spin-orbit coupling

[balabala12s]

Dear nakib

Thank you for developing and maintaining the excellent elphbolt code.

I am writing to inquire if elphbolt currently has the capability to include spin-orbit coupling (SOC) effects in its calculations, particularly for the phonon-drag thermopower.

Any information on this topic would be greatly appreciated.

Best,
Lei

[nakib]

Dear Lei,

Thank you for your kind words.

Including the SOC is a little bit tricky at the moment, but I think it can be done. Please take a look at the answer here: #76

Note the use_ws is not yet supported.

Best,
Nakib

[balabala12s]

Dear Dr. Nakib,

Thank you for your reply. I hope this email finds you well.

I am using elphbolt to calculate transport properties for a 2D material, MoSi₂N₄. When I run the calculation to find the chemical potential (runlevel 0), the result is unphysical. The calculated chemical potential is more than 4 eV away from the reference energy (enref), which I believe is incorrect.

Could you please help me check if there is an error in my input ? I have attached my input.nml file and the relevant output for your reference.

Thank you for your help.

Best regards,
Lei
input.zip
output.zip

[nakib]

Dear Lei,

I can't see anything obviously wrong with your input or output. But how do you know the result is wrong? What is the intrinsic carrier concentration of this material? What is the real band gap? What is the DFT band gap? Why do you need so many Wannier bands? A quick google search shows that there is a single isolated valence band. As such, I don't see why you would need to include 29 Wannier bands. This will make the transport calculations ridiculously expensive. Could you please double check that you have numbered your bands correctly? Could you share your QE and Wannier bands plotted on top of each other?

Best,
Nakib

[balabala12s]

Dear Dr. Nakib,

Thank you for your reply.

I am writing to let you know that I have successfully solved the problem. The issue was resolved by adding the following line to the EPW input file:

bands_skipped = 'exclude_bands = 1:3'

This parameter excluded the redundant bands under Fermi level.

Best regards,
Lei