EPW Forum

Dear Developers

Can I know what are the 11 columns in a2F and a2F_tr output files please ?
If I do not have experimental data, how can i confirm my simulation for a2F Vs frequency ?

Thank you

Jaya

good post

Hi!

jaya wrote:Can I know what are the 11 columns in a2F and a2F_tr output files please ?

I believe they contain the following data:

1 column: phonon frequency
2 - 11 columns: a2F(omega) at differing phonon smearing values, smearing values can be found in the bottom of the file (search for "Phonon smearing (meV)" in prefix.a2f.smx files), phonon smearing, known to the code as degaussq, according to the comments in the code and documentation is the.

smearing for sum over q in e-ph coupling

You would have as many prefix.a2f.smx files, as there are degaussw values used in your calculation. degaussw is a k-specae smearing.

jaya wrote:If I do not have experimental data, how can i confirm my simulation for a2F Vs frequency ?

Difficult question. The one thing that for sure worth checking, is that at least your total a2F has converged with fine grid (nqf*, nkf* variables). Same goes to the form of the spectre, you should be sure, that it doesn't change dramatically, when you increase nqf* and nkf* a little bit.

Hope that helps,
Andrei

Hi Andrei

Thank you so much for your valuable comment.

Yes, I can see different smearing values at the end of the file. In my case I used smearing degaussw=0.1 eV and no degaussq. But a2f_tr has ; 0.0500000 0.1000000 0.1500000 0.2000000 0.2500000 0.3000000 0.3500000 0.4000000 0.4500000 0.5000000

So why these 10 values. I mean, what value I should use for the a2f Vs omega graph?

Thank you again

Jaya

Hi Jaya,

The default value (which can be found in epw_readin.f90) for degaussq is 0.05 meV. In practice that corresponds to the smallest value of smearing used to calculate a2f, the total number of smearings being ngaussq=10 (again, this is the default), spaced by delta_qsmear = 0.05 meV - all consistent with the values you see in your output file. You can choose to plot a2f with any of these smearings - the larger the smearing the smoother the profile, while the convergence will depend on the q grid used in the integration.

Best,
Carla

Dear Carla
Thank you very much.

Jaya