As it approaches its fifth birthday, the National Ignition Facility (NIF), a troubled laser fusion facility in California, has finally produced some results that fusion scientists can get enthusiastic about. In a series of experiments late last year, NIF researchers managed to produce energy yields 10 times greater than produced before and to demonstrate the phenomenon of self-heating that will be crucial if fusion is to reach its ultimate goal of “ignition”—a self-sustaining burning reaction that produces more energy than it consumes.
“This is a very significant achievement, and it’s a very good place to start for going to higher yield,” says Steven Rose of the Centre for Inertial Fusion Studies at Imperial College London.
NIF, situated at Lawrence Livermore National Laboratory in California, aims to reproduce the energy source of the sun and of hydrogen bombs by fusing together nuclei from two isotopes of hydrogen: deuterium and tritium. It does this by heating them to enormous temperatures and pressures with the world’s highest energy laser, so that the nuclei smash together with enough force to overcome their natural mutual repulsion.
Following its completion in 2009, NIF researchers embarked on a 3-year campaign to achieve ignition as quickly as possible. But when that period ended, they were still a very long way from their goal . The U.S. Congress granted the lab another 3 years to carry out a more exploratory series of experiments and identify the problems.
The new results, published today in Nature  and last week in Physical Review Letters , are the first sign that this approach is working. “It’s a nice result,” says Robert McCrory, director of the Laboratory for Laser Energetics at the University of Rochester in New York, another laser fusion lab, quickly adding that NIF is still far from ignition. “People expecting a breakthrough soon will be disappointed,” he says.
To reach the extreme conditions necessary for fusion, some facilities, such as the ITER reactor in France, use powerful magnetic fields to constrain fuel and heat it with particle beams. NIF follows a different approach: blasting a tiny sample of fuel with a laser pulse to make a small fusion explosion. If everything works right, the explosion will have a higher energy than the laser pulse, offering a net energy gain. NIF’s laser, the size of a football stadium, produces 192 ultraviolet beams that can deliver 1.9 megajoules—roughly the kinetic energy of a 2-ton truck travelling at 160 kilometers per hour—in a pulse that lasts just nanoseconds.
The ultraviolet beams are converted into x-rays, which then fall on the fuel capsule, a hollow plastic sphere smaller than a peppercorn that contains 0.17 milligrams of frozen deuterium and tritium. The intense x-ray pulse hitting the capsule causes some of the plastic to blow off; this drives the remaining plastic and frozen fuel in toward the center at high speed. If all goes according to plan, the result is a tiny ball of fusion fuel at 50 million kelvin, 100 times the density of lead—hot enough and dense enough to spark fusion reactions.
NIF’s original plans to reach ignition relied heavily on simulations based on earlier work at Livermore and other labs. Once NIF scientists started firing their shots, the whole process seemed to work, and the simulations predicted NIF should be getting a lot of fusion. But the instruments told a different story: Energy yields were very low. In 2012, Congress ordered an investigation , which ultimately criticized NIF researchers for not being able to explain the divergence between simulation and experiment. In 2013, NIF researchers began to explore the problems more scientifically; there was also a change of leadership  at the lab and new researchers joined the team.
They identified two key problems. The compression of the fuel pellet was often not symmetric and produced a doughnut-shaped blob of fuel; and during the implosion, the plastic capsule was breaking up and mixing with the fuel, making it harder to spark fusion at the end.
To tackle the shape problem, the new team started playing around with the relative energies of the 192 laser beams to push a bit more in some places and a bit less in others, in hopes of getting a more symmetrical implosion.
To prevent the breakup of the capsule, the researchers adjusted the timing of the laser pulse. Earlier shots had run it at a low power for most of its 20 nanoseconds to get the implosion moving without heating up the fuel and then finish with a burst of high power for the final spark. The idea behind this “low foot” approach was that the cool fuel would compress to a higher density at the end. The downside was that the slower speed allowed the capsule time to break up. In low-foot shots, “there are too many deleterious things going on at once, you can’t see what’s going on,” says Stephen Obenschain, head of the laser plasma branch of the plasma physics division at the Naval Research Laboratory in Washington, D.C.
The new NIF team decided to try a pulse that started off with a slightly higher power, to cause the fuel to implode faster, and end the pulse sooner, after just 15 nanoseconds. Although such “high foot” pulses wouldn’t allow them to get as high a density at the end, the researchers hoped it would help control the mixing. A shot carried out on 13 August last year proved them right, with a huge jump in energy yield. Another two shots, on 27 September and 19 November, did even better, producing more energy (14.4 and 17.3 kilojoules) than was deposited in the fusion fuel during the implosion (11 and 9 kilojoules)—the first time that has ever been achieved in a laser fusion experiment. “We took a step back from what had been tried before and that gave us a leap forward,” NIF team leader Omar Hurricane told a press conference this week.
Importantly, the team also saw a self-heating phenomenon that will be vital for increased fusion yield. Fusion reactions produce alpha-particles (helium nuclei) as well as neutrons, and when reactions start in the core of the fuel, the alphas helped by heating the surrounding cooler fuel up to reaction temperature. The NIF team thinks that in their best shots, this alpha-heating doubled their fusion yield. “Alphas really do heat the gas,” Rose says.
Observers also note that in last year’s shots, there was closer agreement between simulations and experimental results. “Doing these less-demanding implosions, results now agree with the codes and that’s very encouraging,” says Michael Campbell, a former NIF director now at Sandia National Laboratories. “They can trust simulations now in a way they couldn’t before,” Rose says.
However, the recent shots are still far from what most fusion researchers consider to be real "gain": more fusion energy out than laser energy in. Although the shots produced more yield than energy into the fuel, much of the laser pulse’s energy is lost when it is converted from UV to x-rays and focused on the fuel capsule. Last year’s best shot produced less than 1% of the energy of the laser pulse.
Opinions are divided about what the NIF team should do now. McCrory does not believe the current approach will eventually lead to successful ignition, so more innovation is needed. “They’re pushing about as far as they can go,” he says. Rose agrees: “I’m not sure they have a route to real gain.” The problem is that the researchers dialed back the final pressure to control the mixing during the implosion; now they have to increase the pressure again to make the fuel dense enough for high yields without letting mixing creep back in. “Yes, we self-limited ourselves to gain this control,” Hurricane said at the press conference. “It’s a point of departure. Now we need to strike out in different directions.”
Despite the uncertainties, researchers are encouraged by the renewed progress at NIF. “These are the right experiments to do,” Campbell says. “Who knows how far they can take this?”
*Correction, 12 February, 5:06 p.m.: This item has been corrected to clarify that Stephen Obenschain is the head of the laser plasma branch of the plasma physics division at the Naval Research Laboratory.