[Pw_forum] Help! phonon calculation at G point.
Stefano Baroni
baroni at sissa.it
Mon Oct 22 08:51:18 CEST 2007
Dear Hailin Yu:
didn?t you know that Quantum Espresso comes with a users' guide?
Well, it is probably not as good as one can expect, and as we would
like it to be, but it does contain the answer to ypur questions. See:
http://www.pwscf.org/guide/3.1.1/other/users-guide-3.1.1.pdf
The above link was found by googling the following keywords:
"negative frequencies " site:pwscf.org
This being said, the first calculation seems to be just affected by
accuracy issues (that are briefly discussed in the users' guide
referenced above). The second calculation, instead, seems to signal
an instability of the optical modes 9such as a ferroelectric
instability) that is not necessarily related to any elastic instability.
Hope this helps
Stefano
On Oct 22, 2007, at 4:23 AM, hailin yu wrote:
>
>
> From: yuhailin_79 at msn.com
> To: pw_users at pwscf.org
> Subject: Help! phonon calculation at G point.
> Date: Wed, 17 Oct 2007 11:08:51 +0800
>
> Dear all:
>
> I have some trouble with phonon calculation at G point.
> Recently, I have calculated two kinds of PtC, rosksalt structure
> and zinc-blende structure. Firstly, I have fully relaxed the
> lattice constants and atomic positions using quantum-
> espressso3.2.3. The results we attained is well agreement with
> these results using by VASP and CASTEP. Following, I calculate the
> phonon frequency at G point of the two structures.
> For zinc-blende structure, we attained the phonon frequency as
> follows:
> q = ( 0.000000000 0.000000000 0.000000000 )
>
> **********************************************************************
> ****
> omega( 1) = 1.086480 [THz] = 36.241319 [cm-1]
> ( 0.340996 0.000000 0.272050 0.000000 -0.549845 0.000000 )
> ;( 0.346068 0.000000 0.276096 0.000000 -0.558022 0.000000 )
> omega( 2) = 1.086480 [THz] = 36.241319 [cm-1]
> ( -0.191604 0.000000 -0.550378 0.000000 -0.391141 0.000000 )
> ( -0.194454 0.000000 -0.558563 0.000000 -0.396958 0.000000 )
> omega( 3) = 1.086480 [THz] = 36.241319 [cm-1]
> ( 0.582777 0.000000 -0.340135 0.000000 0.193129 0.000000 )
> ( 0.591444 0.000000 -0.345194 0.000000 0.196001 0.000000 )
> omega( 4) = 16.225674 [THz] = 541.233828 [cm-1]
> ( -0.026894 0.000000 0.048005 0.000000 0.029340 0.000000 )
> ( 0. 430439 0.000000 -0.768321 0.000000 -0.469592 0.000000 )
> omega( 5) = 16.225674 [THz] = 541.233828 [cm-1]
> ( -0.036407 0.000000 0.009944 0.000000 -0.049641 0.000000 )
> ( 0.582692 0.000000 -0.159157 0.000000 0.794513 0.000000 )
> omega( 6) = 16.225674 [THz] = 541.233828 [cm-1]
> ( 0.042894 0.000000 0.038539 0.000000 -0.023738 0.000000 )
> ( -0.686516 0.000000 -0.616817 0.000000 0.379926 0.000000 )
>
> **********************************************************************
> ****
> I don't know why omega(1), omega(2) and omega(3) are not equal to
> zero? while the optic modes is well agreement with the results
> attained by VASP. I want to know how to improve the accuracy. Can
> you tell me? Thanks a lot!
>
> For roskslt structure, we attained the phonon frequency as follows:
>
> q = ( 0.000000000 0.000000000 0.000000000 )
>
> **********************************************************************
> ****
> omega( 1) = -16.323420 [THz] = -544.494287 [cm-1]
> ( 0.057067 0.000000 0.023583 0.000000 0.002155 0.000000 )
> ( -0.921868 0.000000 -0.380961 0.000000 -0.034805 0.000000 )
> omega( 2) = -16.323420 [THz] = -544.494287 [cm-1]
> ( -0.015312 0.000000 0.032457 0.000000 0.050295 0.000000 )
> ( 0.247345 0.000000 -0.524309 0.000000 -0.812467 0.000000 )
> omega( 3) =&nbs p; -16.323420 [THz] = -544.494287 [cm-1]
> ( 0.018065 0.000000 -0.046988 0.000000 0.035822 0.000000 )
> ( -0.291827 0.000000 0.759046 0.000000 -0.578678 0.000000 )
> omega( 4) = 0.284260 [THz] = 9.481955 [cm-1]
> ( 0.369808 0.000000 -0.578558 0.000000 0.160509 0.000000 )
> ( 0.371847 0.000000 -0.581748 0.000000 0.161394 0.000000 )
> omega( 5) = 0.284260 [THz] = 9.481955 [cm-1]
> ( 0.598650 0.000000 0.340881 0.000000 -0.150560 0.000000 )
> ( 0.601951 0.000000 0.342761 0.000000 -0.151390 0.000000 )
> omega( 6) = 0.284260 [THz] = 9.481955 [cm-1]
> ( -0.045937 0.000000 -0.215223 0.000000 -0.669940 0.000000 )
> ( -0.046191 0.000000 -0.216410 0.000000 -0.673634 0.000000 )
>
> **********************************************************************
> ****
> Why the phonon frequency is negative? As a test, I have caculated
> the elastic constants of this structure and find it satisfy formula:
> C44>0, C11>|C12|,C11+2C12>0. So we don't know what was wrong. Can
> you tell me? Thanks!
> The input file as shown in the follow:
> RS.scf.in:
>
> &control
> calculation='scf'
> restart_mode='from_scratch'
> tstress = .true.
> tprnfor = .true.
> prefix='RS'
> etot_conv_thr=1.0D-5
> forc_conv_thr=1.0D-4
> & nbsp; pseudo_dir = '/home/yhl/espresso-3.0/pseudo/'
> outdir='/home/yhl/tmp/'
> /
> &system
> ibrav= 2, celldm(1) =8.98, nat= 2, ntyp= 2
> ecutwfc =100.0
> ecutrho =500.0
> degauss =0.02
> smearing='mp'
> occupations ='smearing'
> /
> &electrons
> conv_thr = 1.0d-12
> mixing_beta = 0.7
> /
> ATOMIC_SPECIES
> Pt 195.08 Pt.pbe-nd-rrkjus.UPF
> C 12.01 C.pbe-rrkjus.UPF
> ATOMIC_POSITIONS
> Pt 0.00 0.00 0.00
> C 0.5 0.5 0.5
> K_POINTS AUTOMATIC
> 12 12 12 1 1 1
>
> RS.ph.in
>
> phonon of PtC
> &inputph
> tr2_ph=1.0d-14,
> prefix='RS',
> ldisp=.true.,
> nq1=4,nq2=4,nq3=4
> amass(1)=195.08,
> amass(2)=12.01,
> outdir='/home/yhl/tmp/',
> fildyn='RS.dyn',
> /
>
>
> Discover the new Windows Vista Learn more!
>
> Discover the new Windows Vista Learn more!
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
---
Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center -
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)
Please, if possible, don't send me MS Word or PowerPoint attachments
Why? See: http://www.gnu.org/philosophy/no-word-attachments.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20071022/d13babe1/attachment.html>
More information about the users
mailing list