I am trying to figure out a problem. When hydrogen and nitrogen are mixed, the entropy of the gas tends to mix them despite the large difference in mass while exposed to the acceleration of gravity. As I visualize it at a micro scale the gas molecules are translating at a velocity represented by its individual kinetic energy, these molecules collide with adjacent molecules of whatever type and find a new trajectory and kinetic energy from the properties of the collision. Due to the random nature of the collision the new trajectory can be in any direction this is the direct cause of the mixing of gasses, what I want to understand is: is my understanding of this effect correct? At what point does the effect of gravity or acceleration overcome the effect of entropy and cause the separation of the gasses? And at a micro scale how would that be represented? as an intellectual analysis of the problem gives me, I believe that this is the mechanism that will separate the gasses: a molecule of nitrogen with a as a result of a collision velocity of x1,y0,z0 when exposed to acceleration x0,y1,z0 the new result of the collision as exposed to acceleration would be (if I used the right method) x1,y.5,z0 this would show that if the acceleration is high enough that there will be a point that the normal distribution of the vector would favor a separation despite the mixing tendency of the collision. Now as I think about it in a macro scale I am not familiar enough with the units of measure as applied to thermodynamics to produce any methods to solve what I have described in the text above without much effort. But I am still thinking about it. Any education, thoughts, advice, and ridicule will be appreciated. Thank you.