Your concept of electron actions in a collection of micro/nano particles needs to be adjusted.
Electrons and holes are oscillating like a large ball and a small ball connected by a spring. This is dipole oscillation. Heat feeds the strength of this oscillation. When the electrons move far enough, they move off the particle boundary and spin in a vortex current. Fit your thinking into this experimentally proven framework. On Sat, Jul 27, 2013 at 11:22 AM, Eric Walker <[email protected]> wrote: > I have two questions for Robin (or anyone else who might know). First, > some context. > > In a metal lattice, normally the electrons are highly localized around the > lattice sites. In the example of palladium, if you look at a density map > of the electron orbitals, they are highly clustered around the palladium > atoms, and there is a very low density in the interstitial sites. In > relatively unloaded palladium, hydrogen occupies the octahedral sites. I > recently read a calculation looking at the electron density, and it > suggested that almost none of the electrons were located in the vicinity of > the octahedral sites. The implication was that there was little to no > shielding. They were arguing against a reduction in the Coulomb barrier > for two deuterium nuclei occupying adjacent sites (they might have been > considering the highly loaded case, where some tetrahedral sites are also > occupied; I don't remember). They found the effect of electrons in the > host lattice to be negligible on the tunneling rate, if I understand the > calculation. That was a calculation for equilibrium conditions (e.g., not > much going on). They thought they could draw conclusions on the basis of > it for nonequilibrium conditions as well. > > In this context I note that the ionization cross section is a function of > energy. At higher energies, you're likely to get ionization, and at lower > energies you're likely to get an excited (Rydberg) state, but not > ionization. When there is no ionization, the electron steps up into an > orbital of increased reach -- a Rydberg state. It goes out further. Now > imagine a host lattice with a population of Rydberg-excited electrons. The > density map of the lattice would look somewhat different. There would > still be the tightly bound inner shell electrons clustered around the > lattice sites, but in the rest of the lattice, the density would be > evened-out somewhat. Perhaps there would be an overall increase in the > level of screening, even out in the octahedral sites (note that nickel is > also fcc). Presumably, as the energy of the hydrogen population is > increased there will be increasing Rydberg excitations of the lattice site > electrons, and perhaps this would not necessarily get to the point of > ionization because the energy is still quite low in relative terms. > > My two questions for Robin (or anyone else): > > - Do you have a sense of how tunneling would be affected at the > locations that hydrodgen/deuterium pairs are likely to be if a significant > population of nickel electrons were excited into Rydberg states? I think > we can assume Ron's mechanism is also at play, but perhaps not. (If we get > set aside Ron's mechanism, we have gammas to deal with.) > - Is there a basic distinction between tunneling and Coulomb > penetration, or are they both ultimately reducible to the amount of time > that the two nuclei spend in proximity to one another? In the latter case > this would presumably be because they approach so close. > > One interesting point to add -- I am not arguing for the importance of an > fcc lattice. I think there are some very interesting things that could > happen with this screening phenomenon and Ron's mechanism in cracks, for > example (e.g., a dense plasma focus). > > Eric > >
