Hello QE developers and users, I am trying to calculate phonon dispersion using ph.x and matdyn.x routines of QE and I am stuck woth negative frequencies for a while now.
I have a 1-D chain of polyethylene which has 6 atoms in the unit-cell. The polymer chain is aligned in z-direction and I have added vacuum on x- and y- direction. I did the structure relaxation and I am getting total force after relaxtion as 0.000020, which I guess is sufficiently small. After relaxation, when I am trying to calculate phonon dispersion using ph.x and matdyn.x I am getting netting frequencies for some of the phonon modes. I searched online and on QE mailing list and tried couple of different things: 1. The negative frequencies are appearing for non-gamma point as well, so I guess its not a ASR issue. Anyway, I am using ASR but even after this I am getting negative frequencies. 2. I reduced my tr2_ph and conv_thr to 1.0d-14 and 1.0d-15 which I guess are very small as suggested on QE FAQs. 3. I checked phonon mode shapes for modes with negative frequencies and they do not look like rotational mode. 4. I tried changing number of k points from 6 to 12 in z direction but this is also not the cause. I am really stuck here and not sure about the cause of negative frequencies. With vc-relax, my structure is getting relaxed which I guess means my structure is stable. But then I am not sure why I am getting negative frequencies. I really appreciate any help or suggestions. Thanks, Ankit Jain IIT Indore, Indore, India
