A British urban myth has it that sometime in the 19th Century, a (rich) gentleman wanted to find out how suspicious the public really was. So he stood on London Bridge with a tray around his neck full of gold sovereigns, offering to sell a 'Pound for a Penny'. Not one sovereign was sold.

In the vast majority of cases, when a bargain looks too good to be true, it is. So it seemed when, in 1989, two electrochemists called Martin Fleischmann and Stanley Pons promised a revolutionary way of extracting huge amounts of energy from heavy water without the need for anything more complicated than a palladium electrode and a battery. The scientific furore which then erupted over the alleged phenomenon of cold fusion ruined a number of reputations and sharply polarized the scientific community. It still reverberates today, and provides an object lesson in the pitfalls of challenging scientific orthodoxy with unreliable evidence.

How Nuclear Fusion Works

Nuclear fusion is the process by which stars shine. It is the opposite of nuclear fission, in which huge amounts of energy are generated by atoms splitting apart. In fission, large unstable nuclei release energy when they fall to bits. In fusion, very small atomic nuclei join together to release energy. Typically, in stars, hydrogen is transformed into helium and some heavier elements. Stars are able to do this because the incredibly high pressures and temperatures in their cores can squash together positively charged particles. This overcomes their electrostatic repulsion to the point where the forces that bind the nuclei together begin to operate between the nuclei as well as within them. The forces are very strong so, when the nuclei fuse, huge amounts of energy get released.

Exactly how difficult it is to achieve this depends upon the nature of the nuclei involved. As the nuclei get heavier, the process tends to become easier as they contain neutrons which dilute their positive charge. The simplest and lightest of nuclei, the hydrogen nucleus, is a naked proton, and fusing two of these together is next to impossible. Stars manage it in their core, albeit with difficulty. Humans have never managed to achieve proton-proton fusion. They have managed it with two heavier, neutron-laden isotopes of hydrogen, deuterium and tritium¹, but it generally takes an atomic bomb² to get the reaction going. This minor drawback would seem to quite decisively rule out nuclear fusion as a convenient and controllable energy source.

In the Red Corner

Undaunted, certain scientists were determined to prove that fusion could be a controllable and sustainable energy source. If correct, they would have solved the energy needs of the planet at a stroke: a single cubic kilometre of seawater contains enough deuterium to match the energy content of all combined global fossil fuel reserves, burned and unburned. The attraction doesn't stop there; the waste product is helium, which is completely harmless³: compare this with the nasty and insidious fallout from atomic disasters, both unintentional (Chernobyl) and deliberate (Hiroshima). If any technological prize was worth pursuing it was that of contained fusion, and so huge amounts of money, time and brainpower were poured into its pursuit over decades.

Most of these efforts concentrated upon reproducing the conditions in the heart of the Sun. Huge 'tokamaks'⁴ - gigantic toroidal (doughnut-shaped) reactors wrapped with strong electromagnets - confined hot plasma and squeezed the atoms together until fusion started. Some sleight-of-hand was employed in that deuterium and tritium⁵ were used as fuel instead of common-or-garden hydrogen. Even under these favourable conditions, self-sustaining fusion, where the energy output exceeded the input into the reactor, proved unattainable.

Nevertheless, there have been some notable milestones passed on the way to this objective. In 1991, the JET project produced 1.7 megawatts of fusion power, although for a short time. Then, in 1997, the same reactor produced 16MW of power, which in itself went some way to fulfilling the energy input needed to continuously heat the plasma to the 10 million°C needed for sustainable fusion. But not all the way and not for a long time yet: the most optimistic forecasts is that fusion power will not become a viable energy source until 2050 at the earliest.

In an effort to shorten this timescale, the next-generation fusion reactor, ITER, is being built by an international consortium and with a budget exceeding $2.5 billion. ITER is designed to generate five times as much power as it consumes. Other methods of achieving fusion are being investigated, such as the Z-pinch device. But fusion science resolutely remains Big Science, characterized by extended international collaboration, long timescales and huge budgets.

In the Blue Corner

Enter Fleischmann and Pons. Both of these scientists were reputable electrochemists, a profession that, on first impression, is as far removed from that of the fusion scientist as a car battery from the Sun. Electrochemists deal with chemical reactions on a modest energetic scale and with simple and cheap equipment, easily obtainable and simply deployed. In 1989, both these scientists were working at the University of Utah when they shocked both the chemistry and physics community with their announcement that they had made nuclear fusion work in a chemical laboratory at room temperature, using nothing more than a simple electrochemical cell, deuterium oxide (heavy water to the rest of us) and a palladium electrode.

Just as startling as the nature of their claims was the manner in which they were made public: instead of subjecting them to the accepted process of peer review in a scientific journal, they were announced by fax and press conference. Fleischman and Pons claimed to observe excess heat generated in these cells significantly over and above that of the electrical input, even to the extent of the heavy water boiling away. If true, then this simple process would have invalidated the billions of dollars and countless hours spent by fusion scientists on achieving their goal: from the point of view of the physicists, the electrochemists would have 'shot their fox'. The phenomenon was dubbed 'cold fusion' by the media community.

The reaction from the 'hot fusion' community was predictably hostile, and not without good justification. Criticisms were founded on two bases. Firstly, there were the theoretical objections. Fusion reactions take place between atomic nuclei, and at energy levels vastly exceeding those encountered in chemistry. The suggestion that the overwhelmingly electronically-influenced chemical environment could mediate in fusion reactions was preposterous and without any physical basis. Secondly, there were reservations about the experimental methodology used. Fleischmann and Pons had not run any 'control' experiments using ordinary water, which wouldn't undergo D+D fusion. The excess heat observed in the electrochemical cell could not be unambiguously attributed to the presence of fusion of deuterium from heavy water as it hadn't been shown as absent in cells containing ordinary water. Also, deuterium fusion produces highly energetic helium nuclei which rid themselves of the excess by ejecting either neutron radiation or producing protons (hydrogen nuclei) and tritium. Neither was present.

A damning report published in 1989 by conventional fusion researchers who tried to reproduce the results of these chemists concluded that there was no basis for their claims. The results were not at all reproducible and there was no evidence for this effect. The Utah electrochemists had been the victims of their own self delusion and sloppy experimental practice: 'case for the prosecution closed, m'lud.' The physics community breathed a sigh of resigned relief and got on with the job of investigating hot fusion. Cold fusion was sent to that area of the Gulag reserved for pariah science, such as N-rays, antigravity and ghost molecules.

Muddying the Heavy Waters

Under normal circumstances it would probably it would have ended there, with the chemists retreating to the undergrowth to lick their wounds. However, these were extraordinary circumstances which had generated an extraordinary amount of interest. Some scientists were prepared to give the 'Utah Two' the benefit of the doubt, possibly seeking comfort in the epistemological argument that 'absence of evidence is not evidence of absence', and also intrigued by the prospect that the case had not been conclusively settled, the headlong rush-to-judgment seeming rather over-eager⁶.

So, throughout the 1990s a trickle of scientific papers (most cagily avoiding the use of the 'f-word') appeared in respectable journals noting unusual effects when transition metals⁷ were exposed to deuterium. The results of the Utah Two were reproduced, although unreliably. Bursts of neutrons were observed when deuterium gas was loaded into titanium. Tritium formation was observed under a variety of conditions. Helium-4 production was also observed. Between 1989 and 2003 over 180 papers were published detailing strange effects in deuterium-loaded metals. Cold fusion may have been easily killed off as a serious scientific phenomenon, but laying its ghost to rest proved more difficult. The ghost even came to haunt 'hot fusion' research, with allegations of conflicts of interest being bandied.

Finally, a report published by US Navy researchers who had been performing investigations in their spare time over a period of ten years concluded that the effect was real, if not easily reproducible or explainable. In certain cells, significant excess heat could be generated, but not easily. Helium-4 was produced in cells where excess energy was also produced. Light-shielded photographic films positioned close to the cathode fogged over, indicating the presence of penetrating radiation. Control experiments using ordinary water show none of these effects. Besides, the rate production of helium-4 was directly correlated with the rate of generation of excess heat, and the amount produced per helium atom was similar to that expected for D+D fusion. Something highly unusual was going on by any reckoning, even if it wasn't room-temperature nuclear fusion.

It's tempting to conclude from this that cold fusion is, after all the hoo-ha, a real physical phenomenon. But, pace Bill Clinton, this depends upon what one's definition of the word 'real' is. Does a phenomenon have to be totally or partially reproducible to be real? As far as science is concerned, the answer is 'totally'. Reproducible phenomena imply reproducible and well-understood conditions, which then gives the theorists something to get their teeth into. Then, given a decent theoretical model, one can then prove it by extending it to other circumstances and conditions. Cold fusion science is nowhere near this state of maturity.

So, the question of whether cold fusion actually exists is edging towards a resolution, but isn't there yet. It hovers in a state of limbo, the vast grey expanse of 'reasonable doubt' between disproof and proof. Scientists hate ambiguity and so have split into two camps: the vast majority thinks it's a load of rubbish, whereas the believers can continue to console themselves with the ongoing trickle of seemingly positive results. Whatever view one takes on the phenomenon, it's hard to dispute that Fleischmann and Pons certainly did themselves no favours, tripping over their own egos when they tried to steal a march on competitors by dodging the peer review process and, in so doing, neglecting to carry out some basic checks of their own methodology. If anything concrete has emerged from this saga, it's that, in science, one bucks the system at the peril of both one's own reputation and the credibility of the research itself. And, in taking on scientific orthodoxy with extraordinary claims, you need not just extraordinary but extraordinarily reproducible evidence.