[molpro-user] MRCI problem

[David Wilson]

Hi,

I've been having some problems with SA-MCSCF / MRCI geometry 
optimization calcs, which is possibly a program bug (I'm using 2002.6 
at apac.edu.au). Or maybe the input requirments have changed from 
2002.3 to 2002.6? It seems that no matter what wave function I specify 
in the CI calc, that the wave function that used for the geometry opt 
is the first listed in the SA-MCSCF calc. For example,

multi;
wf,23,1,1;wf,23,2,1;wf,23,3,1;wf,23,4,1;
ci;
wf,23,2,1;
optg;

will optimize the geometry of the wf,23,1,1 state! Similarly,

multi;
wf,23,2,1;wf,23,3,1;wf,23,4,1;
ci;
wf,23,4,1;
optg;

will optimize the geom of the wf,23,2,1 state. It seems that the MRCI 
wave function is not being stored/recognised by the optg code?

I've attached an example output illustrating this problem. Running 
2002.3 I don't see this error.

Best wishes
Dave

------------------------------------------------------------------------
David Wilson, PhD
Postdoctoral Fellow
Department of Chemistry      Ph:     +47 2285 5432
University of Oslo                Fax:    +47 2285 5441
Postbox 1033 Blindern         Email: d.j.wilson at kjemi.uio.no
N-0315 Oslo, Norway          www: http://www.uio.no/~davidwi