This thesis basically consists of two parts. In the first part is discussed photoluminescence experiments performed on p-type silicon doped with B, In, or Tl. These substitiutional dopants were believed to form complexes with interstitial Fe and thought to be the source of some very intense luminescence in the silicon. Also presented are the results of photoluminescence experiments performed on luminescence features which were thought to be due to an isolated Fe complex observe in n-type Si:P. These spectra generally consisted of one or two sharp no-phonon lines followed by a subsidiary peak approximately 9 - 10 meV below the no-phonon line which was identified as a phonon replica of the no-phonon line. This phonon mode was thought to be a local vibrational mode of the Fe. Isotope shift experiments were performed on these luminescence features by diffusing 54Fe and 56Fe into the silicon samples to see whether a change in the phonon energy or a shift in the no-phonon line could be observed (as predicted by theory) and thus more conclusively identify the center. No isotope shift was observed in the case of any of the Fe related centers studied. Similar experiments which were performed on luminescence features attributed to (Cu,Cu) pairs in silicon (using 65Cu and 63Cu) are also described in the first part of this thesis. An isotope shift of the no-phonon line and a change in the characteristic 7 meV phonon mode, seen in this spectrum, were observed. This confirmed the identification of these luminescence features as being due to the presence of Cu. Also this result confirmed that the lack of a shift in the Fe case was real. Possible explanations for the null result in the Fe case are discussed.
The second part of this thesis consists of optical investigations of GaAs/AlAs heterostructures. In these experiments the transport of electrons past a thin (50 Å to 200 Å) AlAs barrier sandwiched between thick GaAs cladding layers was studied by measuring the voltage developed across these structures as a function of the wavelength of light illuminating the sample. Calculations to model the optical absorption in these structures were also carried out. Based on the data and the calculations the following explanation for the observed photo- voltages was proposed. Electrons, in the degenerately doped GaAs, are optically excited by free carrier absorption to energies greater than that presented by the AlAs barrier and flow from the illuminated side of the AlAs barrier to the back side of the barrier. The driving force for this flow would be the difference in the concentration of optically excited electrons on either side of the barrier. This difference results from the light intensity difference on either side of the barrier as modeled by the calculations. These experiments were conducted for samples with different AlAs layer thicknesses, GaAs layer thicknesses and dopings, and at various temperatures. Further work involved applying a constant dc bias to the structure while measuring the photovoltage spectrum. This, it was found, increased the photovoltage signal (by several orders of magnitude in some cases) and caused some shifts in the spectrum to slightly longer wavelengths. These effects were explained in terms of charge redistribution in the sample, that is, accumulation and depletion on either side of the barrier and the effective band gap narrowing in the GaAs due to the large electric fields that these dc biases can create in the depleted areas of the GaAs.
