We’ve been trying for a long time to make a tiny Sun on Earth, one that would sustainably produce energy by nuclear fusion of hydrogen or similar atoms. Come to think of it, I’d like one for my basement.
Fusion requires quite a bit of heat to get going, but once it does, it starts producing its own heat. If you can keep that system contained so it doesn’t expand too much or allow too much heat to escape, further fusion happens. If it reaches a point where self-heating becomes the primary driver of fusion, you have yourself a “burning plasma”.
The actual Sun has a pretty easy time sustaining fusion because of the crushing gravity at its center, but we Earthlings need to be a bit more creative to achieve that here at home (because we don’t have any 4-nonillion pound weights handy). The burning plasma, indeed a tiny star, is one of the key milestones on the path to usable nuclear fusion. Until now, no one had ever made such a thing.
Then, on January 26, in the journal Nature, via the National Ignition Facility (NIF) at Lawrence Livermore National Laboratory:
After decades of fusion research, here we achieve a burning-plasma state in the laboratory.
Let me just channel a reaction on behalf of nuclear fusion scientists and engineers everywhere:
A slightly more-complete summary:
“In these experiments we achieved, for the first time in any fusion research facility, a burning plasma state where more fusion energy is emitted from the fuel than was required to initiate the fusion reactions, or the amount of work done on the fuel,” said lead author Annie Kritcher.
We aren’t quite going to be dining by nuclear-fusion candlelight this Valentine’s Day, because there are several more steps along the way. But given that the one thing nuclear fusion has sustainably generated over the years is a lot of oh-you’ll-nevers, it’s very satisfying to see one of these key steps reached.
I’ve written a couple of times in the recent past about nuclear fusion, and all the excitement around our first realistic shot at it in 2025, so I won’t try to go back over all of the fundamentals here. But what I wondered primarily about this result was: Why had no one achieved the burning-plasma state before? What was different this time?
We can’t let a fusion plasma expand too much because it will get diluted and the fusion will stop. So we have to confine the plasma somehow. There are two main ways to do that: magnetic fields and inertia.
The magnetic-field approach is used in the giant donut-shaped reactors called tokamaks, where the plasma goes around and around without touching the walls. That requires powerful magnets and very good superconductors, and it’s delicate and finicky. I liken it to turning on a hose and trying to get the stream of water to go in a nice, tight circle around your yard by blowing on it with thousands of strategically placed fans. Whoa, that would take decades to figure out! Yes. Yes, it would.
But you can also use plain-old inertia to contain a plasma you make (“inertial confinement fusion”, or ICF), and that’s what was done in this work.
To use inertia, you only run fusion on a blob of compressed plasma long enough for it not to start expanding. You can do it in a small space, so it’s very manageable. For ICF to be practical on a large scale, you’d have to use a lot of little blobs of plasma, or pellets, and indeed that idea has been around for some time.
In this study, they used only one pellet at a time, and they wanted to demonstrate that the primary source of heating for fusion was the fusion itself, not energy they were adding, even if only briefly.
So how do we actually get fusion to happen in our pellet? We simulate the intense gravity of the Sun by making our pellet implode … with lasers!
The really important thing about this is symmetry. We generally don’t want to just shine lasers right at the pellet, because the heating would be uneven. So we have to be a bit craftier. We put the pellet inside a little capsule called a “hohlraum” (German for “cavity”), and the lasers are distributed so that they strike the inside of the hohlraum in a uniform arrangement.
The inner wall of the hohlraum, made of a heavy metal like gold or uranium, absorbs the laser energy and re-emits much of it as X-rays. Heavy metals like these have nice, big gaps between some of their electron energy levels, so when their electrons get excited by the lasers, then fall back to their initial states, they can emit powerful X-rays. The nice thing about this setup is that the X-rays get emitted much more uniformly onto the pellet than if we’d tried to just nail the pellet with lasers or X-rays directly. Let’s look at our pellet:
The reason we like X-rays here is that they are ionizing radiation, meaning they’re energetic enough to knock electrons right off of their atoms. Those X-rays are going to hit the pellet from all sides, and the surface of the pellet, made of a material called an “ablator”, is going to get ionized. So free electrons are going to fly around within the ablator and hit other atoms, and that’s going to cause the ablator to heat up and rapidly expand.
If (negatively charged) electrons get knocked out of the ablator material completely, you can also have a Coulomb explosion, where positively charged nuclei are left sitting near each other, and boy, do they hate that.
A couple of common ablator materials are high-density carbon and beryllium spiked with copper. We like those because they have a pretty high electron density and yet aren’t super-dense materials. Under X-ray radiation, that’s a good combination for an explosion!
