[Pw_forum] Can not find minimum free energy by QHA

[GAO Zhe]

 Dear QE Developer and Users
I am using DFPT+QHA to predict model's properties at high temperature. According to the paper,Ab initio lattice dynamics and phase transformations of ZrO2, Phy Rev B 71, 064301, free energyv.s. volume figure can be plotted at certain temperature for EOS-fitting and so on.
Having determined the equlibrium lattice parameter, 8.18 a.u. , of TiC(FCC) by both EOS-fitting and vc-relax, I chose 20 different lattice paramters (7.98~8.38a.u.) and calculated phonon dispersion and thermodynamic properties. However, the free energy kept decreasing as a monotone decreasing function of lattice parameter, even when imaginary frequency began to occur at 8.88a.u..
My DFPT calculation script shows below:
cat >TiC_${INT}.scf.in <<EOF
&control
     calculation = 'scf' ,
     prefix = 'TiC_${INT}' ,
     outdir = '$TempDIR/' ,
     pseudo_dir = '$PseudoDIR/' ,
     disk_io = 'low' ,
     etot_conv_thr = 1.0d-9 ,
     tstress = .true. ,
/
&system
     ibrav = 2 ,
     celldm(1) = $OP ,
     ecutwfc = 42.5 ,
     ecutrho = 400 ,
     nat = 2 ,
     ntyp = 2 ,
     occupations = 'smearing' ,
     smearing = 'mp' ,
     degauss = 0.02 ,
/
&electrons
     conv_thr = 1.0d-9 ,
     diagonalization = 'cg' ,
     startingpot = 'atomic' ,
     startingwfc = 'atomic+random' ,
     mixing_mode = 'plain' ,
/
ATOMIC_SPECIES
  Ti  47.900  Ti.USPP_PW91.UPF
   C  12.011  C.USPP_PW91n.UPF
ATOMIC_POSITIONS crystal
  Ti  0.00  0.00  0.00
   C  0.50  0.50  0.50
K_POINTS automatic
  11  11  11   0  0  0
EOF
mpirun -n 6 pw.x -npool 2 <TiC_${INT}.scf.in >TiC_${INT}.scf.out

cat >TiC_${INT}.ph.in <<EOF
Phonon of TiC at $OP Bohr
&inputph
     outdir = '$TempDIR/' ,
     prefix = 'TiC_${INT}' ,
     nq1 = 3 ,
     nq2 = 3 ,
     nq3 = 3 ,
     reduce_io = .true. ,
     ldisp = .true. ,
     tr2_ph = 1.0d-13 ,
     fildyn = 'TiC_${INT}.dyn' ,
/
EOF
mpirun -n 6 ph.x -npool 2 <TiC_${INT}.ph.in >TiC_${INT}.ph.out
QHA is calculated by Prof. Isaev's code which released with QE package.
Any suggestion will be welcome.
Thanks.



--
GAO Zhe
CMC Lab, MSE, SNU, Seoul, S.Korea
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20110822/92810b29/attachment.html>