The same process that powers the Sun could one day generate clean electricity on Earth. Here's the science behind it, explained from first principles.
Every second, the Sun converts roughly 600 million tonnes of hydrogen into helium, releasing energy that sustains life across our solar system. Fusion power aims to replicate a version of that process here on Earth — not by building a miniature star, but by recreating the conditions under which lightweight atomic nuclei merge and release energy.
Atoms are made of protons, neutrons, and electrons. The nucleus — the dense core of protons and neutrons — is held together by the strong nuclear force, one of the four fundamental forces of nature. When two light nuclei combine to form a heavier one, the resulting nucleus has slightly less total mass than the two originals. That missing mass is converted directly into energy, following Einstein's famous equation E = mc2.
The most studied fusion reaction on Earth uses two isotopes of hydrogen: deuterium (one proton plus one neutron) and tritium (one proton plus two neutrons). When heated to extreme temperatures, these nuclei overcome their natural electrical repulsion and fuse to form a helium-4 nucleus plus a free neutron, releasing 17.6 million electron-volts of energy per reaction.1
Atomic nuclei are positively charged, so they repel each other electrically. To force them close enough for the strong nuclear force to take over, the fuel must be heated to temperatures exceeding 100 million degrees Celsius — roughly six times hotter than the core of the Sun.2 At these temperatures, matter enters a fourth state called plasma: a superheated gas in which electrons are stripped away from nuclei, creating a soup of charged particles.
The Sun achieves fusion partly through its immense gravitational pressure, which we cannot reproduce on Earth. Instead, engineers must find other ways to confine plasma long enough and at high enough density for a meaningful number of fusion reactions to occur.
Magnetic confinement uses powerful magnetic fields to trap the plasma in a doughnut-shaped chamber called a tokamak, or in alternative geometries like stellarators and compact spherical tokamaks. The magnetic field keeps the ultra-hot plasma away from the vessel walls, which would otherwise melt or cool the plasma instantly.3
Inertial confinement takes the opposite approach: rather than holding the plasma steady, it compresses a tiny fuel pellet so rapidly — using intense lasers or particle beams — that fusion occurs before the fuel can fly apart. In December 2022, the U.S. National Ignition Facility demonstrated scientific ignition using this method.4
Magneto-inertial and other hybrid approaches combine elements of both strategies, seeking to reduce the extreme requirements of each.
In a future fusion power plant, the energy released by each reaction would be captured and converted to electricity. In a D-T reactor, most of the energy is carried by fast-moving neutrons, which escape the magnetic field and strike a surrounding structure called a blanket. The blanket absorbs the neutrons' kinetic energy as heat, which is then used to produce steam and drive conventional turbines — much like existing thermal power plants.
The blanket serves a second critical purpose: it contains lithium, which reacts with incoming neutrons to breed new tritium fuel. Since tritium is radioactive with a short half-life (12.3 years) and does not occur naturally in useful quantities, this breeding cycle is essential for a self-sustaining reactor.5
Fusion offers an almost limitless fuel supply — deuterium can be extracted from ordinary seawater, and lithium is abundant in the Earth's crust. A fusion reactor produces no greenhouse gases during operation, generates no long-lived radioactive waste comparable to fission, and cannot undergo a runaway chain reaction. These qualities make it one of the most compelling long-term energy solutions humanity has ever pursued.