Update Diatomic molecule desorption logic

[Alexsp32]

trafficstars

Desorption frame logic in OutputDesorptionTrajectory now uses the H-H distance and H2-surface distance to determine the desorption transition state as the moment where the H-H distance decreases below the H-surface distance for the first time before desorption, where desorption is defined as the H2 centre of mass being above a distance threshold from the highest non-H atom in surface normal direction.

This was motivated by postprocessing of trajectories turning up a bunch of simulations where the desorption state was being incorrectly assigned. I'm now attempting to fix the logic another way while still avoiding any additional potential evaluations (which would otherwise be the smartest choice).

To show this works sufficiently well, I'll upload a comparison between transition states by this method and the potential energy later.