temperature, density, and confinement time. A practical power system would additionally need to capture energy, manage heat and radiation, supply or breed fuel, and operate reliably. A successful laboratory reaction is therefore one part of a much larger engineering challenge.

At the temperatures used for fusion experiments, fuel cannot be contained by an ordinary solid vessel without rapidly cooling and damaging the vessel. Magnetic-confinement systems use magnetic fields to guide charged particles and keep much of the hot plasma away from material walls. The tokamak is a widely studied configuration that confines plasma in a ring-shaped chamber using a combination of magnetic fields. A stellarator also uses magnetic fields in a ring-shaped arrangement, but its coils are designed to create the confining geometry without relying on the same plasma current used in a tokamak. Each approach has technical trade-offs involving stability, control, construction, and operating conditions.

In inertial-confinement research, a small fuel capsule is compressed rapidly, often using powerful lasers, so its center reaches conditions in which fusion occurs before the material flies apart. The pulse is brief, and the implosion creates a different experimental regime from magnetic confinement. Both approaches seek to create a sufficient combination of temperature, density, and confinement, but they do so in distinct ways. Neither a successful magnetic plasma experiment nor a successful laser-driven fuel shot alone demonstrates a complete power plant.

Fuel and reaction products

Deuterium is an isotope of hydrogen with one proton and one neutron in its nucleus. It occurs naturally in water in small quantities and can be separated from water. Tritium has one proton and two neutrons. It is radioactive, has a half-life of about 12.3 years, and is scarce in nature. A deuterium-tritium reaction is attractive to many experimental programs because it can occur at lower temperatures than some alternative fusion reactions and releases substantial energy. The main products are a helium nucleus, also called an alpha particle, and a fast neutron.

The charged helium nucleus remains influenced by magnetic fields and can transfer energy to the plasma. The neutron carries most of the reaction energy out of the plasma because it has no electric charge. In a future power plant, neutrons could deposit heat in surrounding materials, and a heat-transfer system could use that heat to generate steam or drive another electricity-conversion cycle. The energetic neutrons also create major design challenges: they can damage materials, alter their properties, and make some components radioactive through neutron activation.

A proposed source of tritium for a power plant is a breeding blanket containing lithium around the fusion chamber. Neutrons interacting with lithium can produce tritium, which would then need to be recovered, processed, and returned to the fuel cycle. This is an engineering goal, not a feature already demonstrated at the scale and reliability required for a commercial fleet. A plant would need to produce enough tritium to replace what it consumes, account for losses and processing delays, and maintain the material safely. The availability of suitable materials and a dependable fuel cycle are central questions for commercial fusion.

How scientists measure success

The phrase “energy gain” can refer to different comparisons. In inertial-confinement experiments, scientific gain may compare the energy released by fusion with the laser energy delivered to the target, or with a more specific energy measure at the target. In magnetic-confinement research, a common plasma measure compares the fusion power produced with the external heating power supplied to the plasma. These measurements are not the same as comparing all energy produced with all energy consumed by an entire facility. Lasers, magnets, cryogenic systems, pumps, control equipment, and other plant systems require energy, and any future power station would need to deliver more useful energy than its complete operating system consumes.

For this reason, a milestone described as “scientific breakeven” should not be read automatically as electrical breakeven or commercial viability. Scientific breakeven is often used for a defined comparison at the plasma or target level. Engineering breakeven would need to account for the full device and its energy systems; net electricity would require the facility to deliver more electricity to the grid than it uses. Different announcements may use different boundaries, so the key is to identify the numerator and denominator, where each is measured, and what equipment is included.

Other important measures include how long a plasma can be maintained, how stable it remains, how frequently a machine can repeat an experiment, and how much energy the device can handle as heat and neutron radiation. A laboratory can demonstrate a reaction without running continuously or generating electricity. A power plant needs systems that work together over many operating cycles: fuel injection, plasma control, heat extraction, component replacement, maintenance, safety systems, and conversion of heat into electrical output. Milestones should be understood in relation to these separate requirements rather than as a single ladder with one universally decisive step.

From experiment to power station

A fusion power station would have to convert reaction energy into a useful and dependable output. In a deuterium-tritium design, neutrons would deposit heat in a surrounding blanket. A coolant would carry that heat to a power-conversion system, which could use a turbine or another technology to produce electricity. Such a plant would also need to protect magnets and other sensitive equipment, manage tritium, maintain vacuum and cooling systems, monitor radiation, and allow safe access for inspections and repairs. These functions make fusion power an integrated engineering problem, not just a matter of reaching a high plasma temperature.

