NIF fusion ignition: the lasers, targets and measurements behind the gain

Energy gain at the target

The National Ignition Facility at Lawrence Livermore National Laboratory in the United States achieved fusion ignition in December 2022. Its 192 laser beams initiated reactions yielding 3.15 million joules, about 1.5 times the laser energy delivered to the target. Five papers appeared online on 5 February 2024. X-rays from a roughly 3-million-degree-Celsius hohlraum drove a diamond capsule inward at about 400 kilometres per second. Powering the lasers required approximately 350 million joules.

Independent physicists Peter Norreys (Oxford) and Carolyn Kuranz (Michigan) praised the engineering precision. LLNL’s Annie Kritcher described the challenge of maintaining spherical symmetry within 1%. The team raised laser energy from 1.92 to 2.05 megajoules, thickened the diamond shell about 7% and adjusted beam wavelengths by angstroms. A July 2023 experiment reached target gain of 1.9. Those gains did not measure useful electricity.

A capsule’s microscopic imperfections

Diamond’s high density made it the material of choice for NIF’s capsule shells, or ablators, offering higher implosion velocity and ablation pressure than earlier plastic materials. Diamond capsules can contain surface pits, internal voids and foreign-material inclusions. During an implosion, hydrodynamic instabilities can amplify microscopic defects into much larger distortions. LLNL researchers identified capsule imperfections as an important obstacle when experiments failed to reproduce an earlier high-yield result. Thickness variations could also drive motion of the central hot spot.

By April 2023, LLNL, General Atomics and Diamond Materials in Freiburg, Germany, were developing a shared manufacturing sequence. Some shells received their base layer at Diamond Materials, a tungsten-doped layer at Livermore, further polishing and coating in Germany, and quality assessment and fill-tube attachment at General Atomics. Finished capsules then returned to Livermore for placement in hohlraums. The tungsten absorbed X-rays that could otherwise preheat the cryogenic fuel. These travelling batches were an effort to improve capsule quality, rather than evidence that every shell met the increasingly demanding specifications.

More laser energy, with controlled delivery

NIF’s repeated ignition experiments used different input energies. On 8 October 2023, a shot delivered 1.9 megajoules of laser energy and produced 2.4 megajoules of fusion energy. On 30 October, the facility delivered 2.2 megajoules to an ignition target for the first time, producing 3.4 megajoules. That was its fourth ignition experiment. The laboratory’s November announcement tied the higher-energy capability to years of work protecting optics and controlling the laser pulse, rather than to energy increases alone.

Two mitigations completed in June 2023 were especially important: fused-silica debris shields on two-thirds of the beamlines and metal shielding on 32 lower-hemisphere beamlines. The lower optics received particularly heavy debris exposure. The High-Fidelity Pulse Shaping system improved pulse precision and power balance. Refurbished optical fibres also strengthened the signals used to measure laser delivery after repeated neutron exposure. Jean-Michel Di Nicola, a co-program director for laser engineering, identified both energy and precise control as requirements for improving implosions.

Protecting and maintaining the optical system

The laser pulse begins at low energy in the master oscillator room before substantial amplification. A three-year modernisation completed in 2022 upgraded that system to meet tighter performance requirements. Precise timing, energy balance across the beams and frequency broadening helped control the delivered pulse. LLNL’s fiscal 2023 assessment also documented a third protective layer for the final optics: a fused-silica debris shield placed between the disposable shield and the valuable grating debris shield.

Installation began in 2021. The laboratory reported a 98% reduction in damage sites on the grating debris shield, a result specific to that component and measure. Inspection and repair processes helped make the higher-energy shots possible. The same annual assessment described a multiyear sustainment plan with 30 identified refurbishment and recapitalisation activities. These addressed ageing equipment, debris and radiation tolerance, including work on the Final Optics Damage Inspection telescope. The plan sought to sustain performance through the facility’s design lifetime to 2040; it was a maintenance programme, not a completed extension.

Examining the whole capsule

General Atomics’ 4Pi Integrated Metrology System changed how candidate capsules were selected. By September 2024, the system was characterising an average of 30 targets a week. Its instruments examined the entire capsule surface and the alignment of the shell’s inner and outer layers. Four pi steradians describes the full solid angle around a sphere. The name therefore reflected whole-surface measurement rather than a small set of representative views. Surface pits and misalignment could both contribute to hot-spot drift.

