In fact, a battle was brewing. A number of physicists at the laboratory were extremely doubtful of the bubble fusion research, and their doubt triggered a flurry of activity behind the scenes. Shortly after Taleyarkhan submitted the manuscript—with Oak Ridge’s permission—to Science, the skepticism in the lab began to mount. Lab officials apparently asked two other Oak Ridge scientists, Dan Shapira and Michael Saltmarsh, to repeat the bubble fusion experiment. Saltmarsh was a fusion scientist who had testified before Congress about the cold-fusion affair. Shapira studied exotic fusion reactions induced by high-energy beams of ions. Both had the expertise to find neutrons from bubble fusion—if those neutrons existed.
Calling for other scientists to repeat an experiment before publication was an extremely unusual step, and it likely struck Taleyarkhan as a vote of no confidence, but Oak Ridge insisted. The lab seemed determined to avoid becoming the center of another cold-fusion fiasco. So, with Taleyarkhan’s assistance, Shapira and Saltmarsh set up an exact copy of the bubble fusion experiment, except for one detail: they used a bigger and better neutron detector. Not only was it physically larger (making it more sensitive, because more neutrons could strike it), but it also had more sophisticated electronics. Unlike Taleyarkhan’s detector, it could tell the difference between neutrons and gamma rays.
When Shapira and Saltmarsh analyzed the data they had gathered, the results were damning. They found no sign of fusion, no evidence for neutron emission from the bubbling deuterated acetone. They did not try to verify Taleyarkhan’s findings of tritium, but noted that if the tritium had been produced by fusion, the bubbling solution should have produced a million neutrons per second, and that level of activity should easily have been picked up by the neutron detector. According to their equipment, though, nothing was happening in the bubbling liquid, just the expected number of chirps caused by stray neutrons produced by cosmic rays and the like. (And since the team members were making bubbles by zapping the tank with neutrons, a heck of a lot of those particles were skittering about in the background.)
Oak Ridge was in a bind. They were about to look foolish. One of their researchers was about to publish what they considered a bad piece of research that would spark a second cold-fusion fiasco. And they were increasingly powerless to stop it. The lab had already given Taleyarkhan permission to seek publication, and Science had already accepted and reviewed the manuscript. Yet Oak Ridge seemed to have an experiment that blew the Taleyarkhan discovery out of the water. They were rapidly running out of options.
Scrapping the paper ceased to be a possibility the moment Taleyarkhan had sent the paper to the journal. However, Oak Ridge’s objections had slowed publication by a few weeks. In that time, the lab moved fast to try to reduce the impending damage. Shapira and Saltmarsh quickly typed up their results in a short report and sent it over to Science, hoping the two papers would be published side by side. The negative report, if accepted, would at least force readers to cast a skeptical eye on the claims of bubble fusion; Oak Ridge wouldn’t look quite so bad when other researchers poked holes in Taleyarkhan’s work. Unfortunately, that wasn’t an option, either. Any scientific manuscript in Science had to be peer reviewed, and the Shapira-Saltmarsh paper was no exception. There was no way that a new paper could be sent to reviewers, receive comments, and be revised in time to make the March 8 issue. (And it was becoming increasingly clear that holding beyond March 8 would be impossible; word of Taleyarkhan’s paper was beginning to leak out.) Oak Ridge had no options left. The world would soon learn about bubble fusion, even though Shapira and Saltmarsh had shown that it was almost certainly a fiction.
I was back on the case on the afternoon of Wednesday, February 20. The Science editors had decided to go ahead and publish the article on March 8, and while formal approval had not yet come through from Oak Ridge, they had assurances that it would come shortly. (And it did.) I was given the green light to begin reporting again, but I was warned to tread carefully to avoid leaks. I immediately e-mailed Taleyarkhan again and set up an interview. That part was easy. The hard part was figuring whom else to talk to.
I needed to speak to outside researchers, people not in Taleyarkhan’s research group. Only then would I get a reasonably objective opinion on the quality of the paper. At this stage, I couldn’t show the manuscript to anyone who hadn’t yet seen it; I couldn’t be responsible for a leak this far ahead of publication. So I had to figure out who had already seen the paper—I had to find the paper’s reviewers.
Nobody at Science would tell me who they were. The reviewers are kept confidential, even from the reporters who work for the same magazine. But I could guess. The Taleyarkhan paper crossed two fairly established disciplines, sonoluminescence and fusion. Just a few groups had been studying sonoluminescence for years. Lawrence Crum led one at the University of Washington, Seth Putterman led another at the University of California, Los Angeles (UCLA), and Ken Suslick ran a third at the University of Illinois. I was fairly certain that at least one of these scientists had been a reviewer. The fusion side of the paper was tougher. It was a bigger field, with many more researchers. I figured that the most likely candidates were those who knew about the nitty-gritty of neutron detection. If anyone would be able to bolster or tear down Taleyarkhan’s work, it would be a neutron expert. In fact, if I were to pick reviewers for the manuscript, I would choose some of the physicists who had dissected the cold-fusion papers. They would certainly approach the paper with a skeptical eye, and if they were convinced, the paper would automatically get a huge amount of credibility.
