← carlofanara.meDataphysics ↗
NUCLEAR ENERGY — INTRODUCTION

Two Ways to Split the Atom's Ledger

Fission takes big nuclei apart; fusion pushes small ones together. Both are accounting tricks against the same binding-energy curve — but the fuel you pick for either – fission or fusion changes the physics entirely, as well as the engineering problem .

An informal primer, written for people who already own a periodic table! Introduction to two series on Nuclear Physics, one for fission, the other for fusion.

Every nuclear reactor, fission or fusion, is running on the fact that nuclei are held together by a binding energy (BE). Looking at that curve in figure 1, where BE is plotted against the mass number A (total number of nucleons in a nucleus) the question is which direction you'd rather travel in order to harvest that energy – or at least a fraction thereof - and what would it cost to get there. This is where two major processes can be identified.

Fission vs. Fusion, in one breath each

Fission

Split heavy nuclei with neutrons

A slow (thermal) neutron hits 235U or 239Pu, the nucleus splits, and each split releases roughly 200 MeV — plus 2–3 fresh neutrons to keep the chain going. Energy is carried away by the fragments within the fuel zone. This is the physics behind every commercial reactor running today. Alternatively, a fast neutron hits 238U or Pu to split the atom. Other possibilities exist; we’ll cover those in the fission series.

Fusion

Fuse light nuclei under extreme heat

Force two light nuclei — preferentially deuterium and tritium — close enough for the nuclear strong force to take over. A D-T fusion releases about 17.6 MeV. Per reaction that's an order of magnitude less than fission — but per kilogram of fuel, fusion wins by roughly 4×, because the fuel itself weighs almost nothing.

That per-kilogram comparison is the whole reason fusion is worth chasing despite forty years of "still twenty years away" jokes: a fusion power plant sips fuel. The catch, as we'll get to, is that persuading light nuclei to fuse takes temperatures upward of tens of millions of degrees — plasma, not solid fuel rods.

Why either direction pays out: the binding-energy curve

The reason both routes work is the same one hump-backed curve. Binding energy per nucleon (BE/A) rises steeply for the lightest elements, peaks around iron-56 and nickel-62 — the most tightly bound nuclei in nature — and then declines slowly out toward uranium. Any nuclear reaction that moves away from that peak releases energy: Fission walks down from the heavy end; fusion climbs up from the light end. Both are downhill in binding-energy terms, just approached from opposite ends of the mass table.

Binding energy per nucleon versus mass number A, showing the fusion region at low A and the fission region at high A converging toward the peak near iron-56.
Fig. 1 Binding energy per nucleon (BE/A) vs. mass number A. Fusion draws its energy from the steep climb on the left (H, He, Li isotopes); fission draws from the gentle descent on the right (U, Pu). Iron-56, at the peak, is the most stable nucleus in nature. Edited from a standard nuclear-physics reference chart.

You may notice how much steeper the curve is on the fusion side. Looking at the light elements – mostly the Hydrogen isotopes, their BE is lower, but they are light elements. This explains fusion's per-kilogram advantage.

Many more details on several aspects of Fission and Fusion in the ongoing series of episodes for each, here we just sketch the two routes.


Why fission, why now?

Answering that means going back to the fissions in more details, checking how splitting heavy nuclei – and releasing about 200 MeV per event in the process – is dependent on the mechanics: what does it take to keep that splitting going, on its own, once you've started it? Note we are referring here to rounded numbers for the most popular of the fission reactions, the one involving 235U. Later in the series on fission, we will consider alternative routes and deal with the problem of fuel and waste disposal. For now we content ourselves with the observation that energy is expressed in MeV (106 eV) – Conversion should be easy if you recall that 1 eV = 1.602 × 10−19 J.

The neutron bookkeeping

A fission chain reaction is an accounting exercise for neutrons. Each fission releases two or three new neutrons (noticeable differences here between the different reactor zoology within this interval – we’ll discuss those later in the series); if, on average, exactly one of them goes on to cause another fission, the reaction sustains itself indefinitely. That average is expressed by the multiplication factor k:

Subcritical — k < 1

Each generation of fissions produces fewer neutrons than the last. The chain reaction dies out on its own.

Critical — k = 1

Exactly one neutron per fission goes on to cause the next one. A steady chain reaction — a running reactor.

Supercritical — k > 1

More neutrons every generation. The reaction rate grows in time — the regime a reactor briefly visits to raise power, and never wants to stay in unchecked.

So the next series on fission will dive deeper in the calculation of k, the core and reactors sizes, the neutron materials absorption and the job of energy extraction - and will consider the routes towards miniaturized fission reactors and their deployment for the steep rise in energy demand. The first question it will pose is one of size: how big must a core be to sustain the chain, and how do that size and its dimensioning depend on whether the reactor is thermal or fast?