So the ablator material blasts outward like a rocket, and the force of that blast presses hard on the deuterium-tritium on the inside of the pellet, compressing it with an acceleration about 10 trillion times that of the Earth’s gravity. The implosion collapses inward at speeds reaching 800,000 miles per hour within nanoseconds, and the interior of the pellet reaches about 20,000,000°F. The deuterium and tritium atoms within the pellet turn into a plasma — that is, their electrons are blown off and move around freely.
The bare nuclei are thus slamming into each other really hard, and that’s exactly what they need to do in order to fuse together. Protons are positively charged, so they normally want to repel each other, but if they are shoved to within a quadrillionth of a meter of each other, a much more powerful force called the strong nuclear force takes over, and then, strangely enough, protons become very attracted to each other. That’s how a nucleus with multiple protons stays together, and it’s also how two atomic nuclei can fuse together if they fly at each other fast enough.
When deuterium and tritium nuclei fuse, you get a helium nucleus (two protons, two neutrons) plus an extra neutron:
All the little balls add up, but what doesn’t add up is the mass. When two nuclei fuse, the resulting mass is a little less than the sum of the parts. That lost mass m has to go somewhere, and it’s related to energy E by ... E = mc2. (What, you were expecting something else?) The speed of light c is a huge number, so even a little bit of m means a whole lot of E. That’s the potential of fusion.
So when fusion starts happening, suddenly you have brand-new helium nuclei and neutrons flying around in the compressed pellet, carrying that extra E as kinetic energy (you know, speed). The helium nuclei slam into free electrons in the plasma, cranking up their momentum, and they in turn slam into deuterium and tritium nuclei to speed them up and hence keep fusion going. (The neutrons, for their part, don’t contribute much to this, but they can still be used for other things, like making tritium from lithium.)
If your pellet implosion is very uniform (so nothing goes squirting out the sides), you compress the core so well that the energy produced by fusion is greater than the kinetic energy of the compression. And there you have a burning plasma!
From November 2020 to February 2021, experiments at Lawrence Livermore National Laboratory achieved this for the first time, and although the pellet and hohlraum are on the millimeter scale, the overall apparatus most certainly is not:
People have been doing this sort of thing for a while, so why the burning-plasma state now? There were a couple of key factors:
1) The improved hohlraum design allowed more of the laser energy to be delivered to the target, so the same amount of laser energy could be used to compress a larger pellet. That means more compressed material and more opportunities for helium nuclei to transfer their energy before escaping.
2) The symmetry of the X-ray emission off the hohlraum walls was improved by strategically varying the energy input of the laser beams rather than keeping them all constant. That allowed a more symmetrical implosion of the pellet.
So how do we test a pellet implosion to see whether the fusion inside it was primarily sustained by self-heating? How do we know we have a tiny star, short of drying some microscopic salmon fillets with it?
We know how much laser energy we put in, and we know that 92-95% of it is absorbed and directed to ablation, so we know how much energy went in. We need to compare that to the energy we got from the fusion. This kind of experiment has been simulated and carried out enough times that the relationship between the temperature reached and the reacton rate is pretty well-known, so we have accessible ways to measure energy out.
We want to see that energy out/energy in ratio get above 1 so that we know we reached a burning-plasma state.
Below are selected experiments leading up to now, and even with uncertainties in measurements, it’s clear the last few have the Eα/KEfuel ratio — which is alpha heating (heat provided by helium nuclei) divided by kinetic energy derived from the fuel; that is, implosion energy — of well above 1. The authors also measured this in more-rigorous ways, but this one seems most understandable:
Those burning-plasma states we see in the last four experiments only lasted for about 100 picoseconds (100 trillionths of a second), because after that expansion from the reaction dilutes things down and the fusion fizzles out.
But despite their fleeting nature, they are the first documented mini-Suns we’ve ever made on Earth.
The next step, which the authors contend we are very close to, is ignition, where the reaction keeps going even after the external energy source is taken away completely. In fact, in August of last year, the results of this study led to record energy yields that were just short of ignition. Those results will be published soon, but you can still see Livermore Lab’s summary of that here. Briefly:
The [August, 2021] experiment was enabled by focusing laser light from NIF — the size of three football fields — onto a target the size of a BB that produces a hot-spot the diameter of a human hair, generating more than 10 quadrillion watts of fusion power for 100 trillionths of a second.
I realize this will be a dumb and totally unfair comparison, but I’ll make it anyway: If you could sustain 10 quadrillion watts for just a minute and a half, that would be enough energy to power the entire United States for a year.
It’s entirely possible that an experiment has been done in the meantime that has indeed achieved ignition, and we just don’t know it yet. As always, stay tuned…...
Now, whatever you may think of the prospects of practical power production presently presented (or my injudicious use of alliteration), I for one am happy to bask for a bit in the fact that we have managed to make the first mini-star on Earth, just for a flash. With all the nasty crap constantly going on around us, you bet I’ll take it.
Little game of “Where’s Waldo?”: See if you can spot Morris Day in the audience