Published online 10 January 2006 | Nature | doi:10.1038/news060109-5


Desktop fusion is back on the table

Physicist claims to have definitive data, but can they be replicated?

Imploding bubbles, caught on film emitting light. Are they emitting energy from fusion too?Imploding bubbles, caught on film emitting light. Are they emitting energy from fusion too?© D. Flannigan and K.S. Suslick, University of Illinois at Urbana-Champaign

Can the popping of tiny bubbles trigger nuclear fusion, a potential source of almost unlimited energy? This controversial idea is back on the table, because its main proponent has new results that, he claims, will silence critics. But others say that the latest experiment simply comes with its own set of problems.

The idea is simple enough. Blast a liquid with waves of ultrasound and tiny bubbles of gas are created, which release a burst of heat and light when they implode. The core of the bubble reaches 15,000 °C, hot enough to wrench molecules apart. Physicists have even suggested that the intense conditions of this sonoluminescence could fuse atomic nuclei together, in the same process that keeps our Sun running.

Physicist Rusi Taleyarkhan of Purdue University in West Lafayette, Indiana, published the first evidence1 of this 'sonofusion' in 2002; he has been dogged by sceptics ever since.

The underlying physics behind the idea is valid, says Ken Suslick. An expert in sonoluminescence at the University of Illinois in Urbana-Champaign, Suslick tried and failed to replicate Taleyarkhan's first results. If the bubbles' collapse is sufficiently intense, it should indeed be able to crush atoms together. Taleyarkhan just hasn't done enough to prove it, says Suslick.

Needle in a haystack

Taleyarkhan's first experiments were conducted while he was based at Oak Ridge National Laboratory in Tennessee. His idea was to use liquid acetone in which hydrogen atoms had been replaced by their heavier brethren, deuterium. When deuterium nuclei fuse together, they emit a characteristic burst of neutrons. But critics pointed out that Taleyarkhan was using an external source of neutrons to 'seed' the bubbles, and that these were swamping his measurements of neutrons produced by the fusion reaction itself.

"This time round there are no external neutrons," he explains. Instead, his team loaded a mixture of deuterated acetone and benzene with a uranium salt. As the uranium undergoes radioactive decay it releases alpha particles, which can also seed bubble formation, says Taleyarkhan.

"In this experiment we use three independent neutron detectors and a gamma-ray detector," he adds. The results from the four instruments prove that fusion is happening inside his experiment, asserts Taleyarkhan.

Although uranium can release neutrons during fission reactions, Taleyarkhan rules them out because the neutrons he finds bear the energetic hallmark of having come from the fusion of two deuterium nuclei2.

Taleyarkhan's test reactor still puts out a lot less energy than it takes in, making it impractical for generating power. "We have a way to go before we break even," he admits. But in the meantime, he adds, it could be a cheap source of neutrons for analysing the structure of materials. The results are to be published in Physical Review Letters in a few weeks' time.

Unreliable sources

There is one big problem, however: the experiment doesn't always work, and the group is not sure why. Seth Putterman, a physicist at the University of California, Los Angeles, who has also tried to verify some of Taleyarkhan's experiments, notes that the paper does not reveal how many failed runs were required before the team saw a trace of fusion neutrons. "As a paper it doesn't convince me," says Putterman.

Putterman notes that the team did not continuously monitor background neutron levels. Although the neutron count doubles at some points in the experiments, Putterman says that neutrons produced in random showers of cosmic rays, rather than fusion events, could be responsible. But Taleyarkhan points out that the neutron count was smaller in detectors further from the reaction chamber.

To prove that the neutrons are coming from fusion as bubbles burst, Putterman and Suslick suggest that the team closely monitor exactly when the neutrons appear. The current experiment simply counts up the number of neutrons detected over minutes, so correlations with bubble bursts cannot be seen. "The key to improving the signal is timing," says Putterman.

Finding proof

Another obvious way to confirm that fusion is happening would be to look for tritium, a heavier isotope of hydrogen produced by fusion reactions. Tritium leaves a telltale signature of high-energy electrons when it decays and Taleyarkhan claimed to see this in similar previous experiments1,3. But in the current tests, tritium's signature is overwhelmed by ?-decay from the uranium, making it impossible to spot.


Given that Suslick and Putterman have both investigated Taleyarkhan's past claims, they think it odd that they were not consulted by the editors of Physical Review Letters about the paper. "There are other people who are very knowledgeable about this," comments Martin Blume, editor-in-chief of the American Physical Society.

Taleyarkhan says that Suslick and Putterman are welcome to visit his lab to see the results for themselves. Both are eager to go as soon as possible. "We look forward to seeing the experiment run," says Putterman. 

  • References

    1. Taleyarkhan R.P., et al. Science, 295. 1868 - 1873 (2002).
    2. Taleyarkhan R.P., et al. Phys. Rev. Lett. in the press, (2006).
    3. Taleyarkhan R.P., et al. Phys. Rev. E, 69. 036109 (2004).