[Pw_forum] Pt bulk phonon calculated by PWSCF 1.2.0

[Stefano de Gironcoli]

Phonon anomalies in metals are related to Fermi surface nesting.
You therefore need to describe correctly the Fermi surface.
This imply looking for the limit where degauss->0 (mantaining good convergence
with respect BZ sampling)
I suggest you keep a reasonable cutoff (35 or less should be OK for Pt) and  study
instead the phonon dispersion behaviour with respect the degauss value.
Proceed as follow:
For each value of degauss you should increase the k point mesh until frequencies are
stable.
Look for the behavior of the converged frequency as you reduce degauss.

For not anomalous frequencies things should be rather insensitive to these details but
for the anomalous ones you should see the difference.

Have a look to the following two references:

S. de Gironcoli, Lattice Dynamics of Metals from Density-Functional Perturbation Theory,
Phys. Rev. B 51, 6773 (1995).
C. Bungaro, S. de Gironcoli, and S. Baroni, Theory of the anomalous Rayleigh dispersion at
H/W(110) surfaces,   Phys. Rev. Lett. 77, 2491 (1996).

all the best

stefano

"Hong, SamPyo" wrote:

> Hello, dear PWSCF users
>
> Recently I calculated the bulk phonons of Pt along Gamma-K direction.
> I attached the results in pdf format, where B,C and D are the ones calculated using
> PWSCF 1.2.0 version.
> The open circles are done by a mixed-basis code. The comparison shows that the
> frequencies obtained by
> PWSCF are showing some lowering of frequencies but oveall it's failing to reproduce the
> experimentally-observed anomaly in bulk Pt dispersion
>
> along Gamma-K line while a mixed basis program is doing good job.
> I used  ecut of 35 ryd for H/Pt US potentials and 72 special k-points. I increased the
> ecut to 40 ryd but just very slight changes in frequecies.
>
> I also tried higher cut-off for charge density up to 10 times of ecut but just very
> small changes in frequencies.
>
> What else can I try to get better frequencies for bulk Pt? Could any of you give me some
> informations ? You will be very appreciated.
>
> Sampyo
>
> <<ptbulk2.pdf>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20030321/7c1468a8/attachment.htm