Cold Fusion Confusion
By Robert F. Crease and N. P. Samios
The New York Times

September 24, 1989

Letters to the editor in response to article

Out of Utah came the promise of the century: cheap, abundant, safe nuclear energy. But then came the hard questions.

Victims of self-deception? Martin Fleischmann, left, and Stanley Pons told Congress they had achieved fusion at room temperature.

Photo: Diana Walker/Gamma Liaison

On March 23, Dr. B. Stanley Pons of the University of Utah and Dr. Martin Fleischmann of England's University of Southampton announced that, working together, they had created nuclear fusion at room temperature in a test tube.

Fusion, the process that powers the sun, ordinarily takes place at hundreds of millions of degrees. The announcement that so-called cold fusion had been achieved stunned the scientific community. We, like most of our colleagues, followed the story as it unfolded in the academic and popular press. With a mixture of excitement and skepticism, we learned that Pons and Fleischmann claimed to be generating four times as much energy as they were putting in, using a small device built with $100,000 out of their own pockets. Scaled up, the process could conceivably power cities. Small wonder that businessmen and politicians took notice, for such a device would revolutionize the creation and distribution of energy - and with it, the shape of international politics.

"Today we may be poised on the threshold of a new era," declared the chairman of the House Committee on Science, Space and Technology the following month, welcoming Pons and Fleischmann to a Congressional hearing devoted to their achievement. "If so, man will be unshackled from his dependence on finite energy resources."

Alas, by July, the new era had already stumbled and begun to fall on the threshold.

Today, with their findings almost universally discredited, Pons and Fleischmann continue to cling to their assertion that they have found something new. "We are absolutely sure of our results," Pons told The Wall Street Journal earlier this month. Earlier, Fleischmann had said, "If the amount of heat is so large you can't account for it by chemical reactions, then what else are you going to believe?" In fact, there may be another factor at work - self-deception.

Pons and Fleischmann apparently fell victim to the experimental scientist's worst nightmare. Usually, self-deception is quickly and relatively painlessly cleared up, either through the scientist's own labors or those of a neighboring lab bench. What made the case of Pons and Fleischmann different was their meteoric and public rise to celebrity, and their equally spectacular and public downfall.

Nuclear energy is produced when an atomic nucleus is split, in the process called fission, but it is also produced when two light nuclei are squeezed into one heavier nucleus, a process known as fusion. Fusion, at least theoretically, has several advantages over fission as an energy source. It can be fueled by deuterium, which is easily extracted from ordinary seawater; a fusion reactor would produce less dangerous types of radioactive wastes and would be immune to meltdown. As troubles with fission-generated nuclear power mounted, the promise of fusion power beckoned.

But it was not to be easily achieved. Atomic nuclei are positively charged, and hence repel each other. Bringing two positively charged deuterons - deuterium nuclei made up of a proton and a neutron - close enough together for fusion to occur means overcoming an immense electrical repulsion called the Coulomb barrier. The conventional way of overcoming the barrier is to heat up a dense swarm of deuterons and other particles to several hundred million degrees. At those temperatures, the particles move very rapidly, resulting in collisions violent enough to overcome the Coulomb barrier. The scientific and engineering problems involved are enormous, however, and until last spring practical fusion energy seemed to be decades in the future.

Then came Pons and Fleischmann's sensational Salt Lake City press conference. They claimed, essentially, to have come up with a trick to overcome the barrier. Beginning with a mixture of lithium and heavy water (a form of water made with deuterium rather than ordinary hydrogen), they then inserted two electrodes, a negatively charged electrode made of palladium and a positively charged electrode made of platinum, and passed an electric current through the solution. In their experiment, Pons and Fleischmann found that the temperature of the water rose by an unexpected amount. Calculations showed a fourfold increase in energy output, they said, compared to what they were putting in.

