Sun in a Bottle Read online

  The bubble fusion story was about more than a scientific paper. It had become a story about the way science is done—and how the scientific peer-review process sometimes fails. The original ending to my piece grimly emphasized the point:

  Taleyarkhan and his team, in good faith, submitted their paper to peer review and passed. Science magazine subjected the paper to scrutiny in accordance with their procedures, and once the paper was accepted, refused to allow outsiders to influence their publication process. And outside scientists, worried about the quality of the paper, tried to prevent the embarrassment of a second cold fusion fiasco. All three parties had the best of intentions—and now the road ahead is paved.

  This ending was scrapped in favor of a less-editorializing one. Yet even as I penned those words, the misunderstanding built. Bob Park, the tireless critic of pseudoscience (particularly cold fusion and its proponents, such as Thomas Valone, the patent examiner), featured bubble fusion in his weekly What’s New newsletter, which is distributed on Friday afternoons:

  BUBBLE FUSION: A COLLECTIVE GROAN CAN BE HEARD

  A report out of Oak Ridge of d-d fusion events in collapsing bubbles formed by cavitation in deuterated acetone, is scheduled for publication in the March 8 issue of Science magazine. . . . Although distinguished physicists, fearing a repeat of the cold fusion fiasco 13 years ago, advised against publication, the editor has apparently chosen not only to publish the work, but to do so with unusual fanfare, involving even the cover of Science. Perhaps Science magazine covets the vast readership of Infinite Energy magazine.

  Infinite Energy, of course, was the cold-fusion activist Eugene Mallove’s publication. Park was uncharacteristically wrong about the cover; bubble fusion was never going to be a cover story. However, his comment meant that the hitherto private controversy was about to become very public.

  Over the weekend, the rumors flew, as did the press releases. Science distributed one as part of the weekly embargoed notification to journalists:

  FUSION IN A FLASH? SCIENCE RESEARCHERS REPORT NUCLEAR EMISSIONS FROM TINY, SUPER-HOT COLLAPSING BUBBLES

  The dramatic flashing implosion of tiny bubbles—in acetone containing deuterium atoms—produces tritium and nuclear emissions similar to emissions characteristic of nuclear fusion involving deuterium-deuterium reactions. This finding was reported in the 8 March issue of the peer-reviewed journal Science, published by the American Association for the Advancement of Science.

  Shock wave simulations also indicate that temperatures inside the collapsing bubbles may reach up to 10 million degrees Kelvin, as hot as the center of the sun. Although the high temperatures and pressures within the bubbles would be sufficient to generate fusion, the overall results of the study only suggest, but do not confirm, nuclear fusion in the bubbles’ collapse. . . .

  The experiment’s entire apparatus is well within the bounds of “table-top physics,” about “the size of three coffee cups stacked one on top of the other,” says Taleyarkhan. . . .

  Currently, the level of neutron emissions with the characteristic fusion energy appears to be lower than would be expected from the tritium signals observed in the experiment. Further tests are needed to account for this discrepancy, and to verify the observed relations between the neutron emissions, tritium production, and bubble collapse.

  If fusion is confirmed in further tests, these bubbles would still have a long way to go before they could be considered as a possible energy source with any commercial value, says Science co-author Richard T. Lahey Jr. of Rensselaer Polytechnic Institute. First of all, the bubble reaction would have to demonstrate net energy gain—that is, it should produce more energy than the energy needed to drive the reaction itself. Second, scientists would have to find a way to make the reaction perpetuate itself in a chain reaction, without constant input from a neutron source. . . .

  It was optimistic, but not outrageous. The biggest problem, in my mind, is that it did not mention the Shapira and Saltmarsh manuscript. That was a major oversight, though I am not sure whether the press office was even aware of the paper at the time.

  Reporters who received Science’s press package could get the Taleyarkhan paper, but the information was embargoed until 2 PM on Thursday, March 7. Only then would science reporters be allowed to print their stories.

