unable to use Check Topological Mat. for vasp-spinor example

[aali28]

Dear Stepan

I performed DFT calculation in irrep-master/examples/C2B2Gd-vasp-spinor/dft_inputs with VASP6.3 and PAW PPs.

The command " irrep' -spinor -code=vasp > out " gives a trace.txt file. When I upload thin on Check Topological Mat. I get " The traces for some symmetry operations at some maximal k-vector do not correspond to any direct sum of irreps in the space group given. Sometimes the problem can be solved increasing the resolution in the calculations."

I have tried other examples of Bi2Se3 and I get the same thing. Kindly help me with the problem and am I doing something wrong in the calculation steps?

Regards

trace.txt

[MIraola]

Dear @aali28 ,

I have looked into the trace.txt file you attached. The reason why it does not work when you upload it to Check Topological Mat. is a typical problem of VASP (and probably other DFT codes), not of IrRep: the convergence of the last bands is usually poor. There are 2 ways to solve this problem:

• Calculate more bands with VASP by increasing the parameter NBANDS in the INCAR file.
• Run IrRep by setting a value for IBend which leaves out of the calculation the poorly converged bands. In your case, if you use -IBend=61, the problem is solved (I attach the trace.txt you would get this way, so that you can check that it would work).

trace.txt

Regards

[aali28]

Dear @MIraola

Thank you for the help. Following your suggestion I was able to use the trace.txt file for examples.

However, I am still facing the same issue in some cases. I have went up to NBANDS = NumOccupied+15-20 and -IBend = NumOccupied+5. But not able to use the calculated trace.txt on Check Topological Mat. I am not able to figure out what other changes are required in VASP calculation.

I am also attaching the trace files. There are few partially filled bands (18 valence electrons). Kindly let me know if you need the input files.

Regards

traceIBend16.txt traceIBend20.txt traceIBend25.txt traceIBend30.txt

[aali28]

I was able to use the trace file only for IBend =14 (without SOC) and IBend=20/22 (with SOC). I still do not understand why does it work for specific number of bands while for other numbers it gives warning.

Regards

[MIraola]

Dear @aali28 ,

the reason why the trace.txt file might not work for some particular number of bands (-IBend) is that this number should always be chosen in such a way that bands are included consistently in the calculation of the traces. For example, let's imagine that my space group has at all maximal k-points only irreps of dimension 2. Then, if I choose -IBend=21, the last band is not properly included in the calculation of the traces. -IBend would have to be even.

Regards

[aali28]

Thank you I will further look into this.

Regards