Nuclear Fission & Fusion
Both processes convert a tiny fraction of mass directly into energy — only their direction on the binding energy curve is reversed.
Nuclear Fission
When a slow neutron strikes a uranium-235 nucleus, it can split into two medium-mass fragments — releasing two to three new neutrons along with it and an enormous amount of energy. Because ²³⁵U sits beyond the binding-energy maximum at iron, the resulting fragments are more tightly bound than the original nucleus — the difference is released as energy.
Because each fission event produces several new neutrons, which can in turn split further nuclei, a self-sustaining chain reaction can result — controlled inside a nuclear reactor, uncontrolled in a nuclear weapon. For the reaction not to die out, a minimum amount of fissile material must be present: the critical mass.
Nuclear Fusion
At the other end of the binding energy curve lies fusion: when two light nuclei such as deuterium (²H) and tritium (³H) merge into helium, the resulting nucleus is more tightly bound than both starting nuclei combined — again the difference is released as energy, and per nucleon it is considerably more than in fission.
The problem: both nuclei are positively charged and repel each other electrostatically. Only at temperatures of several tens of millions of degrees — as inside the Sun or in a fusion reactor — do the nuclei move fast enough for some of them to overcome the Coulomb barrier (with a little help from quantum-mechanical tunneling). Mastering this problem is the reason controlled nuclear fusion on Earth is still not commercially viable today.
E = mc² compared
Both fission and fusion release a tiny mass deficit as energy — but on a scale that dwarfs any chemical reaction: one kilogram of uranium-235, fully fissioned, yields roughly as much energy as two to three million kilograms of coal burned.
Power plant versus fusion reactor
Above 100 million degrees, no solid or liquid fuel can exist anymore — only plasma, a gas of free, electrically charged atomic nuclei and electrons. No known material could form a vessel wall that withstands this temperature by direct contact; any direct contact would instantly cool the plasma and vaporize the wall. The solution exploits precisely the electric charge of the plasma: charged particles can move almost freely along magnetic field lines, but only with great difficulty across them — the Lorentz force forces them onto tight spiral paths around the field lines. If these field lines are bent so that they form closed loops, for example in the shape of a torus (a donut shape), the plasma can circulate along these paths without ever touching a wall — an invisible "magnetic bottle." This requires enormous, mostly superconducting magnetic coils arranged around the vacuum vessel, generating field strengths of several tesla — ten thousand to a hundred thousand times the Earth's magnetic field.
Technically, two different designs have established themselves for this magnetic confinement: in the tokamak — Russian for "toroidal chamber with magnetic coils" — ring-arranged coils together with an electric current induced in the plasma itself generate the necessary magnetic field; the design is comparatively simple and symmetric, but requires a plasma current that can only be sustained in pulses. The international research reactor ITER in southern France, currently the world's largest experimental fusion reactor, follows this principle.
The stellarator, by contrast, dispenses with an induced plasma current and generates the necessary twisted magnetic field solely through complexly shaped, three-dimensionally wound coils — technically far more demanding to build and calculate, but in principle suited to continuous operation. Germany's Max Planck Institute operates the world's largest stellarator, Wendelstein 7-X, in Greifswald.
Top-down view of a tokamak: the plasma circulates in the torus, held by magnetic coils; charged particles can barely leave the field lines, instead spiraling around them.
Inertial confinement fusion takes a completely different approach: instead of magnetic fields, extremely powerful lasers — as at the National Ignition Facility (NIF) in the US — focus their energy within a few nanoseconds onto a tiny fuel pellet, compressing and heating it instantaneously from all sides so that fusion ignites before the material can fly apart again — in 2022, the NIF achieved a net energy gain from the fusion reaction itself for the first time. Alongside this, privately funded companies such as Commonwealth Fusion Systems (compact high-field tokamaks using high-temperature superconductors) and Helion Energy (pulsed, magnetized inertial fusion) are increasingly pursuing their own, sometimes hybrid technical approaches, aiming to realize commercial fusion power plants faster than state-funded research.
The Sun has been running exactly the fusion process that research teams have been trying to recreate on Earth for decades — for 4.6 billion years — only at far lower pressure than in the solar core, which is why much higher temperatures are needed on Earth to reach the same fusion rate.