ICF Target Capsule

A hollow sphere smaller than a peppercorn, holding the coldest solid hydrogen on Earth, about to become the hottest thing in the solar system.

An ICF target capsule stores the fusion fuel — a thin layer of frozen deuterium-tritium (DT) ice lining the inside of a sub-millimeter shell — and provides the ablator material whose explosive outward evaporation drives the fuel inward. In the indirect-drive scheme used at the National Ignition Facility (NIF) and the Laser MégaJoule (LMJ), the capsule sits inside a small gold or uranium cylinder called a hohlraum. Laser beams enter the hohlraum through holes at each end and convert to X-rays; those X-rays bathe the capsule from all directions with nanosecond-scale pulses. The ablator surface blows off at hundreds of kilometers per second, and by Newton's third law the remaining shell and fuel accelerate inward, reaching pressures exceeding 100 billion atmospheres and temperatures above 100 million kelvin at the center — sufficient to ignite a self-sustaining thermonuclear burn.

The capsule design must simultaneously satisfy several competing requirements. The ablator must be dense enough to absorb X-rays efficiently and drive a strong implosion shock, but smooth enough — to within 200 nanometers — that Rayleigh-Taylor instabilities do not shred the shell during compression. The DT fuel layer must be uniformly thick (to within roughly 1 μm) and free of surface roughness so that the hot spot forms symmetrically. NIF's current workhorse capsule is high-density carbon (synthetic diamond), about 2 mm in diameter with an ~70 μm ablator shell and a ~60 μm DT fuel layer, produced by plasma-assisted chemical vapor deposition around a silicon carbide mandrel. The hohlraum surrounding it is typically a depleted uranium or gold cylinder about 5.75–6.2 mm in diameter. A tiny fill tube, just 2 μm in diameter in recent designs, injects the DT gas that condenses into the ice layer when the whole assembly is chilled to about 18 K. On December 5, 2022, a NIF implosion with 2.05 MJ of laser energy delivered to the target produced 3.15 MJ of fusion yield — a target gain (Gtarget) of 1.54, the first time any laboratory fusion experiment produced more energy from the fuel than was delivered to it by the driver. By April 7, 2025, a continuous-gradient-doped diamond capsule yielded 8.6 MJ from 2.08 MJ of laser energy, a target gain exceeding four. What those numbers do not include: the NIF facility consumed roughly 300 MJ of wall-plug electricity to charge the capacitors for that 2.05 MJ laser shot — a wall-plug gain far below one, and a fundamental barrier that current laser technology cannot bridge.

In plain terms

Imagine you squeezed a grape so fast and so hard from every direction at once that its center turned into a tiny star for a hundred-billionth of a second. The grape skin is the ablator, the juice is the deuterium-tritium fuel, and the squeeze is done not by your hands but by X-rays moving at the speed of light hitting the skin from all sides simultaneously.

Also called: fusion target, hohlraum capsule, DT pellet, cryogenic target.

Why it matters

The December 2022 NIF shot was the first controlled laboratory demonstration of scientific ignition — energy out from the fuel exceeding energy in from the driver — ending a 60-year quest and confirming the physics basis for inertial confinement fusion.
The precision of target fabrication directly sets the upper bound on fusion yield; surface defects at the 1-μm scale or chemical contamination at the part-per-million level have been shown experimentally to reduce yield by factors of two to eight.
ICF targets provide the only laboratory platform capable of studying matter at stellar interior conditions — relevant to nuclear weapons physics, astrophysics, and fundamental plasma science simultaneously.
Progress from 1.54× to 4× target gain between 2022 and 2025 was driven almost entirely by target quality improvements (capsule surface finish, doping uniformity), not by laser upgrades — showing that fabrication science is now the controlling variable.
A credible laser-fusion power plant would require target gains of roughly 50–100 combined with laser wall-plug efficiencies of 10–20%, neither of which has been demonstrated; the current results establish scientific feasibility but leave a large engineering gap.

Where it lives

National Ignition Facility (NIF), LLNL, California — 192-beam, 2.05–2.2 MJ UV laser; first ignition December 2022; record yield 8.6 MJ (April 2025).
Laser MégaJoule (LMJ), CEA CESTA, France — 176-beam facility designed for 1.3–1.4 MJ on target; operates in indirect-drive; first fusion experiments with DT in a hohlraum achieved in 2019.
OMEGA Laser, University of Rochester LLE — smaller-scale facility (30 kJ) used for direct-drive and cryogenic target experiments that inform NIF design.
General Atomics (San Diego, CA) — primary fabricator of NIF diamond and Be capsules, responsible for the 4Pi metrology system that screens capsule surface quality.
Diamond Materials GmbH (Freiburg, Germany) — partner fabricator for the continuous-gradient-doped diamond capsules that set the 2025 yield record.

