search

quanshengwu / wannier_tools

WannierTools: An open-source software package for novel topological materials. Full documentation:
http://www.wanniertools.org
GNU General Public License v3.0
257 stars 142 forks source link

Strange output crystal structure for a trigonal lattice #98

Open liliu910 opened 2 years ago

liliu910 commented 2 years ago

Hello everyone, I met with a problem in calculation the surface state of a trigonal system (SG. 166, R-3m), with the hr file from Quantum Espresso. I have attached the relative files and also extracted the most relative parts below.

The LATTICE and ATOM_POSITIONS cards are like: LATTICE Angstrom 2.7085077130 1.5637576572 4.7348282132 -2.7085077130 1.5637576572 4.7348282132 -0.0000000000 -3.1275153144 4.7348282132

ATOM_POSITIONS 7 !Number of atoms for projectors Direct ! Direct or Cartisen coordinate Ni 0.0000000000 0.0000000000 0.5000000000 Ni 0.0000000000 0.5000000000 0.0000000000 Ni 0.5000000000 0.0000000000 0.0000000000 In 0.0000000000 0.0000000000 0.0000000000 In 0.5000000000 0.5000000000 0.5000000000 Se 0.7200000000 0.7200000000 0.7200000000 Se 0.2800000000 0.2800000000 0.2800000000

and MILLER_INDEX card like: MILLER_INDEX 1 1 1

That is, I adopt the R-type unit cell with a=b=c, \alpha=\beta=\gamma and surface plane of [1 1 1] normal to the z-axis, which corresponds to the H-type unit cell with surface plane [0 0 1].

I am trying to calculate the surface-projected band structure. The output of arc_bulk gives the expected result with correct symmetry, so I expect the hr file is of acceptable quality (The system involves complicated fatbands so the Wannierization is rather hard, and I suppose with SOC included, wannhr_symm is not available). However, neither arc_l nor arc_r is symmetric, with strange surface states (see the attached files).

When I examine other output files of wt.x, I find the both the POSCAR-rotated and POSCAR-slab strange too (see the attached files). Neither of them gives the correct surface plane. Moreover, the POSCAR-rotated file gives incorrect Se atoms positions (0.72, 0.72, 0.84 in direct coordinate, while the correct one should be 0.72, 0.72, 0.72). The POSCAR-slab file includes strange atoms in the vaccum layer, which may explain the strange surface states.

So how do these occur and what can I do to get it over? Please inform me if any more information is needed. Thanks a lot!

wt.in.txt arc_bulk arc_l arc_r POSCAR-rotated.txt POSCAR-slab.txt

WT.out.txt

daiqho commented 2 years ago

hello, was you using the latest version of WT btw?

liliu910 commented 2 years ago

hello, was you using the latest version of WT btw?

Hello, I am using Version 2.5.1. I thought it was the latest version before (I have also posted the WT.out file now for reference). While I did not seen any bug fix regarding the POSCAR generation, I may retry with the newest version. Still struggling with the installation of the newest version which requires ARPACK...

galvao97 commented 10 months ago

Hello,

I've faced issues regarding latest version 2.6.2 installation because of ARPACK requirement as well. You may try version 2.6.0, which does not need ARPACK library (see: https://www.wanniertools.org/installation-guide/)

Regarding the calculation of surface states, I think it may be related to dangling bonds on the surface of interest. Take a look at this discussion: https://github.com/quanshengwu/wannier_tools/issues/31

Best regards