[Pw_forum] lattice instability and local minimum energy postion

[明文美]

Dear users:
 
        I have calculated the phonon dispersion of zinc-blende CrSb. Before calculating phonon dispersion, I relaxed the CrSb lattice and got the optimized lattice parameters,which agreed well with that in previous literature, and after ph.x calculation I found that the two lower frequences at X(-1.0 0.0 0.0) are imaginary,which indicates the lattice instability of two modes.The frequencies are shown as following
 
    q =      -1.0000      0.0000      0.0000 **************************************************************************     omega( 1) =      -2.790113 [THz] =     -93.068763 [cm-1] (  0.000000   0.000000    -0.403189   0.000000     0.754599   0.000000   ) (  0.000000   0.000000     0.456614   0.000000    -0.243973   0.000000   )     omega( 2) =      -2.790113 [THz] =     -93.068763 [cm-1]  (  0.000000   0.000000    -0.754599   0.000000    -0.403189   0.000000   ) (  0.000000   0.000000    -0.243973   0.000000    -0.456614   0.000000   )     omega( 3) =       3.418041 [THz] =     114.014317 [cm-1]  (  0.000000   0.000000     0.000000   0.000000     0.000000   0.000000   ) ( -1.000000   0.000000     0.000000   0.000000     0.000000   0.000000   )     omega( 4) =       3.580039 [THz] =     119.418057 [cm-1] (  1.000000   0.000000     0.000000   0.000000     0.000000   0.000000   ) (  0.000000   0.000000     0.000000   0.000000     0.000000   0.000000   )     omega( 5) =       5.338839 [THz] =     178.085673 [cm-1] (  0.000000   0.000000    -0.817099   0.000000     0.000000   0.000000   ) (  0.000000   0.000000     0.000000   0.000000     0.576498   0.000000   )     omega( 6) =       5.338839 [THz] =     178.085673 [cm-1] (  0.000000   0.000000     0.000000   0.000000    -0.817099   0.000000   ) (  0.000000   0.000000     0.576498   0.000000     0.000000   0.000000   ) **************************************************************************
 
Usuallly, an instability implys that the atoms are not located  in local minimum energy postion,thus next I displaced atoms in the cell accroding to the eigen-displacements shown above,and got a cubic cell containg 8 atoms. But I found the energy rised with the increase of atom displacements, which was totally contrary to what I expected. In fact, only the decrease of energy with atom displacments will be consistent the results from phonon self-consitsten calculation.
 
The electronic structure calculation input file and phonon input file are list as : 
 
 
 
 
###    electronic structure calculation 
####
 &CONTROL                       title = CrSb ,                 calculation = 'scf' ,                restart_mode = 'from_scratch' ,                prefix='CrSb',                  pseudo_dir = '/home/phymwm/PWSCF/espresso-3.2.3/pseudo' , / &SYSTEM                       ibrav = 2,                   celldm(1) = 11.55,                         nat = 2,                        ntyp = 2,                        ecutwfc=60,                        ecutrho=600,                 occupations = 'smearing' ,                     degauss = 0.025 ,                    smearing = 'methfessel-paxton' ,                       nspin = 2 ,   starting_magnetization(1) = 3.0,   starting_magnetization(2) = 1.0, / &ELECTRONS            electron_maxstep = 200,                    conv_thr = 1.0D-8 , /ATOMIC_SPECIES   Cr   52.00000  Cr.pw91-sp-van.UPF   Sb  121.80000  Sb.pw91-n-van.UPFATOMIC_POSITIONS alat   Cr      0.000000000    0.000000000    0.000000000   Sb      0.250000000    0.250000000    0.250000000K_POINTS automatic  8 8 8   0 0 0 
 
#####
###
phonon dispersion for CrSb &inputph  tr2_ph=1.0d-16,  prefix='CrSb',  fildvscf='dv',  amass(1)=52.00,  amass(2)=121.8,  outdir='./',  fildyn='CrSb.dyn',  trans=.true.,  ldisp=.true.  nq1=4, nq2=4, nq3=4 /
 
 
 
Thanks in advance !
 
Wenmei Ming 
 
_________________________________________________________________
MSN 中文网，最新时尚生活资讯，白领聚集门户。
http://cn.msn.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20080731/3c3f73db/attachment.html>