Pump-probe measurements on bulk InGaAsP optical amplifiers are compared with numerical calculations based on simplified density matrix equations. We find that the gain dynamics is well described by the combined effect of carrier heating, spectral holeburning, and two-photon absorption.
The ultrafast dynamics of gain and phase in bulk semiconductor optical amplifiers are numerically investigated over a broad band of about 200 nm based on the semiclassical method of density matrix equations. The gain dynamics and phase dynamics are compared in four different bands for 200 fs Gaussian pump pulse. It is predicted that the recovery time of phase may be longer or shorter than that of gain depending on probe wavelengths. Furthermore, phase recovery overshoot may appear at wavelengths shorter than the pump wavelength. These interesting features result from the competition between spectral-hole burning and carrier heating over the whole band.
Using the Foreman effective mass Hamiltonian for strained In/sub x/Ga/sub 1-x/As--In/sub y/Ga/sub 1-y/As/sub z/P/sub 1-z/ quantum wells, the propagation of subpicosecond pulses in a 1.55-/spl mu/m optical amplifier was calculated. The multisubband carrier dynamics as well as the polarization dynamics were taken into account. Carrier heating and coherent light-carrier interactions as well as the interplay of these nonlinear processes and the amplifier dispersion are studied. Strong Rabi oscillations occur in the optical field of a propagated pulse, its frequency chirp, as well as in the carrier density and temperature. While the Rabi oscillation imposes negative frequency chirp and hence red-shifts the pulse spectrum, positive frequency chirp can occur due to the local gain dispersion, where the higher frequency components of the pulse have larger gain. Due to the Rabi oscillation, the spectrum of the amplified pulse is considerably distorted and sidebands emerge. For a linearly chirped input pulse, the spectrum of the output pulse can be either red-shifted or blue-shifted with respect to its center frequency, depending on its initial chirp. For strong pulse propagation, a pronounced pulse break up occurs when a 175-fs pulse propagates in the gain regime, while a significant pulse compression occurs when the pulse propagates at the transparency point.
We have performed ultrafast cross gain modulation experiments on a gain clamped semiconductor amplifier. Time-resolved observation of the lasing mode intensity after excitation with a strong, detuned pump pulse of 150 fs duration reveals the instantaneous formation of a dark pulse that repeats itself with a period given by the cavity round- trip time. A simultaneously measured cw probe signal at gain maximum shows pronounced relaxation oscillations. The experimental results are in reasonable agreement with numerical simulations.
The gain and phase dynamics of InAs∕GaAs quantum dot amplifiers are studied using single and two-color heterodyne pump probe spectroscopy. The relaxation of the wetting layer carrier density is shown to have a strong effect on the phase dynamics of both ground and excited state transients, while having a much weaker effect on the gain dynamics. In addition, the dynamical alpha factor may also display a constant value after an initial transient. Such behavior is strongly encouraging for reduced pattern effect operation in high speed optical networks.
Introduction In this chapter we will present the theory of spectral gain in optical amplifiers and the related optical power distribution. The concepts are general, as they apply to every optical amplifier configuration, independent of the specific architecture and implementation technology. We will configure the optical amplifying medium as a traveling-wave device, where the optical field undergoes optical amplification as it propagates from the input to the output sections. Generally speaking, optical amplification is achieved by a net stimulated emission process. Referring to a generic section z of infinitesimal thickness dz , optical gain is determined by the imbalance between net stimulated emission and cavity losses, including extrinsic absorption, reflection and radiated fields. As we know from the theory of light amplification in the three-level atomic system presented in Chapter 5, stimulated absorption of photons is governed by the same transition rate as stimulated emission, and is included in the net stimulated emission process. With this in mind, we note the identity between stimulated emission and the stimulated absorption rates, Γ 12 = Γ 21 , and the net stimulated photon term Γ[ j 2 ( t ) – j 1 ( t )] p ( t ) in the rate equation. In order to develop the traveling-wave model, the photon density becomes a function of the longitudinal position 0 ≤ z ≤ L along the amplifier medium. Accordingly, we will set p ( t , z ) as the relative density of stimulated photons per unit volume (0 ≤ p ( t , z ) ≤ 1) at time t and section z . When photons with appropriate energy proceed along the amplifying medium, they are amplified by means of net stimulated emission. At some section z = z 0 , the steady-state photon density P ( z ) can approach the saturation density P sat ( z ).
We report on the characterization of the ultrafast gain and refractive index dynamics of an InAs/InGaAsP self-assembled quantum dot semiconductor optical amplifier (SOA) operating at 1.55 mum through heterodyne pump-probe measurements with 150 fs resolution. The measurements show a 15 ps gain recovery time at a wavelength of 1560 nm, promising for ultrafast switching at >40 GHz in the important telecommunications wavelength bands. Ultrafast dynamics with 0.2-1.5 ps lifetimes were also found consistent with carrier heating and spectral hole burning. Comparing with previous reports on quantum dot SOAs at 1.1-1.3 mum wavelengths, we conclude that the carrier heating is caused by a combination of free-carrier absorption and stimulated transition processes.