Star in a jar? Hints of nuclear fusion found--maybe.

A group of scientists claims to have found evidence of nuclear fusion in a vase-size flask of liquid. The researchers say they created tiny bubbles that seemed to have collapsed with enough violence to force atomic nuclei to fuse.

Skepticism about the results outweighs confidence in them. Still, if the observations reported in the March 8 Science by Rusi P. Taleyarkhan of Oak Ridge (Tenn.) National Laboratory and his colleagues are confirmed, scientists will have a new way to study fusion reactions. A far more speculative and exciting possibility is that the tabletop experiments might lead to the long-sought goal of harnessing fusion for generating power.

"That's the ultimate goal--if it's possible to scale to that level," Taleyarkhan says.

The basis of the new energy source would be so-called sonoluminescence--a phenomenon in which bubbles of vapor in a liquid bombarded by sound waves rapidly implode, generating heat spikes and flashes of light in the bubbles (SN: 10/6/01, p. 213). Taleyarkhan and several of his Oak Ridge colleagues collaborated on the research with scientists from Rensselaer Polytechnic Institute in Troy, N.Y., and the Russian Academy of Sciences in Ufa.

Even if the experiments did yield fusion reactions, practical technology based on the phenomenon would be a long way off. However, many scientists have already pronounced the new findings dead wrong.

"They just don't have the evidence," says William C. Moss of Lawrence Livermore (Calif.) National Laboratory, one of several sonoluminescence specialists who have theorized that fusion in collapsing bubbles is feasible.

Other critics say that the most damning indictment of the new work is an unpublished follow-up experiment by a pair of nuclear physicists, also of the Oak Ridge lab.

Several detractors have compared the new Science report to the infamous "cold fusion" announcement made in 1989 (SN: 4/1/89, p. 196). Two electrochemists claimed then to have sparked fusion at room temperature by passing electric current through a bath of water in which ordinary hydrogen is replaced by deuterium, a heavier isotope. However, neither the original pair nor anyone else could reproduce those findings, which have since largely been discredited as a case study of mistaken science (SN: 6/22/91, p. 392).

On the other hand, scientists have produced tabletop fusion, for instance by zapping small clusters of atoms with high-powered lasers (SN: 3/27/99, p. 196).

In the new work, Taleyarkhan and his collaborators used bursts of neutrons to fabricate clouds of short-lived, but extraordinarily large, sonoluminescence bubbles in acetone, the solvent in many nail-polish removers. In some tests, the researchers filled the flask with ordinary acetone, whose molecules each contain six hydrogen atoms. In other tests, they used deuterated acetone, in which deuterium atoms replace the hydrogen ones.

Under extreme pressure and at temperatures of millions of degrees, such as at the center of the sun, deuterium atoms fuse in a reaction whose products include tritium--hydrogen's radioactive heavy isotope--and neutrons. However, sonoluminescence flashes typically occur at temperatures of thousands of degrees, not millions. "Such high temperatures are unlikely to occur" in the bubbles of the Oak Ridge setup, notes Lawrence A. Crum of the University of Washington in Seattle.

Using detectors, the Oak Ridge research team looked for surges of neutron emissions that correlated with the light flashes of collapsing bubbles. The researchers report that flasks containing deuterated acetone show significant excess of such events compared with background levels, but flasks of ordinary acetone didn't. Moreover, the team observed a buildup of tritium only in the deuterated acetone.

In the current setup, creating sonoluminescence takes far more energy than the bubble collapse gives off, even if fusion is taking place, Taleyarkhan says. However, he adds, mathematical models of the process suggest that much greater energy production may be possible. Of more immediate concern to him and his colleagues is the task of convincing other scientists that their evidence of fusion is sound.

Bristling at comparisons to the cold-fusion drama, the Oak Ridge researchers say that their findings withstood extensive peer review before being published. The cold-fusion claim in 1989 was announced to reporters before being submitted for publication.

The new signs of fusion in bubbles were so extraordinary that Lee L. Reidinger, the Oak Ridge lab's deputy director for science and technology, commissioned two of the lab's nuclear physicists, Dan Shapira and Michael J. Saltmarsh, to monitor the sonoluminescence setup using different detectors.

In a five-page internal document, Shapira and Saltmarsh report finding no correlation between neutrons with the energy expected from deuterium-deuterium fusion and light flashes from collapsing bubbles.

In response to that report, Taleyarkhan and his colleagues produced a 14-page rebuttal, in which they claim that their would-be spoilers made mistakes. When those errors are corrected, Taleyarkhan says, the Shapira-Saltmarsh findings actually agree with the original results.

Not so fast, Saltmarsh counters. "We stand by our data," he says.