I feel as though I have been here before. In March 1989, a press conference held by the University of Utah announced that two scientists there had discovered a way to achieve nuclear fusion, the same source of energy that allows the sun and the stars to shine for billions of years. They had done so with a beaker-sized apparatus that operated at room temperature.

You may remember the sad end of that story. No one else was able to reproduce the results. No one else could find the telltale neutrons that would confirm the fusion process. Two scientists, who until that time had enjoyed distinguished careers, were discredited and slowly faded from the academic scene. It is easy to say that they were the big losers, but in fact we all were. Few things could be more useful to humanity than a cheap, small, safe and endless source of energy.

Now it is happening again. The same small tabletop apparatus. The same claim for nuclear fusion. The same immediate controversy. The players, however, are different, as is the physical process involved.

Last time around, fusion supposedly occurred because of an odd property of the element known as palladium. Palladium is a steely-white metal with a natural affinity for hydrogen, an affinity so strong that palladium can absorb nine hundred times its own volume of hydrogen. It can just as well absorb an equal amount of "heavy" hydrogen, also known as deuterium, which is hydrogen with an extra neutron in its nucleus. In ordinary water, about one part in six thousand is "heavy water," containing deuterium rather than hydrogen. Heavy water is readily separated from ordinary water.

Now for the process. A palladium rod, placed in a container of heavy water, supposedly absorbed so much heavy hydrogen that some of the nuclei fused together. The result should have been helium, plus heat, plus neutrons. Unfortunately, far too neutrons were observed to support the claim of fusion.

This time around the research workers are at the Oak Ridge National Lab, one of thirteen major facilities operated by the U.S. Department of Energy. The phenomenon that squeezes nuclei so close together that some of them fuse is an interesting but well-known one called sonoluminescence. As the name suggests, sonoluminescence can produce light using sound waves.

A sound wave is a pressure wave. If you send intense sound waves through a liquid, you can produce tiny regions of compression and rarefaction where the pressure changes are enormous. Bubbles form at very low pressure, then implode under very high pressure. This generates intense heat, and hence light. The combination of very high temperature and very high pressure is exactly what you need to induce nuclear fusion.

Now for the process. A container is filled with acetone, better known to most of us as nail-polish remover. This, however, is "heavy" acetone, in which the hydrogen has been replaced by deuterium. A sound wave generator on the side of the container generates intense pressure waves within. As a result of these pressure waves, tiny bubbles constantly form and then just as quickly vanish. However - and here comes the controversial step - to encourage rapid bubble expansion and create "super-bubbles" as much as a millimeter across, neutrons from an external source are fired into the acetone. As the biggest bubbles collapse, the heat and pressure in these regions causes fusion of some of the heavy hydrogen nuclei.

If fusion really is occurring, the byproducts should include tritium, which is "super-heavy" hydrogen, hydrogen with two neutrons in its nucleus. Other byproducts would be helium, neutrons, and protons (hydrogen nuclei). The most easily detected and significant markers to indicate that fusion is occurring will be tritium and neutrons. The Oak Ridge group, led by a scientist named Rusi Taleyarkhan with collaborators from Rensselaer Polytechnic and the Russian Academy of Sciences, claims to have detected both tritium and neutrons.

A second group of scientists - shades of the Utah cold fusion experience - looked for neutrons and did not find them. Physicists elsewhere suggested that neutrons would inevitably be found, but only because neutrons had been pumped in. Other critics suggested that the tritium detected came not from fusion, but from contamination by tritium already in the lab. Although you won't find tritium around your home, the contamination idea is less unlikely than it may sound, precisely because the Engineering Science and Technology building at Oak Ridge contains all kinds of nuclear experiments, and is exactly the place one might expect to find almost any form of nuclear materials.

The obvious next steps are being taken. First, the experiments are being repeated at Oak Ridge in a different and better-equipped facility. Second, other labs in other parts of the country are preparing to repeat the experiments. Third, at least one lab will perform the experiment using a laser rather than neutrons to seed the bubbles, thus removing any possibility that the neutrons seen are no more than those provided. Fourth, theorists will tackle the tricky problem of deciding if the pressure and temperatures created are really high enough to permit fusion to occur.

If this sounds like a rerun of the 1989 experience, I want to point out some crucial differences. First, the Oak Ridge process does involve extreme temperatures and pressures, something which all our experience suggests will be necessary for successful fusion. This, if it holds up, will be fusion but it will not be "cold fusion."

Second, and more important, the University of Utah, probably believing that patents of huge value might be involved, provided in its press conference and subsequent releases only a vague description of the work performed. Others, seeking to repeat the experiments, had to guess at many details of the setup. By contrast, the group at Oak Ridge announced that it would help any others trying to replicate the results. Also, the original results were announced not at a press conference, but through the more traditional route of a refereed paper in an established science outlet (SCIENCE magazine). On the other hand, just to make matters more interesting, several of the scientists who reviewed the paper recommended against SCIENCE publishing it.

It is too early to say how all this will turn out. Maybe we really are running a repeat of the 1989 experience. But, oh, don't you wish that this time the results would turn out to be real?

"Borderlands of Science" is syndicated by: