The term cold fusion was first coined in 1983 in relation to work on muon catalysed fusion, but entered the public consciousness six years later during the storm of publicity that surrounded Fleischmann and Pons' claim of excess heat production during an electrolysis experiment. Those results are now generally discredited by the scientific community, but retain popularity with those on the fringe, where even more outlandish claims are advanced. The announcement is also derided for being 'press conference science'- attempting to get the jump on a rival team at Brigham Young University, the University of Utah scientists released their findings to the press before going through a thorough process of peer review. But amidst all the backlash and accusations of pseudoscience, it's worth recalling that both groups had been accepted for publication by Nature and all involved were, at the time, respected scientists. So what went wrong?
Fleischmann, Pons and cold nuclear fusion in condensed matter
The original experiment consisted of performing electrolysis in heavy water with Palladium electrodes. The supply of electricity allows the water molecules to be broken into ions, and thus for a current to flow from one electrode to the other. This process is not completely efficient, so heat is generated as a by-product; surrounding the experimental apparatus with a calorimeter, this heat generation can be measured. So far, nothing controversial- the movement of heat, charge and associated chemical changes are readily explained and balanced by conventional theories. But in Fleischmann and Pons' experiment, the heating rate exceeded, by around 10W, that which would be expected in line with those theories. The inability to explain this through thermodynamics, electrical theory or chemistry therefore led the researchers to a controversial conclusion- that the explanation must be nuclear, with fusion in the heavy water being responsible for the additional power output.
How realistic, then, is such a claim? There are two major reaction paths for deuterium fusion, of roughly equal probability. They are as follows.
d + d --> t + n +4.0MeV
d + d --> 3He + p + 3.3MeV
Where are the neutrons?
Thus, if energy is being produced, then the same should be true of neutrons, tritium, helium, and preferably all three. The neutron output that would correspond to a 10W heating effect would be staggering, implying a production rate of over 1013 neutrons per second. To place that in context, a typical university Physics facility for students would not allow the use of sources churning out more than about 109 neutrons per second. The 'cold fusion' device would therefore have to be 10,000 times more potent- running unshielded, this would probably have offed the researchers long before they could announce any results. Moreover, such a rate simply wasn't observed, with the measured yield being around 104/s.
Where's the Helium?
Perhaps, however, the 50/50 distribution between reaction paths was inappropriate in this situation, and there was a disproportionate tendency (for some unexplained reason) to follow the Helium generating path instead. Moreover, unlike the sub-expectation neutron production, several research groups claimed to observe Helium formation. There's also a third possible reaction path, with a greater energy yield and also giving rise to Helium, but that is eight orders of magnitude less probable than the two main reactions. Ultimately though it's irrelevant- careful checks revealed that any observed Helium was the result of air leakage into the experimental setup.
The demise of a theory
With time, the evidence against cold nuclear fusion piled up: Neutron and Tritium production being ruled out by the BYU team, and erratic observations of excess heat ultimately being attributed to poor calorimetric techniques. Fleischmann and Pons both lost their jobs, yet somehow managed to secure (and spend) $30million from Toyota to continue their research in France. When that ran out, Pons admitted defeat, but Fleischmann returned to the UK and continues to work on cold fusion theory.
Muon catalysed fusion
But what of the original claimant to the cold fusion name, the technique now (presumably to maintain some distance) known as Muon catalysed fusion? Simply put, it works, and at room temperature, but, much like the Farnsworth Fusor, requires more energy to sustain than is produced, making it worthless as an energy source (although it makes a fine Neutron source if you have a use for such a thing). In essence, a muon behaves exactly like an electron, except it is around 200 times more massive. Thus Hydrogen with muons in place of electrons allows the nuclei to get around 200 times closer to each other, which drastically increases the probability of fusion: as this decreases exponentially with distance, so a 200-fold reduction in distance equates to a probability some 75 orders of magnitude higher than for regular Hydrogen.
This fusion process takes around 10-8 seconds, which seems blisteringly fast, until you take into account the lifespan of a muon- an average of 2.2 microseconds. Assuming no other limiting factors, this means that any given muon is only good for a couple of hundred fusions, or an energy output of around 4GeV. Sadly, there are limiting factors - although experimental rates of around 150 fusions per muon have been achieved - but more to the point, Muons aren't cheap. Producing them through Proton bombardment of Lithium or Carbon checks in at around 10GeV each, more than double the energy that can then be reclaimed.
So, despite more than its share of crackpots and controversies, cold fusion is not purely in the domain of sci-fi or pseudoscience. But if you're after a revolution in our energy supply, you'd be better looking elsewhere.
Reference
"How Cold is Cold Fusion?"- Janne Wallenius, Royal Institute of Technology (KTH),Sweden: Seminar at the University of Edinburgh Physics Department, 23/11/06.