Regarding internal error in SET_CORE_WF

[chandan_kumarvishwakarma1]

Dear Users
I am trying to perform optical properties calculation for Nd doped NBT using GW and BSE calculations.
For this initially I had selected Nd_3 pseudo-potential and got following error,
"internal error in SET_CORE_WF: core electrons incorrect 49.000000000000062.0000000000000"
Then, I selected other pseudo-potential "Nd" which is available for vasp, for this case previous error is solved but DFT calculation is not converging as it is the very first step to perform GW and BSE calculations.
One more problem is that for my calculations using 'Nd' pseudo-potential the value of " real part of dielectric functions is coming as 1(one) and imaginary part is coming as o(zero) for all the directions and the last line after the dielectric funtions is
"The outermost node in the dielectric function lies at epsilon= -1.0"
Please suggest me how can I resolve these problem.

Thanks
Chandan

[ferenc_karsai]

Please read the forum guidelines and upload the important files according to that guideline.

[chandan_kumarvishwakarma1]

First_case.zip

The First_case.zip file contains:

All the inputs INCAR (DFT, DIAG, GW0, BSE), POSCAR, POTCAR (with Nd_3 pseudopotential which is converging the DFT calculations but in GW and BSE calculation it is giving internal error in SET_CORE_WF), KPOINTS.

A seperate error file is also attached: stderr_file

OUTCAR file for all steps, and stdout_* (* DFT, DIAG, GW0, NONE, BSE)

second_case.zip

The second_case.zip file contains:

All the inputs with POTCAR contains Nd pseudopotential using which DFT calculation is not converging ( stdout_DFT).
---
Chandan

[ferenc_karsai]

I discussed this issue within the group:

There was a bug in the code, which lead to errors in the recalculation of core wave functions for some POTCARs.
This bug is now solved and the bug fix will be included in the next VASP 6.x release at the end of this year/beginning of next year.

[chandan_kumarvishwakarma1]

Dear Users
I am trying to substitute a charged ion in my system, but using NELECT tag I can only remove
or add electrons from the whole system.
How can I remove or add electrons from a particular atom in a system ?
And using NELECT tag I am getting following error:
"very serious problems the old and the new charge density differ"

Please help me out.

Thanks
Chandan

[henrique_miranda]

Please have a look at the documentation here:
https://www.vasp.at/wiki/index.php/ICORELEVEL

Let me know in case something is not clear.

When you change NELECT and restart your calculation from a previous CHGCAR or WAVECAR it's normal that the old and new charge will differ.
You can either ignore the warning and verify carefully if the SCF converges or (preferably) start a new calculation from scratch.

[chandan_kumarvishwakarma1]

Dear henrique_miranda
Thanks for your reply.

I looked at the documentation but I didn't understand completely.
Suppose I want to substitute Li+ ion in an organic compound then,
How can I set my INCAR file?
For this I am using "Li_sv" pseudopotential with 3 valance electrons.

[henrique_miranda]

I looked at the documentation but I didn't understand completely.

What part of the documentation did you not understand?

[chandan_kumarvishwakarma1]

Dear henrique_miranda
In the documentation how can we define the CLZ tag?
could it be positive or negative integer?


---
Chandan

[henrique_miranda]

In the documentation in https://www.vasp.at/wiki/index.php/ICORELEVEL you find this description of CLZ:

The electron count specifies how many electrons are excited from the core. Usually 1 or 0.5 (Slaters transition state) are sensible choices.

Why would you want to use a negative number here?