[molpro-user] How to calculate accurate spin-orbit splitting in sodium

Peterson, Kirk kipeters at wsu.edu
Fri Sep 13 19:22:03 BST 2013


Dear Gunter,

my apologies, I neglected to see that you are actually already correlating all the electrons in your MRCI calculations.  Adding some tight p functions to the basis set might help.

-Kirk

On Sep 13, 2013, at 9:54 AM, "Peterson, Kirk" <kipeters at wsu.edu> wrote:

> Dear Gunter,
> 
> you could certainly try correlating the outer-core 2p electrons (and adding some extra tight p functions to the basis set). This should increase the splitting.  While Molpro can't actually do it,
> it might be that the Gaunt interaction is needed for the most accurate result (check out the Dirac program).
> 
> best wishes,
> 
> -Kirk
> 
> On Sep 13, 2013, at 5:38 AM, Guenter Klatt <Guenter.Klatt at pci.uni-heidelberg.de> wrote:
> 
>> Dear Molpro users,
>> 
>> I'm having a problem with the result of a simple test calculation on a sodium 
>> atom. From my understanding the 2P state of sodium is split by spin-orbit 
>> coupling, which leads to the famous Na doublet at 16,956 cm-1 and 16,973 
>> cm-1,
>> i.e. a splitting by 17 wavenumbers.
>> 
>> My test calculation for the Na atom at CASSCF/MRCI levels yields  splittings 
>> of only 11.8 cm-1 (CAS) and 10.3 cm-1 (MRCI), respectively. Could you please 
>> suggest ways of improving agreement with experiment? I have attached my input 
>> file below.
>> 
>> Thanks very much for any help,
>> 
>> Gunter Klatt, University of Heidelberg
>> 
>> ***,SO calculation for the Na atom
>> geometry={na}
>> basis={spd,na,vtz}                                 !use uncontracted basis
>> 
>> {rhf;occ,3,1,1,,1;wf,11,1,1}                      !rhf for 2S state
>> 
>> {multi;occ,3,2,2,0,2                                  !casscf
>> frozen,0,0,0,0,0
>> closed,0,0,0,0,0                                            
>> wf,11,1,1;wf,11,2,1;wf,11,3,1;wf,11,5,1}        !2S and 2P states
>> 
>> {ci;wf,11,1,1;save,3042.1;noexc}                 !save casscf wavefunctions
>> {ci;wf,11,2,1;save,3062.1;noexc}
>> {ci;wf,11,3,1;save,3072.1;noexc}
>> {ci;wf,11,5,1;save,3082.1;noexc}
>> 
>> {ci;wf,11,1,1;save,4010.1}                  !mrci calculations for 2S state
>> {ci;wf,11,2,1;save,4042.1}                  !mrci calculations for 2P states
>> {ci;wf,11,3,1;save,4062.1}                  !mrci calculations for 2P states
>> {ci;wf,11,5,1;save,4082.1}                  !mrci calculations for 2P states
>> 
>> lsint                                                   !compute SO integrals
>> 
>> text,2P states, casscf
>> {ci;hlsmat,ls,3062.1,3072.1,3082.1;print,hls=2}   !Only 2P states, casscf
>> 
>> text,2P states, mrci
>> {ci;hlsmat,ls,4042.1,4062.1;print,hls=2}          !Only 2P states, mrci
>> 
>> text,2S, 2P states, casscf
>> {ci;hlsmat,ls,3042.1,3062.1,3072.1,3082.1;print,hls=2} !All states, casscf
>> 
>> text,2S, 2P states, mrci 
>> ci;hlsmat,ls,4010.1,4042.1,4062.1;print,hls=2} !All states, mrci
>> ---
>> _______________________________________________
>> Molpro-user mailing list
>> Molpro-user at molpro.net
>> http://www.molpro.net/mailman/listinfo/molpro-user
> 
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user




More information about the Molpro-user mailing list