There has been several unedifying discussions on Slashdot on fusion power, where a number of pseudoscientific ideas have come up again, and again. I've tried to rebut these, although perhaps a bit impatiently sometimes. Out of the layperson's standpoint, the whole matter probably appears rather confusing. So I feel I'll use some journal space here to try and give the 'mainstream' scientific view.
(Addendum 9/3: And now, another one... Just the header is misleading "IEEE warming up to cold-fusion".. IEEE as an organization isn't warming up to anything. It's a report in their newsletter. Besides which, the IEEE is an electrical engineering organization and therefore generally not a good place for a qualified view of nuclear physics.
Same suspects as usual, really, like SRI. (a well known-psudoscience 'research institute'.. they did experiments which 'proved' Uri Geller was for real. Only problem is.. he wasn't. Their tests were found to be very flawed as well. Anyway, it's the usual story.. they'll be making a real breakthrough reeeaaal soooon.)
First, an aside: Plasma physics is not my area. I do chemical physics. I will therefore not get into details, as I don't want to get anything wrong and end up contributing to all the misinformation floating around
Ok, here we go. Some epistemology: One must accept, in order to get anywhere in knowledge, that there is an objective universe around us, and that things about this universe can be known with varying degrees of certainty.
Some of the things we are most certain of are the laws of thermodynamics. The great strength of these is that they are not really bound to any particular physical model. Most physicists would agree that if there is any part of physics which is completely valid, it would be thermodynamics.
(Einstein, a paradigm-shifter in his own right, believed so too. Although I don't generally appeal to authority, I think it seems appropriate here.)
The first law of thermodynamics is Preservation of energy.
If a system is isolated (i.e. not exchanging any energy or matter with it's surroundings) then its energy is constant. Most people accept this as true. It not only has been validated again and again through the history of physics, it also suits 'common sense', which is always a good thing if you want an idea to be accepted.
(Quantum mechanics, on the other hand, doesn't really follow 'common sense', which is what makes it difficult to accept for many)
Ok, and now to the debunking:
Zero-point energy, what is it, anyway?
Well, to begin with, there are several things in physics which are called 'zero-point effects'. But I'll stay with the one which most people are referring to, namely vacuum zero-point energy.
You see, quantum mechanics has a law called the Heisenberg Uncertainty principle. It's very well known, and usually stated as:
dP*dx (smaller-than) h-bar
Which says that the uncertainty of momentum (P) times the uncertainty of position (x) is smaller than a constant (h-bar). This means that if the momentum of a system is precisely defined, then the position must be poorly defined, and vice-versa.
Now, these aren't the only properties like that. In fact, all observable properties are like that. In the case of energy you have:
dE*dt (smaller than) h-bar
Which means that time and energy relate in the same way.
This means that the energy of a system can fluctuate a lot, as long as they do so for a very short time. And the opposite, the longer time a system stays in a certain state, the more precisely defined the energy of that state is.
(Raman spectroscopy is an example of this. Normally, molecules can only be excited to certain specific energy levels. In Raman, a molecule is excited to a non-existent energy level, but only for a very, very short time.)
So, it would actually violate quantum physics if the energy of a system was truely constant.
Here comes the important part: This does not violate the first law of thermodynamics..
Now for some weirdness: The vacuum has infinite energy. (Depending on how you define the vacuum.) This weirdness comes from the fact that energy is a relative property. There is no absolute 'energy', there is only energy relative something else. For instance, if you lift a ball in the air, you have increased its potential energy. But if the ground were then to move up and meet the ball so that it was no longer suspended, then it will have lost its potential energy without changing.
So, it's only meaningful to talk of "energy" as the energy of one thing relative another.
(Or by analogy: It's like money. If you make up a currency, say the "Slashdollar", saying that 'X is worth ten Slashdollars' doesn't mean anything. But if you say 'X is worth ten Slashdollars and Y is worth twenty Slashdollars', then it starts to mean something.)
So, what happens is that whatever you calculate in quantum electrodynamics is that you have the energy of the vacuum (E0) and the energy of your object (E1) and the total energy, E0+E1 is infinite. Then you do the same for some other object E2, and get another infinite total energy.
But the energy of E1 relative E2 is (E0+E1) - (E0+E2) = E1-E2, where the infinite vacuum terms cancel out. Mathematically, this is not very nice, and these problems caused quite some debate among physicists back in the 30's. But in reality, it works. We just accept that the energy of the vacuum is like that, and that whatever we do, it'll cancel out, so we can safely ignore it.
Again, remember: Energy is conserved here.
This is the part where I think the crackpots get confused. They seem to think that this "infinite energy" of the vacuum is a real thing which we somehow can get energy from.
In reality we can't, because it's really just not real. The vacuum doesn't have infinite energy in reality, it has an incalculable energy which turns up as a mathematical infinity. It is wondering what the energy of nothing is relative nothing. Or asking what zero divided by zero is.
(zero divided is always zero, right? But if you divide anything with zero, you get infinity, right? It's simply something which is not defined)
So, anyway, my point is that the first law is not violated by any of quantum electrodynamics or this 'infinite energy', since it's simply not real.
These "infinite energy from the vacuum!!" guys are really just trying to sell the classical Perpetuum Mobile in a new wrapper. And it's just as impossible this time around. But like all PM-builders, they try to obscure that fact as much as possible.
Fusion occurs when the electrostatic repulsion of two atomic nuclei is overcome by the attractive Strong Force. The strong force is strong, hence the name, but it is also very, very, very short ranged.
At a distance of about a femtometer (1E-15 m) from the nuclei, it's practically zero.
Suppose we were to bring two protons (hydrogen nuclei) close to eachother? Just to be nice, I'll put the distance at a long 1E-13 m from the nuclei. How much energy does it take to get the two protons that close?
Coulomb's law gives us U = k*(q^2)/r, which means an energy of 2.3E-15 Joules. Not much, it might seem, but it's quite a lot for a proton!
Divide this by Boltzmann's constant for an ballpark figure of the thermal energy this corresponds to: 167 million degrees!
Now, the last step there is a bit nasty, since obviously you don't need all the protons to fuse at once. (Unless you're building an H-bomb, which does reach that neighborhood of thermal energy.)
But it does illustrate what I want to point out: That it takes a lot of energy to get two nuclei that close. And as the first law of thermodynamics dictates, that energy has to come from somewhere. And none of the 'cold fusion' guys have been able to pointed out where this energy should come from*, or how it works.
The important thing is this: it doesn't work. If it did, the entire physics world would instantly start studying and reproducing whatever experiment did it, and try to figure out how it works and why. It'd be huge thing.
But the way things are today, noone has demonstrated in theory why cold-fusion should work. And noone has come up with an real (i.e. reproducible) experiment which works either.
(*An active reader might point out that you could use the uncertainty principle I stated earlier, to borrow some energy and give it back once fusion started. Unfortunately, that takes too much time. Nuclei don't usually move that fast, if they did, the point would be moot, because they they'd probably have enough momentum to fuse anyway.)