Wrong Cell Parameters When Optimizing BaTiO3 Unit Cell

[毛瑞麟]

Dear CP2K Users:

        I faced some problems while optimizing Tetragona BaTiO3 unit cell.  I use direct cell optimization method, with Cohen-Wu Exchange correlation functionals.  As described in literature, Cohen-Wu functional is the best functional which can be used in ferroelctric systems. However, the cell parameters after optimization EXTREMELY (a=b=4.30 angstroms, c=5.59 angstroms) diverge from  cell parameters measured by experiments (a=b=3.99 angstroms, c=4.04 angstroms). All electron steps converge normally. 

        I also tried normal PBE functional and PBEsol functional, the cell parameters seemed far away from what is normal. 

        So can anyone tell me what's wrong with my codes or how can I modify my code?

Best regards

[Thomas Kühne]

Try to use at least a 4x4x4 k-point mesh. 

Best, 

Thomas Kühne

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/eb8becbc-612b-4032-a07b-053058194a17n%40googlegroups.com.
<0.out><1.inp><3-1.cell>

==============================

Thomas D. Kühne

Dynamics of Condensed Matter

Chair of Theoretical Chemistry

University of Paderborn

Warburger Str. 100

D-33098 Paderborn

Germany

thomas...@upb.de

+49/(0)5251/60-5726

[Rylin Mao]

Dear Prof.  Kühne

Thank you for your reply!

I modified my code by using a diagonalization method with a 4x4x4 Monkhorst-Pack k-point mesh. It indeed makes sense, however, there is still a big gap between the calculated cell parameters and the experimental results.

While using Cohen-Wu functional, the length of a and b axis seems acceptable, however, the length of c axis is about 4.6 angstroms. I think a difference of 0.6 angstroms is a little big. I also tried normal PBE functional. the calculated length of c axis is about 4.2 angstroms, still has a difference of about 0.2 angstrom from the experimental results. How can I improve my codes?

Looking forward to your reply!

Best regards,

Rylin Mao

[fabia...@gmail.com]

Dear Rylin,

You have no correlation in your XC functional. I have never used the Cohen-Wu functional but from skimming the paper I think you should add the pbe correlation to your input file:

&XC_FUNCTIONAL
    &GGA_X_WC
    &END

    &GGA_C_PBE
    &END      

&END XC_FUNCTIONAL

Please make sure to double check this!

Cheers,

Fabian