[molpro-user] BSSE with ECP potential/basis
Amit Sharma
greifsw at gmail.com
Fri May 6 14:21:26 CEST 2016
Hi all,
I want to calculate BSSE where one of the atoms has an ECP potential basis
on it. Should BSSE be computed differently for ECP? There is some
discussion in the literature about switching off the core but I am not sure
how to do that in molpro.
As a test I compute BSSE for two noble gas atoms with AVDZ basis and get a
BSSE of about 47 cm-1. Then I switch the basis on one of the atoms to be
def2-SVPD and I get BSSE which is twice that of the AVDZ result (about 97
cm-1).
However, the CP corrected interaction energy is approximately the same
between the two calculations. 400 cm-1 vs 403 cm-1, so maybe nothing
special needs to be done. Can anyone please help clarify this for me.
Thanks
Amit
***, Ar + he bsse
geometry={Ar
He Ar Rx}
Rx=5.0
!basis=avdz,he=avdz
basis=def2-SVPD,he=avdz
text,calculation for complex
hf !RHF for total system
ccsd(t)
e_Arhe=energy !save energy in variable e_Arhe
text,cp calculation for HE
dummy,Ar !make Ar a dummy
hf !RHF for He
ccsd(t)
e_he=energy !save energy in variable e_he
text,cp calculation for Ar
dummy,he !make he dummy
hf
ccsd(t)
e_Ar=energy !save energy in variable e_Ar
text,separate calculation for he
symmetry,x,y
geometry={he} !geometry for he alone
hf !RHF for He
ccsd(t)
e_he_inf=energy !save energy in variable e_he_inf
text,separate calculation for Ar
symmetry,x,y
geometry={Ar} !geometry for Ar alone
hf !scf for Ar
ccsd(t)
e_Ar_inf=energy !save energy in variable e_Ar_inf
de(i)=(e_Arhe-e_he_inf-e_Ar_inf)*tocm !compute uncorrected interaction
energy
de_cp(i)=(e_Arhe-e_he-e_Ar)*tocm !compute counter-poise corrected
interaction energy
bsse_he(i)=(e_he-e_he_inf)*tocm !BSSE for he
bsse_Ar(i)=(e_Ar-e_Ar_inf)*tocm !BSSE for Ar
bsse_tot(i)=bsse_he(i)+bsse_Ar(i) !total BSSE
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20160506/57f0b192/attachment.html>
More information about the Molpro-user
mailing list