In terms of surface adsorption and surface plasmonics, in “Enhanced Epicatalysis” - if the surface of the catalyst were to naturally form into Casimir sized cavities, due to impact dynamics … and given the millibar pressure level of hydrogen, we would expect to see perhaps only one molecule of hydrogen bouncing around inside any cavity, on average. This would seem to be the ideal way for that atom to benefit optimally from virtual photon acceleration – especially if the metal matrix itself could be composed of the “competing metals” – such as with a mixed matrix instead of an alloy.
It would be ironic if the Moddel patented technology, using Casimir dynamics - which was generally deemed to be a failure, although some small effect was seen: http://www.sciencedirect.com/science/article/pii/S1875389212024959 was instead simply not carried out in an optimum way. Using too much hydrogen pressure is one way to ruin the effect. Which is to say that too many hydrogen atoms, forced into a proper Casimir cavity which will hold hundreds of them, could ruin the expected gain, and that is what happened in Moddel’s scheme ! An interesting way to engineer this particular kind of excess energy device (as a hybrid of Sheehan and Moddel) - for the enigmatic X-prize (which may never materialize) would be to pass a very thin hydrogen gas at millibar pressure through the porous metal, but to have the metal matrix composed of mixed metal powders (sintered), instead of an alloy of both (where one metal is tungsten and the other is a rhenium substitute such as nickel). Yes, the geometry is very specific – 2-12 nanometers. Higher or lower spacing is no good. In fact, buckyballs (C60) are just a bit too small to experience a Casimir effect, but some forms of CNT (nanotubes) can be part of a Casimir anomaly. For comparison purposes, a sphere of this size (diameter of 5 nanometers) could contain about 150-500 atoms.

