Saturday, December 17, 2011

The physics.SE posts

I just came across this post on a site I hadn't noticed up to now:
I'm quite familiar with StackOverflow, a predecessor to this site.  StackOverflow is an essential resource for software developers, and I suspect that this physics site will be helpful here.

The replies to the above post make for interesting reading.  The original poster asked what went wrong in the experiments that underlay claims of cold fusion. Following are some of the main points that people responded with:
  1. Any sort of fusion will require that nuclei overcome Coulomb repulsion.
  2. Calorimetry of the kind Pons and Fleischmann were attempting is difficult to do right, and they were subtracting a large number from a large number.
  3. The results could not be reproduced by other scientists.
One of the longer replies was from Ron Maimon, a member of the site who is more sympathetic to the validity of the LENR experimental results.  He does not have a PhD in physics and is self-taught, and his reply only got one vote.  But his reputation is in the top 0.3 percent, and he seems to have some knowledge of the topic (a lot of which mirrors details in Mallove's and Beaudette's books).  He referred to the process described in Widom-Larsen as "inverse beta decay."  He thought this particular explanation was erroneous for several reasons:
Weak Force Neutron production:  The Widom Larson theory claims that it is possible for a proton and an electron to do inverse beta decay on the surface of a metal, where there are large local electric fields.  This is preposterous, because of the MeV difference in proton and neutron mass.  It requires millions of volts to accelerate an electron to enough energy to be able to do an inverse beta-decay, and such energies are not available on the surface of a metal.  Further, this theory will predict transmutations of plus/minus one mass unit predominanatly, which is not observed, and does not explain how a deuteron can absorb an electron.
While Maimon's explanation was interesting, his review of the existing theories, including Widom-Larsen, seems a little cavalier.  This article about work at Brookhaven National Laboratory talks about the formation of heavy electrons under extreme conditions, which act as though they were hundreds or thousands of times heavier than normal electrons.  My question is whether this detail is relevant to the point Maimon was making; what if, for example, the "coherence domains" Widom and Larsen talk about provide the necessary energy to accelerate the electrons to a mass that will trigger inverse beta decay?  And what does Maimon mean, exactly, when he says that "such energies are not available on the surface of a metal."  I would also like to know more about the objection concerning transmutations.  Presumably the objection about deuterons would not apply to an Ni-H system.

The most popular reply (nineteen votes) included a reference to a 1989 article in Nature by Koonin and Nauenberg discussing how much the electron mass would need to increase for cold fusion to occur at rates corresponding to claimed observations.  What is the connection to electron mass when we're talking not about inverse beta decay but, instead, normal fusion?  Or maybe they were referring to something similar to Muon-catalyzed fusion?

1 comment:

  1. Hi Eric--- the "coherence domains" are an impossible fantasy--- the collective behavior of objects in thermal equilibrium makes it that to get energy E into an object happens with probability exp(-E/kT). For room temperatures, the kT is 1/30 eV, and the energy to make a neutron is 1MeV, so that the ratio is 30,000,000, and the exponential is so close to zero, you would not be able to tell the difference.

    The experiments on cold fusion reveal that there are necessarily high energies involved, in particlar Mosier-Boss measurements at SPAWAR. The only thing you need to understand the phenomenon is that a K-shell hole (an electron ejected from the n=1 ground state), will accelerate protons and deuterons just as easily (more easily, because of the density of states) as it accelerates electrons (the electron ejection due to K-shell excitations is humdrum physics, it is well known for decates. Ejecting a proton or a deuteron at KeV energies is a humdrum corrolary).

    In a metal, the process of ionization + deuteron acceleration+ ionization of another atom means that the K-shells of atoms in deuterated or protenated metals are mixed up into bands. This means that K-shell excitations travel a while in a deuterated metal from atom to atom before decaying by X-ray. The semiclassical trajectory of a deuteron in such a band is to come very close to the nucleus at the moment of full deioniztion, and then to go far apart and ionize another atom, then go close to this new atom and deionize it, so on through the lattice. These bands are easy to describe using existing theory, but they have not been considered, since normally you don't excite X-ray bands at room temperature (you need ionizing radiation to excite these bands, and they only exist in hydrogenated or protenated metals).

    When two such delocalized deuterons at 20KeV meet near a nucleus, they have a very large chance to fuse, because their wavefunctions are concentrated within 100 fermis of each other due to the band dynamics, and the nucleus can pick up the fusion energy electrostatically (if you like, through a virtual photon, but the photon picture is not necessary). The transfer breaks the Pd nucleus and ejects a fast proton/deuteron/alpha/Be8/C12 or heavier nuclear fragment, and the alpha particle shoots off with tens of MeV's.

    The two ejected particles deposit the energy in the K-shell bands as they ionize the metal, and this easily produces a chain reaction. The theory I have just described explains all the trustworthy results in the field, so much so that I am confident that all cold fusion which is not easily explained by this mechanism is total bunk.

    The theory does not postulate miracles. Widom Larsen, in addition to requiring MeV's of energy to magically concentrate in a single particle, requires that the neutrons formed conspiratorially go together to the same nucleus many many times. That's not what neutrons do--- they don't have a brain--- they don't know to follow each other. The explanation is ridiculous, it is only getting traction because it claims to explain the otherwise mysterious experiments with an explanation that isn't fusion, which allows the anti-CF-ers to save face, saying "see, it's not fusion!" (but it is), while it also allows the pro-CF crowd to say "See, Ni-H transmutations are not bunk!" (but they are). The politics is friendly to the theory, but the science is not.

    The theory I propose explains the transmutation spectrum of the CF process, in particular the appearence of peaks at +/-8 and +/- 12 nuclear mass in the Pd/d systems, which look like magical multi-alpha uptake. These are caused by Pd nuclei breaking off a Be or C, which is then absorbed by another Pd.

    The Widom Larsen theory is a total failure, there is nothing positive to say about it, beyond the fact that it is nice that Widom and Larsen didn't ignore the experimental evidence. But ultimately, it is just politics, it's crappy science.