Harnessing bubbles to trigger nuclear fusion
22 January 2005
From New Scientist Print Edition. Subscribe and get 4 free issues.
Thermonuclear fusion in a bottleTHREE years ago, the publication of a single paper turned the world of fusion research upside down. The paper made the jaw-dropping claim that thermonuclear fusion had been observed in a beaker of nail polish remover bombarded with neutrons and sound waves. The newly discovered phenomenon, "sonofusion", seemed to overturn a generation of work on fusion and promised almost limitless cheap energy - a claim that guaranteed the research worldwide media coverage.
Fusion researchers were stunned, but on closer inspection they realised something was wrong. Surprise soon gave way to disbelief and then anger. The paper, cried the physicists, was severely flawed. Sonofusion should be confined to the realms of junk science, they said. And there the story might have ended.
But something remarkable has emerged from the debris. The researchers behind the original claim have slowly begun to convert their critics with new, more persuasive evidence. Could it be that sonofusion - or bubble fusion as it is also known - is real after all? The possibility has guaranteed funding from one government organisation that fears sonofusion may have more sinister uses, and also from venture capitalists keen to cash in. Now sonofusion is about to hit the headlines again as rumours of other successes begin to leak out.
Unlike claims for cold fusion, the idea of bubble fusion is not left-field. The phenomenon at its heart is sonoluminescence, the mysterious flash of light emitted when bubbles are blasted with powerful sound waves. The waves produce regions of high and low pressure in the liquid, and bubbles form in the regions of low pressure, where the liquid simply boils. Then, as the bubbles experience intense pressure, they collapse and emit light - though no one fully understands why.
For many years the eerie glow was little more than a curiosity because the bubbles were so difficult to control. But that all changed in 1992 when a team led by Felipe Gaitan at the Naval Postgraduate School in Monterey, California, discovered how to trap a bubble in a stationary or "standing" wave and vary the pressure so that it would expand and collapse in a controllable way, each time emitting a pulse of light. For the first time it became possible to study sonoluminescence in detail.
Gaitan's team found that the bubbles produce a range of wavelengths of light, including ultraviolet. To create such energetic light, they reasoned, the temperature within the bubbles must exceed 10,000 kelvin, perhaps reaching as high as 1 million kelvin.
Such staggering temperatures raise the possibility of an even more exotic phenomenon - nuclear fusion. At temperatures of 10 million kelvin or so, hydrogen nuclei begin to fuse, releasing huge amounts of energy. To achieve this kind of temperature in a bubble requires the sound energy to be focused by a factor of 1015, a huge amount for sure, but only two or three orders of magnitude greater than is already possible. Could bubbles somehow be coaxed into working a little harder?
Toil and trouble
Seth Putterman of the University of California in Los Angeles, a leading sonoluminescence researcher, has little doubt that the phenomenon will one day be harnessed to trigger fusion. What he and others question is whether it already has been.
The debate erupted in 2002 with the publication of a bubble fusion claim in the prestigious journal Science (vol 295, p 1868). Rusi Taleyarkhan, a nuclear engineer then at Oak Ridge National Laboratory in Tennessee, and his colleagues reported what happened when they used neutrons to seed the bubbles (see "Neutron bubbles"). They fired the neutrons into a beaker of acetone - a solvent commonly used as nail polish remover - in which all the hydrogen atoms had been replaced with deuterium (a heavier form of hydrogen with a nucleus comprising a proton and a neutron). The bombarding neutrons created bubbles that expanded and collapsed when blasted with sound waves.
Unlike others who had tried, Taleyarkhan and his colleagues found evidence for the creation of neutrons and tritium, a radioactive form of heavy hydrogen. Alone, these by-products are a fair indicator that some kind of nuclear process may be taking place. But when found together, they are powerful evidence that fusion has occurred. Deuterium nuclei appeared to be smashing together and fusing to create tritium and free neutrons.
