Re: Ionization potential

Posted by Björn Roos on April 20, 2001 at 09:07:42:

In Reply to: Ionization potential posted by Rogerio Custodio on April 19, 2001 at 16:38:32:

Dear Rogerio,
The easiest question first: Yes you can use
fractional nuclear charges.
Follow the input instructions in SEWARD.

About Core ionized states. There is no real good
way to this with MOLCAS today. You might try
the following. In the CASSCF program you can
include a core orbital in the active space and
freeze it using the SUPSYM input (note that you
need to move the corresponding starting orbital
above all the inactive orbitals). If you now can
find the root that has this orbital singly occupied
you may be able to converge to the ionized state,
but with a non-optimized core orbital. This might
not be such a serious approximation becuase there
is not much relaxation of a core orbital anyway.
I don't know what happens if you do CASPT2 for
such a reference function. Why don't you
experiment a little on some small molecule and
see what happens? Note: if you have symmetry equivalent
nuclei, try to localize the hole to one atom. I
don't know if this is going to work, but it might
be worth a try. Good Luck!
Björn Roos

Follow Ups:

Post a Followup

Name:
E-Mail:

Subject:

Comments:
: Dear Rogerio, : The easiest question first: Yes you can use : fractional nuclear charges. : Follow the input instructions in SEWARD. : About Core ionized states. There is no real good : way to this with MOLCAS today. You might try : the following. In the CASSCF program you can : include a core orbital in the active space and : freeze it using the SUPSYM input (note that you : need to move the corresponding starting orbital : above all the inactive orbitals). If you now can : find the root that has this orbital singly occupied : you may be able to converge to the ionized state, : but with a non-optimized core orbital. This might : not be such a serious approximation becuase there : is not much relaxation of a core orbital anyway. : I don't know what happens if you do CASPT2 for : such a reference function. Why don't you : experiment a little on some small molecule and : see what happens? Note: if you have symmetry equivalent : nuclei, try to localize the hole to one atom. I : don't know if this is going to work, but it might : be worth a try. Good Luck! : Björn Roos