EPW Forum

Dear users and developers,

In epw.x calculation, the below problem is occured.
I think that it is because of wrong input in epw.in file.

I used the values
for phonon calculation,
12x12x12 k-mesh & norm-converge PP in scf.in
2x2x2 q-mesh in ph.in

for epw calculation,
12x12x12 k-mesh & norm-converge PP in scf.in
8 k-points in nscf.in
nbndsub =3 for px,py,pz in epw.in (due to localized 3 bands near Fermi level)


Please, could you give me any idea?


==================================================

Writing epmatq on .epb files


band disentanglement is used: nbndsub = 3

Writing Hamiltonian, Dynamical matrix and EP vertex in Wann rep to file


Reading Hamiltonian, Dynamical matrix and EP vertex in Wann rep from file


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine efermig (1):
internal error, cannot bracket Ef
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

stopping ...

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine efermig (1):
internal error, cannot bracket Ef
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

stopping ...

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine efermig (1):
internal error, cannot bracket Ef
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

stopping ...

Finished reading Wann rep data from file

Using uniform q-mesh: 2 2 2
Size of q point mesh for interpolation: 8
Using uniform k-mesh: 2 2 2
Size of k point mesh for interpolation: 16
Max number of k points per pool: 4

Fermi energy coarse grid = 7.312544 eV

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine efermig (1):
internal error, cannot bracket Ef
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

stopping ...

Dear Myung-Chul,

You need to make sure that the band structure after Wannierization is correct. I see that you are only using 8 k-points in nscf.in, this may not be enough. You could try 64 k-points instead.

If your Wannierization is correct, a way to go around the error is to set the Fermi energy in the input epw.in This was discussed here:

viewtopic.php?f=6&t=113&p=504&hilit=bracket+Ef#p504

Best,
Roxana

Dear Roxana,

Thanks for your reply.
To begin with, I tried to set the fermi energy, and then I was able to have a result.
But I should check either the wannierization is correct or incorrect, because I got the strange band structure.


Thank you,
Myung-Chul