[molpro-user] Combination of symmetry and lattice

[Jan Mitschker]

Dear Molpro users,

Using Molpro 2010.1, I encountered some problems with the symmetry when 
a lattice of point charges is present.
I have attached a quite simple example (see below). Here, the 
arrangement of atoms has Cs symmetry which is detected correctly by the 
program.
Adding one point charge (which reduces the symmetry of the whole system 
to C1) does not change the symmetry detected. Of course, this can easily 
be circumvented by using the nosym keyword manually.

But the case of two point charges is very strange. Here, the point 
charge field has Cs symmetry, too. But now the results depend on whether 
symmetry is turned on or off. The Nuclear repulsion energy (including 
the interaction of the atoms with the charges) is the same for both 
cases but the Hartree Fock energy is completely different.

Is symmetry not supported when a lattice of point charges is present? 
The manual just says: "Symmetry is not supported for lattice gradients."

Many thanks,

Jan

==========

Here is the input file:

***,Test calculation
memory,1,m

geomtyp=xyz

angstrom
nosym
orient,noorient
geometry={
He  5. 0. 1.0
He -4. 0.  .500
He  1. -3. 1.
He  1.  3. 1.
}

lattice,infile=testlattice.xyz

hf;


==========

The input of the lattice file (testlattice.xyz) is:

Point charge field in Angstrom
2
    1.0  3.0 -1.0 1. 0
    1.0 -3.0 -1.0 1. 0

==========

I get the following results for the Nuclear repulsion and the HF energy:

2 point charges         C1      4.29533130      -11.42232579
2 point charges         Cs      4.29533130      -11.04730670
1 point charge          C1      3.22649898      -11.42142605
1 point charge          Cs      3.58833566      -10.64707187


-- 

Jan Mitschker, M.Sc.

Theoretical Chemistry
Institute for Pure and Applied Chemistry
Carl von Ossietzky Universität Oldenburg
Ammerländer Heerstr. 114-118
D-26129 Oldenburg

http://www.kluener.chemie.uni-oldenburg.de/