I will write what I think the answer is. I'd like someone to confirm and maybe explain in detail.

I think that since surface hopping dynamics simulates movements on PES, it has to calculate gradients in each step. Since gradient for CASPT2 method can be evaluated only numericaly, CASPT2 quantum mechanical base for surface hopping dynamics would make calculations imspossibly slow.

Is this correct?

  • 2
    $\begingroup$ 1) CASPT2 analytic energy gradients are available in some codes (e.g. in MOLPRO). 2) A quick look at surface hopping dynamics reveals that nonadiabatic coupling terms has to be computed, which looks like an actual issue, since they seem to be not available in codes as of now. $\endgroup$
    – Wildcat
    Sep 27, 2016 at 15:04
  • 1
    $\begingroup$ I'm pretty sure nonadiabatic couplings are available in Molcas 8, since you can optimize conical intersections with Molcas. $\endgroup$
    – Yoda
    Oct 2, 2016 at 9:27
  • $\begingroup$ It probably comes a little late, but you might want to check this out! arxiv.org/abs/1701.02259 $\endgroup$ Oct 16, 2017 at 17:24


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.