error in bond dissociation energy of gas phase molecule

[Matt W]

Just a quick comment - describing these molecules with pseudos is
difficult, probably much harder than many solid state situations, with
less chance for error cancellation.

For instance, from the NIST comp chem data base (http://
cccbdb.nist.gov/), the LANLDZ2 pseudo + basis set gives

H2O2 -> 2OH = 2.01 eV

I guess with the weird oxidation state and small O-O distance (core
relaxation?) errors are likely...how do other semi-local pseudos do?

Matt

On Apr 7, 1:21 pm, Jun <chen... at gmail.com> wrote:
> > In the inputs you gave you use TZV2P-GTH as basis set which is *not*  
> > 6-31g*, using a pseudopotential the wavefunction is different from the  
> > all electron wavefunction, so you cannot use the same basis for the  
> > core part.
> > Aside from this 6-31g* can't even be properly represented with 280 Ry  
> > cutoff.
> > To compare you need to go to the basis set limit in *both* gaussian  
> > and cp2k (all electron calculation will probably have a larger BSSE),  
> > 6-31g* is not a particularly good basis set.
> > You cannot use the same basis set in Gaussian and cp2k (unless you do
> > all-electron GAPW).
> > If really the basiset limit of GPW-GTH is much different from the one  
> > of gaussian, then yes probably it is an effect due to the  
> > pseudopotential.
>
> Sorry for the confusion. The inputs attached are doing usual GPW
> calculations with pseudopotential and TZV2P basis sets. The GAPW
> calculations I meant were indeed all electron calculations, for which
> I haven't attached the inputs. I was certainly concerned about the
> BSSE at the beginning. But our testing calculations in Gaussian showed
> 6-31g* gives converged results.
>
> > Please note that gaussian used to call B3LYP a functional that does  
> > not correspond to the B3LYP functional published (they used a  
> > different VWN uniform density parameterization).
> > I heard that the latest version might have fixed this, but I am not  
> > sure. (still I don't expect this difference to produce such large  
> > energy changes)
>
> I agree this issue on B3LYP is secondary. I was also not entirely
> certain how the boundary conditions affect HFX. My present focus is to
> resolve the discrepancy on GGA functionals.
>
> > using the GTH pseudo both GAP and GAPW should give the same results  
> > (allelectron is different).
> > Please note that
> > 1) sample (regtest) files are built to execute quickly, not to give  
> > accurate results
> > 2) convergence wrt. cutoff is *not* variational
>
> You are absolutely right on this. We just checked the results when
> increasing the planewave cutoff. With a cutoff of 350, the all
> electron calculations give H2O2 bond dissociation energy of 2.63 eV
> (was 2.8 eV with smaller cutoff). This better value is almost the same
> as the one from Gaussian.
>
> Thanks,
> Jun