What was causing the increase? Pons and Fleischmann offer the following explanation: As expected, the electric current causes the water molecules to break up and the positively charged deuterons to migrate to the negatively charged cathode, the one made of palladium. Palladium, like other metals, is a crystal, meaning that its atoms are arranged in a regular, tessellated pattern. Its particular atomic structure allows it to absorb deuterons like a sponge. The deuterons lodge themselves in the interstices between the atoms, eventually crowding in two to a space. Powerful electromagnetic forces produced by the palladium atoms squeeze the deuterons together. This pressure, Pons and Fleischmann postulate, pushes the deuterons past the Coulomb barrier and causes them to fuse, with the concomitant release of energy in the form of heat.

Pons and Fleischmann’s "trick" was to fuse nuclei not through violent collisions but by compressing them in the miniature vise created by the palladium atoms. No need to create a sun in the laboratory, no need for the equivalent of a hydrogen bomb. Just a tub of heavy water, an electrolyte, two electrodes and some current.

Equally surprising was the relative lack of neutrons produced. Neutrons are a radioactive byproduct of all known fusion (and fission) reactions - and their presence in nuclear reactors is a major factor in making them hazardous places to be. They also contaminate structural elements of the reactor, creating waste-disposal problems. If the excess heat that Pons and Fleischmann reported was due to fusion, then they should have found a neutron flux of 103 neutrons per second in their laboratory - comparable to the amount produced in nuclear reactors. But they reported only 104. They attributed this to the fact that a new form of reaction was taking place. Wonderful! No need for tons of lead shielding, no vexing problem of waste disposal, no need to decommission plants because their parts had become too hot to handle.

At first, excitement mounted as teams of experimenters at the Georgia Institute of Technology, Texas A & M, Stanford University and elsewhere reported apparent confirmations of aspects of the Utah results. Pons, Fleischmann, the University and State of Utah submitted joint patent applications; dozens of corporations contracted to take a look. Palladium futures soared.

Peter Hagelstein, of the Massachusetts Institute of Technology, produced a theory to account for the Pons-Fleischmann observations, and submitted four papers on it to a scientific journal. Utah's Senator Jake Garn proudly chartered a plane for Senate colleagues to fly to Salt Lake City to witness the miracle first-hand. The Utah Legislature offered $5 million to Pons and Fleischmann for follow-up studies. And the two went to Washington to seek $25 million for seed money to start a $100 million Center for Cold Fusion Research.

But it wasn't long before some laboratories reported difficulty duplicating the observations; Georgia Tech withdrew its confirmation. Pons and Fleischmann offered little help to the unfortunates struggling to repeat their work; they declined to provide details of their techniques, refused to send samples of their equipment to laboratories for analysis, and withdrew the paper they had submitted to the scientific journal Nature, claiming that they preferred to press on with more urgent work rather than stop to handle the reviewers' criticisms. Hagelstein's papers remained unpublished; Garn cancelled the charter flight. One by one, laboratories dropped out of the cold-fusion quest, and corporations stopped knocking on Utah's door. In mid-July, a scientific advisory panel to the Department of Energy recommended that the agency not support the work of Pons and Fleischmann. "Evidence for the discovery of a new nuclear process termed cold fusion is not persuasive," the panel concluded. (The D.O.E.'s final report is due in November.)

Scientists began to dub any spurious result a "Utah effect." Wags began peddling "cold-fusion kits," each consisting of a test tube, a pair of electrodes and an Alka-Seltzer tablet, for $4.95 - "regular price $5M."

There were scientists who had seen as early as the original press conference that something was amiss, and many more knew after a glance at a pirated copy of the Nature article, which was faxed and refaxed around the world. What tipped them off was the presence of certain tell-tale signs, sometimes known as "symptoms of pathological science." Cold fusion may someday rank among such notorious scientific nondiscoveries as the Martian canals, N-rays and polywater. But it is more than a classic example of pathological science; it also provides a lesson about how experimental science works and why it is enormously difficult.

