Posted by Ignacio on February 21, 2008 at 09:25:53:
In Reply to: EXCITED STATE CASSCF/CASPT2 posted by anupriya kumar on February 20, 2008 at 11:14:24:
: Why are the ground state energies different when I use CiRoot
: 1 1
: 1
: and
: CiRoot
: 2 2
: 1 2
: 1 1
Because the orbitals are different. In the first case the orbitals minimize the ground state, in the second case the orbitals are built to minimize the average of the ground and excited states energies (or something like that). It's more or less like using a different basis set. The CI coefficients are calculated for each state individually in both cases, though.
I'm having another problem. When I optimize the geometry for the ground state, it works fine if use the default values (CIRoot = 1 1 \ 1), but the energy goes up and the geometry is severely distorted if I include higher roots, still optimizing for the ground state (CIRoot = 1 3 \ 1, RlxRoot = 1). Is this the expected behaviour?