Dear all,
Can somebody please help me to define projections for LaSb system in Wannier tools. I have taken it as random, but that is creating a problem in convergence. Can somebody please help me for the same?
I would be very thankful.
Hi Payal,
have you solved this issue? I also used random projections in wannier90 calculations, then I have no idea how to set projectors in wannier_tools. Very few information found on this. Many thanks!
Hello,
If anyone resolves the problem of projections in wannier tools please share it .
Hi Amarjyoti Choudhury Canglang Yao
Suppose you have three atoms, La, As, and Bi in your system. Then, put their outer orbitals as their projections.
begin projections
La : d
As : px; py; pz
Bi : px; py; pz
end projections
Mind that the choice of projections is entirely random. Sometimes, taking projections as their outer orbitals help, or occasionally only a few orbitals near the Fermi level. It also depends on the property you are interested in. At the end, we have to minimize the spread in wannier functions and the imaginary part of the Hamiltonian in hr.dat file must be zero. Accordingly, we have to play with these projections.
While putting this projections flag in wannier90.win, you have to change num_wann and num_bands. Suppose in the above step. i have 5 orbitals of La, 3 orbitals of As, and 3 orbitals of Bi, so in total, i have 11 orbitals, Keep num_wann as 11, and num_bands safely can be kept as its double = 22.
For the SOC case, num_wann and num_bands should be doubled.
and same num_bands should be taken as NBANDS in the INCAR file while running VASP2Wannier90 interface.
Hope this helps!
- Badges
- Science topic
- Similar topics
- Filing