Self-deception in science is very different from fraud. Few working scientists lose sleep over fraud. The reason is simple: The motive for fraud is almost always résumé-padding - adding lines to a bibliography by publishing fabricated or plagiarized work. But this succeeds only when the work is trivial, insignificant or peripheral; if at all noteworthy, it will be exposed when other researchers try to duplicate and explore it. As Daniel Koshland Jr., editor of the journal Science, put it, "The bigger the result, the more quickly it is going to be checked." A successful fraud may contribute to the culprit's career but it rarely alters the course of science.

Self-deception, however - involuntarily or unconsciously being led astray - is a daily threat to every experimental scientist. Nonscientists often write of researchers as if all they need do is switch on their equipment and – presto! - facts appear, as if the real work of science consists of fitting such facts into theories.

But experimental science is far more complicated - and more interesting - than that. When an experiment begins producing results, the experimenters must still make sure that the readings represent a true profile of a scientific phenomenon - that the data have not been produced by something else in the environment or idiosyncracies of the equipment. To guard against these "systematic effects," as they are called, scientists run experiments over and over, making small changes in the equipment to see whether the effects change. Experimenters typically are suspicious of early data runs, and habitually ask themselves such questions as "How good are these data?" and "What else could they be due to?"

Impure materials, for instance, have ruined experiments. A classic example is the case of a reputable 19th century chemist named Chenevix. Studying, as it happened, the recently discovered element palladium, Chenevix became convinced that it was not an element at all but a compound of other substances. He circulated his findings among his peers and the general public, ridiculed those scientists who declared it to be an element - and was subsequently horrified to discover that he had erred, apparently by using chemicals that already contained the palladium he believed he was producing from scratch. Utterly humiliated, he abruptly quit science.

Even when the materials are pure, the equipment performs well and is used skillfully, the results must still be "read" correctly.

In the 1930's, a prominent physicist named Irving Langmuir coined the term pathological science, or what he called "the science of things that aren't so." Pathological science, he said, has a characteristic set of symptoms, and he drew up an informal list based on his own experiences. We have drawn up our own, based on ours. It is, we realize, neither exhaustive nor infallible. Genuine science might display any one of these symptoms; pathological science is usually accompanied by several.

Symptom No. 1: Too many miracles.

Miracles do occur in science. On Nov. 8,1895, the German scientist Wilhelm Roentgen noticed that a fluorescent screen on the table of his laboratory was glowing. Fluorescent crystals glow in the presence of light, but Roentgen had darkened the laboratory, in order better to examine emissions produced by the cathode ray tube he was using. He switched off the instrument and noticed that the glow stopped immediately. Astonished, and not quite trusting his eyes, he embarked on a systematic study of the phenomenon - fluorescence in the absence of light - without mentioning anything to anyone. Finally, on Jan. 1, 1896, he sent a paper to colleagues announcing the discovery of invisible "X-rays," emitted by the cathode ray tube, which were able to pass through paper, wood and even the human body.

Many scoffed; one scientist called the story of the rays a "fairy tale." Every well-stocked laboratory had cathode ray tubes and fluorescent screens, and it seemed highly unlikely that such an effect had not been noticed before. (In fact it had been, but the observations were disregarded.)

Five days later - after scientists had had a chance to read Roentgen's paper - the story of X-rays was splashed across the front page of several newspapers. Within three weeks of the discovery, physicians in Dartmouth, N.H., used them to help them set the fractured arm of a boy named Eddie McCarthy. Roentgen had no "theory" of what the mysterious rays were, hence their name. But when the Nobel Prizes were established in 1901, the first physics award went to him.

