January 26, 2004 16:26 WSPC/Book Trim Size for 9in x 6in b ook2
Coulomb Quantum Kinetics 413
be used to predict the optical spectra in systems with elevated electron–
hole–plasma densities where part of the absorption becomes negative, i.e.,
optical gain is realized. These gain media are the basis for semiconduc-
tor laser operation (see Chap. 17). The calculation of the proper gain
line shape has been a long standing problem since the use of a constant
dephasing approximation leads to the prediction of unphysical absorption
energetically below the gain region. As it turns out, the numerical solutions
of the semiconductor Bloch equations, where dephasing is treated according
to the second Born approximation, provide a solution of the laser line-shape
problem yielding very good agreement with experimental results, as shown
in Fig. 21.7. For more details of the microscopic semiconductor laser and
gain theory see, e.g., Chow and Koch (1999).
21.3 Build-Up of Screening
If a sample is excited by a femtosecond pulse, some time is needed until the
optically created carriers rearrange in order to screen their mutual Coulomb
interaction. The characteristic time in this problem is the period of a
plasma oscillation. For an electron-hole gas, e.g., in GaAs, with a density
of about 10
18
cm
−3
, the energy of a plasmon is in the same order as that
of a longitudinal optical phonon. Hence, the corresponding time scale is
in the order of 100 fs. With modern femtosecond spectroscopy where the
pulses can be as short as a few fs, the regime of the build-up of screening
is accessible in experiments. As a test for the predictions of the quantum
kinetic calculations of the build-up of screening, which we will discuss below,
an optical pump and THz probe experiment is ideally suited to detect
the delayed build-up of a plasma resonance. Other experiments which are
performed within this build-up regime of screening are femtosecond FWM
experiments with and without coherent control. In this ultrashort time
regime, we have to give up the quasi-equilibrium assumptions used in the
previous section and calculate the two-time-dependence of the screening and
of the spectral functions self-consistently. In detail the following elements
have to be included in the treatment:
• Self-consistent calculation of the two-time-dependent screened
Coulomb potential
The two-time-dependent retarded screened Coulomb potential W
r
q
(t, t
)
(21.4) has to be calculated. We restrict ourselves to the RPA polar-
ization function (21.8). This approximation is in general not sufficient