Automation combined up to eight technologies in a common coordinate system, reducing errors associated with moving capsules manually between instruments. LLNL chemical engineer Sal Baxamusa described the resulting measurements as useful both for selecting experimental targets and for understanding upstream manufacturing quality. Every capsule entering the production line could now be measured, rather than only a carefully chosen subset. All five experiments that had achieved ignition by the September 2024 account used capsules characterised by 4Pi, either before or after automation. That historical count belongs to the announcement’s date.

Measurements during the implosion

NIF employed more than 100 diagnostics by September 2024. These measurements tested the experimental designs and the physical models embedded in simulations. A core set followed four properties:

• Implosion shape and symmetry.
• The fuel’s resistance to compression, called its adiabat.
• Implosion velocity.
• The amount of capsule material mixing into the fuel.

The two-dimensional Convergent Ablator platform illuminated a backlighter foil while the capsule imploded. A time-gated pinhole camera recorded the resulting radiographs of the moving capsule. Such measurements supplied information missing from earlier experiments. Diagnostics also revealed unexpected effects, including ice forming on a laser-entrance-hole window and significant motion of the hot spot. General Atomics physicist Joe Kilkenny described experiments as a means of testing and improving theory. The National Diagnostic Working Group coordinated contributions across institutions, while LLNL’s Andrew MacKinnon led its national diagnostics programme. The measurements supported repeated comparison between calculated behaviour and the conditions actually reached in a shot.

Simulations that can be checked against experiments

Two long-running computational tools support the experimental programme. Lasnex development began in 1970 under physicist George Zimmerman. It brought together models of laser absorption, electron conduction, hydrodynamics and fusion burn in two dimensions. Its Basis interface allowed researchers to vary modelling specifications and test sensitivity to different processes. HYDRA development began in 1994 under Robert Tipton, with Marty Marinak subsequently helping develop its inertial-confinement capabilities. HYDRA supported massively parallel two- and three-dimensional simulations of both capsule and hohlraum physics.

The models could generate diagnostic predictions for comparison with measurements. Improved computing and experimental information helped resolve the effects of the membrane, or tent, holding a capsule in place. A change in its orientation improved capsule behaviour. Lasnex and HYDRA also predicted hohlraum reheating: after ignition, the expanding capsule encountered surrounding plasma, creating additional radiation that heated the hohlraum. LLNL physicist Mordy Rosen identified agreement between those predictions and experiments. The simulation tools continued to evolve as later measurements exposed additional details of the implosion and laser–plasma interactions.

A smoother transition inside the diamond shell

A later capsule innovation addressed the distribution of tungsten within the diamond. Hard X-rays can penetrate a capsule and preheat its fuel. Tungsten helps absorb them, but abrupt transitions between differently doped layers can introduce instabilities. Continuous gradient doping instead raised the tungsten concentration gradually, reaching approximately 0.44 atomic percent over 10 micrometres of shell thickness, before reducing it smoothly towards the outside. LLNL engineer Sean Hayes described this gradual profile in April 2026.

The first full-scale test on 23 February 2025 produced 5 megajoules of fusion energy. A second ignition experiment using this target approach, on 7 April 2025, delivered 2.08 megajoules of laser energy and produced 8.6 megajoules, a target gain greater than four. LLNL physicist Dan Clark attributed the benefit principally to reduced mixing of capsule material into the fuel. The later technical explanation connected the result to improvements in capsule fabrication, supporting technologies and simulations, with General Atomics and Diamond Materials participating in production.

Opening windows for radiation experiments

Another experiment used ignition as a radiation source. A Los Alamos National Laboratory-led team working with LLNL achieved a yield of 2.4 ± 0.09 megajoules on 22 June 2025. Its Thinned Hohlraum Optimization for Radflow system, or THOR, modified the ignition hohlraum by adding diagnostic windows. These allowed some high-flux X-rays to escape. The design had to preserve sufficient energy and implosion symmetry for ignition despite those openings, a requirement central to the experiment’s purpose.

Los Alamos physicist Ryan Lester, the campaign lead, linked the result to validation of high-fidelity simulations with the modified platform. The planned next applications involved exposing test materials to the X-rays to examine radiation transport and absorption. Physicist Brian Haines identified further work on window transparency and experiments attached to the windows. Those were subsequent research steps described after the ignition result. The June shot established that the modified arrangement could ignite; the announcement did not claim that all those proposed materials experiments had already been completed.

Printed targets follow a different development path

Other NIF experiments explored polymer targets produced through two-photon polymerisation. This technique uses tightly focused infrared light to cure liquid resin with features as small as 100 nanometres. Printed capsules can combine a thin, leak-tight shell with an internal foam holding liquid fuel. The first such direct-drive experiments, in March and May 2024, used fuel vapour and suspended liquid fuel respectively. In direct drive, lasers strike the target directly, unlike the hohlraum-mediated indirect-drive arrangement used for NIF’s high-yield diamond capsules.

A leak-tight, liquid-filled capsule was achieved in November 2024, followed by a March 2025 experiment combining central deuterium vapour with a liquid-wetted foam layer. Researchers worked to reduce defects from stitching together printed sections and to control the capsule and fill-tube temperatures independently. Significant limitations remained: printed capsules did not yet match traditional target quality or the production speed required for a power plant. These tests explored alternative fabrication and target designs; they were not the diamond capsules responsible for the 8.6-megajoule result.

A research programme for repetitive operation

The Department of Energy’s inertial fusion energy programme addressed requirements beyond isolated ignition shots. On 7 December 2023, DOE announced three research hubs led by Colorado State University, the University of Rochester and LLNL. Their programme, Inertial Fusion Energy Science and Technology Accelerated Research, brought together laboratories, universities and companies. Planned work included high-gain targets, efficient lasers operating at high repetition rates, and manufacturing, tracking and engagement of fusion targets. Workforce development was another part of the programme.

The announcement specified projects lasting up to four years and a total of $42 million: $9 million in fiscal 2023 and $33 million in later funding contingent on congressional appropriations. Selection through competitive peer review still required award negotiations, and DOE retained the option to cancel them. The funding terms therefore accompanied a proposed research effort rather than a guarantee of completed facilities. The technical objective was to move from low-gain, single-shot experiments towards the high gain and repetitive operation envisaged for a potential inertial fusion pilot plant.

An upgrade proposal with defined limits

NIF’s Enhanced Yield Capability project entered conceptual planning after DOE’s September 2024 Critical Decision-0. LLNL’s July/August 2025 assessment described a proposal to increase laser energy from 2.2 to 2.6 megajoules, including additional laser glass in open power-amplifier ports. Fusion yields as high as 30 megajoules were a potential outcome, rather than an achieved result. The project still required further coordination and decisions before moving from planning to implementation. Existing-function sustainment work formed a separate programme.

Early design studies considered approximately 500–600 terawatts and 2.6–3 megajoules. The constrained design scope instead used 450 terawatts and 2.6 megajoules because of uncertainty about laser damage thresholds. Kritcher described options for coupling additional energy into the implosion while retaining similar power limits, including thicker ablators and changes to pulse duration. These were avenues for investigation. The capacity assessment separated a proposed enhancement to the existing facility from longer-term ideas for a new facility with greater capability.

A separate magnetic-confinement result

The Joint European Torus in the United Kingdom provided a different experimental reference. On 8 February 2024, UKAEA announced that JET’s final deuterium–tritium campaign had produced 69 megajoules over five seconds using 0.2 milligrams of fuel. JET was a tokamak, using magnetic fields to confine a doughnut-shaped plasma. More than 300 scientists and engineers contributed to the experiments. Scientific operations ended in December 2023, before the record announcement, with repurposing and decommissioning forming the next phase.

CEA physicist Emmanuel Joffrin, leading EUROfusion’s tokamak exploitation task force, identified tests of reduced heat flow to the exhaust and a stable plasma edge that prevented energy bursts reaching the wall. These operating scenarios were tested in a deuterium–tritium environment to inform future machines. JET’s sustained magnetic-confinement experiments and NIF’s brief laser-driven implosions used different arrangements and durations. The 69-megajoule total described energy released in the JET experiment; it was not a published target-gain ratio comparable to NIF’s, nor a measure of electricity supplied by a power station.