plasmon replica in metals
Posted: Mon Feb 11, 2019 6:23 am
Dear all
I have a general question concerning the application of EPW for metals which may be even not valid. It is about the electron-plasmon interaction and the replica band generated from it. From some literature research, I learned that there are two different types of plasmon in solid: one for the pristine compound where the plasmon energy is between 1 eV - 20 eV; while the other is typically in doped semiconductors where the carrier plasmon energy is around few hundred meV, comparable with the phonon energy. For the former, GW + cumulant expansion seems to be the method of choice. As for the carrier plasmon, EPW demonstrates a strong potential in capturing the essential replica bands produced by it. (Nat. Commu. 9:2305 (2018), Nat. Commu. 8:15769 (2017)).
My question is the following: For the undoped system, no matter it is a metal or a semiconductor, there would always be high-energy plasmons as can be calculated in GW+cumulant.
(1) Similarly, do we also expect carrier plasmons for metals?
(2) If this is a valid question, how would we calculate it in EPW?
In EPW, the calculation of carrier plasmons for doped semiconductors requires the knowledge of the dielectric constant, which becomes infinite for metals. This seems to hinder the direct application of EPW for metals. If some random but finite values are given for "epsiHEG", only zero plasmon self-energy can be obtained, which does not seem to be reasonable.
I would appreciate any explanation on the carrier plasmons in metals, including both theory and the application of EPW on it. Thanks in advance!!!
Sincerely,
Simon
I have a general question concerning the application of EPW for metals which may be even not valid. It is about the electron-plasmon interaction and the replica band generated from it. From some literature research, I learned that there are two different types of plasmon in solid: one for the pristine compound where the plasmon energy is between 1 eV - 20 eV; while the other is typically in doped semiconductors where the carrier plasmon energy is around few hundred meV, comparable with the phonon energy. For the former, GW + cumulant expansion seems to be the method of choice. As for the carrier plasmon, EPW demonstrates a strong potential in capturing the essential replica bands produced by it. (Nat. Commu. 9:2305 (2018), Nat. Commu. 8:15769 (2017)).
My question is the following: For the undoped system, no matter it is a metal or a semiconductor, there would always be high-energy plasmons as can be calculated in GW+cumulant.
(1) Similarly, do we also expect carrier plasmons for metals?
(2) If this is a valid question, how would we calculate it in EPW?
In EPW, the calculation of carrier plasmons for doped semiconductors requires the knowledge of the dielectric constant, which becomes infinite for metals. This seems to hinder the direct application of EPW for metals. If some random but finite values are given for "epsiHEG", only zero plasmon self-energy can be obtained, which does not seem to be reasonable.
I would appreciate any explanation on the carrier plasmons in metals, including both theory and the application of EPW on it. Thanks in advance!!!
Sincerely,
Simon