Thesis (Ph. D.)--Oregon Graduate Center, 1986.

Univ., Diss., 2006--Giessen

Rapid Thermal Annealing has been used to study the electrical activation of a range of donor and acceptor species in ion-implanted GaAs. By varying the time and temperature of the post implant anneal, it was found that the activation processes for most implants can be characterised in terms of two distinct regions. The first of these occurs at short annealing times, where the electrical activity is seen to follow a time-dependent behaviour. At longer annealing times, however, a time-independent saturation value is reached, this value being dependent on the annealing temperature. By analysing the data from Be, Mg, S and Se implants in GaAs, a comprehensive model has been evolved for the time and temperature dependence of the sheet electrical properties. Application of this model to each of the ions studied suggests that the activation processes may be dominated by the extent to which ions form impurity-vacancy complexes. An analysis of the time-dependent regime also shows that, at short annealing times, the mobile species is more likely to be the substrate atoms (or vacancies) rather than the implanted impurities. In the time-dependent region, the values of diffusion energy were found to be between 2.3 to 3.0 eV for all ions, these values corresponding to a diffusion of Ga or As vacancies (or atoms). In the saturation region, activation energies of 0.3 to 0.4 eV and 1.0 to 1.2 eV were obtained for the activation processes of interstitial or complexed impurities respectively.

There has been great interest in using ion implantation for III-V semiconductor device isolation as an alternative to mesa isolation technique. This is attributed to several advantages that implant isolation has over mesa isolation. Mesa isolation exhibits problems such as over/under etching, repeatability issue of etching depth and nonplanarity of the surface of the semiconductor. However implant isolation is advantageous as the surface planarity is maintained and in general, less intrusion under the mask edges is observed. This thesis presents a study on the isolation of both n and p-type InP and InGaAs layers and n-type InGaAsP layers by ion implantation. Several different ion species such as protons, helium, nitrogen and iron were used to isolate these materials. The n and p-type layers were grown by Solid Source Molecular Beam Epitaxy. Conductive n-type InP layers were also formed using multiple energy silicon implantation to create a uniform dopant distribution throughout the n-type region. The effects of ion mass, implantation temperature, damage accumulation, initial carrier concentration of the conductive layer and post-implant annealing temperature were investigated in detail through electrical and structural characterisation. The major part of the work was to develop recipes for the isolation of the individual InP, InGaAs and InGaAsP layers. The effects of implantation temperature and dose were also examined thoroughly. A parallel resistor model was also created to confirm the reliability of the measurements.

This thesis reports the development of a model to explain the electrical properties of Si implanted GaAs. The results show that most of the implanted silicon atoms occupy lattice sites and are electrically active. The net carrier concentration is determined by the relative concentration of silicon atoms on gallium and arsenic lattice sites respectively. The activation mechanism is shown to involve the breaking up of complex defects in the form of substitutional silicon with vacancies. The energy required for this process is about 1 to 1.5 eV. A lower value of activation energy (about 0.5 eV) has also been measured and is suggested to be associated with the site switching of silicon from arsenic to gallium sites, when a gallium vacancy diffuses close to a silicon on an arsenic site. This process has diffusion energy of about 2.5 to 3.0 eV. The activation energy obtained from sheet carrier concentration measurements corresponds to a combination of the two activation mechanisms. Which of these mechanisms is observed in an experiment depends on various parameters, such as the implantation conditions, the quality of the encapsulant and the annealing conditions. The model can explain the variations in activation energy (0.5 to 1.5 eV) reported in the literature.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

No description available.

High-energy Si implantation into silicon creates a net defect distribution that is characterized by an excess of interstitials near the projected range and a simultaneous excess of vacancies closer to the surface. This defect distribution is due to the spatial separation between the distributions of interstitials and vacancies created by the forward momentum transferred from the implanted ion to the lattice atom. This dissertation investigates the evolution of the near-surface vacancy excess in MeV Si-implanted silicon both during implantation and post-implant annealing. Although previous investigations have identified a vacancy excess in MeV-implanted silicon, the investigations presented in this dissertation are unique in that they are designed to correlate the free-vacancy supersaturation with the vacancies in clusters. Free-vacancy (and interstitial) supersaturations were measured with Sb (B) dopant diffusion markers. Vacancies in clusters were profiled by Au labeling; a new technique based on the observation that Au atoms trap in the presence of open-volume defects. The experiments described in this dissertation are also unique in that they were designed to isolate the deep interstitial excess from interacting with the much shallower vacancy excess during post-implant thermal processing.

No description available.

No description available.