PC GAMESS/Firefly-related discussion club



Learn how to ask questions correctly

Re^4: New questions on spin-orbit coupling constant

Andrei V Scherbinin
scherb@classic.chem.msu.su

>From this description, I consider that the 3 frozen core MOs among MOs in
>$VEC1 for S0 and those among MOs in $VEC2 for T1 should be identical to each
>other, that is, the 3 frozen core MOs should have the same orbital energies
>and the same atomic orbital-coefficients. However, I donft know how to
>obtain such MOs.
>   I found that separate MCSCF-GUGA orbital optimizations for S0 and T1
>brought different frozen core MOs, although NMCC=3 was chosen both for S0
> optimization and for T1 optimization (initial GUESS MOs for S0 and T1 were
> RHF/SBKJC(p,d) optimized MOs).  What should I do to obtain the common frozen
> core MOs to S0 and T1?

Dear Dr Fumihito Mohri,

To my knowledge, there are two different ways to generate such MOs during MCSCF.

1) When CISTEP = GUGA in $MCSCF:

use the following flag in $MCSCF group,

         FCORE  = a flag to freeze optimization of the MCC core
                  orbitals, which is useful in preparation for
                  RUNTYP=TRANSITN jobs.  Setting this flag will
                  automatically force CANONC false.  This option
                  is incompatible with gradients, so can only be
                  used with RUNTYP=ENERGY. (default=.FALSE.)

So you first perform a normal (FCORE=.false.) MCSCF run to generate MOs for one multiplicity value,
then you put the converged MOs into the $VEC group, read them in during another MCSCF run for the
states of a different multiplicity, and set FCORE=.true. in the second run.

Finally you get two different sets of MOs for states of different spin multiplicities, but both
MO sets do have identical core MOs (optimized in the interest of the states from the first run).
In the follow-up SO run, you set both $VEC1 and $VEC2 groups.


2) When CISTEP = ALDET in $MCSCF:

use the following flag in $DET group,


         PURES  = a flag controlling the spin purity of the state
                  avaraging.  If true, the WSTATE array pertains
                  to the lowest states of the same S value as is
                  given by the MULT keyword in $CONTRL.  In this
                  case the value of NSTATE will need to be bigger
                  than the total number of weights given by WSTATE
                  if there are other spin states present at low
                  energies.  If false, it is possible to state
                  average over more than one S value, which might
                  be of interest in spin-orbit coupling jobs.
                  The default is .TRUE.

Then you obtain just a single common set of MOs for all multiplicities,
and use only $VEC1 in what follows.


Sincerely,

Andrei Scherbinin

[ Previous ] [ Next ] [ Index ]           Tue Jan 6 '09 2:12am
[ Reply ] [ Edit ] [ Delete ]           This message read 1156 times