Nd-Co-B compound only 'converging' to paramagnetic ground state!

[hillarymitchellwarden]

Hello everyone,

I am trying to calculate the magnetic ground state of a Nd-Co-B compound for which we have experimentally observed magnetic ordering. We suspect that the Nd and Co atoms are ordering antiferromagnetically with respect to each other, leading to a ferrimagnet overall (there is a 1:2 ratio of Nd:Co plus the difference in number of unpaired electrons for these atoms makes this seem likely to us).

I have been attempting to run spin-polarized calculations to see if this or some other model makes sense. However, I'm running into a big issue: if I use the Nd_3 PBE pseudopotential, the calculation always converges to a paramagnetic ground state. This happens no matter how I start my calculation (I've tried large MAGMOMs, smaller ones, a completely ferromagnetic model, only ordering in the Co atoms, and Nd/Co atoms ordered AFM to each other). I've tried this both with and without soc, but still always get a paramagnetic ground state. If I use the Nd PBE pseudopotential, the calculations don't converge. When my calculations reach the end of their iterations, they are not converged but do have magnetic moments on the atoms. Again, I've tried this both with and without soc.

Does anyone have any suggestions for how to deal with Nd correctly - I suspect that it is the cause of these difficulties, so any advice would be helpful. I can also provide more calculation details if that would be helpful.

Thanks!

[merzuk.kaltak]

Hello,
uploading the input files (INCAR, POSCAR, POTCAR, KPOINTS) might be useful for the community to help you with the problem.

[hillarymitchellwarden]

POSCAR:

Code: Select all

0.25(B8 Co8 Nd4)
1.0
        5.7203001976         0.0000000000         0.0000000000
        0.0554216267         6.5457653864         0.0000000000
        0.1492560353         0.3938169622         6.7756239517
   Nd   Co    B
    4    8    8
Direct
     0.634169996         0.518689990         0.254999995
     0.365830004         0.481310010         0.745000005
     0.883989990         0.012360000         0.243870005
     0.116010010         0.987640023         0.756129980
     0.342700005         0.207890004         0.130309999
     0.657299995         0.792109966         0.869690001
     0.106480002         0.389140010         0.380120009
     0.893519998         0.610859990         0.619879961
     0.195960000         0.677739978         0.125870004
     0.804040015         0.322260022         0.874130011
     0.393599987         0.894100010         0.379740000
     0.606400013         0.105899990         0.620260000
     0.429100007         0.907400012         0.076600000
     0.570899963         0.092599988         0.923399985
     0.062199999         0.392199993         0.069200002
     0.937799990         0.607800007         0.930800021
     0.855799973         0.314099997         0.577799976
     0.144200027         0.685899973         0.422200024
     0.638800025         0.810000002         0.572499990
     0.361199975         0.189999998         0.427500010

KPOINTS:

Code: Select all

NdCo2B2
0
Gamma
11 11 9
0 0 0

POTCAR:

Using either Nd/Nd_3 (as mentioned in original post), Co, and B PBE potentials

INCAR:

Code: Select all

System = You have lost this privilege
PREC=Accurate
INIWAV = 1
LREAL = .FALSE.
IWAVPR = 11
POTIM = 0.5
ISYM = -1
LWAVE = .FALSE.
LCHARG = .TRUE.
#IBRION = -1
EDIFF = 1E-5
NBANDS = 200
NCORES = 80
ISPIN = 2
MAGMOM = 0 0 1   0 0 1    0 0 1    0 0 1   0 0 -3   0 0 -3   0 0 -3   0 0 -3  0 0 -3  0 0 -3  0 0 -3   0 0 -3   0 0 0  0 0 0   0 0 0  0 0 0   0 0 0   0 0 0   0 0 0   0 0 0
SAXIS = 0 0 1
LMAXMIX = 6
LSORBIT = .TRUE.
LNONCOLLINEAR = .TRUE.

Any advice or help would be very appreciated!

[henrique_miranda]

Have you tried performing a constrained magnetic moment calculation?
Here is the tutorial:
wiki/index.php/Constraining_local_magnetic_moments

Here is a somehow related post:
forum/viewtopic.php?f=4&t=3237

PS: Note that a bug related to the constrained magnetic calculations was present in vasp 6.1.0 and vasp 6.1.1
This bug is not present in vasp 5.X.X
In case you are using vasp 6 please use the latest version vasp 6.1.2