I'd expect the insulating properties of a gas at normal temperatures and pressures to depend only on the molecular mass. That's 40 for argon, and 28 for nitrogen. This web page (https://www.pressglass.com/noble-gases-and-thermal-insulatio...) says the thermal conductivities are 0,026 W/(mK) for air and 0,018 W/(mK) for argon. 26/18 is about 1.44; 40/28 is about 1.43. So perhaps the conductivity is inversely proportional to the molecular mass? In that's true then it's not a "noble" gas you want for improved insulation, but a dense gas (so not helium). You should be using tungsten hexafluoride! OK, not really: see https://en.wikipedia.org/wiki/Tungsten_hexafluoride for why not.
The heat capacity of the gas depends on both mass and molecular structure. Atoms like argon have no structure, but molecules like O2 and N2 carry extra heat energy in their vibrating bond, so they have more heat capacity per unit mass.
That's a good point. Perhaps the result of that argon/nitrogen comparison was a coincidence, or perhaps heat capacity doesn't much affect thermal conductivity.
Here's an argument that heat capacity shouldn't affect thermal conductivity in a solid: imagine a block of material with a smooth temperature gradient through it and the temperature at any point remaining constant, so there is heat flowing through it. Now imagine you have some tiny beads of some other material with a super-high heat capacity. Distribute those beads throughout the block of material, preheating each bead to the correct local temperature before inserting it. Then intuitively it seems that neither any local temperature nor the rate of flow of heat through the block should change much. Yet the heat capacity of the new composite material is significantly higher than that of the original material. Does that argument make sense? I did physics at school but not since then!
Yes, I don't think they're strongly related in a solid.
But in a gas, heat transfer is mostly driven by convection. Gas near the hot surface heats up, rises, moves to the cold surface, falls, and repeats. So what matters is the specific heat capacity * the flow rate between hot and cold surfaces.
Thanks for that explanation. So the "thermal conductivities" I quoted above, if they don't take account of convection, are perhaps not even particularly relevant. The advantage of argon over nitrogen might be greater than those numbers suggest?
I'm not exactly sure how they choose which particular gas to use, but I'm sure the lighter gasses will leak more easily.
Conceptually, it's like a thermos bottle, and you'd think that you would want a vacuum, but apparently that's hard to keep intact on windows, so a noble gas between the layers works better.
Yes, but not any old noble gas. Helium and neon, for example, have higher thermal conductivities than nitrogen. So it seems like argon is better for the job as well as being cheaper.
(While every noble gas has roughly the same heat capacity by mole/volume. Perhaps some CFCs would have been even better for insulation, but bad for the ozone layer, of course, and rightly banned?)
