Work function > vacuum energy? (EFERMI = MIDGAP)

[christopher_sheldon1]

Glad that we could help. Let us know if you've got any follow-up questions, otherwise we'll close this topic on 5th December.
Best,
Chris

[reynaldo.putra]

Thank you!

For now I do not think I have any follow-up questions.

Have a great day, and best regards,
Reynaldo

[reynaldo.putra]

Hello,

It turns out I have a question related to the determination of the band gap using the band_gap.sh script. When I extracted the HOMO LUMO data, I obtained the following:

Code: Select all

HOMO: band: 76  E= 3.2142 eV
LUMO: band: 77  E= 5.5153 eV
./band_gap.sh: line 11: bc: command not found
GAP (OUTCAR): 0.00 eV

I suppose I am also curious about how to fix the gap calculation section and the error on line 11 ;D However, I would like to know why the gap given from the HOMO-LUMO values would differ to the gap shown in the DOS plots. Here, it can be calculated to be ~2.3 eV, whereas when I plot the DOS after post-processing using VASPKIT, the band gap is much smaller at ~1.6 eV (I used ISMEAR = -5 to ensure no gap shrinking via smearing width).

Thank you again for your time and attention!

Below is the DOS plot:

STON2-DOS.png

[fabien_tran1]

Hi,

The error message "bc: command not found" is quite clear: the program bc (arbitrary precision calculator) is not installed on your system.

Concerning the different band gaps obtained from band_gap.sh and VASPKIT, I suggest that you compare both ways with a very simple system (silicon, MgO, ...). You should figure out: (1) which file is processed by VASPKIT and (2) which DOS is shown on the plot (only one spin component or the sum of both?).

[reynaldo.putra]

Dear Fabien,

1) Thank you for letting me know about the bc command! Apologies, I am not too familiar with what the warnings mean.

2) I see, I will give them a try and let you know. Thank you!

[reynaldo.putra]

Hello,

I have another question, this time related to band alignment via the ionization potential (IP) and electron affinity (EA) and combining the surface and bulk calculations. It is mentioned that it is done by doing the following:

IP-alignment.png

The vacuum level from a surface slab model and the VBM from a bulk model are aligned using the common electrostatic reference levels in the bulk-like region of the surface model and the bulk model.

I am still confused how to obtain the common electrostatic reference levels for the bulk model and the bulk-like region of the surface model. How is this obtained in VASP calculations? What tags should I use in the INCAR, and which output files should I look out for?

Is this related to the ICORELEVEL tag?

Thank you for your time!

[fabien_tran1]

Hi,

In the supplemental material (SM) of A. Grüneis et al. and in Sec. II.B of Y. Hinuma et al. a procedure for calculating the reference level in the bulk region of the bulk and surface calculations is described. The reference level is calculated as the average electrostatic potential at atomic sites and should correspond to what is printed in OUTCAR below "average (electrostatic) potential at core". I suggest that you try to reproduce one or two of the PBE results presented in Table IV of the SM of A. Grüneis et al..