[Pw_forum] Problem in carrying out Phonon calculations

[Ari P Seitsonen]

Dear Saswata Halder,

   Same opening words as in my previous response just some seconds ago: Do 
you understand which kind of pseudo potentials you are using? These are 
norm-conserving ones (in principle no need to specify 'ecutrho' 
explicitly, the default = 4*ecutwfc should do) and definitely the value of 
40 Ry is much too low to converge the properties (in particular the stress 
tensor); already for oxygen I would expect a need of 100 Ry or higher.

   The pseudo potential chosen for titanium does not include the semi-core 
electrons in the valence, and even no "non-linear core correction" is 
included: I do not expect very accurate results for the phonons, and I 
would look for another pseudo potential, if just available (for example 
the pslibrary by Andrea Dal Corso, 
http://www.qe-forge.org/gf/project/pslibrary/)

     Greetings from Rainy-Sunny Montrouge,

        apsi

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
   Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
   Ecole Normale Supérieure (ENS), Département de Chimie, Paris
   Mobile (F) : +33 789 37 24 25    (CH) : +41 79 71 90 935


On Wed, 13 May 2015, saswata halder wrote:

> Dear All,
>
> I am a beginner in quantum espresso. I am trying to perform phonon
> calculation for SrTiO3. I am able to perform the calculations but I am
> not able to reproduce the results for SrTiO3 which are already
> published in PHYSICAL REVIEW B 83, 134108 (2011).
>
> http://journals.aps.org/prb/pdf/10.1103/PhysRevB.83.134108
>
> The input file for relax is given below.
>
> &CONTROL
>    calculation='relax',
>    restart_mode='restart',
>    prefix='ST'
>    pseudo_dir = '/home/quantum/quantum/
> potentials/',
>    outdir='./'
>    tstress=.true.
>    tprnfor=.true.
>    wf_collect=.true.
>    nstep=1000
>    etot_conv_thr=1.0D-8
>    forc_conv_thr=1.0D-6
>    disk_io='high'
> /
> &SYSTEM
>    ibrav = 1, celldm(1) =7.38, nat = 5, ntyp= 3,
>    ecutwfc = 40.0, ecutrho=600, occupations='fixed',
>    smearing='gaussian', degauss=0.01
> /&ELECTRONS
>    diagonalization='cg',
>    mixing_mode='plain',
>    mixing_beta = 0.7D0
>    conv_thr =  1.0D-9
>    electron_maxstep=1000
> /
> &ions
> /
> ATOMIC_SPECIES
> Sr  87.620  Sr.pbe-mt_fhi.UPF
> Ti  47.867  Ti.pbe-mt_fhi.UPF
> O   15.999  O.pbe-mt_fhi.UPF
> ATOMIC_POSITIONS crystal
> Sr 0.00 0.00 0.00 0 0 0
> Ti 0.50 0.50 0.50 0 0 0
> O  0.50 0.50 0.00 0 0 0
> O  0.00 0.50 0.50 0 0 0
> O  0.50 0.00 0.50 0 0 0
> K_POINTS automatic
>  6 6 6   0 0 0
>
> The input file for phonon is given below:
>
> phonons of ST at Gamma
> &inputph
>  tr2_ph=1.0d-14,
>  prefix='ST',
>  epsil=.true.,
>  amass(1)=87.620,
>  amass(2)=47.867,
>  amass(3)=15.999,
>  outdir='./'
>  fildyn='ST.dynG',
> /
> 0.0 0.0 0.0
> The frequencies obtained are given below:
>
>     omega(  1 -  3) =         35.5  [cm-1]   -->  T_1u G_15  G_4- I
>     omega(  4 -  6) =        151.8  [cm-1]   --> T_1u G_15  G_4- I
>     omega(  7 -  9) =        187.0  [cm-1]   --> T_1u G_15  G_4- I
>     omega( 10 - 12) =        213.6  [cm-1]   --> T_2u G_25  G_5-
>     omega( 13 - 15) =        600.7  [cm-1]   --> T_1u G_15  G_4- I
>
> I also carried out phonon calculations using the volume cell
> relaxation. the input files for both vc-relax and phonon are:
>
> &CONTROL
>    calculation='vc-relax',
>    restart_mode='restart',
>    prefix='ST'
>    pseudo_dir = '/home/quantum/quantum/
> potentials/',
>    outdir='./'
>    tstress=.true.
>    tprnfor=.true.
>    verbosity='high'
>    wf_collect=.true.
>    nstep=1000
>    etot_conv_thr=1.0D-8
>    forc_conv_thr=1.0D-6
>    disk_io='high'
> /
> &SYSTEM
> ecutwfc = 40.0, ecutrho=500, occupations='fixed',
>    smearing='gaussian', degauss=0.01
> /
> &ELECTRONS
>    diagonalization='cg',
>    mixing_mode='plain',
>    mixing_beta = 0.7D0
>    conv_thr =  1.0D-9
>    electron_maxstep=1000
> /
> &IONS
>    ion_dynamics='bfgs'
> /
> &CELL
>    cell_dynamics='bfgs',
>    press=2.0,
>    press_conv_thr=0.5,
> /
> ATOMIC_SPECIES
> Sr  87.620  Sr.pbe-mt_fhi.UPF
> Ti  47.867  Ti.pbe-mt_fhi.UPF
> O   15.999  O.pbe-mt_fhi.UPF
> ATOMIC_POSITIONS crystal
> Sr 0.00 0.00 0.00 0 0 0
> Ti 0.50 0.50 0.50 0 0 0
> O  0.50 0.50 0.00 0 0 0
> O  0.00 0.50 0.50 0 0 0
> O  0.50 0.00 0.50 0 0 0
> K_POINTS automatic
>  6 6 6   1 1 1
>
> phonons of ST at Gamma by anup
> &inputph
>  tr2_ph=1.0d-14,
>  prefix='ST',
>  epsil=.true.,
>  amass(1)=87.620,
>  amass(2)=47.867,
>  amass(3)=15.999,
>  outdir='./'
>  fildyn='ST.dynG',
> /
> 0.0 0.0 0.0
>
> the output for phonon frequencies as obtained is given below:
>
>
>     omega(  1 -  3) =       -628.3  [cm-1]   --> T_2u G_25  G_5-
>     omega(  4 -  6) =       -615.5  [cm-1]   --> T_1u G_15  G_4- I
>     omega(  7 -  9) =       -375.6  [cm-1]   --> T_1u G_15  G_4- I
>     omega( 10 - 12) =        190.2  [cm-1]   --> T_1u G_15  G_4- I
>     omega( 13 - 15) =        811.8  [cm-1]   --> T_1u G_15  G_4- I
>
> I have carried out both the relax and volume cell relaxations and have
> obtained two different results in the phonon mode calculations. Both
> the results are not matching with reported results for SrTiO3.
> Could you please suggest as to which of the methods is applicable to
> SrTiO3 system so as to provide me with more accurate results.
>
> Thanking you,
>
> With Regards.
>
> Saswata Halder.
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>