I began making discreet inquiries.68 On the sonoluminescence end, I called Crum and struck pay dirt. (As it turned out, all three of the big names in sonoluminescence—Crum, Putterman, and Suslick—had been reviewers.) I got the distinct impression that the sonoluminescence people were impressed by Taleyarkhan’s technique, if a bit skeptical about his team’s conclusions. Crum, at the very least, seemed particularly interested in Taleyarkhan’s method of creating large bubbles with a beam of neutrons and thought it might open some new opportunities for research. (“I thought, doggone! I’m depressed I hadn’t done that experiment,” Crum told me. “It’s a remarkable result, and I would like very much for this to be true.”) So the sonoluminescence end of the experiment seemed relatively solid, at least from my limited reporting.
By that time, the editors had given me more detailed neutron data from Taleyarkhan’s lab. The new information didn’t assuage my doubts. I was no expert at interpreting such data, but they didn’t look quite right. They were muddy; the shape of the peaks in the graph didn’t appear the way I expected them to. These were just my gut instincts, but they emphasized my need to find a neutron expert.
When the time came for my interview with Taleyarkhan, I found him open and friendly. He was happy to tell me all about the research. I confirmed that he was, quite naturally, enthusiastic about the quality of the results—including the neutrons he was detecting. However, no matter how confident Taleyarkhan was, he was not going to be the person who could assure me about the quality of the research. I still had not found a neutron expert who had already seen the paper, especially since I was still supposed to be very discreet. That problem was about to be made moot.
I had sensed that the editors at Science were getting increasingly tense. By the twenty-sixth, I had heard rumors in the building that somebody was trying to pressure the journal to reject the Taleyarkhan paper, and that Science’s editor in chief, Don Kennedy, was hopping mad. I didn’t know anything more for certain until the morning of February 27, when Coontz passed me a cryptic note. He told me I had to call Princeton’s Will Happer and IBM’s Dick Garwin.
Happer and Garwin were legendary figures in the community. They were the big guns of fusion (and of nuclear weapons). Garwin had helped design Ivy Mike; Happer was a former head of the government’s JASON panel. Both had been at the top of the scientific hierarchy for decades and had been involved in debunking cold fusion. They weren’t r
eviewers of the manuscript—I was pretty certain of that—but clearly they had seen it. And they apparently had some very strong opinions that they expressed to Don Kennedy.69
I didn’t know what Happer and Garwin said, but I knew that Kennedy was furious. He felt that outsiders were trying to disrupt the peer-review process and derail a paper that had already been accepted for publication. “There was certainly pressure from Oak Ridge to delay, if not to kill, the paper,” Kennedy told me when I interviewed him. “I’m annoyed at the intervention, I’m annoyed at the assumptions that non-authors had the authority to exercise constraints on the publication and telling us we couldn’t publish the paper—which they did.”
I called Happer and began to piece together what was going on. (Garwin was in China at the time but I soon got his side of the story, too.) Someone—I never found out for sure who it was—had sent Garwin and Happer each a copy of the Taleyarkhan manuscript and a copy of the Shapira-Saltmarsh paper. Both then e-mailed Kennedy. Garwin was harsh and succinct:
I understand there has been some discussion as to whether Science , having accepted the paper, should nevertheless not print it. I certainly don’t want to enter into such a discussion with you.
But I do want to tell you that I have read both papers carefully, and that I think the odds are extremely high that this “discovery” is simply error and incompetence.
So I caution you to mute the natural enthusiasm of people on your team who want to publish the latest significant discovery.
Happer’s note was longer and more involved, but he, too, urged Kennedy not to hype the results, likening it to cold fusion and polywater:70
I am told that the paper in question, the most recent version of which has the title “Evidence for Nuclear Emissions During Acoustic Cavitation,” will have a place on the cover of “Science” when it is published, and will be accompanied by additional laudatory editorials and sidebars. I may well be misinformed about this, but if I am not, I hope I can persuade you to exercise your authority as Editor in Chief to stop these plans....
Giving the Taleyarkhan paper lots of publicity in “Science” would damage the credibility of the magazine and would do quite a bit of harm to our scientific community. I have seen several examples of similar episodes in my career, for example, polywater and cold fusion. Both episodes caused lasting damage to science.
Happer strongly argued that the Taleyarkhan paper was in error, and said that several other prominent fusion researchers (including the laser fusion scientist John Nuckolls) had concluded that Taleyarkhan was flat-out wrong. But at the same time, Happer insisted that he wasn’t trying to block publication. “I like Science; I’m a member of AAAS,71 and I don’t want them to shoot themselves in the foot—or some other body part,” Happer told me. “All I told him was, for God’s sake don’t put it on the cover of Science.” Garwin was a little more oblique about what Science should do with the bubble fusion manuscript. “It would be unfortunate if Science magazine were to take any position on its correctness,” he said.
