This proposal is also meritless. The problem isn't hydrogen vs vacuum, it is the ratio of each to the density of the oxygen/nitrogen atmosphere. <br /><br />Look at it this way: because the Nitrogen atom is 14 times heavier, and the Oxygen atom is 16 times heavier, given x many molecules of hydrogen vs equal molecules of N2 and O2, H2 molecules have approximately 2/29ths the atomic mass for a given atmospheric pressure than a normal atmospheric mixture of N2 and O2. The difference between 2/29ths and zero is so small that the mass of any structure you would need to maintain an absolute vacuum would be significantly in excess of that difference. It is thus most efficient to use Hydrogen.<br /><br />This also limits you: hydrogen lift gas is molecular H2, while most hydrogen in LEO is hydrogen ions, which are half the mass of H2 molecules. You would need to heat or otherwise turn your lift gas into a plasma of ionized hydrogen. Doing so would also allow you to minimize osmotic losses through the envelope, as you could positively charge your envelope, which would electrostatically bottle your lift gas and may allow you to fly with less mass inside than outside, while at the same time producing an electrostatic field outside that would repel hydrogen outside, possibly creating a larger virtual envelope described by the electrostatic field intensity.