Before moving to fusion, a quick map of the fission side. The first split is by the speed of the neutrons that sustain the chain: thermal reactors slow them down with a moderator, while fast reactors do without one, and in doing so can turn fertile material into new fuel (breeding). Roughly (examples in parentheses):

THERMAL-NEUTRON REACTORS (moderated)
LWR (PWR) — pressurised light water as coolant and moderator; the bulk of today’s fleet, and of light-water SMR designs · (most commercial plants) BWR — the water boils in the core and the steam drives the turbine directly · (Fukushima Daiichi, ABWR) PHWR — heavy-water moderator, so it can run on natural uranium · (CANDU) HTGR — helium-cooled, graphite-moderated; high-temperature heat · (HTR-PM) others: molten-salt, graphite-moderated, …
FAST-NEUTRON REACTORS (unmoderated; can breed fuel)
Sodium-cooled (SFR) — the most mature fast design · (BN-800) Lead-cooled (LFR) — chemically inert coolant, high boiling point · (BREST-OD-300) others: gas-cooled and molten-chloride fast concepts…

A deeper classification is possible, by coolant, fuel form and so on, but in this series we will deal mostly with SMRs and MMRs.

Branching into fusion: how do you actually confine a plasma?

Whilst fission reactors have converged over seventy years, onto a handful of proven architectures, fusion hasn't yet: it's still an open field of competing confinement philosophies, each with its own flagship experiments. In the last 7-8 years, increasingly, private companies have emerged, each betting on a particular approach. Broadly, everything on the fusion side falls into three families, as depicted below (note in parentheses, the names/examples for each type):

MAGNETIC CONFINEMENT (MFE)
Tokamak — magnetic "bottle" shaped like a donut · (JET, ITER) Stellarator — twisted coils, steadier but harder to build · (Wendelstein 7-X) others: z-pinch, spherical tokamaks…
INERTIAL CONFINEMENT (ICF / IFE)
Direct drive — lasers hit the fuel pellet directly · (HiPER) Indirect drive — lasers heat a "hohlraum" that irradiates the pellet · (NIF)
OTHER / HYBRID APPROACHES
Magneto-inertial · (Helion) Sheared-flow-stabilized Z-pinch · (ZAP Energy) Muon-catalyzed, fusion–fission hybrids…
A snapshot, not a scoreboard. The company names above reflect the field roughly as of end-2024 and are just illustrative pointers, neither an endorsement nor a completeness claim: things are moving fast (funding rounds, milestones, and even approaches shift year to year), so treat it as a sketch we'll refresh in a later pass.

Every one of these is solving the same problem: hold a plasma hot and dense enough, for long enough, that fusion reactions happen faster than the plasma leaks its energy away. Magnetic approaches try to do this steadily, for seconds to indefinitely; inertial approaches try to do it explosively, compressing a pellet so fast that it fuses before it has time to fly apart. We'll come back to how you actually measure "success" here — the Q-factor and break-even question — in a later part of this series on fusion.


One step back: which fuel, and why?

Before confinement, there's a more basic choice: what are you actually fusing? Deuterium-tritium (D-T) is and has long been - the textbook answer, but it's not the only candidate, and the alternatives trade away D-T's easy ignition for other advantages — mainly, fewer or no neutrons (aneutronic fusion). The reaction table below reports the most popular reaction candidates together their probability of occurrence and the energy released per fusion event. Again, in nuclear physics, energy is measured in MeV (106 eV).

D + T → α (3.5 MeV) + n (14.1 MeV)17.6 MeV
D + D → T (1.01 MeV) + p (3.02 MeV)50%4.03 MeV
D + D → ³He (0.82 MeV) + n (2.45 MeV)50%3.27 MeV
D + ³He → α (3.6 MeV) + p (14.7 MeV)18.3 MeV
p + ¹¹B → 3α8.7 MeV

Two of these deserve a flag. D-³He and p-¹¹B are aneutronic, or close to it — their primary products are charged particles (protons, alphas) rather than neutrons. That matters enormously for engineering: neutrons activate the reactor structure, degrade materials over time, and can only give up their energy as heat (via a conventional steam cycle), while charged particles can in principle be converted toward electricity more directly. The price is that both reactions are much harder to ignite than D-T.


That's the ledger, at the level of "why bother", “what do I break” and "what do I fuse." in this brief and approximate introduction, we branched along the two main series in which the two routes will be described in several aspects: from the critical mass and volumes for SMRs and MMRs - thermal or fast – and the fuel routes in Fission; and, for fusion, from the reactivity of the different fuel pairs (how much harder than D-T each alternative really is to ignite) to what "break-even" actually means once you get a plasma burning — scientific Q vs. engineering Q, the triple product nτT, and why headline numbers from NIF or JET need a second read before you believe the "energy gain" claim in the press releases.

Dataphysics Limited works on energy-systems modelling and physics-informed AI. → dataphysics.ae

§