determination of symmetry from z-matrix coordinates

[Dave Moore]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

	I am trying to do a PES scan of a two-molecule system using MOLPRO,
and there seems to be some problem with the symmetry determination
schemem in MOLPRO.  At small intermolecular separations, I can run
the calculations with any 'correct' Z-matrix for the system, and it
will work fine.  At longer distances, some Z-matrices work fine, but
others generate symmetry errors, either before ("REQUEST FOR SYMMETRY
ELEMENT <X,Y or Z> CANNOT BE HONORED") of after the calculation of
the integrals ("BASIS SET LINEARLY DEPENDENT OF WRONG S").  It seems
that the latter type of error can always be fixed by reconstructing
the Z-matrix with appropriate use of dummy atoms, however the former
type is harder to correct, and will always eventually come up again
once the intermolecular distance becomes large enough (howver not so
large that the intermolecular interactions are negligible,
unfortunately).  What is worse, using the same Z-matrix with the
NOSYMM argument causes generation of incorrect cartesian coordinates
... it seems that MOLPRO assigns both the X and Z axes as the main
symmetry axis.  I have also checked the "offending" Z-matrices using
GAUSSIAN, and it can use all of them with no problem, so it seems to
be a problem with MOLPRO.  I am wondering if anyone else has
experienced these problems and knows a fix or workaround.  Thanks,

Dave Moore
Chemistry Dept.
University of North Carolina at Chapel Hill

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 7.0.3 for non-commercial use <http://www.pgp.com>

iQA/AwUBO2AgkV5cg+YcPBRAEQKYVQCgvIwBhvMbzkfszdpLglGirhNEE6QAoNOc
sxb7hoNzYLKRIHFXjhnGfmWl
=lo9j
-----END PGP SIGNATURE-----