[Pw_forum] error in phonon calculation with SCAN functional

Claudio Perottoni caperott at gmail.com
Mon Nov 27 11:59:07 CET 2017


Dear all,

  This is just to report I have got some strange results in a phonon 
calculation for hexagonal boron nitride with the SCAN functional and QE 
master branch (as of Nov. 25, corrected for a bug in SCAN). The 
high-energy Raman-active mode is being calculated by ph.x as 2142 cm-1 
(experimental is about 1364 cm-1). The same calculation with the rVV10 
functional yields 1361 cm-1. A finite-differences calculation with the 
SCAN functional yields a much closer to experimental result, 1365 cm-1.

Here is the input files (please note the space group is NOT the correct 
s.g. for hBN - s.g. 186 is being used as part of a test. The structure, 
however, is identical to that of hBN):

&CONTROL
    calculation      = 'scf'
    title            = 'hBN'
    verbosity        = 'default'
    restart_mode     = 'from_scratch'
    wf_collect       = .true.
    nstep            = 400
    tstress          = .false.
    tprnfor          = .true.
    outdir           = './files'
    wfcdir           = './files'
    prefix           = 'hbn'
    disk_io          = 'low'
    pseudo_dir       = '../../../pseudo/'
/
&SYSTEM
    ecutwfc          = 180
    ecutrho          = 720
    input_dft        = 'scan'
    ntyp             = 2
    nat              = 2
    space_group      = 186
    a = 2.4958606275
    c = 6.79980712666
/
&ELECTRONS
    electron_maxstep = 200
    conv_thr         = 1e-9
    mixing_mode      = 'plain'
    mixing_beta      = 0.7
    mixing_ndim      = 8
    diagonalization  = 'david'
    diago_thr_init   = 0.0001
    startingpot      = 'atomic'
    startingwfc      = 'random'
/
ATOMIC_SPECIES
B 10.81 B_ONCV_PBE-1.0.upf
N 14.007 N_ONCV_PBE-1.0.upf

ATOMIC_POSITIONS crystal_sg
B      0.33333333333333       0.66666666666667       0.25
N      0.33333333333333       0.66666666666667       0.75

K_POINTS automatic
9 9 21  0 0 0


************************************************************************

and the input file for ph.x:

&inputph
  tr2_ph=1.0d-14,
  prefix='hbn',
  search_sym = .true.,
  asr = .true.,
  trans = .true.,
  epsil = .true.,
  outdir='./files',
  fildyn='hbn.dyn'
/
0.0 0.0 0.0


************************************************************************

Results for SCAN functional (after dynmat.x/asr='crystal'):

   # mode   [cm-1]    [THz]      IR
     1      0.00    0.0000    0.0000
     2      0.00    0.0000    0.0000
     3      0.00    0.0000    0.0000
     4    955.62   28.6489    0.0000
     5   1030.21   30.8850    0.0000
     6   1030.21   30.8850    0.0000
     7   1498.74   44.9311    2.1168
     8   1886.37   56.5519    0.0000
     9   2071.00   62.0869   31.8359
    10   2071.00   62.0869   31.8359
    11   2142.58   64.2328    0.0000
    12   2142.58   64.2328    0.0000


************************************************************************

Results for rVV10 functional (after dynmat.x/asr='crystal')

# mode   [cm-1]    [THz]      IR
     1      0.00    0.0000    0.0000
     2      0.00    0.0000    0.0000
     3      0.00    0.0000    0.0000
     4     73.08    2.1909    0.0000
     5     73.08    2.1909    0.0000
     6    131.36    3.9381    0.0000
     7    757.03   22.6951    4.0269
     8    808.43   24.2360    0.0000
     9   1360.85   40.7973    0.0000
    10   1360.85   40.7973    0.0000
    11   1361.31   40.8111   54.8204
    12   1361.31   40.8111   54.8204

************************************************************************

Regards,

Claudio


-- 
*********************************************************************
Claudio A. Perottoni

Universidade de Caxias do Sul
Rua Francisco Getulio Vargas, 1130
95070-560 Caxias do Sul - RS - Brazil

http://www.researcherid.com/rid/B-8409-2008
*********************************************************************



---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20171127/dbc7c1aa/attachment.html>


More information about the users mailing list