I hope you're not trying to demonstrate something that doesn't occur. Well mixed air can't and doesn't stratify over a reasonable height.
If you compare the kinetic energy of the molecules (1.5kT/molecule or 1.5RT per mole, optionally ΔT if heat were to separate them) with the gravitation energy (mgΔh, optionally Δm if height shall separate chemical species and isotopes), you notice that the kinetic energy is stronger. It mixes the air more efficiently than gravitation separates it. This is the spontaneous evolution. It's the difficulty of uranium enrichment: to work against the mixing done by kT, one must replace gravitation by much stronger centrifugal force, and it still takes many steps and much time.
One exception is if you consider the whole thickness of our atmosphere. It suffices to separate light from heavy molecules. However, reasoning at it is difficult because our atmosphere is very far from thermodynamic equilibrium: its temperature varies much with the altitude. The atmosphere gets heat mostly from the ground and radiates it at higher altitude.
An other exception, that does occur, is when the air isn't well mixed. Then, big bubbles of warmer or colder air (or of natural gas, or carbon dioxide...) have collectively the difference of density and the force to work against kT, and they can buoy or sink - until kT mixes everything, which can take long.
So maybe you could refine your description of "mixed" liquids? The same happens for them. If they are really mixed to the molecule size, they won't stratify. Neither would that be a useful demonstration for the people watching, as it doesn't happen in the atmosphere. Surely we can propose non-miscible coloured liquids that will separate from a coarsely mixed state, maybe some dye floating in air.