There’s no question in my mind that Happer and Garwin would have been happy if Science had pulled the Taleyarkhan paper. They were convinced (as I was becoming) that bubble fusion was a fiction. However, I think that by the time they contacted Kennedy, they knew that Science was likely to publish the paper. They just wanted to minimize the damage to Oak Ridge and to the journal and, above all, to science. Don’t hype the results. Don’t put it on the cover. Don’t go out on a limb for bubble fusion.
Kennedy, not unreasonably, interpreted the notes slightly differently. He saw it as a last-ditch attempt by Oak Ridge to quash publication. His response to Happer was measured, if indignant. Science editors had taken exceptional care with the manuscript, he insisted, and the review process had led to a firm recommendation to publish:
Concern on the part of research managers from ORNL appeared late in the game, followed by some secondary measurements taken with a different detector which are claimed to show differences with respect to some of the results. That work has not been peer-reviewed. . . .
I am now hearing from you and a few other distinguished physicists arguing that I should now block publication of a paper that has met and passed all our tests. Because I believe that the right way to resolve serious scientific differences is through repetition, peer review, and publication I plan to proceed. I have told the ORNL management that we will be happy to consider a manuscript by those who have a different interpretation.
It was getting more complicated by the minute, but now that the Taleyarkhan paper was circulating around the physics community, I was freed from the high level of secrecy that had hampered my investigations. I made dozens of phone calls to scientists around the country, asking not only about bubble fusion but also about what had happened with the review process at Oak Ridge. I knew about Shapira and Saltmarsh, so I contacted them. I also spoke to one of Oak Ridge’s deputy directors, Lee Riedinger.
I was pretty sure Riedinger thought bubble fusion was garbage—but he never said so directly. “I’m confused,” he told me, when I asked him whether he believed the Taleyarkhan experiment or the Shapira-Saltmarsh one. “There’s an active dialogue back and forth about what could be wrong with either set of measurements.” Riedinger seemed to be walking a very fine line. He was trying to temper enthusiasm for Taleyarkhan’s results without publicly faulting his own employee’s research. (And he went out of his way to compliment Taleyarkhan’s abilities, adding that his work is “very novel and interesting.”)
I found a fusion expert at Livermore, Mike Moran, who had also done sonoluminescence—with deuterated water, no less—and I asked him what he thought of the paper. “The paper’s kind of a patchwork, technically, and each of the patches has a hole in it,” he told me, and pointed out a number of damning flaws that I hadn’t considered. Taleyarkhan’s experiment seemed to be producing some tritium, and the tritium production disappeared when the sound-wave generator wasn’t operating. Moran pointed out that this disappearance was a problem. A deuterated solution that had been irradiated by neutrons should show an increase in tritium levels whether or not the sound-wave generator—which collapsed the bubbles to ignite fusion—was working. Even without the fusion reactions, some of those neutrons would strike deuteriums and stick, creating tritium. In Taleyarkhan’s experiment, this was not the case; the control experiment with deuterated acetone and no sound-wave generator showed no increase in tritium. “If he’s really right, it should have shown up,” Moran said. “It’s an inconsistency in the data.”
I was convinced. Taleyarkhan was wrong: bubble fusion was a fiction. And because of the spurious result, a scientific drama was playing out before my eyes. The officials at Oak Ridge felt that the Shapira-Saltmarsh paper was damning, and they were hoping to avoid embarrassment. Garwin and Happer were trying to prevent another cold-fusion controversy, and Kennedy was trying to preserve the integrity of the peer-review process. Rumors were flying, and they were getting nastier and more paranoid by the minute. Everybody was getting increasingly annoyed with everyone else.
I got a note from Happer on Friday, March 1. He relayed part of a message from an unnamed colleague (not from Princeton or Oak Ridge) who was telling people of rumors that I had been “calling around asking about activities surrounding the publication of the article, not the ‘science’ in the article.” That message continued:
I don’t want to get anyone upset about this, but it does tend to verify the rumors we have heard about Don Kennedy, the current Science editor and former Stanford President, wanting to go after Princeton people for opposing the publication of the research paper in its original form.
The story had just gotten harder. The allegations were absurd—Kennedy had not influenced me at all, so I could hardly be his attack dog. However, it would be impossible to get honest opinions from physicists who believed that I was preparing a hatchet job at the behest of Don Kennedy. Later in the day, Happer sent a final note:
Be care
ful what you write and remember:
“The moving finger writes, and having writ moves on.
Nor all thy piety, nor all thy wit,
Can call it back to cancel half a line
Nor all thy tears wash out a word of it.”
I had been Omar Khayyamed.
Happer need not have reminded me. Of course I was acutely aware of the sensitivity of the situation. People were quite likely to get their first impression about bubble fusion from my article at the front of the magazine. I was skeptical, and I wanted that skepticism to come through, but at the same time I wanted to make sure that everybody’s view was represented fairly: the researchers, the editors, and the skeptics. In my mind, the whole bubble fusion controversy was fueled by these three parties’ tragic mutual incomprehension.
Sun in a Bottle_The Strange History of Fusion and the Science of Wishful Thinking Page 19