Degree Name

Doctor of Philosophy


Department of Physics


The research reported in this thesis involves two main strands, both concerned with optical properties near the band edge. One involves studies of reflectance from shallow quantum wells, while the other is concerned with hydrogen passivation studies of MBE bulk GaAs and AlxGa1-xAs and GaAs related layered structures. For the reflectance work the structures employed were very high quality undoped MBE grown multi-quantum wells consisting of 25 periods of a 20 nm GaAs well and a 50 nm or 20 nm AlxGa1-xAs barrier, with x=0.01, 0.02 or 0.045. Low temperature reflectance spectra of such samples are extremely sharp and detailed, ideal for the study of transition parameters such as oscillator strength per unit area and transition linewidth. Spectra were consistent with full quantum confinement of excitons for all samples. Parameters were extracted by simulation of the reflectance spectra until a very close fit was achieved using both a local, semiclassical model and the full nonlocal transfer matrix of Andreani [146,147]. Both methods were in agreement on oscillator strength per unit area, but analysis by means of the nonlocal method was necessary to achieve accurate fits of excitonic transition linewidth where the intrinsic broadening associated with the radiative lifetime of photocreated excitons was comparable in magnitude to that due to all other sources of broadening. Such an outcome has been theoretically predicted. Transition linewidths were as sharp as 53 meV FWHM, as sharp as any reported. Oscillator strengths per unit area were consistently larger than theoretically predicted, ranging for the Ellh transition from 51±7 to 80±4 x 10-5Å-2, depending on x. The acoustic coefficient of broadening was found to be 3.0±0.4 meV/K for the E11h transition and 4.5±0.4 meV/K for E11h

One aspect of the hydrogenation work involved the study of MBE 𝑝 -type GaAs samples with strong 'KP' lines passivated by hydrogen and donor excitonic PL lines that were undetectable before hydrogenation, but strong afterwards. Selective excitation and other experiments revealed that isolated shallow acceptors were not being passivated, and a model was developed to explain the findings. Another aspect involved the passivation of short period superlattices. It was discovered that hydrogen only improves PL efficiency when free carriers are prevented from migrating out of the superlattice b y confinement in a double heterostructure or by disorder. The exp(T/T0) temperature dependence results give evidence that quenching of PL intensity as the temperature is raised is caused by phonon assisted tunneling or hopping of localized excitons or carriers to sites where nonradiative recombination is likely to occur.