Pseudopotential of Cerium (Ce)

Question on input files/tags, interpreting output, etc.

Please check whether the answer to your question is given in the VASP online manual or has been discussed in this forum previously!

Moderators: Global Moderator, Moderator

Post Reply
Message
Author
fujisaki_takaya1
Newbie
Newbie
Posts: 2
Joined: Mon Mar 09, 2020 5:23 am

Pseudopotential of Cerium (Ce)

#1 Post by fujisaki_takaya1 » Thu May 21, 2020 2:50 am

Hi :)

I try to understand the way to describe the electron configuration of pseudopotential of Cerium (Ce).

I found there are "Ce_h", "Ce_GW", and "Ce_3" folders in potpaw_PBE.54 folder. I think that each folder contains the electron configuration without inner core.

“Ce_h” shows [core=Xe4] s2d1f1. (However, two folders of "Ce_GW", and "Ce_3" show different one describing "[core=Kr4d]".
I am confused about the reason why there is such difference.
I think there should be purpose on each POTCAR file, but I can not get it still. Unfortunately, I could not find any answer I can understand as beginner level on any webpage.

I would appreciate if you give me some hints.
Thank you for reading :D

Best regards,

ferenc_karsai
Global Moderator
Global Moderator
Posts: 422
Joined: Mon Nov 04, 2019 12:44 pm

Re: Pseudopotential of Cerium (Ce)

#2 Post by ferenc_karsai » Tue Jun 02, 2020 10:06 am

Principally the most important difference between POTCAR files for a given atom is in the cutoff for the energy channels and the configuration of the core. For calculations where conduction band states are important (like RPA, GW etc.) more energy channels are required, hence the POTCARs named GW usually have a higher cut off.
It can be also desired that semicore states are put into valence, since semicore states can be spatially to large for the PAW spheres and inaccuracies in the calculation can occur.

fujisaki_takaya1
Newbie
Newbie
Posts: 2
Joined: Mon Mar 09, 2020 5:23 am

Re: Pseudopotential of Cerium (Ce)

#3 Post by fujisaki_takaya1 » Fri Jun 05, 2020 8:44 am

Dear ferenc_karsai

Thank you for your reply.
I try to use "Ce_GW" because of larger number of valence electrons. (I am thinking that POTCAR with "core=Kr4d" does not include electrons of Kr4d)

Can I think that POTCAR having more valence electrons generally provide more accurate results?
(Side effect is to make calculation speed slower)

If I misunderstand, I would really appreciate your advice.

Best regards,
Takaya

Post Reply