[Pw_forum] Confused about vibrational modes of isolated molecule from ph.x

[Jin Zhang]

Hi all,

   I posted this issue one week ago. Now I have relaxed the structure and
got total force less than 5.0D-5. I used a large cubic supercell with 30bohr
alat and ecutwfc up to 50Ry. However, things only went worse. The highest
frequency from ph.x dropped from 3187 cm^(-1) to 1849, while gaussian03
shows it should be around 3300cm^(-1).

I also checked H2 molecule. In this case, I got three negative values and
three positive ones( values are 898,898,3679cm^-1 ). But there should be
only ONE vibrational frequency which corresponds to the stretching mode
along bonding direction. And the highest freq. is also too low (should be
around 43xx).

I really couldn't figure out what's wrong. Can someone help me? Thanks in
advance!

The input files for H2 are:
---------------------------scf_input-----------------(run two times just
like example07 )-----
 &control
    calculation='scf'
    restart_mode='from_scratch',
    prefix='h2',
    pseudo_dir = '$PSEUDO_DIR/',
    outdir='$TMP_DIR/'
 /
 &system
    ibrav=  8, celldm(1) =18, celldm(2)=1.0, celldm(3)=1.0,
    nat= 2, ntyp= 1,
    ecutwfc = 50.0,
    occupations='smearing', smearing='gausssian', degauss=0.01,
    la2F = .false., nspin=1,
 /
 &electrons
    conv_thr =  1.0d-8
    mixing_beta = 0.7
 /
ATOMIC_SPECIES
  H     1.008    H.blyp-vbc.UPF
ATOMIC_POSITIONS (bohr)
H        0.000000000   0.000000000  -0.707244771
H        0.000000000   0.000000000   0.707244771
K_POINTS {automatic}
1 1 1 0 0 0
------------------------------ph_input---------------------------
 &inputph
  tr2_ph=1.0d-8,
  prefix='h2',
  fildvscf='h2dv',
  amass(1)=1.008,
  outdir='$TMP_DIR/',
  fildyn='h2.dyn',
  elph=.false.,
  trans=.true.,
  ldisp=.true.,
  nq1=1, nq2=1, nq3=1
--------------------------------------------------------------------

Regards,
Jin Zhang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20061127/b17f5b38/attachment.html>