Calcul de gain optique de structures a puits quantiques sous contraintes

Strained quantum wells brought new flexibility in the design of optoelectronic components such as semiconductor lasers and semiconductor optical amplifiers. For example, the optoelectronic laboratory of Ecole Polytechnique de Montreal built polarization insensitive semiconductor optical amplifiers using alternate tension and compression quantum wells.; These designs led to new challenges concerning the optical gain modeling of the components. Up to now, our laboratory was using an optical gain calculation based on the parabolic approximation. However, some studies conducted here and other published showed that this approximation is not fully valid in the case of strained quantum wells. In that case, a calculation that takes into account band-mixing is needed.; This work is an answer to the need of developing a new computer model to calculate the optical gain of a strained quantum well. At every step of the calculation (energy band calculation, choice of a linewidth function, methods of calculating the Coulomb interactions between charge carriers), a list of all the possible strategies is presented. In order to determine which models give the best results, a comparison with experimental results is established.; The first achievement of this work is to present a new comparison method between experimental and theoretical spontaneous emission spectra. Using a tool already developed in our laboratory (the two dimensional semiconductor laser simulator LAS2D), we show that it's possible to estimate the charge carrier density generated by photoluminescence in a complex structure consisting of six strained quantum wells. This reduces the number of fitting parameters used in the calculation and is therefore an improvement compared to the comparison methods published previously.; The work then shows that a model using a 6x6 band calculation with a constant intraband relaxation time and a Gaussian linewidth function including many-body effects gives good agreement with experimental results. From this successful comparison, we can now say that the Optoelectronic Laboratory has a reliable tool for the calculation of the optical gain of strained quantum wells. This tool will be useful in the study of semiconductor lasers as well as semiconductor optical amplifiers. It will allow to test different designs on computer, resulting in a save of time and money.