Collisional excitation: selection rules for rotations?

[uselesslemma]

To be specific, I am referring to CO molecules undergoing collisions with H₂, resulting in CO transitioning to an excited vibrational state. I can't seem to find any rotational selection rules for collisions, meaning ΔJ could be essentially anything, as long as energy and angular/linear momentum are conserved. The rigid rotor approximation (i.e., rotational transitions for CO) selection rule is ΔJ = ±1, but apparently this only applies to electronic transitions. Can anyone shed any light on the reason behind this discrepancy?

 

[DrClaude]

The dipole selection rule is directly related to conservation of angular momentum. The photon having spin 1, its spin angular momentum has to be transferred to/from the molecule, hence the $\Delta J = \pm 1$ rule. Note that $\Delta J = 0$ (Q-branch) is also possible for some electronic transitions, where the angular momentum of the electronic state changes.

So there is no discrepancy between absorption/emission and collisions. Conservation of angular momentum is the only thing in play.