Fusion fuel is the raw material heated to extreme temperatures inside a fusion reactor so that atomic nuclei merge together, releasing energy. The leading candidate is a mix of two hydrogen isotopes, deuterium and tritium, which produces roughly 17.6 million electron volts of energy per reaction and generates 20 to 100 million times more energy per unit mass than burning a fossil fuel like coal. Several other fuel combinations exist, each with different advantages and trade-offs in temperature requirements, neutron output, and how easy the ingredients are to obtain.
Deuterium-Tritium: The Front-Runner
Almost every major fusion project today, including ITER, is designed around deuterium-tritium (D-T) fuel. When one deuterium nucleus and one tritium nucleus fuse, they produce a helium nucleus and a high-energy neutron, releasing 17.6 MeV of energy. About 80% of that energy rides out on the neutron, which is what makes the reaction both powerful and engineering-intensive: those fast neutrons carry the energy that will ultimately generate electricity, but they also slam into reactor walls and gradually make structural materials radioactive over time.
D-T wins the front-runner spot because it ignites at a lower temperature than any other fusion fuel combination. “Lower” is relative here. The plasma still needs to reach roughly 100 to 150 million degrees Celsius, but competing fuel cycles demand temperatures several times higher. D-T also has the highest reaction rate at those temperatures, meaning more fusions happen per second in a given volume of plasma.
Where Deuterium Comes From
Deuterium is ordinary hydrogen with one extra neutron in its nucleus. It occurs naturally in water at a small but consistent ratio: every cubic meter of seawater contains about 33 grams of deuterium. That makes the ocean an essentially unlimited reservoir. Extracting it is a well-established industrial process, already done at scale for heavy-water nuclear reactors. For fusion purposes, fuel supply is not a concern on the deuterium side.
The Tritium Problem
Tritium is the difficult half of the equation. It’s radioactive with a half-life of about 12.3 years, so it barely exists in nature. Today’s global supply comes almost entirely as a byproduct of heavy-water fission reactors. Ontario Power Generation operates a tritium removal facility at its Darlington nuclear station in Canada, extracting tritium from the heavy water used to moderate its CANDU reactors. South Korea and Romania also extract tritium from their heavy-water reactors, and the United States produces small quantities by irradiating lithium-containing rods in government-owned light-water reactors, primarily for defense purposes.
The total amount available worldwide is limited. ITER alone will need about 12.3 kilograms of tritium, and supplying even 8 to 10 kilograms for a future demonstration power plant in the 2050s would strain Canada’s production capacity. This is why every serious fusion power plant design includes a “breeding blanket,” a layer of lithium surrounding the reactor. When the high-energy neutrons from D-T fusion strike lithium-6, they split it into helium and fresh tritium, replenishing the fuel the reactor consumes. Making this breeding cycle efficient enough to sustain operations is one of the biggest unsolved engineering challenges in fusion.
Deuterium-Deuterium Fuel
A simpler option on paper is to skip tritium entirely and fuse two deuterium nuclei together. This reaction follows two branches with roughly equal probability. One produces tritium plus a proton and releases 4.03 MeV. The other produces helium-3 plus a neutron and releases 3.27 MeV. The advantage is obvious: deuterium is cheap and abundant, so you never need a breeding blanket or a tritium supply chain.
The disadvantage is equally obvious. D-D fusion requires significantly higher temperatures than D-T, and at any given temperature the reaction rate is much lower. A D-D reactor would need to confine a hotter, denser plasma for longer, which current technology cannot reliably achieve. It also still produces neutrons through one of its branches, so it doesn’t fully escape the materials damage problem. Most researchers view D-D as a possible second-generation fuel once D-T reactors prove the basic physics and engineering work.
Helium-3: The Lunar Fuel
Fusing deuterium with helium-3 produces a proton and a helium-4 nucleus with a combined energy of 18.4 MeV, and critically, no neutrons. That means far less radiation damage to reactor components and virtually no radioactive waste. An even cleaner option is fusing helium-3 with itself, which produces only protons and helium-4 at 12.9 MeV, eliminating neutron-producing reactions entirely.
The catch is supply. Helium-3 is vanishingly rare on Earth. The Moon, however, has been absorbing helium-3 from the solar wind for billions of years because it lacks an atmosphere and magnetic field to deflect it. Apollo soil samples showed concentrations of at least 13 parts per billion by weight, with titanium-rich soils averaging around 20 parts per billion. Near the lunar poles, cold trapping may triple that concentration. These numbers sound tiny, but scaled across the Moon’s surface they represent a meaningful energy resource. Extracting it would require heating large volumes of lunar soil to release the trapped gas, an enormous mining operation that remains firmly in the planning stage.
Proton-Boron: No Neutrons, Much More Heat
Proton-boron-11 fusion is the fuel combination that most excites advocates of “clean” fusion. A proton fuses with a boron-11 nucleus and produces three helium nuclei with no neutrons at all. That means no neutron-activated structural waste and the potential for direct energy conversion rather than the traditional steam-turbine cycle.
The physics are punishing, though. The reaction requires ion temperatures around 300 keV, roughly 3.5 billion degrees Celsius, more than 20 times what D-T needs. At those temperatures the plasma radiates enormous amounts of energy as X-rays (a process called bremsstrahlung), which drains power from the system faster than the fusion reactions can replace it. Some physicists have concluded that straightforward thermonuclear proton-boron fusion is simply infeasible, though newer schemes involving non-thermal plasma distributions are being explored by several private companies.
How Fuel Gets Into the Reactor
Fusion fuel doesn’t flow into a reactor like gas into a furnace. The plasma is magnetically confined at millions of degrees, so fresh fuel has to be shot across the magnetic field lines to reach the core. The standard method, used on most tokamaks and stellarators today, is cryogenic pellet injection. Tiny pellets of frozen hydrogen or deuterium, typically 0.4 to 6 millimeters in diameter, are cooled to around 10 to 20 kelvin (near absolute zero) and launched into the plasma by a light-gas gun at speeds of 100 to 1,000 meters per second. Pellets injected from the inner wall of the machine travel more slowly, around 100 to 300 meters per second, because they have to survive passage through curved guide tubes before reaching the plasma.
The idea of firing solid fuel into a fusion plasma was first proposed in 1954, and it took decades before the technology matured enough for routine use. Today nearly every major fusion experiment has at least one pellet injector, and the technique doubles as a way to control plasma instabilities, not just refuel.
What Comes Out the Other Side
The primary “waste” product of D-T fusion is helium, the same inert gas used in party balloons. A fusion power plant producing roughly a gigawatt of electricity would generate only about 0.6 metric tons of helium per year. That helium is non-toxic, non-radioactive, and commercially valuable. The neutron activation of reactor structural materials does create some radioactive waste over a plant’s lifetime, but it is far lower in volume and radioactive longevity than the spent fuel from a fission reactor. Components would typically need to be stored for decades rather than millennia before they could be recycled or disposed of as conventional waste.
For aneutronic fuels like helium-3 or proton-boron, even this structural activation largely disappears. The exhaust is almost entirely charged helium particles, which can in principle be captured and their kinetic energy converted directly into electricity. That vision of a nearly waste-free power source is what keeps researchers working on these harder fuel cycles despite the daunting temperature requirements.