Questions about the MO in periodic boundary condition and MLWF calculation

Top Page > Browsing

Questions about the MO in periodic boundary condition and MLWF calculation

Date: 2024/05/13 11:44
Name: Lingzhi ZHANG <lingzhi@g.ecc.u-tokyo.ac.jp>: Dear OpenMX community,

I have a question about the MO cube file for the periodic boundary condition.
According to https://www.openmx-square.org/openmx_man3.9/node93.html,
it says that it will generate the cube file for each k-point.

As I know, OpenMX generates the wavefunction in the k-space first and we can generate its distribution in the real space with the wavefunctions of all the k-points.
I would like to know how it works to generate its distribution with only one k-point.

In addition, I am also trying to generate the spin and charge density around each atom.
I think that generating them with the Wannier functions may be more convenient for me.
However, I found that the MLWF calculation in OpenMX is too slow.
For 886 Wannier functions under 32 mpi threads, each step takes more than 10 minutes.
I can also turn to wannier90, but it cannot plot the wannier function without the unk files.

I would like to know how to accelerate it and how to generate the unk file via OpenMX to plot the Wannier functions.

With best regards
Lingzhi ZHANG

Page: [1]

Re: Questions about the MO in periodic boundary condition and MLWF calculation ( No.1 )

Date: 2024/05/21 13:49
Name: Lingzhi ZHANG <lingzhi@g.ecc.u-tokyo.ac.jp>

Dear OpenMX community,

Sorry that I have one stupid question.

The unk file is the distribution of Bloch states in the real space for each band and k-point.
I find that it's very similar to the MO file generated by OpenMX.
May I generate the unk file from the MO files?

With best regards
Lingzhi ZHANG

Re: Questions about the MO in periodic boundary condition and MLWF calculation ( No.2 )

Date: 2024/05/21 19:30
Name: Naoya Yamaguchi

Hi,

>May I generate the unk file from the MO files?

It is possible although phase adjustment due to e^(ikr) is needed.

Regards,
Naoya Yamaguchi

Re: Questions about the MO in periodic boundary condition and MLWF calculation ( No.3 )

Date: 2024/05/22 16:38
Name: Lingzhi ZHANG <lingzhi@g.ecc.u-tokyo.ac.jp>

Dear Prof. Yamaguchi,

Thanks for your reply.
I wonder if there is some reference related to the phase adjustment.

Also, I have another question about the electronic state given by OpenMX.
OpenMX is a LACO-based package, thus I think the ground state is also the linear combination of the atomic basis.
May I generate the charge density for each atom by only adding the contribution from their basis?

With best regards
Lingzhi ZHANG

Re: Questions about the MO in periodic boundary condition and MLWF calculation ( No.4 )

Date: 2024/05/23 15:40
Name: Naoya Yamaguchi

Dear Lingzhi ZHANG,

>I wonder if there is some reference related to the phase adjustment.

It is merely an operation to obtain its periodic part from the Bloch function.

>May I generate the charge density for each atom by only adding the contribution from their basis?

Since the Bloch function is taken in linear combination with the atomic basis, it does not satisfy that linearity for the charge density, which is the square of its norm, and I don't think it can be decomposed so simply.

Regards,
Naoya Yamaguchi

Page: [1]