Fission and Fusion: The Numbers

A single gram of uranium-235, fissioned completely, releases about 8\times 10^{10}\ \text{J} — the same energy as burning roughly 2.5 tonnes of coal. A thimbleful of nuclear fuel out-punches a truckload of the chemical kind by a factor of a few million. And over your head, every second, the Sun fuses some 6\times 10^{11}\ \text{kg} of hydrogen into helium and pours the leftover binding energy out as sunlight. Fission and fusion look like opposites — one splits heavy nuclei apart, the other welds light ones together — yet both hand back enormous energy.

You already met the reason qualitatively in the binding-energy-per-nucleon curve. This page is about the quantities: exactly how much energy comes out, why, and what it takes to get it. We will do the energy accounting three ways, split a real uranium nucleus on paper, fuse deuterium and tritium, count neutrons through a chain reaction to see where "critical mass" comes from, and finish at the temperatures and confinement a star — or a tokamak — needs to make fusion go. One idea runs through all of it: energy is released by any move toward the peak of the binding-energy curve.

The master key: climbing toward the peak

The binding energy B of a nucleus is how much energy you would have to pour in to pull it completely apart into free protons and neutrons. Equivalently — this is the whole trick — a bound nucleus weighs less than its separate parts by the mass defect \Delta m = B/c^2. What matters for energy release is not B itself but the binding energy per nucleon B/A, and how it changes across the curve.

That curve rises steeply from hydrogen, peaks near B/A \approx 8.8\ \text{MeV} around iron and nickel (A \approx 5662), then sags gently down to uranium at B/A \approx 7.6\ \text{MeV}. A nucleus near the peak is the most tightly bound — the most stable. So there are two ways to move uphill toward that peak and release the difference:

Both end up with products whose nucleons are more tightly bound than the reactants'. The energy released is simply the improvement in total binding energy, and by Einstein's relation that is the same number as the mass that "went missing":

The unit conversion is worth memorising: 1\ \text{u} of mass defect is worth 931.494\ \text{MeV} of energy. Every number on this page comes out of that one line.

Two worked examples

Example 1 — splitting uranium-235. A slow neutron strikes a {}^{235}\text{U} nucleus; the compound nucleus wobbles and snaps in two. One typical channel is

{}^{235}\text{U} + n \;\longrightarrow\; {}^{141}\text{Ba} + {}^{92}\text{Kr} + 3n.

Add up the masses (in atomic mass units \text{u}):

The mass defect and its energy are

\Delta m = 236.052595 - 235.866562 = 0.186033\ \text{u}, Q = 0.186033 \times 931.494 \approx 173\ \text{MeV}.

That 173\ \text{MeV} is the prompt energy — mostly the kinetic energy of the two fragments flying apart. The fragments are neutron-rich and radioactive, and as they beta-decay toward stability they release the rest, so the total climbs to the famous \approx 200\ \text{MeV} per fission. Spread over 236 nucleons that is only 200/236 \approx 0.85\ \text{MeV} per nucleon — exactly the little downhill step from uranium's B/A to the fragments' higher one.

Example 2 — deuterium–tritium fusion. The reaction chosen for the first fusion reactors, because it ignites at the lowest temperature:

{}^{2}\text{H} + {}^{3}\text{H} \;\longrightarrow\; {}^{4}\text{He} + n.

Masses again:

\Delta m = 0.018884\ \text{u}, \qquad Q = 0.018884 \times 931.494 \approx 17.6\ \text{MeV}.

Only 17.6\ \text{MeV} per event — one-eleventh of a fission. But look per nucleon: 17.6/5 \approx 3.5\ \text{MeV} per nucleon, more than four times fission's 0.85. Because energy per kilogram of fuel scales with energy per nucleon, D–T fusion delivers far more punch per kilogram than fission does. (The 14.1\ \text{MeV} neutron carries most of that energy away and is what a reactor's blanket must catch.)

Per kilogram, and versus chemistry. One kilogram of {}^{235}\text{U} contains about 2.56\times10^{24} nuclei, so fissioning it all yields

(2.56\times10^{24})(200\ \text{MeV})(1.602\times10^{-13}\ \tfrac{\text{J}}{\text{MeV}}) \approx 8.2\times10^{13}\ \text{J/kg}.

The same sum for D–T fusion gives about 3.4\times10^{14}\ \text{J/kg} — four times more. Compare burning coal at roughly 3\times10^{7}\ \text{J/kg}: nuclear fuel beats chemical fuel by a factor of a few million. That is the whole reason we bother with the enormous difficulty of fission and fusion — the energy density is in a different universe. (This is also why the hook is true: one gram of uranium \approx 8\times10^{10}\ \text{J} \approx 2.5 tonnes of coal.)

Chain reactions and critical mass

Fission is special because each event spits out spare neutrons (three, in our example) — and neutrons trigger more fissions. Whether that snowballs, holds steady, or fizzles is captured by a single number, the neutron multiplication factor k: the average number of neutrons from one fission that go on to cause the next fission. After g generations the neutron population is

N(g) = N_0\,k^{g}.

Sweep the slider below through k=1 and watch the curve flip from a fading decay to a runaway climb. The knife-edge at k=1 is dramatic: a fraction of a percent either side is the difference between "off" and "exponential".

