[molpro-user] SA-CASSCF for S1 state - getting wrong opt geometry

[Artur Nenov]

Hi,

I don't think that there is something wrong in your input. The thing with 
SA-CASSCF calculations is that you practically use the same MO coefficients 
to describe all states, that you average over (in your case S0 and S1). So 
that the algorithm searches for the best MO coefficinets, that describe both 
S0 and S1 as good as possible. This immediately means that the SA minimum you 
find for S1 (and S0 respectively) is not the real minimum. Recall that the 
CO2 minimum is linear on the S0 state and you will understand why you get the 
wrong angle. It's always better to do noSA if it is possible (unless you 
optimize CoIns). 

Artur Nenov
Ludwig-Maximilians-Universität München
Department Chemie
Butenandt-Str. 11, Haus E
81377 München
Germany


>Am Montag, 6. April 2009 21:24 schrieb Anastassia Alexandrova:
> Dear All,
> I am doing geometry optimization for the S1 state with SA-CASSCF. However,
> my results did not seem to make sense. So I tried to test my input on a
> known species: CO2 in the S1 state. But even thought the S1 optimized
> geometry is bent, as expected, I get a wrong valence angle (156 instead of
> 129). Could you please help me identify the source of error in my input
> (below)? Thank you very much.
> Anastassia
>
> ***,co2  opt of exctd state
> memory,300,m
> gprint,civector
> basis=6-31G*
> geometry={angstrom,nosymm;
> c;
> o,1,r;
> o,1,r1,2,a};
> r=1.155
> r1=1.156
> a=170.0
> {uhf; occ,11; wf,22,1,0}
> {casscf;occ,14;closed,8;wf,22,1,0;state,2;
> CPMCSCF,GRAD,2.1,spin=0.0,accu=1.0d-7,record=5102.1}
> {Force
> SAMC,5102.1}
> {optg,startcmd=casscf;method,slnr}
>  ---
>
> ---------------------------------------
>      Anastassia Alexandrova, Ph.D.
>            Yale University
>        Department of Chemistry
>          225 Prospect Street
>        New Haven, CT 06520-8107
>          Phone: 203-432-6068
>           Fax: 203-432-6068
>     anastassia.alexandrova at yale.edu
> http://zarbi.chem.yale.edu/~anastassia/
> ---------------------------------------
> Go one more mile...
>
>
>
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user