MAGATOM: POSCAR section in vasp.wrap and MAGMOM

Alloy Theoretic Automated Toolkit (ATAT)
1

There is a bug when defining the POSCAR section in the vasp.wrap file to specify the magnetic moment on each atom via MAGATOM. If then the POSCAR section is included, then the code just appends the generated new structure to the vasp.wrap file to build vasp.in, Therefore when running runstruct_vasp, the pristine structure of the POSCAR section in the vasp.wrap file is used to create the VASP-POSCAR file, meaning I don’t get the generated structures but always the pristine one. Consequently, the magnetic moments of the pristine sructure are obtained as well. I guess it should be looked at.

This is the vasp.wrap:

[INCAR]
ISTART = 0
ICHARG = 2
PREC = Accurate
ENCUT = 520
ALGO = V
EDIFF = 1E-8
SIGMA = 0.01
ISPIND = 2
ISPIN = 2

MAGATOM =

Geom Opt
IBRION = 2
NSW = 200
ISIF = 3
POTIM = 0.5

NSIM = 2
NPAR = 32
KPAR = 32

KPPRA = 6000

LDAU = .TRUE.
LDAPRINT = 2
LDAUTYPE = 2
LDAUL = -1 2 -1
LDAUU = 0 5.6 0
LDAUJ = 0 0 0

LREAL = .FALSE.
LPLANE = .TRUE.
LSCALU = .FALSE.
ADDGRID = .TRUE.

[POSCAR]
LiCoO2
1.00000000000000
2.810500 0.000000 0.000000
-1.405250 2.433964 0.000000
0.000000 0.000000 14.050000
Direct
0.666667 0.333333 0.593497 O+0
0.333333 0.666667 0.926831 O+0
1.000000 1.000000 0.260164 O+0
1.000000 1.000000 0.739836 O+0
0.333333 0.666667 0.406503 O+0
0.666667 0.333333 0.073169 O+0
0.666667 0.333333 0.333333 Co+5
0.333333 0.666667 0.666667 Co+5
1.000000 1.000000 1.000000 Co+5
0.666667 0.333333 0.833333 Li+0
0.333333 0.666667 0.166667 Li+0
1.000000 1.000000 0.500000 Li+0

This is lat.in:

2.8105 2.8105 14.05 90 90 120
1 0 0
0 1 0
0 0 1
0.6666666670000012 0.3333333329999988 0.5934973479707178 O
0.3333333329999988 0.6666666670000012 0.9268313479707114 O
0.0000000000000000 0.0000000000000000 0.2601643479707146 O
0.0000000000000000 0.0000000000000000 0.7398356520292854 O
0.3333333329999988 0.6666666670000012 0.4065026520292822 O
0.6666666670000012 0.3333333329999988 0.0731686520292885 O
0.6666666670000012 0.3333333329999988 0.3333333329999988 Co
0.3333333329999988 0.6666666670000012 0.6666666670000012 Co
0.0000000000000000 0.0000000000000000 0.0000000000000000 Co
0.6666666670000012 0.3333333329999988 0.8333333329999988 Li,Vac
0.3333333329999988 0.6666666670000012 0.1666666670000012 Li,Vac
0.0000000000000000 0.0000000000000000 0.5000000000000000 Li,Vac

This is POSCAR, vasp.in and INCAR od structure 3:

LiCoO2
1.00000000000000
2.810500 0.000000 0.000000
-1.405250 2.433964 0.000000
0.000000 0.000000 14.050000
3 3 6
Direct
0.666667 0.333333 0.333333 Co+5
0.333333 0.666667 0.666667 Co+5
1.000000 1.000000 1.000000 Co+5
0.666667 0.333333 0.833333 Li+0
0.333333 0.666667 0.166667 Li+0
1.000000 1.000000 0.500000 Li+0
0.666667 0.333333 0.593497 O+0
0.333333 0.666667 0.926831 O+0
1.000000 1.000000 0.260164 O+0
1.000000 1.000000 0.739836 O+0
0.333333 0.666667 0.406503 O+0
0.666667 0.333333 0.073169 O+0

