[Pw_forum] atomic energy levels in pseudopotential generation

[Maosheng Miao]

Dear QE professionals,

This is a question for pseudopotential generation, or rather for atomic calculations of DFT.

I have been using 'atomic' to generate pseudopotentials. However, I found the following trend for energy levels. For example, comparing Rh and Ir orbital energies, I found:
Rh: 5s -0.316 Ry  4d: -0.4366 Ry
Ir:    6s -0.4653 Ry 5d: -0.5253 Ry
I think the unit is in Ry. Am I right?

I wonder whether the 5s level should be lower than 6s and 4d lower than 5d. The HF levels as shown in Harrison's book (Electronic structure and the properties of solids, Dover 1980)  are  in the increasing order while move from 4d to 5d transition metals. 

I attached the input for Rh which is taken from ~atomic_doc/pseudo_library/PBE/SR:

&input
  title='Rh',
  zed=45.,
  rel=1,
  config='[Kr] 4d7.0 5s2.0 5p0.0',
  iswitch=3,
  dft='PBE'
/
&inputp
 lloc=0,
 pseudotype=3,
 file_pseudopw='Rh_ScaRel.UPF',
 nlcc=.true.,
 rcore=0.9,
 author='ADC',
/
4
5P  2  1  0.00  0.00  2.40  2.40  1
4D  3  2  7.00  0.00  1.90  2.20  1
4D  3  2  0.00 -0.30  1.90  2.20  1
5S  1  0  2.00  0.00  2.20  2.20  1

Thanks for your help,
Maosheng



Maosheng Miao, Ph. D
Materials Research Laboratory
University of California Santa Barbara
miaoms at mrl.ucsb.edu