[molpro-user] erratic energies

[Aris Mitropoulos]

Dear Molpro experts,
 
With the 2010.1 version of MOLPRO, I try to compute at the
MRCI-vtz level a potential energy cut of the system Al4 + O2 at its quintet
overall state, which corresponds to the triplet ground states of each
fragment.  In this system CAS is too big
so I have to employ a reduced active space (12orbitals / 20 electrons), which
may be the cause of the following problem.  At and around the minimum of the energy curve the calculation gives reasonable
energies.  However, as the two fragments
separate, the energies become erratic and seem to depend on the  i n i t i a l  g u e s s.  For instance, at the
same geometry two calculations with a different initial guess give energies
differing by as much as 30 mH!  These two
calculations involve configurations with drastically different coefficients and
even different configurations.  Is it
possible to concatenate configuration files from various geometries and initial
guesses into one file and use this file with the “selection” option for all
geometries?  Could this alleviate the
erratic behavior?  I would appreciate any
ideas that could explain and correct this problem.  Many thanks. 
 
Regards,
Aris. Metropoulos
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20130711/5dc1e128/attachment.html>