My Community

Dear VASP users,

I am calculating band structure for Pb, Sn and Te alloy. I made a supercell by doubling the primitive lattice vectors. It has fcc (NaCl) structures. With primitive lattice vectors there are 2 basis vectors for fcc structure (one for Na and one for Cl). Once I have supercell it will have impurity included (in my case Sn doping) in supercell.

Following is my POSCAR file for alloy. I included all the atoms that lie within the supercell. There are 3 Sn atoms to get 20% doping that replaces Pb atoms (Pb_0.8Sn_0.2Te).

SYSTEM = PbSnTe
1.00000000000000
0.00000000 6.55000000 6.55000000
6.55000000 0.00000000 6.55700000
6.55000000 6.55000000 0.00000000
11 3 13
Cartesian
0.00000000 0.00000000 0.00000000 #Pb
6.55000000 0.00000000 0.00000000 #Pb
6.55000000 6.55000000 0.00000000 #Pb
0.00000000 6.55000000 0.00000000 #Pb
3.27500000 0.00000000 3.27500000 #Pb
6.55000000 3.27500000 3.27500000 #Pb
3.27500000 6.55000000 3.27500000 #Pb
6.55000000 0.00000000 6.55000000 #Pb
0.00000000 0.00000000 6.55000000 #Pb
6.55000000 6.55000000 6.55000000 #Pb
0.00000000 6.55000000 6.55000000 #Pb
3.27500000 3.27500000 0.00000000 #Sn
0.00000000 3.27500000 3.27500000 #Sn
3.27500000 3.27500000 6.55000000 #Sn
0.00000000 3.27500000 0.00000000 #Te
3.27500000 6.55000000 0.00000000 #Te
3.27500000 0.00000000 0.00000000 #Te
6.55000000 3.27500000 0.00000000 #Te
0.00000000 0.00000000 3.27500000 #Te
6.55000000 0.00000000 3.27500000 #Te
3.27500000 3.27500000 3.27500000 #Te
6.55000000 6.55000000 3.27500000 #Te
0.00000000 6.55000000 3.27500000 #Te
3.27500000 0.00000000 6.55000000 #Te
6.55000000 3.27500000 6.55000000 #Te
0.00000000 3.27500000 6.55000000 #Te
3.27500000 6.55000000 6.55000000

Following is the INCAR file.


SYSTEM = PbSnTe
ISTART = 0 # 0 start fr. scratch, no wavefunction WAVECAR need,
# 2 start from previous WAVECAR
ICHARG = 2 # 2 start fr. scratch,
# 1 read from CHGCAR,
# 0 Calculate from WAVE function
# 11 read charge density from CHGCAR
#and keep constant during the subsequent run
LREAL=.FALSE. # reciprocal space projection scheme .FALSE.
# automatic real space projection scheme AUTO?
ISPIN = 2 # non spin polarized calculations 1
# spin polarized calculations 2
LSORBIT = .TRUE. # Include spin-orbit intercation
#LORBIT = 11
#LORBMOM = 11
#MAGMOM = 2*1
LMAXMIX = 4
SAXIS = 0 0 1 # Direction of spin axis
EDIFF = 0.0001 # Convergency of ELECTRONIC iteration
EDIFFG = -0.01 # Convergency of IONIC iteration
# "-" mean the condition for FORCE
ENCUT = 250 # Energy cuttoff, need to test your value
LWAVE = .TRUE. # Save wavefunction WAVECAR or not
LCHARG = .TRUE. # Save CHGCAR charge density
NELM = 200 # Maximum number of electronic iteration
NSW = 0 # maximum number of ionic iteration
# 0 or 1 mean there is no relaxation (fix everything)

IALGO = 48
#NGX=24
#NGY=24
#NGZ=24

#GGA = PE # This is defult in case of PAW_PBE.

IBRION = -1 # -1 Ions won't move
# 2 Do relaxation, ions will move after every ionic iteration
#ISIF = 2 # stress, relaxation, volume change
POTIM = 0.10 # time step for ionic-motion
TEBEG = 0.0 # start temperature
TEEND = 0.0 # final temperature
ISMEAR = -5 # 0 gaussian smearing
# -5 tetrahedron method
SIGMA = 0.2 # Width of smearing
SYMPREC = 1e-7
ISYM = 2



Now my question is that I always get an error in VASP run. Error is the following:

POSCAR, INCAR and KPOINTS ok, starting setup
WARNING: wrap around errors must be expected
LAPACK: Routine ZPOTRF failed! 10
LAPACK: Routine ZPOTRF failed! 10

Could you please give me what would be the solution for this?

Thank you.

1) aren't you missing one Te atom?
2) check the compilation/installation of LAPACK, and linking to it in VASP

Danny,

There are 13 Te atoms in the POSCAR file and in total 27 atoms. Is my supercell is ok which I made multiplying by 2 to all lattice vectors for the unit cell?
I did not have this problem before when I had small system (eg. PbTe which has fcc structure as well). Calculation was ok. Could it be still problem with LAPACK?

Thank you.

if your cell is obtained by multiplying the lattice vectors by two the number of atoms in your cell should be an eight-fold (or if you did not multiply all lattice vectors at least a twofold, aka an even number)