[INCAR]
ISTART = 0
ICHARG = 2
PREC = Accurate
ENCUT = 520
ALGO = V
EDIFF = 1E-8
SIGMA = 0.01
ISPIND = 2
ISPIN = 2

MAGATOM =

Geom Opt
IBRION = 2
NSW = 200
ISIF = 3
POTIM = 0.5

NSIM = 2
NPAR = 32
KPAR = 32

KPPRA = 6000

LDAU = .TRUE.
LDAPRINT = 2
LDAUTYPE = 2
LDAUL = -1 2 -1
LDAUU = 0 5.6 0
LDAUJ = 0 0 0

LREAL = .FALSE.
LPLANE = .TRUE.
LSCALU = .FALSE.
ADDGRID = .TRUE.

[POSCAR]
LiCoO2
1.00000000000000
2.810500 0.000000 0.000000
-1.405250 2.433964 0.000000
0.000000 0.000000 14.050000
Direct
0.666667 0.333333 0.593497 O+0
0.333333 0.666667 0.926831 O+0
1.000000 1.000000 0.260164 O+0
1.000000 1.000000 0.739836 O+0
0.333333 0.666667 0.406503 O+0
0.666667 0.333333 0.073169 O+0
0.666667 0.333333 0.333333 Co+5
0.333333 0.666667 0.666667 Co+5
1.000000 1.000000 1.000000 Co+5
0.666667 0.333333 0.833333 Li+0
0.333333 0.666667 0.166667 Li+0
1.000000 1.000000 0.500000 Li+0

[POSCAR]
title
1.
1.405250000 -2.433964000 0.000000000
-2.810500000 0.000000000 0.000000000
0.000000000 0.000000000 -14.050000000
Cartesian
0.000001405 -1.622643478 -2.341671350 Vac
-1.405251405 -0.811320522 -11.708328650 Vac
-1.405250000 -2.433964000 -7.025000000 Li
0.000001405 -1.622643478 -5.711367150 O
-1.405251405 -0.811320522 -1.028024450 O
-1.405250000 -2.433964000 -10.394695800 O
-1.405250000 -2.433964000 -3.655304200 O
-1.405251405 -0.811320522 -8.338632850 O
0.000001405 -1.622643478 -13.021975550 O
0.000001405 -1.622643478 -9.366671350 Co
-1.405251405 -0.811320522 -4.683328650 Co
-1.405250000 -2.433964000 -14.050000000 Co

ISTART = 0
ICHARG = 2
PREC = Accurate
ENCUT = 520
ALGO = V
EDIFF = 1E-8
SIGMA = 0.01
ISPIND = 2
ISPIN = 2
Geom Opt
IBRION = 2
NSW = 200
ISIF = 3
POTIM = 0.5
NSIM = 2
NPAR = 32
KPAR = 32
LDAU = .TRUE.
LDAPRINT = 2
LDAUTYPE = 2
LDAUL = -1 2 -1
LDAUU = 0 5.6 0
LDAUJ = 0 0 0
LREAL = .FALSE.
LPLANE = .TRUE.
LSCALU = .FALSE.
ADDGRID = .TRUE.
MAGMOM= 5 5 5 0 0 0 0 0 0 0 0 0

2

If MAGMOM is specified, then the values are not adapted to accomodate the number of atoms in the generated structures. Also they are do not change the order when atoms are rearranged in the POSCAR in terms of in which line they are placed.

3

The vasp.wrap file should not contain a [POSCAR] section!
That section will be added by runstruct_vasp, based on str.out

Caution: make sure that you specify the parameters of the LDAUL LDAUU and LDAUJ tags in alphabetical order of the chemical species (because species are ordered in that way in the generated POSCAR file).

Otherwise the input files seem ok. You correctly use MAGATOM = with an empty right-hand side
and then the str.out will contain the magmom as Co+5 Li+0 and so forth as the species names.