The projection centre specified by the LOCPROJ tag is outside the unit cell.

[yutoml]

The following specifications were made in LOCPROJ tag in INCAR.

Code: Select all

LOCPROJ = ‘1 2 3 : d : Pr’

However, in the resulting LOCPROJ file, the projection centre has negative coordinates and is outside the unit cell(see ISITE: 2 and ISITE 3:).

Code: Select all

1  1400    30    15  # of spin, # of k-points, # of bands, # of proj
   ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dxy
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dyz
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dz2
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dxz
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :   dx2-y2
    ISITE:     2    R=     -0.3333333     0.3333333    -0.5000000  PAW projector    :    dxy
      ISITE:     2    R=     -0.3333333     0.3333333    -0.5000000  PAW projector    :    dyz
      ISITE:     2    R=     -0.3333333     0.3333333    -0.5000000  PAW projector    :    dz2
      ISITE:     2    R=     -0.3333333     0.3333333    -0.5000000  PAW projector    :    dxz
      ISITE:     2    R=     -0.3333333     0.3333333    -0.5000000  PAW projector    :   dx2-y2
    ISITE:     3    R=      0.3333333    -0.3333333    -0.5000000  PAW projector    :    dxy
      ISITE:     3    R=      0.3333333    -0.3333333    -0.5000000  PAW projector    :    dyz
      ISITE:     3    R=      0.3333333    -0.3333333    -0.5000000  PAW projector    :    dz2
      ISITE:     3    R=      0.3333333    -0.3333333    -0.5000000  PAW projector    :    dxz
      ISITE:     3    R=      0.3333333    -0.3333333    -0.5000000  PAW projector    :   dx2-y2

The output LOCPROJ file I expect is as follows.

Code: Select all

1  1400    30    15  # of spin, # of k-points, # of bands, # of proj
   ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dxy
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dyz
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dz2
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :    dxz
      ISITE:     1    R=      0.0000000     0.0000000     0.0000000  PAW projector    :   dx2-y2
    ISITE:     2    R=     0.6666666     0.3333333    0.5000000  PAW projector    :    dxy
      ISITE:     2    R=     0.6666666     0.3333333    0.5000000  PAW projector    :    dyz
      ISITE:     2    R=     0.6666666     0.3333333    0.5000000  PAW projector    :    dz2
      ISITE:     2    R=     0.6666666     0.3333333    0.5000000  PAW projector    :    dxz
      ISITE:     2    R=     0.6666666     0.3333333    0.5000000  PAW projector    :   dx2-y2
    ISITE:     3    R=      0.3333333    0.6666666    0.5000000  PAW projector    :    dxy
      ISITE:     3    R=      0.3333333    0.6666666    0.5000000  PAW projector    :    dyz
      ISITE:     3    R=      0.3333333    0.6666666    0.5000000  PAW projector    :    dz2
      ISITE:     3    R=      0.3333333    0.6666666    0.5000000  PAW projector    :    dxz
      ISITE:     3    R=      0.3333333    0.6666666    0.5000000  PAW projector    :   dx2-y2

Is there a way to avoid this?
I am using the following as POSCAR.

Code: Select all

Ti3
1.0
   3.9412204462015445    2.2754646855501184    0.0000000000000000
  -3.9412204462015445    2.2754646855501184   -0.0000000000000000
   0.0000000000000000   -0.0000000000000000    2.8178036989832456
Ti
3
direct
   0.0000000000000000    0.0000000000000000    0.0000000000000000 Ti
   0.6666666666666643    0.3333333333333357    0.5000000000000000 Ti
   0.3333333333333357    0.6666666666666643    0.5000000000000000 Ti

[ferenc_karsai]

Please upload your calculation according to the forum guidelines so that we can try to reproduce your problems.

[yutoml]

Sorry, I overlooked the guidelines.

I am attaching the relevant input/output files as locproj.tar.zst.
I am using "Ti" as POTCAR
However, I could not upload the essential LOCPROJ file due to the size limitation.

Thank you in advance.

[ferenc_karsai]

In the LOCPROJ file the coordinates are mapped to -0.5 to 0.5 instead of 0 to 1.

There is no way to avoid that, but you can easily transform them back. Add 1.0 to all values that are smaller than 0.

[yutoml]

I’m sorry, my previous explanation was insufficient.
After this VASP step, I would like to perform further analysis using Wannier90 with the obtained output. In the subsequent Wannier90 process, when plotting Wannier functions (WF), I would like the WF centers to be located inside the unit cell.

Although I do not yet fully understand Wannier90, it seems that the initial WF centers in Wannier90 are the same as those given by the LOCPROJ file output from VASP. Therefore, I thought that if the centers in the VASP output were kept inside the unit cell, this issue could be resolved, which is why I asked this question. Does this mean that, in VASP, it is not possible to keep the centers of localized functions within the unit cell?

[yutoml]

This issue has already been resolved.
I found that by adjusting the way I operate the plotting software, I was able to obtain the plots I wanted.
Thank you very much for your support.

[ferenc_karsai]

Thanks for writing us that you have solved the problem.

I will now lock this issue.