[Pw_forum] the negative phonon mode of ZnS
Sun ShaoRui
sunsr at ihep.ac.cn
Mon Jan 31 08:38:11 CET 2005
Dear all:
I have calculated the phonon of ZnS with zincblende structure under anbient
condition, and I get the pseudopotentials from the pwscf web page. For the
gamma point, the result seems right, but fro the X point, there are some
negative frequency modes in the output file.
The pw input file is shown as follows:
-----------------------------------------------------------------------------
----------------
&CONTROL
calculation = 'phonon',
restart_mode = 'from_scratch',
ndr = 51,
ndw = 51,
nstep = 500,
iprint = 10,
tprnfor = .TRUE.,
dt = 5.0d0,
etot_conv_thr = 1.d-9,
forc_conv_thr = 1.d-4,
pseudo_dir = '/home/wien2k/pw/pseudo',
outdir = '/home/wien2k/pwwork/ZnS'
prefix = 'ZnS-X'
/
&SYSTEM
ibrav = 2,
celldm(1) = 10.231926,
nat = 2,
ntyp = 2,
ecutwfc = 20.0,
smearing = 'f-d'
nspin = 1,
/
&ELECTRONS
mixing_beta = 0.7,
conv_thr = 1.0d-8,
/
&phonon
xqq(1)=1.0, xqq(2)=0.0, xqq(3)=0.0
/
ATOMIC_SPECIES
S 32.0d0 S.pbe-van_bm.UPF
Zn 65.4d0 Zn.pbe-van.UPF
ATOMIC_POSITIONS (alat)
Zn 0.0000 0.0000 0.0000 1 1 1
S 0.2500 0.2500 0.2500 1 1 1
K_POINTS automatic
8 8 8 1 1 1
******************************************
And the ph input file is shown as follows:
*******************************************
phonons of ZnS at X
&inputph
tr2_ph=1.0d-14,
prefix='ZnS-X',
amass(1)=65.39,
amass(2)=32.066,
outdir='/home/wien2k/pwwork/ZnS/',
fildyn='ZnS-X.dyn',
/
1.0 0.0 0.0
*********************************************************
The phonon mode list in the ph output program is shown as follows:
**************************************************************************
omega( 1) = -32.553715 [THz] = -1085.882263 [cm-1]
omega( 2) = -32.447538 [THz] = -1082.340572 [cm-1]
omega( 3) = -32.447538 [THz] = -1082.340572 [cm-1]
omega( 4) = 10.118904 [THz] = 337.532556 [cm-1]
omega( 5) = 10.118904 [THz] = 337.532556 [cm-1]
omega( 6) = 10.391189 [THz] = 346.615057 [cm-1]
**************************************************************************
I am puzzled about the three negative frequency modes for the stable state.
I also tried to use other kinds of pseudopotentials which are produced by
the fhi98PP and uspp, respectively, and increase the energy cut-off to 60
Ry.. But all the results are similar.
I think for the stable crystal system, all the phonon mode should not be
negative. From the former discussion, somebody have talked about the reason
about the negative mode, and I have tried to consider that in my computation.
Who could give me some suggestions and comments?
Best regards!
Shaorui Sun
More information about the users
mailing list