alongd
commented
4 years ago I think there's a problem with the theories we're using. If we neglect it, we assume this mode behaves as an harmonic oscillator, but here you show that it does not. If we include it, we're asking the hindered rotor theory to treat a high barrier torsional vibration mode, I'm not sure how good this assumption is. Since it's problematic either way, I'm neutral to the change (I'm principally OK with the suggested modification). Perhaps we should consult Bill.
I am trying to benchmark some of our calculations. Currently, ARC will neglect a single-well hindered rotor with a high barrier. It is reasonable when the PES shape is close to cosine curve, so that RRHO approximation can work okay. But there are many cases where the shapes are much different, like the one below
(Around 120 degrees, there is a plateau-like region, but not a well). The barrier height is 50 kJ/mol
Whether or not treat this mode as hindered rotor, can cause a difference, especially at high-T. like below
(without [1,2] is not treating the torsional mode above as a hindered rotor which obtained from ARC, and the other curve is by manually add this torsional mode in Arkane calculation)
I think we should still treat it as a hindered rotor. Although this is a single example, but there are plenty of cases like this. If people agree with the idea, I will modify the corresponding part in ARC.