Author Topic: Semicore Pseudopotential  (Read 3999 times)

0 Members and 1 Guest are viewing this topic.

Offline davalenciah

  • Heavy QuantumATK user
  • ***
  • Posts: 26
  • Reputation: 0
    • View Profile
Semicore Pseudopotential
« on: December 15, 2011, 20:03 »
Hello

I am running a structure with Lanthanum for GGA and LDA
exchange correlation function as:

for LDA
#----------------------------------------
# Basis Set
#----------------------------------------
basis_set = [
       HGHLDABasis.Lanthanum_HGH_11_DoubleZetaDoublePolarized
   ]
and for GGA

HGHGGAPBEBasis.Lanthanum_HGH_11_DoubleZetaDoublePolarized

and I got a error message
" HGHLDABasis.Lanthanum_HGH_11_DoubleZetaDoublePolarized,
AttributeError: 'module' object has no attribute 'Lanthanum_HGH_11_DoubleZeta'
"
 
Could you help me to solve this error?


Thanks

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5428
  • Country: dk
  • Reputation: 89
    • View Profile
    • QuantumATK at Synopsys
Re: Semicore Pseudopotential
« Reply #1 on: December 15, 2011, 20:43 »
We don't have predefined basis sets for La with the HGH potentials. Use the standard LDA/GGA basis sets, they should be reasonably good.

Offline davalenciah

  • Heavy QuantumATK user
  • ***
  • Posts: 26
  • Reputation: 0
    • View Profile
Re: Semicore Pseudopotential
« Reply #2 on: December 15, 2011, 23:17 »
Ok but I was confuse because in the file
/share/pseudopotentials/normconserving/HGH/
exist a file LA.GGAPBE.11.zip and in
atkpython/examples$ vi HGHLDABasis.py
the lanthanum is listed too

Thanks

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5428
  • Country: dk
  • Reputation: 89
    • View Profile
    • QuantumATK at Synopsys
Re: Semicore Pseudopotential
« Reply #3 on: December 15, 2011, 23:38 »
True; that's partly a mistake, however, it was put in there during testing, and apparently not taken out for the final release. I believe the reason we didn't include it, is that it gave strange results. You can try it, just import the basis set from the example file, but as I said, I think it might not work well.