[molpro-user] Wrong orbital energy order using MCSCF

Peter Knowles KnowlesPJ at Cardiff.ac.uk
Wed Jul 18 23:14:00 BST 2012


It's not clear to me that there really is a problem here. Apparently what you have calculated and shown are the state-average natural orbitals. The energy numbers printed are the diagonal elements of a (non-diagonal) generalised Fock operator, and I see no particular reason to expect that the diagonal elements for some of the unoccupied virtual orbitals will necessarily be higher than the diagonal elements for all of the active orbitals. What you hope to have, and  likely do have, is the global minimum of the state-average energy functional. Of course whether your active space is adequate for describing the three lowest states in each symmetry should be investigated: try looking at a bigger active space, using natorb,,ci, to see if the extra orbitals are actually needed.

Peter

On 17 Jul 2012, at 10:31, Little, Duncan wrote:

> Dear Molpro users,
> 
> I am trying to optimize MCSCF orbitals for an N2+ calculation for use elsewhere. However, the orbitals for the B1U symmetry are not in energy order which is leading to missing states in the latter parts of my calculation (this latter part is not in molpro). I have tried rotating the orbitals but they switch back when I re-optimize them after the rotation. Rotating and then not optimizing them is giving me incorrect results. Changing the CAS is an option but not ideal as it exponentially increases the size of the latter part of the calculation.  
> 
> Has anyone else experienced this problem or know a solution?
> 
> My input is as follows: 
> 
> ccpv-QZ basis
> 
> {MCSCF
> CONFIG
> OCC,3,1,1,0,3,1,1,0;
> 
> !!!!DOUBLETS
> wf,charge=1,symmetry=1,spin=1;state,3;weight,1;
> wf,charge=1,symmetry=2,spin=1;state,3;weight,1;
> wf,charge=1,symmetry=3,spin=1;state,3;weight,1;
> wf,charge=1,symmetry=5,spin=1;state,3;weight,1;
> 
> 
> }
> 
> Example of problem: 
> 
>  Orb  Occ           Energy
> 
>   1.5  2.00000   -16.123087     0.992537 -0.000157  0.017897 -0.007742  0.001701 -0.001596  0.000105  0.002226 -0.000476  0.000245
>                                             0.000000 -0.000206  0.000000 -0.000611  0.000000 -0.000019  0.000000  0.000156  0.000000 -0.000071
>                                             0.000000  0.000000
> 
>   2.5  1.75592    -1.115031     0.006900  0.344553  0.283574  0.640772  0.331032  0.080486  0.152870  0.157003  0.006566 -0.003090
>                                            0.000000 -0.009470  0.000000 -0.003061  0.000000  0.000340  0.000000 -0.000912  0.000000 -0.000072
>                                            0.000000  0.000000
> 
>   3.5  0.02252     0.671863    -0.045094 -0.328442 -0.368122 -0.876173 -0.337843  0.298956  0.680599  0.721205  0.106564  0.011983
>                                            0.000000  0.012700  0.000000 -0.044164  0.000000 -0.002505  0.000000 -0.000834  0.000000 -0.000632
>                                            0.000000  0.000000
> 
>   4.5 -0.00000     0.069813     0.022204 -0.058912  0.288811 -2.453087-12.086730 -0.053870 -0.114967  2.250168  5.826679  0.018843
>                                            0.000000 -0.200370 -0.000000 -1.201562  0.000000  0.006444  0.000000  0.177451 -0.000000 -0.019938
>                                            0.000000  0.000000
> 
>   5.5 -0.00000     0.512569     0.022862 -0.134958 -0.291055  0.514570 20.560504 -0.036081  0.092508 -3.713471 -7.282255 -0.034450
>                                           -0.000000  0.316381  0.000000  1.359290 -0.000000 -0.012612 -0.000000 -0.209154  0.000000  0.018522
>                                           -0.000000 -0.000000
> 
> Many thanks,
> 
> Duncan Little
> 
> Physics and Astronomy,
> University College London
> _______________________________________________
> 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
Email KnowlesPJ at Cardiff.ac.uk 
WWW http://www.cardiff.ac.uk/chemy/contactsandpeople/academicstaff/knowles.html
Secretary: Alison Thompson, ThompsonAF at Cardiff.ac.uk Telephone +44 29 208 74805







More information about the Molpro-user mailing list