Key numbers

Capsule outer diameter (NIF, typical): ~2 mm (High-density carbon (diamond) ablator shell, ~70 μm thick)
DT ice layer thickness (NIF): ~60 μm (Must be uniform to ~1 μm; cooled to ~18 K)
Scientific gain (Dec. 5, 2022 shot): Gtarget = 1.54 (3.15 MJ fusion yield from 2.05 MJ laser-to-target energy; first ignition)
Record target gain (Apr. 7, 2025): Gtarget > 4 (8.6 MJ from 2.08 MJ laser-to-target; diamond capsule with gradient doping)
NIF wall-plug energy per shot (approx.): ~300–400 MJ (Laser system electrical input; roughly 150× the fusion yield of the 2022 shot)

Surprising facts

The DT fuel layer inside a NIF capsule must be uniform to within roughly 1 micrometer of thickness — about 1/100th the width of a human hair — over a sphere 2 mm across; a single 10-nanometer surface pit can seed an instability that kills the implosion.
The 2022 NIF ignition shot produced 3.15 MJ of fusion energy from 2.05 MJ delivered to the target, a scientific gain of 1.54 — but the NIF facility required roughly 300 MJ of wall-plug electricity to power the laser that delivered those 2.05 MJ, meaning the wall-plug gain was approximately 0.01.
When the capsule ignites, the alpha particles produced by DT fusion deposit their energy locally back into the fuel, heating it further in a runaway cascade; in the 2022 ignition shot this 'alpha heating' contributed more than half of the total energy deposited in the hot spot — the signature of a true burning plasma.
The fill tube connecting the cryogenic DT capsule to the gas-handling system in NIF experiments has shrunk from 44 micrometers in early designs to 2 micrometers today, because even a microscopic tube protruding from the shell surface can disrupt the implosion symmetry and cut yield by a factor of several.
By April 2025, NIF had achieved ignition eight separate times; the highest yield shot (8.6 MJ, target gain >4) used a capsule doped with a gradient of tungsten atoms — increasing from zero to 0.44 atomic percent over 10 micrometers of shell thickness — to tune the X-ray absorption profile without introducing macroscopic density steps that could seed instabilities.

Open questions

Burning plasma shows ion velocity distribution inconsistent with thermal equilibrium: Analysis of neutron spectra from high-yield NIF implosions revealed that the relative kinetic energies of fusing D and T ions do not follow the Maxwell-Boltzmann distribution expected for a thermal plasma at the measured ion temperature. Higher-yield shots show a systematic excess of suprathermal ions — particles with energies more favorable for fusion than the bulk temperature would predict — corresponding to ion temperatures inferred from velocity measurements that are higher than the measured bulk temperature. This discrepancy widens with increasing yield and has not been fully explained; proposed causes include large-angle scattering (non-Maxwellian tails driven by alpha heating) and kinetic effects beyond standard hydrodynamic models. The observation implies the igniting plasma may behave more favorably for fusion than standard theory predicts.
Alpha-particle deposition density nearly double model predictions: Kinetic studies comparing one-fluid and two-fluid models of alpha-particle stopping in the December 2022 NIF burning plasma found that alpha energy deposited into DT fuel ions was roughly double the value expected from standard models — about 30% of the 3.5 MeV alpha energy went to ions (not electrons) in the igniting shot. The two-fluid model further predicts a ~4 keV shift upward in the effective ignition criterion compared to single-fluid models, meaning the plasma was operating in a regime where standard ignition thresholds are quantitatively incorrect. These results suggest the plasma coupling physics in burning conditions differs meaningfully from the pre-ignition extrapolations used to design the experiments.
Four-fold yield increase from target quality alone, without changing laser energy: Between July 2023 and April 2025, NIF's record fusion yield jumped from 3.88 MJ to 8.6 MJ — more than doubling — while the laser energy delivered to the target remained essentially constant (2.05–2.08 MJ). The entire gain came from improvements in capsule surface uniformity and the introduction of continuously graded doping profiles. No plasma physics model predicted in advance that capsule-fabrication advances of this magnitude would translate to yield gains of this size; the observation has forced a reassessment of how sensitively implosion performance scales with microscopic target defects, and makes target quality — rather than laser energy — the current primary design lever at NIF.

A short history

Inertial confinement fusion was proposed independently by Nuckolls, Wood, and colleagues at LLNL in the early 1970s, drawing on classified insights from weapons physics. The NIF was authorized in 1993 and achieved first light in 2009; early implosion experiments consistently fell short of ignition because of unexpected ablator mixing and implosion asymmetry. A key insight from 2020–2021 was that capsule surface quality — pits, voids, and inclusions in the diamond shell — was seeding Rayleigh-Taylor fingers that contaminated the hot spot with high-Z material and quenched the burn. Reducing pit count by a factor of ten and the fill-tube diameter from 10 μm to 2 μm enabled the August 2021 shot (1.37 MJ yield) and then the December 2022 ignition milestone. The LMJ achieved its first DT fusion milestone in 2019 and continues to ramp toward higher beam-bundle counts.

Sources

Abu-Shawareb et al. (Indirect Drive ICF Collaboration) — Phys. Rev. Lett. 129, 075001 (2022) (https://link.aps.org/doi/10.1103/PhysRevLett.129.075001)
Achievement of target gain >1 — Physical Review Letters (Dec. 2022, Lund University abstract) (https://portal.research.lu.se/en/publications/achievement-of-target-gain-larger-than-unity-in-an-inertial-fusio/)
NIF record shots timeline — LLNL National Ignition Facility (https://lasers.llnl.gov/about/keys-to-success/nif-sets-power-energy-records)
Target capsule breakthrough enabling record yield — LLNL (April 2025) (https://lasers.llnl.gov/news/target-breakthrough-enabled-fusion-record-nif)
LMJ facility status and design — JACoW / ICALEPCS 2021 (https://jacow.org/icalepcs2021/papers/fral01.pdf)