Orbital composition of a band

Asked by syed on 2020-06-14

Hi,

I would like to know how to determine the orbital composition of a band (for a particular k-point). I assumed one had to use the (readable) output file of wavefunctions.

So if we consider the following extract obtained from the output of 'readwfsx' utility

 ***********************************************************************
            k-point = 8 0.183552 0.105974 0.000000
     Spin component = 1
 Num. wavefunctions = 41

       Wavefunction = 1
        Energy (eV) = -19.631447
 -----------------------------------------------------------------------
   Atom Species Orb-global Orb-in-atom Orb-type Re(psi) Im(psi)
     1 Mo 1 1 5s -0.202342 0.017192
     1 Mo 2 2 5s 0.441369 -0.012307
     2 S 16 1 3s 0.447549 -0.001274
     2 S 22 7 3pz -0.124095 -0.000321
     3 S 29 1 3s 0.447549 -0.001274
     3 S 35 7 3pz 0.124095 0.000321
 -----------------------------------------------------------------------

then I assumed one has to use the values 'Re(Psi)' and 'Im(Psi)' but I have failed to find any documentation on the exact nature of these values and/or their possible connection to the orbital composition.

Perhaps I am wrong in the approach, either way, I would appreciate more light on this.

Thanks

Question information

Language:
English Edit question
Status:
Solved
For:
Siesta Edit question
Assignee:
No assignee Edit question
Solved by:
syed
Solved:
2020-06-14
Last query:
2020-06-14
Last reply:
2020-06-14
Alberto Garcia (albertog) said : #1

Hi,

You can use the "fat" program in Util/COOP to plot the "fatbands" information, and/or obtain the projections of the band eigenvectors onto specific orbitals. See Util/COOP/README and Util/COOP/Tests/README.

----- El 14 de Junio de 2020, a las 14:31, syed <email address hidden> escribió:

| New question #691307 on Siesta:
| https://answers.launchpad.net/siesta/+question/691307
|
| Hi,
|
| I would like to know how to determine the orbital composition of a band (for a
| particular k-point). I assumed one had to use the (readable) output file of
| wavefunctions.
|
| So if we consider the following extract obtained from the output of 'readwfsx'
| utility
|
| ***********************************************************************
| k-point = 8 0.183552 0.105974 0.000000
| Spin component = 1
| Num. wavefunctions = 41
|
| Wavefunction = 1
| Energy (eV) = -19.631447
| -----------------------------------------------------------------------
| Atom Species Orb-global Orb-in-atom Orb-type Re(psi) Im(psi)
| 1 Mo 1 1 5s
| -0.202342 0.017192
| 1 Mo 2 2 5s
| 0.441369 -0.012307
| 2 S 16 1 3s
| 0.447549 -0.001274
| 2 S 22 7 3pz
| -0.124095 -0.000321
| 3 S 29 1 3s
| 0.447549 -0.001274
| 3 S 35 7 3pz
| 0.124095 0.000321
| -----------------------------------------------------------------------
|
| then I assumed one has to use the values 'Re(Psi)' and 'Im(Psi)' but I have
| failed to find any documentation on the exact nature of these values and/or
| their possible connection to the orbital composition.
|
| Perhaps I am wrong in the approach, either way, I would appreciate more light on
| this.
|
| Thanks
|
| --
| You received this question notification because you are an answer
| contact for Siesta.

syed (smh112) said : #2

Hi Alberto,

Thank you for that. I actually used the fat program to produce the projections, but I couldn't interpret the values of the output files except for the energies. I must have overlooked the part of the README where it said the output files produce BOTH the eigenvalue AND weight.

Huge thanks though