[molpro-user] SAPT-DFT and dkroll=1 option

[jacek klos]

Dear Molpro Users,
I was trying to calculate He-Zn interaction energy using
SAPT-DFT implementation in molpro 2006.2 version.
In my input I used AVQZ-DK basis and I had keyword
dkroll=1 to use Douglass-Kroll hamiltonian
for scalar relativistic effects.
SAPT-DFT results with dkroll=1 option are wrong, as can be
seen here in fragment of table of results:
   RAD     ELST        EXCH        IND         EXIND        
DISP        EXDISP      ETOT
11.00************   0.0007828-182.3092210  -0.0000534  -0.0130822    
0.0000548************

When I removed dkroll=1 from the input, results look OK and agree  
very well with RCCSD(T) interaction energy:

   RAD     ELST        EXCH        IND         EXIND        
DISP        EXDISP      ETOT
11.25  -0.0001165   0.0006433  -0.0000448   0.0000537  -0.0115956    
0.0000471  -0.0110128
11.50  -0.0000780   0.0004314  -0.0000295   0.0000353  -0.0099990    
0.0000319  -0.0096079

Does that mean, that SAPT as currently implemented in Molpro cannot  
work with Douglass-Kroll
based integrals?
Best regards
Jacek

__________________________________
Jacek Klos
Department of Chemistry and Biochemistry
University of Maryland
College Park, MD 20742 USA
__________________________________




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20070629/25bed4e1/attachment.html>