  There was also an Oak Ridge press release, but I didn’t see it until a few days later. Its tone was a little more pessimistic:

  PRELIMINARY EVIDENCE SUGGESTS POSSIBLE NUCLEAR EMISSIONS DURING EXPERIMENTS

  Researchers at Oak Ridge National Laboratory, Rensselaer Polytechnic Institute and the Russian Academy of Sciences have reported results that suggest the possibility of nuclear reactions during the explosive collapse of bubbles in liquid, a process known as cavitation. . . .

  Experiments suggest the presence of small but statistically significant amounts of tritium above background resulting from cavitation experiments using chilled deuterated acetone. This tritium could result from the nuclear fusion of two deuterium nuclei. Tritium was not observed during cavitation of normal acetone, which does not contain deuterium.

  Attempts to confirm these results by looking for the telltale neutron signature of the deuterium fusion reaction have yielded mixed results. While there are indications of neutron emission in the newly published results, subsequent experiments with a different detector system show no neutron production.

  Theoretical estimates of the conditions in the collapsing bubbles are consistent with the possibility of nuclear fusion, under certain assumptions concerning the relevant hydrodynamics.

  These results suggest the need for additional experiments, said ORNL’s Lee Riedinger, deputy director for Science and Technology. In particular, the difference in the two sets of neutron measurements must be clarified. Additional tritium experiments would also allow a better understanding of the tritium observations.

  Until confirmatory experiments are completed, a cautionary view is appropriate, according to Riedinger, who said, “The manuscript has been through external peer review, but the scientific record shows that tritium and neutron measurements at these levels are difficult, and one must do further tests before firm conclusions can be drawn.”

  Like Riedinger’s comments to me, the Oak Ridge press release was as negative and cautious as it could possibly be without directly undermining Taleyarkhan. Everybody at Oak Ridge was carefully watching their words.

  I was, too. Over the weekend, after a few last-minute e-mails, I put the finishing touches on the first draft of the article. I had been asked to send a copy to Don Kennedy, so I did. It was an unusual request. Kennedy, as the editor in chief of Science, had the final say over everything that appeared in the pages. However, like his predecessors, he was pretty hands-off, at least when it came to the news section. We reporters were reasonably insulated from the politics of the magazine.

  On Monday, I heard back from Don Kennedy. He seemed a little annoyed by the tone of my piece. The references to cold fusion and the use of the word tabletop bothered him. I stood my ground, arguing that everyone was, rightly or wrongly, comparing bubble fusion to cold fusion and so we had to use the term. As for tabletop, Taleyarkhan had used the word, as had Science’s own press release. Kennedy immediately relented:

  I’m sorry if my cold-fusion allergy led to a slight grumble on my part. I don’t see how you could have avoided that term, and table-top is certainly okay. . . . I hope it was clear that although I might make a comment on a draft in such a situation, I am absolutely pledged to non-interference.

  The term cold fusion was clearly driving the editors at Science to distraction. On the morning of Monday, March 4, Science’s press office admonished reporters who were even thinking of using the phrase:

  [W]e ask all Science Press Package registrants to note that the peer-reviewed Science paper describes reactions inside bubbles that reach temperatures as hot as the center of the sun—up to 10 million degrees Kelvin. Thus, descriptions of this work as “cold fusion”
are grossly inaccurate. We wish to thank all journalists who are taking the time to read and understand this research, to convey accurate information to the public.

  In the same notice, the press office told of the Shapira-Saltmarsh paper and offered to provide it to reporters, as well as a response by Taleyarkhan’s team. (Oddly enough, Taleyarkhan’s team looked through the raw data provided by Shapira and Saltmarsh and claimed to see evidence of neutrons that the two were ignoring.)

  The situation had already reached a boiling point. By 1:30 PM on Monday, the embargo was blown. The press office lifted all restrictions on using the articles and begged, once more, that journalists write a balanced story.

