My Community

In the potential package, I see many atoms have 'VRHFIN' to get the valence electron configuration. But some atoms don't have the explict line. For example, Gd, its POTCAR says VRHFIN =Gd : [core=Xe4]. It gives us the number of valence electrons, but no their configuration. How can I get the electron configuration?
Thx.

please have a look at the V_RHFIN file which is in the same directory as the POTCAR for each element. This file gives the electronic configuration of the POTCAR explicitely: the last number in the second line gives the number of core electrons
1464.00 .002000 157.25000 125. .10E-04 .080 400FPE 46.00000
---> 46 for the Gd potential you mentioned.
below, you can find all states (including core) with their respective occupancies listed explicitely.
Therefore, skipping the core states 1s, 2s, 2p, 3s,3p,3d, 4s,4p and 4d (which sum up for the 46 core electrons), you fnd the valence electron configuration for Gd (PAW_PBE) to be
5s2 5p6 5d1 6s2 4f7 (18 valence electrons, one of the 4f-electrons promoted to 5d for the electronic ground state of bulk Gd (please note that this may differ from the electronic ground state configuration of the free atom in the gas phase!)

Dear Admin So why Ni has no valence electron configuration but has the following in V_RHFIN file
Ni:
9 28. .002000 58.69000 125. .25E-05 .100 200F91
18.00000
But it should be s2d8 or s1d9 valence electron configuration.
Thanks
Anteneh

admin wrote: ↑Wed Jul 16, 2008 3:26 pm please have a look at the V_RHFIN file which is in the same directory as the POTCAR for each element. This file gives the electronic configuration of the POTCAR explicitely: the last number in the second line gives the number of core electrons
1464.00 .002000 157.25000 125. .10E-04 .080 400FPE 46.00000
---> 46 for the Gd potential you mentioned.
below, you can find all states (including core) with their respective occupancies listed explicitely.
Therefore, skipping the core states 1s, 2s, 2p, 3s,3p,3d, 4s,4p and 4d (which sum up for the 46 core electrons), you fnd the valence electron configuration for Gd (PAW_PBE) to be
5s2 5p6 5d1 6s2 4f7 (18 valence electrons, one of the 4f-electrons promoted to 5d for the electronic ground state of bulk Gd (please note that this may differ from the electronic ground state configuration of the free atom in the gas phase!)

Dear Admin,
I'm using VASP 5.4. There is no V_RHFIN file in the POTCAR folder.
From the PAW_PBE POTCAR file for Gd, ZVAL = 9, which is the valence for each atomic species according to VASP manual. The POTCAR the following information:
VRHFIN =Gd : [core=Xe4]
Atomic configuration
14 entries
n l j E occ.
1 0 0.50 -50087.2354 2.0000
2 0 0.50 -8269.2182 2.0000
2 1 1.50 -7357.3672 6.0000
3 0 0.50 -1833.2681 2.0000
3 1 1.50 -1550.9376 6.0000
3 2 2.50 -1178.7203 10.0000
4 0 0.50 -369.2977 2.0000
4 1 1.50 -276.2600 6.0000
4 2 2.50 -147.6610 10.0000
4 3 2.50 -8.7140 7.0000
5 0 0.50 -49.7259 2.0000
6 0 0.50 -4.1254 2.0000
5 1 1.50 -27.2537 6.0000
5 2 2.50 -2.5337 1.0000

I thought VAL is the number of valence electrons. But it's not 18, as you replied.
Did I misunderstand something? What's the meaning of [core=Xe4] and how is it related to ZVAL?

Thank you in advance!

Hi kw7rruva,

In the list of available PAW potentials or in the POTCAR folder, you will see that there are two POTCAR files available for Gd. Specifically, Gd and Gd_3 with ZVAL=18 and ZVAL=9, respectively. They differ in the treatment of semi-core f-orbitals.

To understand the initial valence-electron configuration, it is useful to run a non-selfconsistent calculation of the isolated atom using ICHARG=12. In the POSCAR file, define a large box and a single atom. You can switch off the symmetry with ISYM=-1, but as periodic boundary conditions are used, there will always be an implicit symmetry according to the defined lattice matrix. Therefore, you may want to set the lattice matrix in the POSCAR file to reflect the site symmetry present in the compound you are interested in. Then, use crystal field theory to judge the resulting valence-electron configuration. The band energies and occupancies are written to the OUTCAR file.

I hope this is helpful.
Best regards,
Marie-Therese

This helps a lot. Thank you!

Hi,

After further investigation, let me now add how you can always exactly infer the valence configuration underlying each PAW potential.

The table, that is written below Atomic configuration in the POTCAR file, lists first the core states and then the valence states. ZVAL is the number of valence electrons. Hence, the final rows those occupancies add up to ZVAL comprise the valence configuration. Note that the rows are not ordered by energy. For instance, for Gd_3 strongly localized, semi-core f electrons are treated as core states despite being higher in energy than other valence states.

The documentation about available PAW potentials has also been updated to include this information.

Best regards,
Marie-Therese