Incompatibility Warning Between POSCAR and POTCAR Files

[Sherif_Yehia1]

Dear VASP Community,

I am currently running a VASP calculation for a Ho2Co17 system. However, I have encountered a warning message that I’m unsure about. The warning message is as follows:
POSCAR found type information on POSCAR HoCo
POSCAR found : 2 types and 38 ions
scaLAPACK will be used
WARNING: type information on POSCAR and POTCAR are incompatible
POTCAR overwrites the type information in POSCAR
from my understanding, this warning suggests that the type information in my POSCAR and POTCAR files are not matching. I understand that the POSCAR file specifies the types and positions of atoms in the system, while the POTCAR file provides the pseudopotential for each atomic species used in the calculation. If the cell contains different atomic species, the corresponding POTCAR files have to be concatenated, in the same order as the atomic species are given in the POSCAR file.

I have checked the order and types of atoms in both my POSCAR and POTCAR files, and they seem to be consistent. However, the warning persists. (I must be doing something wrong)

I would appreciate any advice or suggestions on how to resolve this issue. Is this warning something I should be concerned about? Could it potentially affect the results of my calculation?

Thank you in advance for your help.

Best regards, Sherif Yehia

[jonathan_lahnsteiner2]

Dear Sherif_Yehia1,

Please supply your input files, especially the POSCAR and POTCAR files as described in the vasp forum guidelines.
Otherwise, it is hard to tell from where the warning you are getting originates.

All the best Jonathan

[Sherif_Yehia1]

Dear Jonathan,

Thank you for your request. Here are the POSCAR and POTCAR files for Ho2Co17. as you request according to the vasp forum rules.

Thank you ,

Sherif

[jonathan_lahnsteiner2]

Dear Sherif_Yehia1,

I checked your POSCAR and POTCAR file. In the POSCAR file you have two atom types, Ho Co. When looking at your POTCAR file
I get the following output when using

Code: Select all

grep PAW_PBE POTCAR

:

Code: Select all

  PAW_PBE Ho 25May2022                   
   TITEL  = PAW_PBE Ho 25May2022
  PAW_PBE Ho 25May2022                   
   TITEL  = PAW_PBE Ho 25May2022
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007
  PAW_PBE Co 02Aug2007                   
   TITEL  = PAW_PBE Co 02Aug2007

So this means you have twice the pseudopotential of Ho and the one for Co you concatenated 17 times. For a proper POTCAR file corresponding to the POSCAR file you supplied you should have the potential for Ho and Co once.

I hope this is of help.
All the best Jonathan

[Sherif_Yehia1]

Dear Jonathan,

Thank you for your clarification regarding my mistake with Ho2Co17.

Regarding my first point, I understand now that if I have twice the pseudopotential of Ho and the one for Co concatenated 17 times, it might not correspond correctly to my POSCAR file. In a typical VASP calculation, each element in my system should have its pseudopotential represented once in the POTCAR file, in the same order as they appear in the POSCAR file.

I would also like to understand the difference between Ho and Ho_3. I couldn’t find specific information about Ho_3 in the context of DFT calculations. However, I understand that in general, the subscript often refers to different oxidation states or different configurations of the same element. The choice between Ho and Ho_3 would depend on the specific physical and chemical environment of the Ho atoms in my system.

When it comes to magnetic properties, I understand that the choice of pseudopotential can indeed affect the results. Some pseudopotentials are constructed to better represent magnetic properties. If both Ho and Ho_3 pseudopotentials are available, I should test both and see which one gives results that better match experimental data or other theoretical results.

I would appreciate any further guidance you can provide on this matter.

Thank you once again for your quick and insightful response.

Best regards,
Sherif
PS Please forgive me if I sent this message twice

[michael_wolloch]

Dear Sherif,

Pseudopotentials with an integer suffix (_2 or _3) denote a specific valence state. These potentials are only provided for the Lanthanides. Some 4f electrons for these potentials are put in the frozen core, although they are higher in energy than other valence states. You are completely right that those potentials are only applicable if you know the chemical environment and the valence is correct for your system. Note that magnetism will not be correctly described with those potentials, since some f-electrons, which might spin-polarize, are frozen in the core.

f-electrons, especially partly filled shells, are notoriously difficult to describe in DFT (a failure of the method, and not VASP specifically), so be careful with your calculations and check against experimental data if available.

Some more information on these potentials can be found in the f-electron section of the Available PAW Potentials article on the wiki. We are working on improving the documentation of Pseudopotentials on the wiki and making it more digestible.

The Ho potential has been updated in the latest release (potpaw.64). You are using the new potential, which is great, but you have to be careful when comparing your calculations to older results.

I hope that answers your questions,
Michael