  “Here we go. . . . Fasten your seatbelts,” one editor told me. It was all over the Internet in seconds. My article was being made available to reporters as well, and I started getting phone calls from television producers inviting me to talk on the air about bubble fusion.72

  The press coverage ran the gamut from optimistic and credulous (“Fusion ‘Breakthrough’ Heralds Cleaner Energy,” trumpeted London’s Guardian) to pessimistic and weary (“Here we go again; Table-top fusion,” sighed the Economist). Most were in the middle. My impression was that television reporters (as usual) were more keen on bubble fusion than their print counterparts, but few went overboard. After an intense burst of interest for a week or so, the media frenzy began to calm down. But an undercurrent of bad feeling remained within the science community.

  I knew that Science would be vulnerable to attack because of the bubble fusion paper, but I was surprised by the source of the most damning criticism. A week after the story broke, three of the reviewers of the paper—Putterman, Crum, and William Moss, a sonoluminescence theorist at Livermore—told the Washington Post that Science had published the Taleyarkhan paper over their objections:

  “I reviewed the paper twice, I rejected it twice,” said William Moss, a physicist at the Lawrence Livermore National Laboratory in California.

  “I told Science you can’t publish it because it’s not right,” said Lawrence Crum, a physicist with the Applied Physics Lab of the University of Washington at Seattle.

  “They say it was subject to stringent peer review, but does that mean it passed peer review?” asked Seth Putterman, a physicist with the University of California at Los Angeles, who also rejected the article.

  Now this—this surprised me. I had been so busy looking for the neutrons that I didn’t spend a lot of time with sonoluminescence physicists. And what I heard from them had been complimentary, if cautious. (After all, Crum had even used the word “doggone” when describing the beauty of Taleyarkhan’s idea!) Shortly after the Washington Post story, Ken Suslick, too, chimed in. In the beginning of April, Putterman, Suslick, and Crum wrote a short criticism of the Taleyarkhan paper, arguing that it had been “unready for publication” and suffered from “substandard experimental techniques.”

  The public criticism was not coming from fusion scientists, but sonoluminescence people,73 and from those I thought were reasonably supportive of Taleyarkhan’s technique. When I had interviewed Crum, and he admitted that he was a reviewer, he seemed positive enough that it didn’t occur to me to ask whether he suggested rejecting the paper. I had completely missed it. The three reviewers were also heaping criticism on Science’s review process. Later in the year, Putterman challengedScience to publish the positive reviews: “Somewhere out there is a positive report from someone,” he told Nature, Science magazine’s main rival. “Science should publish that report because then we’ll see what kind of information they went on to overrule four negative reviewers.” Of course, Don Kennedy refused. “We maintain our end of the confidentiality bargain about peer review, so I can’t discuss the process specifically, except to say that the positive reviews outweighed the negative ones. Why else should we publish the paper?”

  Did Science overrule the virulent objections of its reviewers and deliberately publish a bad story? Were the sonoluminescence people jumping on the anti-bubble-fusion bandwagon after it got hammered by Shapira, Saltmarsh, and other fusion scientists? Unfortunately, I don’t know for sure.

  Science is a peer-reviewed journal. But it is also a magazine. And magazines, especially those that run advertisements and classifieds, are always trying to boost their circulation. Peer-reviewed journals like to publish provocative and spectacular results in their pages to get extra attention. Sometimes this leads to bad science; even the best peer-reviewed journals occasionally publish substandard manuscripts in their pages. (Nature’s letters section, for example, is notorious for occasionally publishing attention-grabbing but dubious research.) The editor who received the bubble fusion manuscript might have been influenced by the spectacular nature of the claims, but at the same time, I don’t think that he (or other editors at Science) were consciously gaming the peer-review system to accept a manuscript that its reviewers had clearly rejected.

  Though I have never gotten my hands on them, I believe that the reviews of Taleyarkhan’s paper were mixed—skeptical and admiring at the same time—and that there was enough innovation in the experiment’s technique that the editor in charge felt that the reviews were sufficiently positive to merit publication. However, even today, I have to shrug my shoulders when asked what happened behind the scenes during Taleyarkhan’s peer review. Frankly, when the bubble fusion frenzy diminished, I was relieved.

