Lattice constant deviation

Message

IBRAHIM · #1 Post by **IBRAHIM** » Mon Nov 08, 2021 6:51 am

Dear All,

I doped Fe perovskites by Co, however the lattice constant is strangely expanded by one Co atom but then it is normally contracted by increasing number of Co atoms. I tried PBE and PBEsol.

I am using POTCAR: PAW_PBE Fe_pv 02Aug2007 and PAW_PBE Co 02Aug2007.

why the lattice constant is expanded by one Co atom?

I will also try different POTCARs.

Many thanks in advance,
Ibrahim

IBRAHIM · #2 Post by **IBRAHIM** » Mon Nov 08, 2021 7:28 am

Details for my post

#3 Post by **martin.schlipf** » Mon Nov 08, 2021 7:44 am

This seems to be more of a physics question, which we cannot really address. I can suggest having a look at the magnetic moments, structural stability, dependence on xc functional, ...

Perhaps another user has some experience in this or other similar systems and can give more specific advise.

IBRAHIM · #4 Post by **IBRAHIM** » Fri Nov 12, 2021 4:03 am

Thank you very much.

May be also it is related to the oxidation states of those cations.
Could that happened using VASP?

Best,
Ibrahim

#5 Post by **martin.schlipf** » Fri Nov 12, 2021 8:01 am

Yes changes to the oxidation state can impact the size of the cell. However note that in general electrons don't sit exactly on a single atom so there will be always a bit of arbitrariness to which atom you assign a certain charge.

For a qualitative picture, it may be enough to use the LORBIT tag, possibly tweaking RWIGS to get the best possible results. If you want to make a more systematic approach, you can do a Bader type charge analysis. A quick Google search found this tool that claims to work on VASP output, though that software is not developed by us, so I cannot comment on its robustness or precision.

IBRAHIM · #6 Post by **IBRAHIM** » Fri Nov 12, 2021 8:51 am

Thank you very much for the detailed explanation.

Indeed, I did the Bader charge analysis and found a low charge on Co in that system compared to pure systems and the other doped systems.

Thanks for confirming that the oxidation states can impact the cell size using VASP.

Best,
Ibrahim

#7 Post by **martin.schlipf** » Sat Nov 13, 2021 9:55 pm

Just to be completely clear, if the oxidation state changes it is a physical origin not an artifact of the calculation. Essentially the interactions change and therefore the distances can as well.

IBRAHIM · #8 Post by **IBRAHIM** » Wed Nov 17, 2021 4:01 am

Thank you very much.

My Community