Last February, about an hour's drive from San Francisco, a pinhead-sized sun of sorts flared into fleeting existence.
In less than a billionth of a second, it shrugged off 170 kilojoules — the equivalent energy of nine 9-volt batteries, or about half a chocolate-coated biscuit.
And while that doesn't seem like much, this came from 200 millionths of a gram of fuel.
Plus the miniature sun, and three others like it, reached a point where they could self-heat their own reactions.
This marks a milestone on the way to nuclear fusion power — and perhaps sustainable electricity generation — using lasers.
The experiments, conducted in late 2020 and early 2021 at the Lawrence Livermore National Laboratory in the US, are reported in the journal Nature today.
The study comes after the same group announced results from another mini-sun-type experiment in August last year.
That one flung out even more energy — nearly four chocolate biscuits' worth.
Physicist Alex Zylstra, study co-lead author, led all five experiments.
"The understanding that we gained based on these [published] experiments led to the one in August," Dr Zylstra said.
"We're working on a publication now on that experiment."
Omar Hurricane, who also co-led the study, said these experiments were all about energy density.
How to get out more than you put in
These nuclear fusion experiments, and others like them, aim to generate more energy than what's pumped into them.
Unlike nuclear fission, where energy comes from splitting heavy chemical elements such as uranium, nuclear fusion takes two light atoms such as hydrogen, merges them into a heavier element like helium, and kicks out a bit of energy in the process — no dangerous nuclear waste either.
We feel energy from nuclear fusion every day. It generates heat and light in stars including our Sun.
It can happen there because the immense gravitational pull in a star's core creates the hot, crushing circumstances needed for fusion reactions to take place.
But recreating those conditions on Earth is easier said than done.
One way is to use powerful lasers to implode a minute package of hydrogen fuel, which is exactly what the crew at the Lawrence Livermore National Laboratory's National Ignition Facility did.
The facility's 192 powerful lasers were trained onto a tiny golden cylinder.
Inside the cylinder was a hollow polished diamond sphere, just a millimetre across, which, in turn, contained fuel: two types of hydrogen called deuterium and tritium.
When the lasers fired and heated the inside of the cylinder, the diamond sphere heated too, and quickly expanded.
This rapidly compressed the fuel inside, which reached speeds of up to 400 kilometres per second.
And within 10 billionths of a second, the fuel was crushed to a fraction of its volume and its centre reached 50 million degrees Celsius.
"These implosions are extremely high pressure, hundreds of billions of atmospheres, and exceeding the pressures of the centre of the Sun," Dr Hurricane said.
Electrons were stripped from the deuterium and tritium atoms, creating a plasma soup of atomic nuclei and electrons.
When a deuterium nucleus (one proton and one neutron) merged with a tritium nucleus (one proton and two neutrons), they made a helium nucleus (two protons and two neutrons), a free neutron, and kicked out a little bit of energy as heat.
But what the National Ignition Facility team also did was create a "hotspot" of self-burning plasma — that is, a region that produced enough heat to propagate more fusion reactions.
The hotspot was small, though — only around 10 to 15 per cent of the fuel mass — and couldn't sustain ongoing reactions.
But the next step, Dr Zylstra says, is to get plasma to reach what's called ignition.
More reactions means more energy produced and, potentially, a sustainable source of electricity, Dr Zylstra said.
But getting to that point will require some more finessing.
Where's my nuclear fusion power plant?
Matthew Hole, a physicist and mathematician at the Australian National University who was not involved in the study, said the paper marked a milestone for this approach to nuclear fusion.
But, he added, laser-driven plasma burning is "extremely transient" and unlikely to be used in power plants.
"Those experiments, they might run one a week or something like that," Professor Hole said.
"But you need to go to 10 every second to turn that into a power plant."
The main focus of the National Ignition Facility is not clean energy production, but national security, he added.
"It's interesting physics, but it's not funded for energy purposes. It's funded for a different reason."
In an accompanying News and Views article, University of York physicist Nigel Woolsey wrote that such experiments will, "be key to national security, because the [National Ignition Facility] is funded as part of the US programme to improve understanding of nuclear weapons and extreme environments.
"It remains unclear whether this research will lead to a viable future power source."
The second main nuclear fusion technology being used around the world is being developed with sustainable energy in mind, Professor Hole says.
It uses powerful magnets to corral a hot, spinning mass of plasma in a special chamber.
This "magnetic confinement fusion" is also still in an experimental phase, and a handful of fusion facilities have been or are being built to refine the process and iron out any wrinkles.
But, he added, the joke that nuclear fusion is always 30 years away may be getting more serious.
ITER, a facility under construction in southern France, is likely to start operating in a few years and running actual experiments in the early 2030s, says Professor Hole, who is also an ITER science fellow.
"And China has a very ambitious project to develop an electricity pilot plant, and most of these designs promise power to the grid around 2040 or so."
And is this realistic?
"I think probably closer to the middle of the century is more realistic," Professor Hole said.
"But in the scheme of things, for such a massive energy-changing technology, that's fairly near term.
"So is it still within 30 years? Probably — it probably really is 30 years now."
