I have been trying to obtain a plot of the phonon dispersion of a BCC metal using Quantum Espresso. In the literature it seems that a commonly reported path is Gamma-H-P-Gamma-N and so I am attempting to use the same. However, when I look for the k-points (q-vectors) in fractional coordinates so that I can specify them in the interpolation part of the calculation (using matdyn.x) it seems that the following coordinates are reported (I also see the following when using xcrysden, etc):

Γ   0.0000000000    0.0000000000    0.0000000000
H   0.5000000000   -0.5000000000    0.5000000000
N   0.0000000000    0.0000000000    0.5000000000
P   0.2500000000    0.2500000000    0.2500000000

Using these values my dispersion looks completely wrong (too many bands in some segments, etc). On the other hand, I have seen in various tutorials which do bandstructure calculations of BCC metals, the following coordinates are reported

  0.000  1.000  0.000   !H   
  0.500  0.500  0.000   !N
  0.000  0.000  0.000   !gamma
  0.500  0.500  0.500   !P
  0.000  1.000  0.000   !H

using these (or related ones for the slightly different BCC basis I'm using) gives much better looking plots.

I'm wondering if anyone can reconcile these observations for me? I'm at a loss to connect the two things and am unsure if it is a fundamental misunderstanding on my part or a particularity of how matdyn.x represents k-points.

  • 1
    $\begingroup$ +1 and welcome to our new community! Thank you for contributing your question here and we hope to see much more of you in the future!!! Also, your comment would be more appropriate as an answer. If the community disagrees with it, they can comment on it. The way you've done it, you won't be getting feedback (at least in the way that's intended on this site). $\endgroup$ Commented Feb 26, 2023 at 15:54
  • $\begingroup$ Hi, thanks for the welcome and your comment! Ah, I see - I'll switch my comment to an answer in that case. Cheers! $\endgroup$ Commented Feb 26, 2023 at 20:30

1 Answer 1


Actually, I think I may understand now. I believe that the first set of coordinates that I listed are the coordinates of the high symmetry points in the basis of primitive reciprocal lattice vectors (either in the literature or in the input to xcrysden). On the other hand, it would seem that matdyn.x requires the Cartesian coordinates (in $\frac{2π}{a}$ units). The latter set of coordinates make sense from that point of view.


You must log in to answer this question.

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