[molpro-user] simultaneous treatment of states with different spins and symmetries in multi-state CASPT2 and MRCI‏

Peter Knowles KnowlesPJ at Cardiff.ac.uk
Fri Oct 25 07:38:06 BST 2013

The simple answer is that in both CASPT2 and MRCI, a linear equation system is solved. This means that different symmetries (spatial or spin) are not coupled and therefore separate calculations can be done for each symmetry. This is in contrast to state-averaged MCSCF where the energy functional whose minimum defines the orbitals comes from several states of different symmetry.

There is, however, a slight complication: the contracted multireference pair basis that defines the wavefunction is dependent on the reference function, and sometimes one wants to explicitly define a basis that is the union of those arising from several reference states of the same or different symmetries. See http://www.molpro.net/info/2012.1/doc/manual/node280.html and the relevant literature.


On 24 Oct 2013, at 19:11, Yang Yang <yang13404 at hotmail.com> wrote:

> Hi, Dear Molpro community,
> is there a way to treat states with different spins and symmetries in multi-state CASPT2 and multi-state MRCI simultaneously? If so, how to specify them? I tried to use the same command as CASSCF but it didn't work:
> {rs2
> mix=3
> wf,28,1,5
> state 1
> wf,28,2,5
> state 1
> wf,28,2,3
> state 1}
> it turns out the calculation always defaults to only one symmemtry and one spin (i.e.,wf,28,1,5), even if different spins and symmetries were already used in the preceding CASSCF calculation.
> thanks a lot
> sincerely
> yang
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user

Prof. Peter J. Knowles             
School of Chemistry, Cardiff University, Main Building, Park Place, Cardiff CF10 3AT, UK
Telephone +44 29 208 79182 Email KnowlesPJ at Cardiff.ac.uk 
WWW http://www.cardiff.ac.uk/chemy/contactsandpeople/academicstaff/knowles.html

More information about the Molpro-user mailing list