EN 日本語
← Back to archive

Nuclear Fusion

The power source of stars · February 8, 2026

Summary

Nuclear fusion is when small atoms combine to form larger ones, releasing enormous energy. It's how the sun works. Scientists are trying to recreate it on Earth for unlimited clean energy — we're getting close but not quite there yet.

1
Squishing atoms together
Elementary school

Everything around you is made of tiny, tiny pieces called atoms. They're so small you can't see them even with a magnifying glass.

The sun is super hot — so hot that atoms crash into each other really, really hard. When they crash hard enough, they stick together and become a bigger atom.

When atoms stick together like this, they release a HUGE amount of energy. That's why the sun is so bright and warm!

Scientists want to do the same thing here on Earth to make electricity. It's like having a tiny piece of the sun in a special container.

2
From alchemy to starlight
High school

Fusion is literally creating new elements — something medieval alchemists dreamed of. When two hydrogen atoms fuse, they become helium plus energy.

The challenge: atoms are positively charged and repel each other like magnets with the same pole. To overcome this, you need extreme conditions:

  • Temperature: 150 million°C (10x hotter than the sun's core)
  • Pressure: Or extreme pressure (like inside stars)
  • Confinement: Keep the hot plasma contained long enough to fuse
Historical context

1920s: Scientists realize the sun runs on fusion. 1952: First hydrogen bomb proves fusion works (uncontrolled). 1958: First fusion reactor designs. 2022: First "ignition" — more energy out than laser energy in (NIF). 2025: Multiple startups race toward commercial reactors.

The promise: One gallon of seawater contains enough hydrogen for fusion energy equivalent to 300 gallons of gasoline.

3
D-T reactions and plasma physics
Undergraduate

The most accessible fusion reaction uses deuterium (D) and tritium (T), both hydrogen isotopes:

D + T → He⁴ (3.5 MeV) + n (14.1 MeV)

This releases 17.6 MeV per reaction — about 4x more energy per nucleon than fission. The neutron carries most of the energy, which is captured as heat.

Two main approaches to confinement:

  • Magnetic confinement (Tokamak): Uses powerful magnetic fields to contain plasma in a torus shape. ITER, the world's largest, aims for Q=10 (10x energy out vs. in).
  • Inertial confinement: Lasers compress fuel pellets so fast that fusion happens before the plasma expands. NIF achieved ignition in 2022.
The Lawson criterion

For net energy gain, you need: n × τ × T > threshold

Where n = plasma density, τ = confinement time, T = temperature. This "triple product" must exceed ~3×10²¹ keV·s/m³ for D-T fusion.

4
Engineering Q vs. scientific Q
Graduate student

The fusion community distinguishes between different definitions of "breakeven":

  • Scientific Q (Qplasma): Fusion power / heating power absorbed by plasma
  • Engineering Q (Qeng): Electrical output / total electrical input (including magnets, cooling, etc.)
  • Wall-plug Q: Net electricity delivered to grid / electricity consumed

NIF's 2022 "ignition" achieved Qplasma > 1, but Qeng was about 0.01 (the lasers consumed 300 MJ to deliver 2 MJ to the target). Commercial viability requires Qeng > 10-20.

Material challenges

14.1 MeV neutrons cause severe radiation damage to reactor walls. Options include:

• Reduced-activation ferritic-martensitic steels (RAFM)

• Silicon carbide composites

• Liquid metal blankets (Li, PbLi) for neutron moderation and tritium breeding

Tritium is rare (12.3 year half-life). Reactors must breed it: Li⁶ + n → T + He⁴. Achieving tritium self-sufficiency (breeding ratio > 1) is a critical unsolved challenge.

5
Beyond D-T: Advanced fuels and compact approaches
Expert

While D-T offers the lowest ignition threshold, advanced fuel cycles avoid the neutron problem:

  • D-He³: Aneutronic but He³ is scarce (lunar regolith, gas giants)
  • p-B¹¹: Fully aneutronic, abundant fuels, but requires ~300 keV ion temperatures — currently out of reach
  • D-D: Abundant fuel, mixed neutronic/aneutronic, could be stepping stone

Novel confinement concepts gaining traction:

  • Spherical tokamaks: Higher beta (plasma pressure / magnetic pressure), more compact. ST40 (Tokamak Energy) achieved 100M°C in 2022.
  • Field-reversed configurations (FRC): TAE Technologies claims efficient ion heating via neutral beam injection
  • Magnetized target fusion: General Fusion's pistons compress magnetized plasma
  • High-field magnets: REBCO superconductors enable 20+ Tesla fields, shrinking reactor size. Commonwealth Fusion's SPARC aims for Q>2 by 2025.
Economic considerations

Levelized cost targets: <$50/MWh to compete with renewables+storage. Key variables:

• Capital cost (dominant — current estimates $5-15B for first plants)

• Capacity factor (tokamaks: pulsed vs. steady-state operation)

• Availability (neutron damage limits component lifetime to ~5-10 years)

Private investment exceeded $6B by 2024, with 40+ companies pursuing various approaches.

The consensus timeline has shifted from "30 years away forever" to potential pilot plants by 2030-2035, with commercial deployment possible by 2040 if current progress continues.

Companies Innovating in Fusion

Sources