Many scientists were disturbed by how much existing scientific knowledge had to be suspended in order to swallow the discovery, and were not surprised when laboratories in Germany and Britain had difficulty reproducing N-rays, or when some laboratories retracted their early confirmations. The rays were exposed as false most dramatically when an American physics professor visiting the lab secretly pocketed a supposedly crucial piece of the apparatus while Blondlot and his assistant continued their observations unperturbed. Blondlot had clearly fallen victim to a grand case of self-deception. The numerous miracles his finding required were an early sign.

Cold fusion, too, required too many miracles. The first was that an utterly unknown way of achieving fusion had escaped the attention of generations of nuclear physicists. The second was that deuterons could be squeezed closely enough together inside palladium for fusion to occur. The third was that the fusion produced so few neutrons. Each miracle, taken separately, was plausible. But the simultaneous appearance of three was strong circumstantial evidence of pathology at work.

Symptom No. 2: The "discoverers" are outsiders.

In the early 1970's, an astrophysicist named Ian McCusker decided to step into particle physics in search of single quarks - fundamental particles with electric charges a fraction of those of the electron and proton. Particle physicists were fairly sure that they came only in trios or in pairs (consisting of a quark and an antiquark) with a total integral charge. Single, isolated quarks had been hunted for years without success. McCusker decided to look anyway - and promptly announced, at a conference in Hungary in 1974, that he had found them.

Again, most scientists first learned of the discovery through the newspapers, and eagerly awaited details. McCusker visited several laboratories, armed with cloud-chamber photographs of particles whose tracks were sparser than usual, and asserted that such tracks could only be produced by a particle with a fractional charge.

At Brookhaven National Laboratory, particle physicists asked him whether he had measured the density of each of the tracks in his pictures, whether made by a quark or other particle. He hadn't felt it necessary, he said. When further experiments, replicating McCusker's work, looked at all the tracks, including those of the kind McCusker had ignored, scientists could see that the "quark" tracks had been produced by normal particles and that their sparseness was the result of systematic effects.

McCusker's phantom quarks are but one instance of outsiders making discoveries that turn out to be false. Outsiders start with the disadvantage of hostility from insiders - which they invariably interpret as, at best, the inertial resistance of a hidebound community to bright new ideas, and at worst as xenophobia. But a general skepticism about outsiders is still well-founded. An insider knows the traps, has made the mistakes, and is aware of the temptations. Discoveries are usually made when a person intimately familiar with the field learns or does something and the implications suddenly snap into place. They do happen when someone playing around in unfamiliar terrain stumbles across something that the locals have overlooked - but this is very rare.

Outsiders, however, have the advantage of appearing to the public as Davids in a contest against Goliaths. Pons and Fleischmann, both chemists equipped with the equivalent of slingshots, strode into the province of nuclear physicists, who inhabited huge laboratories well-funded by the Federal government. Add to that romantic image one final touch: the well-known chauvinism physicists tend to have about their field. Small wonder the press cheered, and that editorials in prominent newspapers scorned the "compulsive naysaying" of the establishment's "academic mentality" that dared to question their feat.

As newcomers, Pons and Fleischmann seem to have fallen into several traps that experienced physicists might have avoided. For example, old hands know that where you place the thermometer in a mixture can affect temperature readings. Some physicists have claimed the mixture Pons and Flelschmann worked with should have been stirred to get a uniform temperature; it's unclear whether Pons and Fleischmann did that. Then too, physicists say, Pons and Fleischmann may have encountered difficulties in counting neutrons and gamma rays.

In the long run, scientists paid far more attention to the work of Steven E. Jones, a physicist at Brigham Young University, who at the same time as Pons and Fleischmann's announcement had made a much more modest claim about cold fusion. Jones claims that he detected fusion, but at a level that, as he said, bore the same relation to Pons and Fleischmann's measurements as a dollar bill to the national debt. His colleagues took him seriously not because he was one of their own, nor even because he showed up at all the important meetings to defend his work. Rather, it was because his work betrayed an awareness of potential pitfalls. When Jones reported detecting cold fusion, though at rates far too low to be commercially practical, it was counted as a genuine scientific claim - and still is.

