4
$\begingroup$

In a post from the Quantum Espresso mailing list, a user suggests adding k-points with weight zero to an SCF calculation as a workaround to allow bandstructure to be computed using a hybrid functional.

I'm not super familiar with how to interpret k-point weights. What does it mean to add k-points with weight zero?

Improve this question
$\endgroup$

2 Answers 2

Reset to default
3
$\begingroup$

According to the pw.x input description:

In a non-scf calculation, weights do not affect the results. If you just need eigenvalues and eigenvectors (for instance, for a band-structure plot), weights can be set to any value (for instance all equal to 1).

So, in your K_POINTS card, the card option should be set as tpiba_b or crystal_b (for band structure plot). So, you can set the card as follows (just an example):

K_POINTS {crystal_b}
5
    L       0
    gG      0
    X       0
    U       0
    gG      0

I cannot help you with how this works, or why we need this for hybrid functional, but if you are looking for how to set the weight to zero, this should work. An alternative version of the above card is given below where each of the k-point is written explicitly:

K_POINTS {crystal_b}
5
  0.0000 0.5000 0.0000 0 
  0.0000 0.0000 0.0000 0 
  -0.500 0.0000 -0.500 0
  -0.375 0.2500 -0.375 0 
  0.0000 0.0000 0.0000 0
Improve this answer
$\endgroup$
2
  • $\begingroup$ So when you put zeros in the 4th column of k_points {crystal_b} that's the way you set the weight of k-points to zero $\endgroup$
    – Camilla
    Mar 23 at 11:02
  • $\begingroup$ @Camilla yes. You can find the syntax here. $\endgroup$
    – Abdul Muhaymin
    Mar 23 at 11:18
5
$\begingroup$

Each k-point contributes with a different 'weight' to your calculation.

Say you have a calculation featuring an N number of k-points, each Ni k-point will contribute differently rather than with a 1/N. This is due to the symmetry of your crystal (or model, generally speaking).

This could help you!

Improve this answer
$\endgroup$
2
  • $\begingroup$ Can you give me an example of an scf input file for quantum ESPRESSO with this type of k-points please? $\endgroup$
    – Camilla
    Mar 22 at 15:31
  • $\begingroup$ What 'type' of k-points are you actually referring to? If you mean diffently weighted k-points, you can take whatever simple crystal (Si, graphene etc), run a calculation selecting a certain grid of k-point and look at the weight of each in the output file. Depending on the coordinates of the k-points (that should be reported in the reciprocal space), you can assess which k-point contributes more and which one less. $\endgroup$
    – Laura
    Mar 22 at 16:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .