Why La element has no La_2/_3 pseudopotential？

[hong-hui_wu1]

Why does La element have no La_2/_3 pseudopotential, while other rare earth element Ce-Lu has _2/_3 pseudopotential.

[fabien_tran1]

Hi,

If the formal valency of 3 is considered (supposing that it is the most commone one), then the number of f electrons that are kept frozen in the core should be 3-3=0 (the electronic configuration of La outside the Xe shell is given as 5d6s^2). Thus, no f electrons to freeze into the core.

[hong-hui_wu1]

Thank you for your reply. My current system is 79Fe1X(La-Lu). May I ask why I am using the pseudopotential of PAW-PBE of Ce-Lu, and the structural optimization always fails to converge?

[fabien_tran1]

In order to provide some help, we would need the input files as explained at forum/viewtopic.php?f=4&t=17928

[hong-hui_wu1]

thank you for your reply. This is a bulk system of 79Fe1Tb, the pseudopotential used is PAW_PBE, but no convergent results are obtained

[fabien_tran1]

If you want some help for your convergence problem, then you need to provide the input files and the OUTCAR of your calculation on 79Fe1Tb. Also, you should mention a few more things, like how many attempts you have made or if you have already tried to improve convergence by modifying the INCAR files, etc.

[hong-hui_wu1]

Thank you for your reply. I have tried to remove the magnetism, modify the convergence criteria, and encut, but I have not obtained the result of convergence.

[fabien_tran1]

At the end of OUTCAR, there is the message "ZBRENT: fatal error in bracketing please rerun with smaller EDIFF, or copy CONTCAR to POSCAR and continue". Have you tried to do it and continue the calculation?