[QE-users] TiO2 negative frequencies

[Connor Wilson]

Hi everyone,

I would like to use Quantum ESPRESSO to calculate the Raman intensity in rutile TiO2. The Raman-active modes I find agree very well with experiment, but ph.x also predicts a handful of negative-frequency modes which I cannot eliminate. I have tried increasing ecutwfc and the density of k-points in the SCF calculation, to no effect. I have attached my input/output files below. The lattice parameters/atomic coordinates result from a vc-relax calculation with a 7x7x11 k-mesh.

If anyone could suggest a fix for this, I would greatly appreciate it.

Thank you for your time,


Connor Wilson
Condensed Matter and Materials Science
Brock University
St. Catharines , ON, CAN


! pw.x
&CONTROL
  calculation='scf',
  restart_mode='from_scratch',
  prefix = 'TiO2',
  etot_conv_thr = 1.0d-6,
  forc_conv_thr = 1.0d-6,
  pseudo_dir = '/home/cw279/pseudo',
  outdir = '/home/cw279/projects/def-crandles/cw279/TiO2/ph',
/

&SYSTEM
  ibrav = 0,
  celldm(1) = 8.6924328,
  !celldm(3) = 0.6433300,
  nat = 6,
  ntyp = 2,
  ecutwfc = 200,
  ecutrho = 800,
  occupations = 'fixed',
/

&ELECTRONS
  mixing_beta = 0.7,
  conv_thr = 1.0d-10,
/

ATOMIC_SPECIES
Ti  47.867   Ti.pz-hgh.UPF
O   15.9994  O.pz-hgh.UPF

K_POINTS {automatic}
4 4 6  1 1 1

CELL_PARAMETERS {alat}
   0.982600556   0.000000000   0.000000000
   0.000000000   0.982600556   0.000000000
   0.000000000   0.000000000   0.629408068

ATOMIC_POSITIONS {crystal}
Ti            0.5000000000        0.5000000000        0.0000000000
Ti            0.0000000000        0.0000000000        0.5000000000
O             0.6959241507        0.6959241507        0.5000000000
O             0.1959241507        0.8040758493        0.0000000000
O             0.8040758493        0.1959241507        0.0000000000
O             0.3040758493        0.3040758493        0.5000000000


! ph.x
&INPUTPH
  tr2_ph = 1.0d-18,
  prefix = 'TiO2',
  amass(1) = 47.867,
  amass(2) = 15.9994,
  outdir = '/home/cw279/projects/def-crandles/cw279/TiO2/ph',
  epsil = .true.,
  trans = .true.,
  lraman = .true.,
  fildyn = 'TiO2.dyn',
/
0.0 0.0 0.0


! dynmat.in
&INPUT
   fildyn = 'TiO2.dyn',
   asr = 'simple',
/


! dynmat.out
# mode   [cm-1]    [THz]      IR          Raman   depol.fact
    1   -165.11   -4.9500  132.3930         0.0000    0.3695
    2   -113.43   -3.4004   91.9463         0.0000    0.6303
    3   -113.43   -3.4004   91.9463         0.0000    0.5188
    4    -87.89   -2.6347    0.0000         0.0000    0.2925
    5     -0.00   -0.0000    0.0000         0.0000    0.4869
    6     -0.00   -0.0000    0.0000         0.0000    0.3389
    7    0.00     0.0000    0.0000         0.0000    0.3544
    8    122.17    3.6624    0.0000         0.3837    0.7500
    9    372.09   11.1551    0.0000         0.0000    0.7144
   10    375.77   11.2654    3.9843         0.0000    0.3068
   11    375.77   11.2654    3.9843         0.0000    0.4022
   12    421.88   12.6475    0.0000         0.0000    0.7072
   13    447.56   13.4176    0.0000   239.3889    0.7500
   14    447.56   13.4176    0.0000   239.3889    0.7500
   15    480.43   14.4028   14.4178         0.0000    0.5190
   16    480.43   14.4028   14.4178         0.0000    0.7318
   17    601.42   18.0300    0.0000  1203.1157    0.2824
   18    815.02   24.4338    0.0000         5.3306    0.7500
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20230101/edfc2434/attachment.html>