EPW Forum

Dear all,

As in my title, after EPW interpolation, there are imaginary frequencies of phonon dispersion. While there is no imaginary frequency from the QE calculated phonon dispersion.

The two plot could be found at https://www.dropbox.com/s/egyjr93ux9nsf1v/phonon%20comparison.png?dl=0

The red line is from QE calculation and the black circles are from EPW interpolated.

Thanks in advance,

Dear Fanchem,

If you have the latest QE package (6.2.1), you can use the EPW tag lifc=.true. to use the ifc generated by q2r.x directly with EPW. Please see the documentation for lifc to see how it is done.

Cheers,

Vahid

Vahid Askarpour
Department of Physics and Atmospheric Science
Dalhousie University,
Halifax, NS, Canada

Thank you Vahid.

It some kind helps. But there is still discrepancy between QE calculated and EPW interpolated phonon dispersion. How can I fix it? Any suggestions are welcome.
https://www.dropbox.com/s/hrptck6insda388/phonon-withifc.png?dl=0

Thanks,
Fanchen

Vahid wrote:Dear Fanchem,

If you have the latest QE package (6.2.1), you can use the EPW tag lifc=.true. to use the ifc generated by q2r.x directly with EPW. Please see the documentation for lifc to see how it is done.

Cheers,

Vahid

Vahid Askarpour
Department of Physics and Atmospheric Science
Dalhousie University,
Halifax, NS, Canada

I would check a few things:

1. Are the masses of the two atoms consistent in all the QE and EPW runs?
2. Is the ifc generated by q2r.x converged w.r.t the q-grid? In other words, if you increase the q-grid in PHONON calculation, will matdyn.x result in the same phonon dispersion? This would have nothing to do with EPW.
3. Is the asr consistent in all calculations?

If you have already checked these, then perhaps other users or the moderators may have better ideas.

Cheers,

Vahid