IBTE for a metal

[Vahid]

Dear EPW users,

I have calculated the el-ph matrix elements for a metal with EPW-5.4.1 and have restarted to calculate the scattering rates and the IBTE. The following tags are set:

Code: Select all

 
  elph        = .true.
  epbwrite    = .false.
  epbread     = .false.
  epwwrite    =  .false.
  epwread     =  .true.
  wannierize  = .false.

  scattering      = .true.
  scattering_serta = .true.
  iterative_bte   = .true.

  efermi_read = .true. 
  fermi_energy = 2.8

I would like to use IBTE to find the mobility at the given Fermi energy.

The EPW run ends with the following lines:

Code: Select all

      Progression iq (fine) =        900/       999
      300.000      0.0000       0.966545E-08
      
     epmatkqread automatically changed to .TRUE. as all scattering have been computed.

No scattering rates is printed and the output shows no mobility values.

Does IBTE work for metals in EPW?

Thank you,
Vahid

Vahid Askarpour
Department of Physics and Atmospheric Science
Dalhousie University
Halifax, Canada

[Shashi]

Hi Vahid,

IBTE works for metals in EPW. To enable it, please make sure to include the tag assume_metal = .true. in your EPW input file.

Let me know if you have any questions.

Best regards,
Shashi

[Vahid]

Hi Shashi,

Thank you for your input. Using assume_metal = .true. did result in getting the IBTE conductivities for different iterations.

I was wondering what should the tag "int_mob" be for IBTE in metals? I set it to false and am getting ~ 54% of the expected conductivity. In addition, the input settings of the following lines

efermi_read = .true.
fermi_energy = 2.8249

are changed by the code to E_F=2.8637eV. Is this related to int_mob setting?

Thanks,
Vahid

[Vahid]

Dear Shashi,

I ran an IBTE calculation on the Pb (no SOC) example in the EPW folder. Attached are the input and output EPW files. The Fermi energy is set to 11.7176eV in the input. In the output near the end where the iterative conductivities are printed, the Fermi level is changed by EPW to 11.5752eV. Is this a bug or am I missing something in the input? Setting int_mob to .true. does not remedy this discrepancy.

Thanks,
Vahid

[Shashi]

Dear Vahid,

Thanks for sharing the input/output files. The shift in Fermi energy printed during the IBTE calculation is expected as EPW recalculates the Fermi level at each temperature. When assume_metal=.true., the Fermi energy at the given temperature is estimated assuming a hole-type carrier.
The input Fermi level is used to determine the states within the fsthick window.

Setting `int_mob = .true.` doesn't override this behavior. It only affects how intrinsic mobilities are computed:

If int_mob = .true. and carrier = .false., EPW computes the intrinsic mobility using a single Fermi level, ensuring equal electron and hole concentrations.

If int_mob = .true. and carrier = .true., EPW uses two Fermi levels to compute intrinsic mobilities for electrons and holes separately, with each having a carrier concentration of ncarrier.


Regards,
Shashi