Hi, OI and Jamin, Many thanks for your reply. I also guess it is the intruder
state problem and MCSCF converged to another state. By inspecting the natural
orbitals, I found the converged active orbitals are different from the space I
defined in the input. Do you have any suggestion to avoid the intruder
states?
Thanks again
Sent to CCL by: "Ol Ga" [eurisco1]*[pochta.ru]
Dear Roy,
I suggest you to look closer on convergence process - how much changes the
energy during MCSCF procedure. If you observe significant large oscillation of
energy, your choice of active space does not correspond with chemical problem of
your interest. I think your MCSCF is plagued by intruder states problem. Hence,
you obtain different values of energy.
MCSCF is very sophisticated method in its applications. In general, Choice of
active space is not an obvious task. The usual advice for newbie in MCSCF is
"Look on orbitals".