Symptom No. 3: The discoverer has not tried to kill the discovery.

The microbiologist Ludwik Fleck, in "Genesis and Development of a Scientific Fact," a landmark book on the practice of science, compared scientists like Roentgen to Columbus; they set out for an "India" and suddenly run into "America" Inevitably, the discoverer initially is bewildered by the appearance of the unexpected. But most scientists are not as fortunate as Roentgen, and the unexpected turns out not to be a new continent but something produced by faulty or misread equipment. Therefore, the first instinct of a good experimenter, when confronted with the unexpected, is to try to "kill" it - to track down every possible conventional explanation. Anyone prepared to announce the discovery of a new continent should be certain it's not familiar territory misidentified. A true scientific phenomenon is invariant; it will show up in many different kinds of equipment in different circumstances, somewhat the way a real object can be seen by many different people and from many different angles. Often, an object can momentarily appear to be something it's not - the way, for instance, a photograph might seem to show George Bush with a tourist on a Washington street corner, whereas in reality "Bush" is a cardboard prop. In trying to kill an unexpected result, scientists vary the systematics of the experiment, and make sure to collect plenty of statistics.

Pons and Fleischmann were unable to convince their colleagues that they had made a genuine attempt to kill their results. To all appearances, they had not varied those parameters that one ordinarily varies for studying systematic effects. They hadn't performed the experiment, for instance, with ordinary water instead of heavy water. Nor had they adequately checked the reliability of the neutron detectors. Georgia Tech withdrew its confirmation when it discovered, through systematic checks, that its neutron counters were temperature sensitive.

To be sure, the history of science does contain cases of "overkill," in which good scientists have ignored new phenomena staring them in the face. Leon Lederman, former director of Fermilab and winner of the 1988 Nobel Prize in physics, once wrote a paper about "the big ones that got away." In attempting to kill unexpected results by pushing conventional explanations as far as possible, he explained away real breakthroughs, including several that later won their discoverers Nobel prizes. Still, he is in fine company; fellow Nobel laureate Irène Joliot-Curie, to name but one, realized - after a rival discovered it - that she had observed nuclear fission. But she had convinced herself that the phenomenon could be ascribed to things she already knew.

Symptom No. 4: Inability to repeat the experiment is met by ad hoc excuses.

An experimental scientific paper is in some sense a recipe. It does two things. First, it testifies that the experimenters have a result they want others to know about; second, it provides directions for others to follow. It need not describe every detail - where the bowls were bought, the brand names of the ingredients - but assuming a background set of concepts and a standardized set of laboratory practices, a scientific paper warns, "This is the tricky part. Watch out for that." Scientists establish credibility and priority for discoveries by laying out their work as openly as possible. A scientific paper with an inadequate recipe is a tip-off that the authors' understanding of their work is incomplete.

Blondlot, for instance, was never able to provide a sufficient recipe for N-rays. When his colleagues said that they couldn't detect the rays, Blondlot produced additional details, accompanied by dark suggestions that those who couldn't repeat his work were incompetent. A few of Blondlot's champions even began claiming that only individuals of Latin descent had good enough eyesight to see evidence of the rays.

Pons and Fleischmann followed this pattern. When someone claimed that it was not possible to produce colt fusion, the two Utah scientists would add more instructions - that the palladium cathode had to be prepared in a certain way, or the electrolyte had to be of a certain concentration. As Robert Park, head of the Washington office of the American Physical Society, remarked, "Any time someone did the experiment with no result, they would say, 'You didn't do the experiment right,' and offer up another tidbit."

Why does pathological science occur? The motivation may be a desire for fame, immortality, wealth - or a genuine desire to help mankind. Fleischmann once stated that the motivation for the Utah team's work was social "We recognize the desperate energy needs that will confront the 21st century," he said. But if such desires interfere with one's disinterested evaluation and testing, the outcome may be pathological science.

