[Wien] BerryPI

[Oleg Rubel]

The GGA+U calculation is likely affected by pi wrapping (-pi = +pi).
So -p +(-pi) = -2*pi can be different from what you expect -pi + pi =
0.

Also, I would not make conclusions based on a single calculation of
polarization for a structure. All observables (spontaneous
polarization, effective charge, piezoelectricity) are related to the
polarization change (difference). Therefore calculations often require
at least two structures.

Please do not apply BerryPI to calculation of polarization when the
band occupancy is metallic.

Thanks
Oleg


On Sun, Jan 3, 2016 at 2:58 PM, nilofar hadaeghi <n.hadaeghi at gmail.com> wrote:
> Dear all
>
> I have studied the polarization of some metal and insulator cases using the
> latest version of the berryPI code. We performed the scf calculations by
> Wien2k14.2.  One of Our calculations for one of my cases using GGA method
> show that this case is an insulator in agreement with experimental reports.
> The structure of this compound is centrosymmetric in the Pm3m space group.
>
> So based on the theoretical background of berryPI code it is expected that
> the polarization value for this compound must be zero. After running the
> code for this compound using this command
>
> berrypi -k6:6:6
>
> the results has been achieved as what was expected and copied below:
>
> ---------------------------------------------------------------------------------------
>
> Electronic polarization (C/m2)     sp(1)  [ 0.000000e+00,  0.000000e+00,
> 0.000000e+00]
>
> Ionic polarization (C/m2)          sp(1)  [ 1.366237e-11,  1.366237e-11,
> 1.366237e-11]
>
> Tot. spin polariz.=Pion+Pel (C/m2) sp(1)  [ 1.366237e-11,  1.366237e-11,
> 1.366237e-11]
>
> ---------------------------------------------------------------------------------------
>
> TOTAL POLARIZATION (C/m2)          both   [ 1.366237e-11,  1.366237e-11,
> 1.366237e-11]
>
> ---------------------------------------------------------------------------------------
>
> We also performed calculations for another case. We have used the GGA+U
> (U>6eV) method for this case  to obtain the suitable gap compared to
> experiment.  The GGA method leads to zero gap in contrast to experiment. The
> same as previous studied case, this structure is centrosymmetric(space group
> :Pm3m)
>
>
>
>  So we expect the zero polarization vector for this compound too, but the
> results show the non-zero polarization for the mentioned method in contrast
> to theory. Could you help me to solve this problem, please?
>
>
>
> In the GGA+U method and ferromagnetic polarization (with  SP calculation),
> after running the command:
>
> berrypi -o -k10:10:10
>
> ---------------------------------------------------------------------------------------
>
> Electronic polarization (C/m2)     sp(1)  [-5.253369e-01,  5.253369e-01,
> -5.253369e-01]
>
> Ionic polarization (C/m2)          sp(1)  [ 3.304038e-11,  3.304038e-11,
> 3.304038e-11]
>
> Tot. spin polariz.=Pion+Pel (C/m2) sp(1)  [-5.253369e-01,  5.253369e-01,
> -5.253369e-01]
>
> ---------------------------------------------------------------------------------------
>
> Electronic polarization (C/m2)     sp(2)  [ 0.000000e+00,  0.000000e+00,
> 0.000000e+00]
>
> Ionic polarization (C/m2)          sp(2)  [-5.253369e-01, -5.253369e-01,
> -5.253369e-01]
>
> Tot. spin polariz.=Pion+Pel (C/m2) sp(2)  [-5.253369e-01, -5.253369e-01,
> -5.253369e-01]
>
> ---------------------------------------------------------------------------------------
> TOTAL POLARIZATION (C/m2)          both   [-1.050674e+00,  6.700285e-11,
> -1.050674e+00]
> ======================================================================================
>
> We also performed our calculations using the GGA method without spin
> polarization. after running the command
>
>  berrypi -k10:10:10
>
> ---------------------------------------------------------------------------------------
>
> Electronic polarization (C/m2)     sp(1)  [-1.470943e-01, -9.456064e-02,
> 2.626685e-01]
>
> Ionic polarization (C/m2)          sp(1)  [-5.253369e-01, -5.253369e-01,
> -5.253369e-01]
>
> Tot. spin polariz.=Pion+Pel (C/m2) sp(1)  [-6.724312e-01, -6.198976e-01,
> -2.626685e-01]
>
> ---------------------------------------------------------------------------------------
>
> TOTAL POLARIZATION (C/m2)          both   [-6.724312e-01, -6.198976e-01,
> -2.626685e-01]
>
> ---------------------------------------------------------------------------------------
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>