[molpro-user] Jensen basis sets in molpro ( pc-3 problem for oxygen)

[Gerald Knizia]

On Sunday 02 January 2011 00:34, Teobald Kupka wrote:
> I am trying to calculate water harmonic and anharmonic vibrations
> (VSCF and VCI) in molpro. It works fine with Dunning basis sets.
> Jensen pc-2 and pc-4 also works OK (pasted from EMSL in molpro
> format). However, both energies and frequencies calculated with pc-3
> (and also aug-pc-3) basis set are significantly worse.
> When I repeated calculations for H2 molecule all basis sets work fine
> but for O2 does not (pc-3 problem).
>
> Could it be the EMSL converter or molpro problem ?

We are pretty sure that the integrals in Molpro are okay...
If you would provide some additional information (e.g., a failing input) we 
might have a closer look.

In general: Two things to look out for when importing sets from EMSL are that
(a) sometimes the generated syntax is kind of bogus (e.g., c,1.000; lines; I 
have not seen real errors yet, however)
(b) the 'optimize general contraction' option is enabled by default, but this 
is something you usually do not want to have when using Molpro. Particularly 
for real generally contracted basis sets, using this option might lead to 
worse Hartree-Fock initial guesses than if actual AOs were used.
-- 
Gerald Knizia