[molpro-user] CAS with wrong configuration of HF

[Peterson, Kirk]

Dear Sunghwan,

what version of Molpro are you using?  I just tried with a cc-pVDZ basis set and it looks reasonable to me. I got an energy of -1938.60740959 for a singlet state with an orbital occupation of  19   9   9   3   9   3   3   0  .

If you don’t like the orbitals that are occupied, you can of course use the rotate command (put an orbprint,4 or so in the previous step) to attempt to reorder the orbitals for an initial guess.

best regards,

-Kirk


> On Nov 18, 2014, at 7:25 AM, SungHwan Choi <sunghwanchoi91 at gmail.com> wrote:
> 
> Dear all, 
> Hi, I am now trying to compute the energy difference between high spin state and low spin state of octahedral metal complex with CAS level. 
> 
> The problem that I am struggling is that the HF method does not give a proper configuration even in singlet state. The d orbitals are completely occupied and also some ligand orbitals are above them. It is difficult to generate the electronic structure with the correct configuration with HF. In this case, may I set the active space of CAS as fully occupied orbitals with 6 electrons? or is there any other way to make the correct electronic configuration with HF? 
> 
> The case that I am trying is quite simple one. Here is my target geometry
> 
> 13
> 
> Fe       0.000000    0.000000    0.000000
> C        0.000000    0.000000    2.003402
> C       0.000000    2.003402    0.000000
> C      -2.003402    0.000000    0.000000
> C       0.000000   -2.003402    0.000000
> C       2.003402    0.000000    0.000000
> C       0.000000    0.000000   -2.003402
> O       0.000000    3.187907    0.000000
> O       3.187907    0.000000    0.000000
> O       0.000000    0.000000    3.187907
> O      -3.187907    0.000000    0.000000
> O       0.000000   -3.187907    0.000000
> O       0.000000    0.000000   -3.187907
> 
> 
> Sincerely 
> Sunghwan Choi
> 
> 
> 
> 
> 
> 
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user