Analysis of Quantum Dot-based optical amplifiers: a time-domain traveling-wave approach

Chirped and tapered Quantum Dot semiconductor optical amplifiers (SOAs) are a valid solution for the direct amplification of ultra-short pulsed sources such as mode-locked lasers, thanks to their broad gain bandwidth and fast recovery times. Traditionally, the SOAs are operated in a so-called single-pass configuration: the radiation enters from one facet, it is amplified and exits from the opposite side. However, it was found out in literature that a double-pass configuration, where the radiation is extracted from the same facet it is injected into, can lead to enhanced performances, increasing gain and output power, with a limited degradation of the amplified pulses duration, making this configuration a possible future standard for the amplification of light sources. The main objective of my thesis work consists in the enhancement of an existing time-domain traveling-wave (TDTW) model to allow the simulation of QD layers with different optical properties ("chirped layers") and to include stimulated emission from the second QD excited state, which is typically neglected in theoretical laser models. Those improvements were implemented in an existing TDTW simulator in MATLAB. The reliability of the simulator to predict the behaviour of chirped QD SOAs is validated through the direct comparison of the numerical results with the experimental data measured by the Photonics group at Heriot-Watt University, Edinburgh, on a chirped, tapered and multi-section QD SOA. After a brief introduction to the problem and a state-of-the-art review in the first chapter, in the second one I concisely examine the peculiar aspects of Quantum Dot physics, with particular attention to their optical properties. Then, in the third chapter, I derive and discuss the equations constituting the TDTW model. In the fourth chapter, I present the device under test alongside the experimental setups considered. In the fifth chapter I extensively describe the main details of the MATLAB implementation of the TDTW model and, finally, in the last chapter, I report and discuss the most relevant simulations results. The main results presented in my thesis are the following. I found out that this simulator is able to predict the tunable spectral asymmetry affecting the output power spectra at the rear and front facets of the considered SOA under continuous wave biasing conditions, in which the device works as a SLED. Three main spectral components are registered at 1210 nm, 1240 nm and 1280 nm. Their relative weights can be tuned by changing the front and rear section driving currents. Results are in good agreement with the reference measurements. Moreover, I performed simulations with an external optical excitation in order to test the amplifying capabilities of the device, in both single and double pass amplification configurations. In agreement with the experimental data, considering a source pulse with average power of 2.5 mW and pulse duration of 2.3 ps, the SOA under test in double pass configuration offers an enhancement of signal gain up to almost 4.1 dB with respect to the same device in single pass configuration. This comes without significant pulse degradation: the pulse duration varies between 2.4 ps and 2.7 ps.