Forum EPW

Problem with number of electrons
Page 1 of 1

Author:  eliephys78 [ Tue Apr 16, 2019 4:23 pm ]
Post subject:  Problem with number of electrons

I am trying to calculate the electron phonon coefficient (lambda) in a 2D structure. I have been supplying the Fermi energy value in the epw input files using the flag : efermi_read=.true.
In the 'crystal.fmt' file, the number of electrons in my system is "22.000". However, when I run the calculations for lambda, I get :

"The Fermi level will be determined with -30.00000 electrons." Why is it -30 and not 22.. I have checked both the electronic and phonon dispersion bands and they look pretty much the same when compared to those obtained directly from QE. Would this "wrong" number of electrons affect lambda values?


Author:  eliephys78 [ Wed Apr 17, 2019 1:34 pm ]
Post subject:  Re: Problem with number of electrons


can anyone respond to my question whenever you guys are free.


Author:  carla.verdi [ Tue May 14, 2019 8:48 am ]
Post subject:  Re: Problem with number of electrons


If you set efermi_read=.true. the Fermi level used in EPW is the one you provided. The message 'The Fermi level will be determined with -30.00000 electrons' is a bit misleading, as it is written also in the case efermi_read=.true.
It doesn't look like the wrong value of nelec should affect your result for lambda, since you read the Fermi level in. However, it is strange that you get a value of -30, and it might indicate there are some further problems in your EPW run. As a note, the value of nelec printed is not the one you have in crystal.fmt, but it is strictly speaking nelec - fac * nbndskip (where fac=1 if noncolin=T, otherwise it's 2).


Page 1 of 1 All times are UTC
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group