After calculating the band diagram of a structure, is there a way to identify localized surface states?

I am working with Si nanowires with different cross sections. I want to determine the band gap variation between them, and so saturated the surface bonds using H atoms. However I want to confirm that the resulting structures are free of any localized surface states. I have attached two sample band diagrams below.

Note: the K-path varies from the Gamma point along the length of the structure.

Structure 1

Structure 2

  • 3
    $\begingroup$ I would try and project the DOS onto the surface atoms thus creating the so-called "fatbands". This will show you the weight of the surface atoms on the individual bands. $\endgroup$
    – zeroth
    Oct 28 at 7:40
  • $\begingroup$ Thank you. I shall try that. $\endgroup$
    – PBH
    Oct 28 at 8:40

Identification of bands belonging to specific atoms is generally easiest done by using the so-called fatbands.

Basically the fatbands is a projected DOS analysis for each individual band. The projected DOS can then be summed over specific orbitals to see the weight on each atom. Generally states are wide-spread and so determining whether a state has a large weight on specific sites may be difficult to quantify due to the relative nature of the states.
In some cases the weight is very clearly localized in which case the projected weight on the atoms should be 1 (fully localized).

Something to consider when comparing projected weights for sites is that a system of N sites (orbitals) will have a weight of 1/N for a fully delocalized state.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.