The physics of fusion
Fusion releases energy because of a fundamental property of nuclear physics: the mass of two light nuclei is slightly greater than the mass of the nucleus they form when they combine. This mass difference (the mass defect) is converted to energy according to Einstein's equation E = mc². Because c² is enormous (~9 × 10¹⁶ m²/s²), even a tiny mass difference produces a huge energy release.
The most practical fusion reaction uses deuterium (²H) and tritium (³H):
²H + ³H → ⁴He + n + 17.6 MeV
For comparison, burning one kilogram of hydrogen in oxygen releases about 120 MJ. The same mass of deuterium-tritium fuel, fully fused, releases ~338,000 MJ — a factor of nearly three million more.
The challenge is not the energy yield — it is confinement. Fusion requires temperatures of ~150 million °C, at which matter exists as a plasma (a gas of bare nuclei and free electrons). No material container can hold this. Two methods have dominated: magnetic confinement (tokamaks) and inertial confinement (laser implosion).
For the atomic nuclear structure context — protons, neutrons, binding energy — see our atomic structure guide and the isotope glossary entry.
The NIF ignition milestone
The National Ignition Facility uses 192 laser beams delivering 2.05 MJ of ultraviolet light in a pulse lasting a few billionths of a second. The beams converge on a gold cylinder (the hohlraum) containing a 2 mm diameter capsule of frozen deuterium-tritium. X-rays generated inside the hohlraum compress the capsule to densities and temperatures sufficient for fusion — a technique called inertial confinement fusion (ICF).
On 5 December 2022, NIF achieved fusion ignition for the first time: 3.15 MJ out from 2.05 MJ laser input, a gain of 1.54. Subsequent shots in 2023 and 2024 have confirmed and exceeded this, with gains reported up to 2.4 by late 2024.
However, ignition here refers to energy out exceeding laser energy in — not total electrical energy in. The NIF laser system requires ~300 MJ of grid electricity to deliver 2 MJ to the target. Commercial ICF power would require gains of ~100, not 2. The physics problem is solved; the engineering problem is enormous.
ITER and magnetic confinement
The ITER project (International Thermonuclear Experimental Reactor), under construction in Cadarache, France since 2010, takes a different approach. A tokamak — a doughnut-shaped chamber wound with superconducting magnets — confines a plasma using magnetic fields, keeping it away from the reactor walls.
ITER's 23,000-tonne machine is designed to achieve Q = 10 (ten times more fusion energy out than heating power in) and to sustain fusion for 400–600 seconds. First plasma is targeted for the late 2020s; deuterium-tritium operations for the 2030s. The successor demonstration plant, DEMO, is planned to feed electricity to the European grid in the 2050s.
Several private companies are racing ahead of the government programmes. Commonwealth Fusion Systems plans to demonstrate a compact high-field tokamak (SPARC) in 2025 and a pilot commercial plant (ARC) in the 2030s, using high-temperature superconducting (HTS) tape magnets that achieve 20 tesla — double what ITER uses.
The chemistry of fusion fuel
Deuterium (²H, or D) contains one proton and one neutron. It occurs naturally in water at about 1 in 6,400 hydrogen atoms — meaning every litre of seawater contains roughly 33 mg of deuterium, an effectively inexhaustible resource.
Tritium (³H, or T) contains one proton and two neutrons. It is radioactive (half-life: 12.3 years, beta emitter) and does not occur in significant natural abundance. Fusion reactors breed their own tritium by irradiating a lithium-6 blanket with the neutrons produced by the fusion reaction:
⁶Li + n → ⁴He + ³H + 4.8 MeV
This tritium breeding makes a fusion reactor self-sustaining in fuel, with deuterium from the sea and lithium from the Earth's crust as the only inputs. For context on nuclear equations and balancing them, see our balancing equations guide and our element 119 article on the current limits of nuclear physics.