At this point, in 2000, the research was being kept under wraps at Oak Ridge while the work was reviewed. "They gave me one heck of a hard time," says Taleyarkhan. Managers at the lab even asked a different pair of experimenters, Mike Saltmarsh and Dan Shapira, to conduct the same experiment with much of the same equipment. "More than 50 reviewers had a crack at us," he says. The reviewers posed numerous questions and asked for loose ends to be tied up before publication.
Despite this attention to detail, it was clear after the paper appeared that all was not well. Taleyarkhan had included a reference to the repeat experiment at Oak Ridge. To many physicists' astonishment, Saltmarsh and Shapira had found no evidence of fusion even though the experiment was almost identical. "Taleyarkhan's measurements were null and void," says Shapira.
Then holes began to appear in Taleyarkhan's method. Much of the criticism focused on the way he detected neutrons. According to nuclear theory, the neutrons produced by fusion have five times less energy than the neutrons used to trigger the formation of bubbles, so in principle the two types should be easy to distinguish. "But neutrons are pesky things," says Shapira. They have a habit of hanging around, rattling around in the experiment and turning up fractions of a second later with a lower energy. "You have to be extremely careful."
What Taleyarkhan did was inject the neutrons and then wait a fraction of a second for the arrival of what he believed were the fusion by-products. What many others think is that Taleyarkhan was simply picking up reflected neutrons that he had released moments before.
Other criticisms emerged, among them a joint paper by Putterman, Larry Crum at the University of Washington in Seattle and Ken Suslick at the University of Illinois in Urbana-Champaign (www.arxiv.org/abs/cond-mat/0204065
). Their verdict was damning.
Not only had Taleyarkhan failed to prove that the neutrons were generated by fusion, they said, but the evidence that tritium had been produced was also unconvincing. Taleyarkhan found the tritium in swabs taken around the lab. As far as Putterman, Crum and Suslick were concerned, this was not evidence of tritium production but of contamination in the lab. How it got there should have been discussed in Taleyarkhan's paper, but was not.
Finally, the group said that the neutron burst itself was highly suspicious because it was so strong. Putterman says the figures equate to 10,000 neutrons being created in each bubble collapse. "That's a mind-bogglingly huge signal," he says. The mystery is how a bubble seeded by a single neutron can generate so many. "If you had one neutron coming out for each neutron that went in, that would be interesting. Taleyarkhan has 10,000," says Putterman.
Bit by bit, Taleyarkhan's claim was falling apart. His team didn't take all this criticism lying down. He re-analysed the Saltmarsh and Shapira results, claiming they show exactly what you'd expect if fusion were taking place.
But the criticisms stuck. By the beginning of 2004, few people working in the field were taking Taleyarkhan seriously. "Everyone who sees the Science paper says it is a travesty," says Shapira. Others agree. "It was severely flawed," says Putterman.
The ferocity of the criticism took Taleyarkhan's team aback. Dick Lahey, an engineer at Rensselaer Polytechnic Institute in Troy, New York, and a member of the team, believes that some of the mud-slinging was politically motivated. At the time, the US Congress was deciding whether to fund the multibillion-dollar ITERproject, a reactor to test the feasibility of generating fusion power by heating hydrogen isotopes to huge temperatures while they are confined in a magnetic field. The funding of many fusion scientists depended on this project. "People were worried that congress would fall in love with bubble fusion and cancel ITER," says Lahey. As it turned out, Congress backed ITER, although the project is now stalled amid international bickering over where to site it (New Scientist, 4 December 2004, p 3).
The next step was clear to Taleyarkhan's team. "Some of the criticisms were valid," Lahey concedes. "We decided to address them systematically." The result of this second experiment appeared in the peer-reviewed journal Physical Review E in March last year (vol 69, p 036109). It has fundamentally changed the way people think about bubble fusion.
The new experiment was different in a number of crucial ways. Instead of looking for fusion neutrons shortly after firing the initial burst, the group monitored their arrival continuously throughout the experiment. The results show that the initial burst used to seed the bubbles gradually dies down. Then, after a short time, a couple of peaks appear. These peaks, claims Taleyarkhan, are the bursts of neutrons generated in fusion reactions. "It's pretty solid evidence," says Michel Laberge, a physicist at General Fusion, a start-up in Vancouver, Canada, who also works on fusion research.
