Additional basis function on the ghost atom

Daniel Svozil svozil at uochb.cas.cz
Thu Sep 25 13:44:42 BST 2003

```Hi,

I need to calculate dipol-bound DB anionic states using Molpro, but I can't figure
out, how to perform such a computation. I set up simple system (H2) and
I am comparing the results with those from Gaussian, which I already can
use for the DB calculation.

Here is the simplified Molpro input:

geomtyp=xyz
geometry={
3
H2
H1       0.000000000    0.000000000   -0.500000000
H2       0.000000000    0.000000000    0.500000000
Q       0.500000000    0.000000000    0.000000000
}

basis
DEFAULT=6-31G
S,Q,0.7;
S,Q,0.07;
S,Q,0.007;
P,Q,0.7;
P,Q,0.07;
P,Q,0.007;
end

uhf;
wf,3,1,1;

So what am I basically doing is I am putting the set of SP functions
with the 0.7, 0.07, 0.007 coefficients on the dummy center and I specify
the number of electrons=3 in the wf card. This is similar approach as
for Gaussian, where the basis set is augmented in th ghost atom Bq:

#UHF/6-31G massage

H2

-1 2
H       0.000000000    0.000000000   -0.500000000
H       0.000000000    0.000000000    0.500000000
Bq      0.500000000    0.000000000    0.000000000

3 Nuc 0.0
3 SP 0.7
3 SP 0.07
3 SP 0.007

The problem is, that I am getting completely different results from
Gaussian (E=-1.0816352H) and from Molpro (E=-0.61054645). The results
from Molpro are actually as if the dummy center Q was not present at
all.

It seems to me that it is not possible to place basis functions on
the dummy center atoms in Molpro (as it is also not possible to do this
in Gaussian with XX dummy atoms, one has to use ghost atom Bq instead).

Is there any equivalent to Bq atoms in Molpro? How should I perform the
above described calculations?

Dan

--
Daniel Svozil, PhD
Institute of Organic Chemistry and Biochemistry
and Center for Complex Molecular Systems and Biomolecules
http://www.molecular.cz/
Czech Republic

phone: +420-2-20 183 263

```