[Pw_forum] negative frequency

[Eyvaz Isaev]

Dear Ghosh,

Well, first of all, negative modes near q=2/3(111)  means QE works correctly!!!
As bcc Zr is a high temperature phase, but DFT deals with T=0K, so, there is no direct way to get stable modes for bcc Zr (Ti, Hf). Instability near q=2/3(111) means the possiblity to transform to \omega phase. 

At high temperatures phonon-phonon interaction (anharmonicity) becomes important, but it is quite difficult to implement it directly. In the paper you mentioned they do account anharmonic terms (see the papar).  

What you can do? Of course, you can perform MD calculations at high temerature and even get  Phonon DOS, but in this way you can not calculate phonon dispersions. 
Not sure, D3 code can do this (phonon dispersions).
The only code that deals with phonons at high temperature is so called SCAILD code. But it is not linear response method, it is based on direct (supercell) method.  
I am thinking about is it possible to implement it in QE (sure, it will take time). 
For reference see PRL 100, 095901 (2008).

Bests,
Eyvaz.

-------------------------------------------------------------------

Prof. Eyvaz Isaev, 
 


DeTheoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia,partment of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden 

isaev at ifm.liu.se, eyvaz_isaev at yahoo.com

--- On Fri, 5/21/10, partha sarathi ghosh <parthasarathi13 at gmail.com> wrote:

From: partha sarathi ghosh <parthasarathi13 at gmail.com>
Subject: [Pw_forum] negative frequency
To: pw_forum at pwscf.org
Date: Friday, May 21, 2010, 1:41 PM

Dear all,
                           Thank you xirainbow and Michael J. Mehlfor for your reply & suggestion .
 I was trying to calculate longitudinal phonon dispersion of Zr along [111] direction. 
I wanted to validate my calculation with experimental results for bcc-Zr (not hcp) given in the paper :Phy.Rev.B,vol-29,page 1575-1587(1984). 

But I am getting negative frequencies near q=2/3(1 1 1) and the over all appearance of the curve is different from the experimental one.  I optimized the system w.r.t  ecutwf, ecutrho , k-point sampling , volume ,degauss, conv_thr. I have also tried with different q-point sampling in ph.in file (like 4 4 4, 8 8 8) with tr2_ph=1.0d-12.

My questions are :
1. Is DFTP suitable for studying particular phonon mode like in Zr ??
2. If yes then how to make my calculations efficient ?? 
3. If not then any suggestion of other methods ??
 Any one can please help me.

Thanks in advance.
P. S. Ghosh
B.A.R.C.
INDIA  


-----Inline Attachment Follows-----

_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20100521/cf37c600/attachment.html>