error in convergence
Posted: Fri Jul 13, 2012 6:52 am
Dear Abinit Users
I am working on phase change materials which have composition (Ge2Sb2Te5). I have carried out structure relaxation successfully on supercell having 58 atoms (32 Te atoms, 13-Germanium atoms, 13 Antimony atoms and 6 vacancies) for these materials. This material has rock salt structure. But when I dope this materials with Tin, replacing three germanium atoms with tin atoms, there is problem in convergence in some time steps. It converges for first two iterations where as it fails to converge for next iterations with Total energy like
ETOT 1 -483.27013332199 -4.833E+02 2.862E-05 4.034E+03 3.926E-02 6.952E-02
ETOT 2 -483.34344528548 -7.331E-02 4.228E-06 1.258E+03 2.389E-02 4.563E-02
ETOT 3 -483.37978241776 -3.634E-02 3.850E-06 1.234E+02 3.487E-02 1.327E-02
ETOT 4 -483.38271544981 -2.933E-03 6.860E-07 3.887E+01 1.093E-02 1.617E-02
ETOT 5 -483.38359859723 -8.831E-04 2.608E-07 2.760E+01 8.591E-03 2.272E-02
ETOT 6 -483.38402102748 -4.224E-04 5.501E-08 7.162E+00 1.557E-03 2.376E-02
ETOT 7 -483.38105168408 2.969E-03 3.121E-08 1.146E+02 1.723E-02 2.491E-02
ETOT 8 -483.38415045183 -3.099E-03 4.436E-09 2.847E+00 1.504E-02 2.399E-02
ETOT 9 -483.38404951578 1.009E-04 9.700E-10 6.358E+00 2.252E-03 2.442E-02
ETOT 10 -483.38411689736 -6.738E-05 6.430E-12 4.016E+00 8.496E-04 2.437E-02
ETOT 11 -483.38414970944 -3.281E-05 1.563E-10 3.017E+00 1.334E-03 2.427E-02
ETOT 12 -483.38411435640 3.535E-05 6.753E-11 3.815E+00 1.624E-03 2.446E-02
ETOT 13 -483.38414977061 -3.541E-05 9.369E-11 2.626E+00 6.814E-04 2.439E-02
ETOT 14 -483.38415914727 -9.377E-06 1.289E-10 2.263E+00 1.877E-04 2.437E-02
ETOT 15 -483.37672877615 7.430E-03 2.417E-10 2.983E+02 1.355E-02 2.694E-02
.............................................................................
My input file is
acell 3*21.00
diemix 0.333333d0
ecut 45
enunit 2
intxc 0
ntime 60
nline 6
nnsclo 4
nsym 1
occopt 1
ortalg -2
prtcml 1
prtvol 0
rprim 1 0 0
0 1 0
0 0 1
kptopt 1
nshiftk 1
shiftk 0.5 0.5 0.5
ngkpt 1 1 1
#Definition of the atom types
ntypat 4
znucl 32 50 51 52
#Definition of the atoms
natom 58
typat 4 4 4 4 4 . . . . .
xcart 0.0000000000E+00 0.0000000000E+00 0.0000000000E+00
1.0500000000E+01 0.0000000000E+00 0.0000000000E+00
0.0000000000E+00 1.0500000000E+01 0.0000000000E+00
1.0500000000E+01 1.0500000000E+01 0.0000000000E+00
0.0000000000E+00 0.0000000000E+00 1.0500000000E+01
1.0500000000E+01 0.0000000000E+00 1.0500000000E+01
..........................................................
toldff 5.0d-6
iprcel 45
npulayit 16
# Definition of the SCF procedure
nstep 200
diemac 35
optcell 2 ionmov 3
ndtset 2
#Definition of the atomic positions without relaxing the lattice parameters
optcell1 0
ionmov1 3
#Definition of the basis of the geometry determined by dataset 1 (so using getxcart 1)
getxcart2 1
optcell2 2
ionmov2 3
dilatmx2 1.2
ecutsm2 0.5
I m using Troullier-Martins pseudo-potentials. Is this error related to pseudo-potentials. Kindly help.
