❓ QUESTION: Should BandsData also contain the Fermi level?

[mbercx]

Currently the BandsData does not contain the Fermi level, as far as I know. Since this is quite an integral quantity to interpreting the band structure, and typically band structures are plotted by aligning the Fermi level to zero, maybe we should add the Fermi level to the BandsData as an attribute?

I think that for most codes, the calculation of the band structure involves the calculation of the charge density on a grid, so we can obtain the Fermi level from this step in the work chain?

A couple of side notes, unrelated to the common workflows:

• One of the additional use cases for this would be the viewer widgets for AiiDAlab. Here typically the viewer just takes one node as an input, so currently the viewer widget for the band structure doesn't align the Fermi level to zero, making interpretation more difficult.
• There is a move to using better representations of nodes for the aiida-core data types. I'd prefer to e.g. have a _repr_html_ method for the BandsData that automatically plots the band structure when it is returned in a Jupyter notebook. Here having the Fermi level available would also be a boon for these representations.

Once we've decided on this topic, we can migrate this issue to the aiida-core repo.

[PhilippRue]

Coming from KKR calculations I am actually a bit surprised why the standard is not to measure energies relative to the Fermi level. After the self-consistency this should be known.

But maybe in other codes this is not so simple.

[bosonie]

Thanks for opening this issue. A related issue is already present in aiida-core. aiidateam/aiida-core#2847

The BandsData is also used for phonons and there are situation where also electronic bands are better not shifted to have the zero at the Fermi energy (Ef). However I agree that an optional attribute containing the Ef would be useful.

[giovannipizzi]

I left some comments in aiidateam/aiida-core#2847

[sphuber]

See my comment on the relevant issue on aiida-core: aiidateam/aiida-core#2847 (comment)