But the real beauty of the work is the control experiment that Taleyarkhan used to validate the results. He changed the timing of the initial neutron pulse, firing it when the pressure in the liquid was at its highest. Under high pressure, a bubble cannot form and so fusion cannot take place. Sure enough, the peaks that Taleyarkhan says are proof of fusion do not appear and so cannot be caused by neutrons from the initial burst. "It's very compelling data," says Putterman.
Another author of the paper that tore Taleyarkhan's original claims to pieces, Crum, has also been won over by the work. "I've changed my view," he says.
What caught Crum's eye is the apparent delay between the creation of the bubble and the subsequent bursts of neutrons. Crum thinks that sound waves first break up the original bubble to form a cloud of bubbles. When this bubble cloud collapses, the outer bubbles send an additional shock wave towards the centre of the cluster, generating the higher temperatures needed for fusion. "Bubble cluster formation is much more likely to lead to fusion," he says.
The most recent work on bubble fusion seems to confirm the idea that a cloud is involved - but not in the way Crum suspects. Lahey says that his analyses and simulations indicate that the initial burst of neutrons does not create a single bubble that breaks up, but an entire cloud of bubbles in one go. "The optimum number of bubbles in the cloud is around a thousand," says Lahey. This explains why neutrons are so important and why other methods of initiating the bubbles with laser triggering and bubble injection do not work. Only a burst of neutrons can create the required cloud - other methods simply inject a single bubble into the experiment.
It could also explain why the experiment produces such a strong signal. Lahey and his colleagues have calculated that each bubble produces a handful of neutrons, as might be expected, but the fact that a thousand bubbles take part explains the magnitude of the signal.
The new results have caused a sea change in the perception of bubble fusion. In Grass Valley, California, a start-up company called Impulse Devices is looking for ways to commercialise the effect, bankrolled by $3.5 million of venture capital funding. Its chief scientist is Gaitan, a founder of the field.
If the reaction can be made self-sustaining, then sonofusion could be a rich source of cheap neutrons for analysing the structure of materials. Beyond that, if the device can be scaled up, the big hope is that sonofusion will generate large amounts of cheap power. "It's too early to make predictions about how far away energy from sonofusion really is," says Ross Tessien, president of Impulse Devices.
In the background, watching quietly while the public slanging matches rage back and forth, is the US Department of Defense's research agency, DARPA, which funds high-risk projects with big potential pay-offs. DARPA is the main reason the debate has moved so quickly. It has financed Taleyarkhan's work since 1998 but also funds many of the other researchers in the field, including Putterman, Crum and Suslick.
Just why is DARPA so interested? The obvious answer is the potential for unlimited energy. But there is another reason. In theory, sonofusion could generate large amounts of radioactive tritium, which is highly dangerous inside the human body. Terrorists could use it to spike a water supply, for example. If sonofusion can be created with technology easily available to a rogue nation, DARPA needs to know, says Crum. "It has an obligation not to be surprised by technological change."
Now DARPA wants to organise an experiment to end the bickering. The idea is for Taleyarkhan to collaborate with his critics to recreate the original experiment in Putterman's lab later this year. By forcing the collaboration, DARPA hopes that everyone will agree with the results, whatever they show. "People have tried to reproduce the experiment but none in exactly the same way. There's always a get-out somewhere," says Crum. Shapira agrees: "The experiment will end the debate."
It is too early to tell whether the upcoming experiment will back Taleyarkhan's ideas, but tantalising evidence that he could be right is already emerging from other labs. Laberge believes he has evidence for fusion in the form of neutrons.
And at Purdue University in West Lafayette, Indiana - which now employs Taleyarkhan - an independent group led by Lefteri Tsoukalas has recreated the experiment. Although reluctant to talk about the results, Tsoukalas plans to present evidence for tritium at a conference in France in October.
His evidence is by no means conclusive but it is potentially revolutionary. Later this year, the world of fusion research could be turned upside down again and this time the outcome could be very different.