The effect of different concentrations on electron–phonon scattering. #197
-
|
Dear Nakib, I’m calculating the effect of different doping concentrations on electron–phonon scattering. After running with runlevel = 0, I obtained the corresponding chemical potentials (chempot). When I proceed to runlevel = 1, do I only need to update the chempot to the value corresponding to each doping concentration, or are there additional settings I need to adjust? |
Beta Was this translation helpful? Give feedback.
Replies: 3 comments
-
|
I encountered an unusual phenomenon in my electron-phonon scattering calculations. When simulating n-type doping, the electron-phonon scattering rates and thermal conductivity agree well with literature values. However, under p-type doping conditions, the scattering rates drop abruptly, inconsistent with published results.The colors blue, gray, and red represent carrier concentrations ranging from 1×10¹⁸ cm⁻³ to 1×10²⁰ cm⁻³, respectively. |
Beta Was this translation helpful? Give feedback.
-
|
Dear ashore-oss, Thanks for your kind words and your questions. The carrier concentration depends only on the temperature and the chemical potential. So those are your only two knobs to tune. The plots you shared are the phonon-electron scattering rates. The plots have no labels, but it is clear to me that the horizontal axis gives the phonon mode energies in eV and the vertical axis gives the mode resolved phonon-electron scattering rates. The vertical axis is likely in log-scale and the numbers on the ticks are probably the exponents. I am guessing the units are THz. Could you please confirm? You shared two plots. Which one is for n-doping and which one for p-doping? For your p-doping calculation, are you making sure you have a large enough Fermi window? For "large" band gap materials, the usual method is to do two different sets of calculations, one for the n-type transport and the other for the p-type. In the former, one would set the reference energy close to the conduction band edge, whereas in the latter it would be near the valence band edge. The transport active bands should be set accordingly in the input file. I can comment further if you can share your input and output files for the p-doped calculation. I hope this helps. Best, |
Beta Was this translation helpful? Give feedback.
-
|
Dear Nakib, Yesterday, I tested this approach using silicon (Si) as an example. I set the reference energy (enref) at the midpoint between the maximum energy of the valence band (VBM) and the minimum energy of the conduction band (CBM). This choice yielded results consistent with expectations. Below are my input and output files, with both n-type and p-type doping concentrations set to 1.0×10²⁰ cm⁻³. Best, |
Beta Was this translation helpful? Give feedback.
Dear ashore-oss,
Thanks for your kind words and your questions.
The carrier concentration depends only on the temperature and the chemical potential. So those are your only two knobs to tune.
The plots you shared are the phonon-electron scattering rates. The plots have no labels, but it is clear to me that the horizontal axis gives the phonon mode energies in eV and the vertical axis gives the mode resolved phonon-electron scattering rates. The vertical axis is likely in log-scale and the numbers on the ticks are probably the exponents. I am guessing the units are THz. Could you please confirm?
You shared two plots. Which one is for n-doping and which one for p-doping?
For your p-doping c…