[QE-users] Either NEGATIVE FREQUENCY or CONVERGENCE problem in PHONON

[Tamas Karpati]

I have to admit that an almost well relaxed structure may have no
negative freqs. at all.
I would like to add that the 6 lowest are still far from zero in this case which
is an indicator, albeit not very clear (a few hundreds 1/cm can still be OK).
Thanks for clarification.

On Wed, Dec 9, 2020 at 2:03 PM Stefano Baroni <baroni at sissa.it> wrote:
>
> I beg to differ. Whatever the specific reason may be in this case, negative frequencies do *not* (necessarily) indicate incomplete relaxation, but the fact that the nearest equilibrium structure is unstable (i.e. it is a saddle point, rather than a minimum). If the current configuration is close enough to a minimum, the computed frequencies should be positive no matter whether you are at an equilibrium configuration or not. SB
>
> —
> Stefano Baroni - Trieste —  http://stefano.baroni.me
>
>
>
>
> On 9 Dec 2020, at 12:39, Tamas Karpati <tkarpati at gmail.com> wrote:
>
> apart from the convergence issues negative frequencies from converged
> calculations suggest
> that your structure -as opposed to your standpoint- is not stable or
> the geom. optimization was not
> complete (many negative freqs. may indicate not being at the exact
> energy minimum).
> you can reoptimize (both cell and ions, ie. via vc-relax) the geometry
> and next redo the phonon
> sim. to check it.
>
> On Wed, Dec 9, 2020 at 11:47 AM Sohail Ahmad via users
> <users at lists.quantum-espresso.org> wrote:
>
>
> Dear QE experts
> I am trying to calculate the phonon of NiCrSi using QE 6.5.
>
> Either i get negative frequency or it does not converge
> I tried with different values of convergence criteria, q mesh, alpha_mix in ph.in and k mesh in scf.in
>
> tr2_ph = 1.0d-14, or
> tr2_ph = 1.0d-12, or
> tr2_ph = 1.0d-16,
>
> nq1 = 4, nq2 = 4, nq3 = 4, or
> nq1 = 3, nq2 =3, nq3 = 3, or
>  nq1 = 2, nq2 = 2, nq3 =2,
>
> alpha_mix(1) = 0.7, or
>  alpha_mix(1) = 0.3,
>
> K_POINTS
> 12 12 12 or
> 16 16 16
>
> What could be the possible reason of getting negative frequency if the structure is considered as stable ? Appreciate if someone point out the possible mistakes
> Best Wishes
> ---------------------------------------------------------
>
> #phonon of NiCrSi structure
> &inputph
> prefix = 'NiCrSi',
> tr2_ph = 1.0d-14,
> ldisp = .true.,
> nq1 = 4, nq2 = 4, nq3 = 4,
> amass(1) = 51.99,
> amass(2) = 28.08,
> amass(3) = 58.69,
>  alpha_mix(1) = 0.7,
> outdir = './OUT',
> fildyn = 'NiCrSi.dyn',
> /
> &control
>  calculation = 'scf',
>  restart_mode = 'from_scratch',
>  prefix = 'NiCrSi',
>  pseudo_dir = '/home/sohail/pseudo',
>  outdir = './OUT',
> /
> &system
> ibrav = 2, a = 5.4559000675, b = 5.4559000675, c = 5.4559000675,
> cosAB = 0.0, cosBC = 0.0, cosAC = 0.0,
>  nat = 3, ntyp = 3,
>  ecutwfc = 75,
>  ecutrho = 300,
>  nbnd = 25,
>  occupations = 'smearing', smearing = 'gaussian', degauss = 0.001,
> /
> &electrons
>  mixing_beta = 0.3,
>  conv_thr = 1.0d-9,
> /
> ATOMIC_SPECIES
> Cr  51.99  Cr.pbe-sp-van.UPF
> Si  28.08  Si.pbe-n-van.UPF
> Ni  58.69  Ni.pbe-nd-rrkjus.UPF
> ATOMIC_POSITIONS  {alat}
> Cr  0.00000000  0.00000000  0.00000000
> Si  0.50000000  0.50000000  0.50000000
> Ni  0.25000000  0.25000000  0.25000000
> K_POINTS AUTOMATIC
> 12 12 12 0 0 0
>
> ---------------------------------------------------------------------
>
> Sohail Ahmad
>
> Department of Physics
>
> Faculty of Science, P. O. Box - 9004
>
> King Khalid University
>
> Abha, Saudi Arabia
>
> --------------------------------------------------------------------
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users