Questions about TB2J calculations and window selection for a metallic system

6 views
Skip to first unread message

Iram Bel

unread,
Oct 27, 2025, 5:32:19 AM (7 days ago) Oct 27
to TB2J

Dear HeXu and TB2J users,

I performed a TB2J calculation, but the isotropic exchange interactions I obtained seem to have the wrong sign, so I suspect this might be due to an improper choice of energy windows.

I consulted the TB2J documentation as well as the WannierTools tutorial (https://www.wanniertools.org/tutorials/high-quality-wfs/) to understand how to choose the disentanglement and frozen windows, but I am still having some difficulty.

For the projectors, I would like to include the d orbitals for Co and the f orbitals for Pr, since these are mainly responsible for the magnetism in my compound. I have attached my band structure and DOS, and I would greatly appreciate it if you could suggest reasonable values for the disentanglement and frozen windows for my metallic system.

I also have a small additional question: should the Wannier bands be plotted before or after choosing the windows for comparison with the VASP bands, Is the comparison used to choose the windows or to judge the quality of my Wannier functions?
Finally, is the anisotropy package complete and ready for production use, or is it still under development?

Thank you very much for your help.

Best regards,
Iram


TB2J.pdf

Iram Bel

unread,
Oct 28, 2025, 12:27:08 PM (6 days ago) Oct 28
to TB2J
Hello,
Could you please provide the energy ranges for the frozen and disentanglement windows?
I would say :
dis_win_min = -6 +Ef
dis_win_max = 6 +Ef
dis_froz_min = -4+Ef
dis_froz_max = 4+Ef

I would also like to know whether the calculations of the MAE and are finalized and ready for production, or if they are still under development.

Thank you !
Iram

--
You received this message because you are subscribed to the Google Groups "TB2J" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tb2j+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/tb2j/24ab8094-6ceb-49fb-abbb-5fd95e3029c5n%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Xu He

unread,
Oct 29, 2025, 4:23:40 AM (5 days ago) Oct 29
to Iram Bel, TB2J
Hi, 
From the band structure I am not sure there is enough number of bands in the DFT calculation, as on the top some of the bands are not smooth and looks like a hard cut.  So my suggestion is to increase the number of bands, replot the DOS, then optimize the energy window. 
Sometimes the Pr d bands can be included to improve the fitting, as they are not very far from the f bands, sometimes even within the energy range of the f-bands. 

The MAE is implemented but not for Wannier functions. It can work with ABACUS and SIESTA.  I am not sure what you mean by TC. 

Best regards ,
HeXu


Reply all
Reply to author
Forward
0 new messages