Page 1 of 1

Negative frequencies in linewidth claculation

Posted: Mon Apr 17, 2017 6:10 am
by gururaj
dear sir ,
i am running phonon linewidth calculations of graphene on 8x8x1 grid but i am getting negative frequencies and following error is coming up.i would be so greatful if you would help me..
thank you.

ERROR:

xq = 1.089753 0.619101 0.000000
freq = -0.2700E-07 0.2815E-05 0.7042E-05 0.6876E-04 0.1886E-03 0.1906E-03

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine freq_phq_safe (1):
cannot continue with negative frequencies
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

stopping ...
--------------------------------------------------------------------------

Re: Negative frequencies in linewidth claculation

Posted: Tue Apr 18, 2017 12:04 pm
by sponce
Dear gururaj,

I think this error should not be in an EPW calculation.

Are you using the phonon code? If so the best way to report issues for pw.x or ph.x is through the quantum espresso mailing list.

However, I can suggest the following:
- be sure to have correctly relaxed your atoms and unit cells to their minimum
- be sure to use a large enough k-grid and energy cut-off.

In principle, the code should not crash even with negative phonons though...

Best,

Samuel