[Pw_forum] Questions about PP generation
a.ugolotti at campus.unimib.it
Mon Jul 18 13:50:15 CEST 2016
I'm trying to generate a new pseudopotential for Pt atom which is based on
optB88 xc functional and after reding the notes and the slides i have
1) atomic.x does not seem to support the generation of a PP including
vdw-df1 correction (error message: "non-local xc not implemented"). Do you
have a workaround for this? Do you think that creating a PP with just the
xc terms without the van der Waals term will introduce a significant error?
2) when i try to create a US PP i have to specify at the end of the input
file the same orbitals twice, but with different energies. I'm not looking
for unbound states, but if i set both to 0 (all-electron results) i find
some inversion errors while calculating the beta coefficients, instead if i
set some energy i get the error of dealing with unbond states. Why do i
have to do that?Is there a reference for that procedure?In RRKJ and US
papers i don't remember such requirement.
3) what is the effect of using all-electron pseudized wavefunctions for the
local part instead of using one of the orbitals as reference?
4)i did not understand how i can properly check the final PP, for example
against ghost states (other than trying it for a bulk scf or dos for
Thanks in advance,
Università of Milano-Bicocca
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the users