This thread is locked.Only browsing is available.
Top Page > Browsing
Macroscopic polarization calculation by Berry's phase
Date: 2020/10/28 20:11
Name: HOI

Dear all,

I'm using OpenMX Ver. 3.8. I tried to calculate spontaneous polarization for MonoChalcogenides. But I did not get the desired results, and i found out Electric dipole Value in "SnTe.out" file and "SnTe.scfout" file (it's read by polB.c tool) is different. As follows:
- Electric dipole Value of SnTe.out:
"***********************************************************
***********************************************************
Dipole moment (Debye)
***********************************************************
***********************************************************

Absolute D 6.02972395

................. Dx.............. Dy ............. Dz
Total......... 3.84744046 ... 4.64270700 .. -0.00667878
Core.......... 659.809221 ... 664.654149 ... 2507.39149218
Electron..... -655.961780 .. -660.011442 .. -2507.39817096
Back ground.. -0.00000000 .. -0.00000000 .. -0.00000000 "


- Electric dipole Value of SnTe.scfout:
"*******************************************************
Electric dipole (Debye) : Berry phase
*******************************************************

Absolute dipole moment 2674.59107678

........ Background ... Core ............... Electron ..... Total
Dx .. -0.00000000 .. 659.80922139 ... 0.00007039 .. 659.80929177
Dy .. -0.00000000 .. 664.65414977 .. -8.09011978 .. 656.56402999
Dz .. -0.00000000 .. 2507.39149218 .. 0.00014077 .. 2507.39163296 "

I don't know why can you give me some advice?

Best regards,
Hoi Nguyen
メンテ
Page: [1]

Re: Macroscopic polarization calculation by Berry's phase ( No.1 )
Date: 2020/10/29 01:00
Name: Naoya Yamaguchi

Hi,

The behavior seems natural because OpenMX calculates the dipole moment by using the electronic density on the real space grid, while polB calculates with Berry phases.
So, the electronic contribution looks different, and in periodic crystals, the electric polarization obtained from an integral over the real space is ill-defined, while the Berry phase expression of the polarization is well-defined.

Regards,
Naoya Yamaguchi
メンテ

Page: [1]