Why a minimum mass exists. Neutrons are only useful if they hit another nucleus before escaping through the surface. Neutron production happens throughout the volume, so it scales as r^3; neutron leakage happens at the surface, so it scales as r^2. The ratio of loss to gain goes as r^2/r^3 = 1/r — it shrinks as the lump grows. Below a critical radius too many neutrons leak and k < 1; above it, production wins and k \ge 1. The corresponding mass is the critical mass.

Two more pieces make a controllable reactor. A moderator (water, graphite) slows the fast fission neutrons to low "thermal" speeds, where {}^{235}\text{U} is vastly more likely to grab them — this is what lets natural or lightly enriched uranium reach k=1 at all. And crucially, a small fraction (\sim0.65\% for {}^{235}\text{U}) of fission neutrons are delayed, emitted seconds later by decaying fragments rather than promptly. Reactors are run so that they are subcritical on prompt neutrons alone and reach k=1 only with the delayed ones. That stretches the effective generation time from microseconds to seconds — slow enough for control rods (and human operators) to keep k pinned at 1. Without delayed neutrons, a reactor would be uncontrollable.

Making fusion go: stars and laboratories

Fusion gives more energy per nucleon, so why is it so much harder than fission? Because before two nuclei can fuse, their positive charges must be pushed together against a fierce electrostatic repulsion — the Coulomb barrier. For two protons at the range of the strong force (\sim10^{-15}\ \text{m}) that barrier is of order \sim1\ \text{MeV}. To slam nuclei together that hard by sheer thermal motion needs temperatures around 10^{7}10^{8}\ \text{K}, where matter is a fully ionised plasma.

Remarkably, the Sun's core is "only" about 1.5\times10^{7}\ \text{K} — the typical particle energy (\sim1\ \text{keV}) is a thousand times below the barrier top. Fusion still happens because of two quantum rescues: nuclei in the high-energy tail of the thermal distribution, and quantum tunnelling straight through the barrier. The two effects pull in opposite directions with energy, and their product peaks at the Gamow peak — a narrow window of energies where almost all stellar fusion actually occurs.

In stars, hydrogen becomes helium two ways. In the Sun and cooler stars the proton–proton chain dominates; in hotter, heavier stars the CNO cycle uses carbon, nitrogen and oxygen as catalysts. Both have the same bottom line:

4\,{}^{1}\text{H} \;\longrightarrow\; {}^{4}\text{He} + 2e^{+} + 2\nu_e + \text{energy}, \qquad Q \approx 26.7\ \text{MeV}.

That 26.7\ \text{MeV} is again just a mass defect: four hydrogen atoms outweigh one helium-4 by about 0.7\% of their mass, and that 0.7\% is what lights the sky.

On Earth, we cannot use gravity to confine a plasma the way a star does, so a reactor must both heat the fuel and hold it together long enough and densely enough to get net energy out. The condition is the Lawson criterion, usually stated as a triple product of density n, confinement time \tau and temperature T:

n\,\tau\,T \;\gtrsim\; 3\times10^{21}\ \text{keV}\cdot\text{s}\cdot\text{m}^{-3} \quad(\text{for D–T ignition}).

Two engineering routes chase that number. Magnetic confinement (the doughnut-shaped tokamak) traps a thin, hot plasma in magnetic fields for seconds at a time — low n, large \tau. Inertial confinement crushes a tiny fuel pellet with lasers to enormous density for a few nanoseconds — huge n, tiny \tau. Both are aiming at the same triple product from opposite ends.

Because the road to iron gets harder the closer you get, and eventually stops paying. Fusion in stars stalls at iron: fusing iron into anything heavier would go down the B/A curve, so it absorbs energy rather than releasing it. A massive star that builds an iron core has no fuel left to hold itself up — the core collapses, and the rebound is a supernova. Elements heavier than iron aren't made by ordinary fusion at all; they're forged in the neutron-rich fury of supernovae and neutron-star mergers, by rapid neutron capture. So iron is a peak, not a destination: the universe is slowly drifting toward iron, but the timescales are astronomical and most matter is still hydrogen. The gold in a ring and the uranium in a reactor are both "uphill" of iron — leftovers from cosmic events violent enough to climb the wrong way.

No — and treating it as a universal constant is a classic error. The "critical mass of {}^{235}\text{U} is about 52\ \text{kg}" figure you see quoted is for one specific case: a bare sphere of pure metal. Change the situation and the number moves a lot:

"Critical mass" is a property of a whole configuration — geometry, density, purity, surroundings — not a fixed label on an element.

It's a harmless slogan but a misleading picture. Nothing gets annihilated in a fission or fusion reaction — you have exactly the same protons and neutrons before and after. What changes is how tightly they're bound. A more tightly bound system sits in a deeper potential well, and a deeper well means less total energy — and, since E = mc^2, less mass. The "missing" mass is just the binding energy that was carried off by the fast fragments, neutrons and photons. So the honest statement isn't "mass turned into energy"; it's that the products are more tightly bound, and the released binding energy shows up as kinetic energy and radiation — while the whole system's mass drops by exactly Q/c^2. Mass and energy were never two different things being swapped; they're two labels for the same conserved quantity.