Why Bubble Geometry Doesn't Explain Copper (But Might Explain Batteries)

July 2026

Generated via Grok.

If the equilibrium bubbles that hold atoms together really do pack into geometric structures, a natural next question is whether some packings offer more direct flow paths than others — and whether that maps onto which real materials conduct well. It’s a real, calculable question, worth testing properly rather than assuming the answer. If the answer were yes, SC/BCC/FCC channel-openness would line up with which metals actually conduct or order magnetically. It doesn’t.

The calculation. For simple cubic (SC), body-centered cubic (BCC), and face-centered cubic (FCC) sphere packings, the maximum open straight-channel radius was computed along the standard low-index crystallographic directions. Cross-checked two ways: against hand-calculation for the simplest case, and independently against the real, established ion-channeling literature (the physics behind Rutherford backscattering). Both checks passed. FCC’s most open direction really is ⟨110⟩, matching decades of real channeling data for copper, gold, and silver; SC is most open along ⟨100⟩; BCC along ⟨100⟩ and ⟨111⟩. The geometry itself is correct.

Ruled out, decisively, for electrical conductivity — and the comparison doesn’t even require ranking which packing is “most open.” HCP metals — zinc, cadmium, titanium, cobalt — have the identical packing fraction and coordination number as FCC: identical local bubble geometry, in every sense this hypothesis cares about. Yet FCC’s copper, silver, and gold are the best elemental conductors there are, while the HCP metals are not comparably good. If open channel geometry were the reason, that shouldn’t happen. It does. Lead is FCC too — the same packing family as copper — and nowhere near it as a conductor, which kills the “FCC openness means good conductor” reading without even needing the HCP comparison. The real, established reason: conductivity comes from electron band structure — a free-electron-like band, low effective mass, weak electron-phonon scattering — not geometric channel width. Conduction electrons are delocalized quantum waves with mean free paths of tens of nanometers, far larger than the spacing between atoms. They don’t pick a path through the gaps the way a classical particle would.

Ruled out the same way, even more cleanly, for magnetism. The three room-temperature elemental ferromagnets — iron, cobalt, nickel — span three different crystal structures: BCC, HCP, and FCC. More decisive still, iron’s magnetic ordering is structure-dependent — ferromagnetic in its everyday BCC phase, not a simple ferromagnet in its high-temperature FCC phase, and fully suppressed in the high-pressure HCP phase — same element, same electron count, different structure, different magnetic behavior entirely. The real mechanism is the Stoner criterion: a material goes ferromagnetic when the density of electron states at the Fermi level, times an exchange-interaction strength, exceeds 1. Iron, cobalt, and nickel satisfy it because their d-electron bands are narrow and partially filled. Channel geometry has nothing to do with it. (Why a magnet keeps its order once formed is a separate question — domain structure and pinning, not the origin of the order itself; see Why Some Magnets Last and Others Don’t.)

A real geometric factor for magnetism does exist — just not this one. The Bethe-Slater curve plots magnetic exchange strength against the ratio of interatomic spacing to d-orbital radius, not channel openness. It’s negative (favoring opposite-spin alignment) for chromium and manganese, positive for iron, cobalt, and nickel, and falls off again toward the noble metals — which is exactly why manganese, despite having a large atomic magnetic moment, orders antiferromagnetically instead of ferromagnetically. This gives a real, measured mechanical lever: pressure changes interatomic spacing directly, moving a material along this curve. Iron, cobalt, and nickel’s Curie temperatures all drop under pressure, and iron loses its magnetic moment entirely at high enough pressure. That’s a genuine, testable, structure-based prediction — just not the one this hypothesis started with.

Where the mechanism survives. Ions are heavy particles, closer to classical particles moving through real physical gaps — much closer to what the bubble-matrix picture actually describes than delocalized conduction electrons are. Open channel geometry is a real, established factor in ionic conductivity: solid electrolytes, battery materials, and zeolite-like framework conduction all depend on exactly this kind of channel structure. That’s the defensible target for this mechanism.

Bottom line: a real hypothesis, tested properly, and cleanly ruled out for the two things it was first aimed at — electrical conductivity and the origin of magnetism — with the real, established mechanisms named in both cases. Not a failure to hide: a real result, pointing at two genuinely promising, more precisely matched directions instead — ionic conductivity, and pressure as a measurable lever on magnetism.