Materials are a major challenge because the reactor’s inner structures would face intense neutron bombardment, high heat loads, and demanding mechanical conditions. Neutron damage can make materials brittle, swell them, or change their performance. Activation occurs when materials become radioactive after exposure to neutrons; its extent depends on the materials and exposure. Selecting materials that can withstand the environment, and designing components that can be replaced or recycled, are important research tasks. These requirements affect cost, maintenance schedules, waste handling, and plant availability.

The fuel cycle also has to work in an operating facility. Tritium must be contained, accounted for, extracted from a breeding blanket if one is used, and supplied back to the plasma. The blanket must protect components and capture heat while allowing tritium production and recovery. Engineering the system involves competing requirements, and experimental demonstrations must show that it can function under relevant conditions. A plant also needs to meet regulatory, safety, and environmental requirements, and its construction and operation must be economically competitive. Technical feasibility and commercial viability are related but separate questions.

Safety and environmental considerations

Fusion is not a chain reaction like the fission process in a conventional nuclear reactor. The conditions needed to sustain fusion are difficult to maintain, and if the plasma loses its required conditions, the reaction stops. This limits the possibility of a runaway chain reaction of the kind associated with fission. However, the absence of that particular risk does not mean fusion facilities have no hazards. High-energy equipment, cryogenic systems, powerful magnets, tritium, activated components, and stored energy all require engineered controls and regulatory oversight.

Deuterium-tritium fusion produces neutrons that activate exposed materials. The resulting radioactive waste profile depends on the materials selected, the design, operating history, and applicable regulatory standards. Fusion is often described as avoiding the long-lived high-level spent fuel characteristic of some fission fuel cycles, but it would still generate activated components and other waste that must be managed. Statements about waste should specify what materials are involved, how long they remain radioactive, and what disposal or recycling assumptions are being made. The details depend on the design and are not uniform across all fusion concepts.

Fusion does not emit carbon dioxide from the nuclear reaction itself. A full assessment of climate and environmental impact would also consider construction, materials, fuel processing, power-system integration, and the emissions associated with those activities. Like any energy technology, fusion would have impacts across its lifecycle. A potential advantage in operational emissions would need to be assessed alongside cost, reliability, siting, water use where relevant, supply chains, and the time needed to build and maintain facilities.

How to interpret fusion claims

When reading about a fusion result, first identify the method and experiment: magnetic confinement, laser-driven inertial confinement, or another concept. Then check what was measured. Was the claim about plasma temperature, duration, a particular reaction, energy released by the fuel, energy delivered to the target, or overall facility output? A technically accurate result can sound broader than it is if the measurement boundary is omitted. Reliable accounts explain both the achievement and the context needed to interpret it.

Next, separate experimental performance from power-plant performance. A facility may produce a record-setting result without generating electricity, operating continuously, or demonstrating a complete fuel cycle. Ask whether the system repeated the result, what equipment and energy inputs were included, what materials and maintenance requirements remain, and whether the claimed next step has been demonstrated or is still a design objective. These questions do not diminish scientific progress; they make it possible to compare progress across different stages.

Finally, distinguish a research milestone from a forecast. A company or research institution may describe a target design or a planned date, but plans are not operational plants. Commercial fusion requires physics performance, durable engineering, fuel supply, regulatory approval, construction, and economic performance to come together. Public reporting is clearest when it identifies which requirements have been tested, which remain under development, and what evidence would show that an experimental system can deliver reliable net electricity.

References

U.S. Department of Energy, Fusion Energy Sciences: https://science.osti.gov/fes

International Atomic Energy Agency, Fusion Energy: https://www.iaea.org/topics/energy/fusion

U.S. National Ignition Facility, National Ignition Campaign: https://lasers.llnl.gov/science/national-ignition-facility

What is the difference between fusion and fission?

Fission splits heavy atoms such as uranium, which is how today's nuclear plants work. Fusion joins light atoms such as hydrogen isotopes, the process that powers the Sun.

Has fusion produced net energy?

In December 2022 the US National Ignition Facility produced more fusion energy than the laser energy delivered to the fuel, but not more than the whole facility consumed.

When will fusion power plants exist?

No commercial fusion plant is operating. The ITER research reactor in France and several private companies are working toward demonstration plants, with timelines that have repeatedly slipped.