# Inflating a bike tire with CO2, why does it go flat by the next day?

Typically when I get a flat I use a CO2 cartridge on the road. The tire is always flat and needs to be pumped up with air the next day. What is the difference in molecule size between CO2 and air (mostly N2 and O2) Why does it escape through the rubber siding of the tire or cap quicker?

• I presume you have fixed the flat before inflating with CO2? It might be nice to specify that... Commented Jun 15, 2017 at 22:57
• The answer is here and to do with not just diffusion but the solubility of gases in rubber. madsci.org/posts/archives/1998-05/895552329.Ch.r.html Commented Jun 16, 2017 at 5:48

Looking in the NIST Doc regarding permeability of gases through rubber this document does show the permeability to be higher for CO2 than air. See table 9 and table 10:

 Perm. of Air/Perm. of CO2 = 0.10


Read Section II for a discussion as to why - but it's an old document and may not contain a current understanding.

(Edit) See page 360 - the diffusion is driven by the partial pressure on the other side of the barrier. In the case of air, the other side of the barrier is identical. But with the CO2, the difference in concentration is much larger.

$\mathrm{CO}_2$ is a larger, heaver, molecule than $\mathrm{N}_2$ and $\mathrm{O}_2$. It is also smaller than $\mathrm{H}_2 \mathrm{O}$ (probably, due to the fact that water has a dipole moment, and carbon dioxide only has a quadrupole moment), though heavier.

You probably just have a slow leak. Does the same thing happen if you use an air compressor or an ordinary bicycle pump? See, when the $\mathrm{CO}_2$ gas expands it should cool, raising the pressure even higher when it warms back up. The opposite happens with a compressor/pump.

I would be surprised to learn that there was something special about the chemistry of rubber inner-tubes that allowed them to preferentially pass any of the gases discussed above in a significant way.

• This would be my intuition too, but what you think of the other answer? Commented Jun 15, 2017 at 23:58
• I upvoted it - it had real data, and a plausible mechanism (difference in partial pressure). Commented Jun 16, 2017 at 0:18
• Surprising, no? I've just gone out and bought a few $CO_2$ squirters to replicate the experiment. Commented Jun 16, 2017 at 0:19