Phonon Density of States (PDOS) at Gamma Point

[RaquelYR]

Hello everyone!

I'm trying to replicate phonon density of states (PHDOS) diagrams for some solids using Quantum Espresso. The usual way I do it is the following one:

1. scf calculation at minima (pw.x)
2. Calculation of dynamical matrix in reciprocal space with nq=1 or 2 (ph.x)
3. Calculation of interatomic force constants in real space (q2r.c)
4. Calculation of phonon frequency (matdyn.x)

The frequencies I obtained differ with the ones of the example. The only significant difference that I find is that they perform the calculation at Gamma Point. I have tried to carry out my procedure at gamma point but it seems to be incompatible.

Does any one know how can I get the phonon density of states at gamma point?

Thank you in advance!

[krollina8]

As i know instead of using ph.x to calculate the dynamical matrix with a non-zero q-point, you will use q2r.x to calculate the interatomic force constants in real space and then matdyn.x to compute the phonon frequencies at the gamma point. It's worth noting that the frequencies at the gamma point may behave differently from those at other non-zero q-points due to the degeneracy at the gamma point. The PHDOS at the gamma point will typically show different characteristics compared to the PHDOS obtained at other q-points.

[carterodell]

I believe the dynamical matrix with a non-zero q-point is best derived using q2r.x, the real space interatomic force constants, and matdyn.x for the gamma point phonon frequencies. Be careful, the gamma point's degeneracy can affect frequencies differently than at other non-zero q-points. Typically, PHDOS derived at the gamma point will differ from those derived at other q-points.

[watsicagaylord]

q2r.x and matdyn.x are the best ways to figure out dynamical matrices at non-zero q-points. Note: Gamma-point degeneracy can change frequencies, and PHDOS at gamma may be different from q-points at other places.

[ChenBeier]

Thread is almost 1 year old.

[Reatele]

I think this is a very difficult problem.

In essence, performing a Gamma-point PHDOS calculation is essentially focused on using only the Gamma point throughout, which requires adjusting the SCF, dynamical matrix, and skipping the q2r.x step. This method should align your results more closely with the Gamma-only example you’re aiming to replicate.

If you encounter specific issues with settings in ph.x or matdyn.x, let me know, and we can troubleshoot further.

[ChenBeier]

Old thread again.