problem in calculation of mobility for heavily doped semiconductor

General discussion around the EPW software

Moderator: hlee

Post Reply
Posts: 14
Joined: Mon Jul 11, 2022 8:48 am
Affiliation: PhD student

problem in calculation of mobility for heavily doped semiconductor

Post by souravrudra »

Hello developers,

I have tried to do mobility calculation for GaN using the input files available on the EPW website. In case of hole concentration with 1E20, the code is not running, writing "The carrier density is 0". I looked some discussion with the same context in the forum, So to get mobility with such high carrier concentration I need to change the input code where it is calculating the fermi level using T and fermi_energy from epw input. But I am not understand whinch .f90 file in the src directory I have to change. I found a subroutine nammed fermicarrier of 461 line in utilities.f90 file and fermi energy related some codes in ephwann_shuffle.f90 and ephwann_shuffle_mem.f90. Which file i have to modify and what lines in that file i have to comment out?

Also, In BN tutorial i found that with the given input files of school calculated hall mobility with Bx=0, By=0.0 and Bz=1E-10 Tesla is very small of the order of E-12, where as the drift mobility is of the order of 1E3. Why hall mobility is so small that should not be? Also how to choose the magnetic field components in case of hall mobility calculation? Can I choose a large value like 2 Tesla?
In papers I have seen that mobility vs temperature curve, but EPW gives mobility tensor of having six components, So which component should be taken as mobility value? is it the avg of sigma_xx sigma_yy and sigma_zz ? are the off-diagonal components important?

Please help me and rectify my doubts.
I am using QE version V_7.1 and epw V_5.5

Thanks with regards,
Sourav Rudra
PhD student
JNCASR, Bengaluru, India

Post Reply