If Pons and Fleischmann actually achieved what they claimed, to get the scientific community to change its mind about cold fusion they will have to start over from scratch - by publishing in the scientific literature, demonstrating expertise in the procedures, showing that they have tried and failed to kill their results, and providing a comprehensive set ofinstructions that others can follow.

In the movie "Back to the Future" the time-traveling DeLorean is powered by a small device that runs on beer. The label reads:

Mr. Fusion
Home Energy Reactor

The scene effectively captures, and sends up, the way science seems to be constantly discovering, simplifying, standardizing and miniaturizing complex and powerful phenomena in such a way that they can be handled by ordinary people with no scientific training at all. The world may someday contain fusion devices that will be as convenient as automatic coffee makers. But if so, there will be no shortcuts in the scientific process leading from here to there.

Robert P. Crease is an assistant professor of philosophy at the State University of New York at Stony Brook and historian at Brookhaven National Laboratory. N. P. Samios is director of the Brookhaven National Laboratory.

Letters to the Editor

Go to original

November 5, 1989

I was amused, then not so amused by ''Symptom No. 2: The 'discoverers' are outsiders.'' The ridicule of outsiders in their quest for truth is the hallmark of orthodoxy. The function of orthodoxy, however, is to preserve the status quo. Creativity comes from outsiders, like two bicycle mechanics from Dayton, or a lonely patent clerk in Switzerland.

STEPHEN WARD, M.D.
Abington, Pa.

Go to original

November 5, 1989

Your authors contend that ''good'' scientists always try to ''kill the discovery.'' That is straight out of Karl Popper's philosophy of falsification, and it simply doesn't describe the behavior of real scientists. No scientist kills his own discovery - you verify it to your own satisfaction, publish it and let the skeptics do their worst. If the discovery holds up, well and good. If not, tough. The fact is that only a very small fraction of published discoveries hold up to long-term scrutiny.

ROBERT ROOT-BERNSTEIN
East Lansing, Mich.

Go to original

November 5, 1989

The remarks of Robert P. Crease and N. P. Samios, in their article ''Cold Fusion Confusion'' (Sept. 24), add to a long history of scientific ''spin control'' - attempts by established science to protect its professional solidarity from outside threats, real or imagined. Brookhaven National Laboratory is an important center of conventional fusion research. As its director and its historian, your authors have an obvious stake in Dr. Pons's and Dr. Fleishmann's being exactly wrong about their fusion-in-a-bottle experiments. The ''mixture of excitement and skepticism'' with which they say they followed the story no doubt stemmed, in part, from anxiety over new and possibly powerful competition. ''There were scientists who had seen as early as the original press conference that something was amiss,'' they write. The smug we-knew-it-all-along tone betrays their relief that, as it seems, cold fusion has fizzled.

WADE ROUSH
Cambridge, Mass.

Go to original

November 26, 1989

We are writing to correct Wade Roush's letter (Nov. 4) in response to our article ''Cold Fusion Confusion'' (Sept. 24). Although Brookhaven National Laboratory is a distinguished, multipurpose facility with many programs in numerous areas of science, it is not, as Mr. Roush claimed, ''an important center of conventional fusion research.'' Thus, the charge that the article was self-serving is misplaced. America's major fusion facilities are at Princeton and Lawrence-Livermore.

ROBERT P. CREASE
N. P. SAMIOS
Brookhaven National Laboratory
Upton, L.I.

(In accordance with Title 17, Section 107, of the U.S. Code, this material is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. New Energy Times has no affiliation whatsoever with the originator of the original text in this article; nor is New Energy Times endorsed or sponsored by the originator.)

"Go to Original" links are provided as a convenience to our readers and allow for verification of authenticity. However, as originating pages are often updated by their originating host sites, the versions posted on New Energy Times may not match the versions our readers view when clicking the "Go to Original" links.