[Pw_forum] Help on converting from fhi to UPF form!

[Paolo Giannozzi]

On Tuesday 14 April 2009 05:19, shangy wrote:

> When I run the fhi2upf.x, it asked the "Wavefunction label and occupancy" 
> [...] I tried it like this
> ---------------------
> Wavefunction # 1: label, occupancy > 4s 2
> Wavefunction # 2: label, occupancy > 4p 2
> Wavefunction # 3: label, occupancy > 4d 0
> Wavefunction # 4: label, occupancy > 4f 0
> ---------------------
> but I don't know if it is right?

the FHI format doesn't contain - to the best of my very limited
knowledge - any information on the electronic configuration
used to generate the pseudopotential. Atomic orbitals are 
actually used only to produce the starting wavefunctions 
and charge density for self-consistency. You should specify
the labels of l=0 to lmax valence orbitals (in the order:s,p,d,f,..)
and a reasonable value for their occupancy. You have Ge with 
4 valence electrons, so your choice seems to be correct.

Paolo
-- 
Paolo Giannozzi, Democritos and Udine University