This thread is locked.Only browsing is available.
Top Page > Browsing
Exchange coupling parameter program (jx)
Date: 2016/05/13 18:47
Name: Joao Amaral   <jamaral@ua.pt>

Dear all,

It would be very useful if the jx program could automatically calculate the J coupling parameters between all atom pairs in the structure.

Would a bash-type script suffice for this?

Thank you for your help,
Jo&#227;o Amaral
メンテ
Page: [1]

Re: Exchange coupling parameter program (jx) ( No.1 )
Date: 2016/05/15 10:42
Name: T. Ozaki

Hi,

This can be easily done by a bash-type scrip.

Regards,

TO
メンテ
Re: Exchange coupling parameter program (jx) ( No.2 )
Date: 2016/05/24 18:35
Name: Joao Amaral  <jamaral@ua.pt>

Dear Dr. Ozaki,

Thank you for your reply!

We have used the Fe BCC example from the installation (Febcc2.dat), and magnetic interaction (J) parameters are easily calculated with the jx executable.

We get a quite high value of J (70 meV), while typical values are ~20 meV from TB-LMTO (PRB B 64 174402) or ~14 meV from SPR-KKR PBE calculations.

Is the resulting J value from openmx jx calculations the sum of all equivalent atom pairs in the structure? That would justify the higher value, but it does seem strange.

Best regards,
Joao Amaral
メンテ
Re: Exchange coupling parameter program (jx) ( No.3 )
Date: 2016/06/05 10:24
Name: T. Ozaki

Hi,

We have also noticed the same issue, and been investigating the origin.
Once we understand, we may post again.

Regards,

TO
メンテ
Re: Exchange coupling parameter program (jx) ( No.4 )
Date: 2016/06/06 18:00
Name: Joao Amaral  <jamaral@ua.pt>

Dear Dr. Ozaki,

Thank you again for your reply. I will be grateful for further information.

Best regards,
Joao Amaral
メンテ
Re: Exchange coupling parameter program (jx) ( No.5 )
Date: 2016/08/18 23:33
Name: Joao Amaral  <jamaral@ua.pt>

Is there an update on this matter?

Best regards,
Jo&#227;o Amaral
メンテ
Re: Exchange coupling parameter program (jx) ( No.6 )
Date: 2016/09/06 16:58
Name: Hongkee Yoon  <hongkeeyoon@kaist.ac.kr>

Hi

The jx.c code included in OpenMX calculates J(q=0) as described in [M. J. Han: PRB-2004, Mn molecular system].
In bcc Fe system, there exist 8 nearest neighbours (atom 1 and 2).
So J between 1 and 2 would be J(q=0)/8 (85 meV/8 => 10.6 meV) which is a reasonable number.
We are working on to calculate J(q) which computational cost is much heavier than J(q=0).
It’s on testing stage for now.

Regards,
Hongkee Yoon
メンテ

Page: [1]