Ok - I will assume the role of vortician "shrink" for a while ... and will try to convey what it is that can be witnessed with angstrom bifocals, so to speak.
This attempt at verbalization of an iterative sequence of events at the nano level may not far removed from Horace's AEH, but it is a mash-up of Mills, electron degeneracy, 4-deuteron fusion (to 8Be -> 2He) and is different enough to be have a descriptive name, and will be called "musical chairs" fusion, for now. The visual image intended is that there will be one cell in the center of nanoparticle that cannot escape the musical squeeze of the THz "beat", and will pay the price (in identity loss). The idea is that the active nanoparticles, in the size range of a few hundred atoms of an alloy of nickel with ~16% palladium - when fully loaded, have two significantly different diameters over very short time intervals, based on kinetic pulsation and thermal vibration in the terahertz range: an "expanded size" and a "compressed size". We already know that near this size range, Planck's law begins to fail, and eventually jumps from a fourth power law to a fifth. Max Planck, even 100 years ago suspected that his eponymous law was not valid at scales comparable to the wavelength of thermal radiation. Recently verification of the shift in the power law has been reported. http://www.physorg.com/news168101848.html Essentially what this means is that there can be an exponentially greater range of physical vibration in particles of very small size. First additional assumption: There is a natural "preference" for paired hydrogen even in a confined matrix situation where the electrons are "shared". With a three to one (3:1) loading, then - and the assumption that the individual matrix cavities prefer multiples of two, on average half of the cavities will have 2 deuterons and half will have four. These pairs "slosh around" easily *except* in one and only one geometric spot. The fusion g-spot so to speak. This matrix loading situation will result in a tight fit even with 2:1 ratio, and in any case is continually being pumped so that it is relaxed a bit when the nano-particle is in the expanded size. When the geometry drops into the compressed size ... that is when the fusion is poised to occur. But more is needed. Second assumption: Spherical convergence in the nanoparticle. Spherical convergence is a rather amazing feature in that it multiplies energy towards the focal point. In the Farnsworth Fusor, for instance, spherical convergence in the thin plasma allows the fusion threshold to drop from 40,000 applied volts to 10,000 volts. There is effectively a 4:1 energy multiplication in the Furor - yet there it is in a gas, not a solid. The energy multiplication level should be much higher in "condensed matter". Third assumption: Overpotential = effective pressure - which can be multiplied in the center of the sphere. Fourth assumption: electron "orbital reduction" below ground state, even if it occurs in a markedly different way than Mills suggests (and it does) - this would operates to push the more tightly bound deuterons towards the center the nanoparticles. Conclusion: The effective pressure in the center of the nanoparticles will on occasion result in four deuteron fusion to beryllium, then immediately to two alphas. But prior to this, the dynamics of the process will demand that the electron-effective-radius has been reduced by some exothermic method (in multiples of the Hartree energy) and that they have already given up much of the expected energy of fusion by the time that they actually fuse. This is a QM "time reversal" situation and most of the energy to be released has already been transferred away as downshifted UV, by the time of fusion. In fact, QM fusion itself depends on a "quantum probability field" which has statistically risen - due to these interacting dynamics - to the point that fusion can occur at an exponentially enhanced rate, and without the normal indicia of fusion (gammas). Signed, Siggy the Trick-Cyclist Nano-vision Specialist