  The story continued to simmer in the background. In July 2002, I covered an article by Suslick that appeared in Nature. Suslick and a colleague had used fluorescent dyes to measure the by-products of sonoluminescence in water and compared them to theorists’ expectations. “They’re saying, ‘We understand what’s going on inside the bubble,’ and if this is what you believe the science is, you should be suspicious of the Taleyarkhan paper,” Crum told me. (I wasn’t able to reach Taleyarkhan for comment by deadline, unfortunately.) After my article ran, Putterman dropped me a note saying that I was taking the theoretical work too seriously. I didn’t agree entirely, but the point was valid if a little odd. He seemed not to want to dilute the direct criticism of Taleyarkhan’s Science experiment with less-direct criticism from a theoretical perspective.

  Taleyarkhan had also been busy. Oak Ridge had included him on a team of scientists assembled to attempt an experiment that would end the controversy once and for all, but the Department of Energy refused to pony up the necessary money. In 2003, Taleyarkhan quit his post at Oak Ridge in favor of a named chair in Purdue University’s School of Nuclear Engineering. In 2004, he published a paper in Physical Review E, a high-level peer-reviewed physics journal, that seemed to confirm his original findings. I didn’t think it added much to the debate, so I didn’t cover it, despite an overheated press release from Purdue:

  EVIDENCE BUBBLES OVER TO SUPPORT TABLETOP NUCLEAR FUSION DEVICE

  Researchers are reporting new evidence supporting their earlier discovery of an inexpensive “tabletop” device that uses sound waves to produce nuclear fusion reactions. . . .

  ... Whereas data from the previous experiment had roughly a one in 100 chance of being attributed to some phenomena other than nuclear fusion, the new, more precise results represent more like a one in a trillion chance of being wrong, Taleyarkhan said.

  The New York Times covered the paper, as did a few other outlets, but it didn’t spark a huge amount of discussion, even though Crum told the Times that the new work was “much better” than what had appeared in Science. Nor did another confirming paper, written by Adam Butt and Yiban Xu, that appeared in a lesser journal, Nuclear Engineering and Design, raise any eyebrows. In this case, Purdue’s press release stressed the independence of the new work:

  PURDUE FINDINGS SUPPORT EARLIER NUCLEAR FUSION EXPERIMENTS

  Researchers at Purdue University have new evidence supporting earlier findings by other scientists who designed an inexpensive “tabletop” device that uses sound waves to produce nuclear fusion reactions.

  The technology, in theory, could
lead to a new source of clean energy and a host of portable detectors and other applications. . . .

  Xu and Butt now work in Taleyarkhan’s lab, but all of the research on which the new paper is based was conducted before they joined the lab, and the research began at Purdue before Taleyarkhan had become a Purdue faculty member. The two researchers used an identical “carbon copy” of the original test chamber designed by Taleyarkhan, and they worked under the sponsorship and direction of Lefteri Tsoukalas, head of the School of Nuclear Engineering.

  Taleyarkhan saw this as a great achievement, and in 2006 he wrote about how his results had been “independently confirmed.” This claim, too, was about to be challenged.

  In 2005, I resigned from Science magazine to become a professor of journalism at New York University. So when the bubble fusion affair exploded again, I was watching from the sidelines. It was a much more comfortable position.

  The bubble fusion affair began very differently from the cold-fusion fiasco. Unlike Pons and Fleischmann, Taleyarkhan and his team had not sought publicity until their research had been peer-reviewed by a major journal. Once Science stamped its imprimatur on the work, then the group could claim they were doing the right thing, at least according to the traditions of science—they were steering debate about the work into the scientific literature. Of course, there were questions about their competence as well as their conduct. For example, why hadn’t they withdrawn their paper after the devastating counterexperiment by Shapira and Saltmarsh?