Posted by Steven Vancoillie on January 09, 2015 at 10:54:23:
In Reply to: Root flipping in PCM /SA-CASSCF calculation posted by Stefano Crespi on December 10, 2014 at 17:41:57:
: I'm currently working on a rigid scan of a bond with SA-CASSCF (10 states, no symmetry) in gas and solvent, using PCM model for the last.
: When I try to do the solvent single point using coordinates that are far from equilibrium, the scf gives an error due to the root flipping: is it possible to avoid this from happening?
rasscf will try to follow roots in the CI, but in doesn't really work. There is a more reliable option to follow a certain root, i.e. CISElect (see documentation). In order for this to also work with PCM, you can additionally use the undocumented keyword CIRF, which takes a number that is the relative CISelect root you wish to follow.
You can email/post a working input and we can try to help you further.
grtz
Steven