The plasmon concept arises from the consideration of an interacting electron gas described by the following Hamiltonian:
|
(2.125) |
where the first sum gives the kinetic energy and the second one arises from the Coulomb interaction between electrons. This Hamiltonian can be rewritten
using the random phase approximation [31,32,33,34] as follows:
|
(2.126) |
where each contribution can be represented in the second quantized form [35]. The first term
is the kinetic
energy of the electron gas:
|
(2.127) |
where
and
are the electron creation and annihilation operators, respectively, and spherical and parabolic dispersion is
assumed.
The second term
gives the contribution from a two-electron screened Coulomb interaction:
|
(2.128) |
As can be seen, this term accounts for scattering of two electrons with the initial quasi-momenta
and
and the final
quasi-momenta
and
respectively. Screening is taken into account through the cut-off wave vector
which separates short- and long-range parts of the Coulombic term.
The third term
describes a non-interacting plasmon gas, that is, the quantized oscillations of the electron
gas2.34:
|
(2.129) |
where
,
are the plasmon creation and annihilation operators and
is the plasmon energy.
The forth term
represents the electron-plasmon interaction:
|
(2.130) |
where two terms in the second sum can conveniently be treated in terms of absorption and emission of a plasmon in the same way as it has been shown above
for phonons.
The possible plasmon-phonon coupling [36,37,38] is not considered in this work as it plays an important role only in polar
semiconductors where in the degenerate case the frequencies of the charge density fluctuations are comparable to the optical frequencies.
S. Smirnov: