...DIPOL Flag in Berry`s phase calculations...

Message

brocks · #1 Post by **brocks** » Tue Feb 13, 2007 5:14 pm

Dear Users,

i`m performing some calculations to look at the polarization of HoMn2O5.
I have a centrosymmetric structure, experimentally resolved.
As first step of my study i take this structure and i relax it.
After a relaxation of the unit cell I get a non-centrosymmetric structure:
the current unit cell allows the system to be ferroelectric.
I have two structures with different lattice parameters: one centrosymmetric and
not ferroelectric and the other one ferroelectric.
I want now to calculate the polarization (P) of HoMn2O5 and i want calculate it
by Berry's phase.
I think to get its value (P) i have to do two separate calculations (for the two different structures i have) following what it written in the VASP manual.
So, for both the structures (this means using the two different POSCAR), i could make the CHGCAR file and WAVECAR following the Calculation 1 of the NaF in the manual.
Then, always for both the structures, i could make the Calculations 2,3,4 as described in the manual. Naturally i should choose in the INCAR file the right "parameters".
Then i could collect the results following what it is written in the manual.
Does it work like this? Am i wrong?
If this is the way to calculate the P, i have an other problem because
i`m not able to understand the words of the manual to describe the DIPOL flag:
"This tag specifies the origin with respect to which the ionic contribution to the dipole moment in the cell is calculated. When comparing changes in this contribution due to the displacement of an ion, this center should be chosen in such a way that the ions in the distorted and the undistorted structure remain on the same side of DIPOL (in terms of a minimum image convention)"
Practically what does it mean?

Then i would ask an other thing.
The values of DIPOL flag i have to give to the sets of my calculations (because i have a POSCAR for centro and non-centrosymmetric structures), i believe, must correspond to the same point in cartesian coordinates.
Is this true?

Replies are welcome.
Thank you.
Gianluca.

#2 Post by **admin** » Thu Mar 08, 2007 8:21 am

1) you are right, you have to do the complete set of calculations for both of the phases.
2) this comment about the DIPOL means that it has to be set in such a way that the direction of the dipole moments must not revert before and after the displacement of the atoms which you have to introduce to get a certain polarization mode. If your cell geometries do not differ too much, the centers of the dipoles may be the same.
However, in principle, DIPOL has to be set for each structure separeately.