Wrong results about spin-polarized band structure in MoS2 monolayer

[WEIW]

Dear all,

I used ypp -s b to interpolate band of MoS2 monolayer.
It is known that valence band maximums (VBM) at different valleys (that is, K valley and K' valley) have an opposite spin polarization.
However, the output 'o.spin_factors_DN_interpolated' displays the wrong results. The figure attached show that VBMs at K and K' valleys have the same spin polarization.
The output files and figure plotted by the yambo data are attached.

Best,
Xu

[Davide Sangalli]

Dear Xu,

the interpolation procedure is just interpolating the values computed on the available k-points.
Since your simulation is with SOC, K and K' should be both computed, e.g. they are not related by symmetry.
Then you can just check the values in your '.spin_factors_UP_built_in' and 'spin_factors_DN_built_in' files.

Add this flag to the input file to generate the latter

Code: Select all

BANDS_built_in                 # Print the bands of the generating points of the circuit using the

Can you attach the output files here?
Are you sure you are taking K and K' and not two equivalent K ?

Best,
D.

[WEIW]

Dear Davide,

Thanks for your rapid reply!

I saw that the report file in yambo contains a tag showing 'Symmetries: 12 [spatial+T-rev]'. That says the system have time-reversal symmetry. So, I wonder if the interploation process consider the T symmetry.

If the valleys are not related by symmetry, the flag 'force_symmorphic' in QE will be set to False, right? If do that, I think the number of k-points will be large and the simulation is considerably time-consuming. Is there any other method to compute the K and K' valleys simultaneously.

I follow your instruments to uncomment the 'BANDS_built_in' flag, and the outputs are attached.

Are you sure you are taking K and K' and not two equivalent K ?

According to the band circuit set in band.in, the K is -0.333 0.666 0 and K' is 0.333 0.333 0. I think these two point are not equivalent in MoS2 monolayer.