I have a surface (paraffin) and a vapor of atoms (alkali). When atom get the wall, it is sticked for some time, exchange it energy and momentum with wall and then release. I want to know the atomic velocity distribution (both in directions orthogonal and parallel to the surface) after atom releases. I guess the velocity distribution in direction parallel to the surface should be Maxwellian. But what about the orthogonal direction? More common, will atoms go out preferably in direction orthogonal to the surface, or preferably along the surface, with small angles between their velocity vectors and surface? If somebody know something about this topic (ideas, references et.c) --- could you help me?