Take for instance the Sun's gravitation potential at our present distance to it. Earth moves at 30km/s, which means each neutron has -7*10-19J gravitation+kinetic energy, or 200 times 300K. So temperature does not suffice to mix the isotopes evenly over the Sun's gravitation field, and the elements even less so.
Many more cases exist in astronomy at different scales: galaxies, planets... On Earth, the top atmosphere has slightly more of the lighter isotopes than the bottom, but Earth is a bit too light and the atmosphere too shallow for that. Jupiter would be far better.
At the microscopic scale, you can consider that an elastic collision between atoms tends to redistribute the kinetic energy evenly as a mean result, including between atoms of different mass, but at the same distance from the Sun or any massive object, atoms would need the same speed to stay on the same orbit. So the heavier atom falls down.
Though, this is a dynamic process. Whether gravitation had the opportunity to segregate the elements and isotopes depends on how long they take and how much time they had.
My quick-and-dirty answer is: over the early evolution of a gas clump, we're likely to find an age where density is still small enough that atoms go through the cloud with few collisions; this is when segregation has chances to occur. I didn't model these conditions; an astronomy student on a forum was interested at the project. Maybe this will be the bingo for him.
One example is the formation of a proto-star. Oxygen is known to vanish as the cloud begins to collapse, but reappears as the star is formed. That would fit.
Present models of gas cloud collapse want it to radiate heat, but evaporation (from the gravitation well) of the lighter elements and nuclides could help condensate (fall down) the heavier ones.
If isotope (and element) segregation has been important in our Solar system, it would rock very seriously many models that use isotopic abundance as a proof for the origin of varied objects.
Intergalactic hydrogen clouds have too little deuterium.
So in fact, tests exist already - we just need to check if segregation is a plausible explanation.
Or even... Stars with more elements beyond helium reside in the deeper gravitation wells of galaxies: the bulge, the globular clusters... Commonly explained through the age of the stars. If segregation contributes to explain the proportions, a little bit, or a little bit more... this would lead to re-tune the models a little bit, with many implications.
I suggested to measure the 2H/1H ratio at Jupiter's big spot, which is said to result from ascending wind. Maybe H2O or NH3 rotations and vibrations can be measured in the GHz or IR domains.
Segregation can result in anything between
- Small effect at few exotic objects with no other implication
- General mechanism at all scales that needs to re-tune all explanations of elements and isotopes abundances, and every deduction made from them.
Fun, isn't it?
Marc Schaefer, aka Enthalpy