EPW Forum

Dear all

I just started my PhD and very new for EPW. I need to calculate thermal conductivity of metallic materials. I know only EPW can do that. I looked any examples for thermal conductivity calculation but failed to find. Can someone guide me on this please?

Thanks for your time.
I appreciate any hint very much.

Daya

Dear Daya,

Actually no. EPW computes electron-phonon coupling.
For thermal conductivity, you might want to look at phonon-phonon coupling. Codes like phono3py can help you do that.

Best,
Samuel

Dear Dr. Samuel

It is sad to hear. I am not sure about phono3py also because it gives only phonon contribution. But for metallic systems electron contribution is the dominant.

I know BoltzTrap code can calculate electrical conductivity in the units of tau (sigma/tau).
Also I saw in your Lead example, there is a graph for electrical conductivity Vs T. But I am not sure whether it is a general method to use for any metallic material as in that paper it is related with superconductivity. I do not have good theory knowledge at this moment to understand that.

I went through several post regarding relaxation time calculation and then the thermal conductivity. But again I need to read more for those theories for better understand them.

Can you please clarify me about following ( because I feel little doubt now if i am wasting time on a wrong path)

1. Can we use EPW for calculate electrical conductivity as a function of temperature for any metallic material up to their melting point as shown in Pb example or is it only for special material related with superconductivity?
2. Is there a way to calculate electron relaxation time related with electron scattering?

Thank you very much for your valuable time and wish you successes for your good works.

Daya

Dear Daya,

Yes to both questions. But electrical conductivity is different from thermal conductivity (your original question).

Therefore yes EPW can help you compute electrical conductivity in metals and semiconductors. Not limited to superconductivity.
And it can also give you the electron relaxation time.

Best,
Samuel

Dear Dr. Samuel

Thank you for your reply.
Yes. But I think it is possible to convert electrical conductivity to thermal conductivity (wiedemann-franz law?).

I am working on Diamond and Pb examples and trying to learn to calculate sigma(e.conductivity). Is that correct if I follow the same method for Diamond as for Pb to find sigma_diamond as a function of temperature?
I couldn't find any examples for relaxation time calculation. Do you have any instructions for that?

I have another question regarding the output files. I could plot linewidth Vs k-path for diamond from linewidth.phself file. However, in linewidth.elself format is different. I copied the headings of the columns below.

# Electron lifetime (meV)
# ik ibnd E(ibnd) Im(Sgima)(meV)


Could you please tell me where the linewidths for electrons are written?
Thank you and sorry for asking many questions. As a beginner, I am still investigating basics of the code.

Daya

Dear Daya,

I see. The resistivity (or conductivity) formula presented in the 2016 EPW paper from Ziman is based on the LOVA approximation.
This is only valid in the case of metals (like Pb).

To compute mobility you can use the Boltzmann transport equation. This has been very recently implemented into EPW and will be in the next release of EPW along with example.

The imaginary part of the electron-phonon self-energy (=Sigma) is the lifetime of the electron.

Best,
Samuel