Cheers
Dr Singh
I am working on phase change materials which have composition (Ge2Sb2Te5). I have carried out structure relaxation successfully on supercell having 58 atoms (32 Te atoms, 13-Germanium atoms, 13 Antimony atoms and 6 vacancies) for these materials. This material has rock salt structure. But when I dope this materials with Tin, replacing three germanium atoms with tin atoms, there is problem in convergence in some time steps. It converges for first two iterations where as it fails to converge for next iterations with Total energy like
ETOT 1 -483.27013332199 -4.833E+02 2.862E-05 4.034E+03 3.926E-02 6.952E-02
ETOT 2 -483.34344528548 -7.331E-02 4.228E-06 1.258E+03 2.389E-02 4.563E-02
ETOT 3 -483.37978241776 -3.634E-02 3.850E-06 1.234E+02 3.487E-02 1.327E-02
ETOT 4 -483.38271544981 -2.933E-03 6.860E-07 3.887E+01 1.093E-02 1.617E-02
ETOT 5 -483.38359859723 -8.831E-04 2.608E-07 2.760E+01 8.591E-03 2.272E-02
ETOT 6 -483.38402102748 -4.224E-04 5.501E-08 7.162E+00 1.557E-03 2.376E-02
ETOT 7 -483.38105168408 2.969E-03 3.121E-08 1.146E+02 1.723E-02 2.491E-02
ETOT 8 -483.38415045183 -3.099E-03 4.436E-09 2.847E+00 1.504E-02 2.399E-02
ETOT 9 -483.38404951578 1.009E-04 9.700E-10 6.358E+00 2.252E-03 2.442E-02
ETOT 10 -483.38411689736 -6.738E-05 6.430E-12 4.016E+00 8.496E-04 2.437E-02
ETOT 11 -483.38414970944 -3.281E-05 1.563E-10 3.017E+00 1.334E-03 2.427E-02
ETOT 12 -483.38411435640 3.535E-05 6.753E-11 3.815E+00 1.624E-03 2.446E-02
ETOT 13 -483.38414977061 -3.541E-05 9.369E-11 2.626E+00 6.814E-04 2.439E-02
ETOT 14 -483.38415914727 -9.377E-06 1.289E-10 2.263E+00 1.877E-04 2.437E-02
ETOT 15 -483.37672877615 7.430E-03 2.417E-10 2.983E+02 1.355E-02 2.694E-02
.............................................................................
My input file is
acell 3*21.00
diemix 0.333333d0
ecut 45
enunit 2
intxc 0
ntime 60
nline 6
nnsclo 4
nsym 1
occopt 1
ortalg -2
prtcml 1
prtvol 0
rprim 1 0 0
0 1 0
0 0 1
kptopt 1
nshiftk 1
shiftk 0.5 0.5 0.5
ngkpt 1 1 1
#Definition of the atom types
ntypat 4
znucl 32 50 51 52
#Definition of the atoms
natom 58
typat 4 4 4 4 4 . . . . .
xcart 0.0000000000E+00 0.0000000000E+00 0.0000000000E+00
1.0500000000E+01 0.0000000000E+00 0.0000000000E+00
0.0000000000E+00 1.0500000000E+01 0.0000000000E+00
1.0500000000E+01 1.0500000000E+01 0.0000000000E+00
0.0000000000E+00 0.0000000000E+00 1.0500000000E+01
1.0500000000E+01 0.0000000000E+00 1.0500000000E+01
..........................................................
toldff 5.0d-6
iprcel 45
npulayit 16
# Definition of the SCF procedure
nstep 200
diemac 35
optcell 2 ionmov 3
ndtset 2
#Definition of the atomic positions without relaxing the lattice parameters
optcell1 0
ionmov1 3
#Definition of the basis of the geometry determined by dataset 1 (so using getxcart 1)
getxcart2 1
optcell2 2
ionmov2 3
dilatmx2 1.2
ecutsm2 0.5
I m using Troullier-Martins pseudo-potentials. Is this error related to pseudo-potentials. Kindly help.
Cheers
Dr Singh