[molpro-user] RHF collapsed due to scfocc patch?
Tsuneo Hirano
hirano at nccsk.net
Mon Jan 3 15:57:19 GMT 2005
Dear Molpro users,
After I applied the scfocc patch to the MOLPRO 2002.6, MOLPRO becomes to
fail for an rhf open-shell job.
Even if I specified the electron configuration by using 'open' command as
--------------------------
rhf; occ,8,3,3,1; open,8.1,3.2;
sym,1,1,1,1,1,1,1,1,2,1;
wf,28,2,2;
--------------------------
the output still complains as follows.
-------------------------------------------------------------------
.......
Singly occupied orbitals: 8.1 3.2
Extra symmetries of mos in symmetry 1: 1 1 1 1 1 1 1 2 1
Orbital guess generated from atomic densities. Occupancy: 8 3 3 1
Molecular orbital dump at record 2100.2
ITERATION DDIFF GRAD ENERGY 2-EL.EN.
DIPOLE MOMENTS DIIS
1 0.000D+00 0.000D+00 -1381.46180982 1049.816576
0.000000 0.000000 -2.117146 0
sorry, failure in assigning trial configuration
assumed occupation
1- 4 8 3 3 1
assumed closed shell orbitals
1- 4 7 2 3 1
You should specify open shell configuration
more precisely using CLOSED or OPEN commands
ERROR EXIT
CURRENT STACK: MAIN.
---------------------------------------------------------------------
My machine is operated by SunOS-5.8, and this failure did not happen
before. I would like to know whether this is a machine dependent failure
or a bug in the patch.
Thank you in advance.
Tsuneo Hirano
P.S. My input file will follow.
***,CoH as Co+ 3d8s0 (3F) + H-(1s2)(1S), Roos(Co) and aVQZ(H)
memory,42,m;
gprint,orbital;
rcoh = 1.52 ang;
basis={
spdfg,Co,Bauschlicher-ANO;c;
spdf,H,avqz;c;
}
geomtyp=xyz
geometry={
2
CoH z-axis is the molecular axis
Co, 0.0, 0.0, 0.0
H, 0.0, 0.0, rcoh
}
!=== RHF initial MOs for Triplet-Phi (3d8s0, B1)
text,rhf, to obtain ^3Phi (3d8s0, B1) initial orbitals
rhf; occ,8,3,3,1;open,8.1,3.2;
sym,1,1,1,1,1,1,1,1,2,1;
wf,28,2,2;
---
**********************************
Tsuneo Hirano
Emeritus Professor of Chemistry, Ochanomizu University
Visiting Professor at Hiroshima University
e-mail: hirano at nccsk.net
**********************************
More information about the Molpro-user
mailing list