The novel idea of GaSb quantum dots (QDs)1 quantum rings (QRs) stacked layers single junction solar cells have been investigated for the examination and enhancement of the infrared photo response. Initially the investigation used photoluminescence to probe the optical properties of a type-II material interface between GaSb/GaAs using optimum growth temperature for QDs/QRs with two different growth modes (Stranski-Krastanow (SK) and exchange growth); and two different GaSb deposition thickness (1.5ML and 2.IML). The photoluminescence spectra of the stacked epilayers confirmed that the dominant radiative recombination mechanism was band-to-band in the GaSb QDs/QRs stacked layers. Excellent structural quality is observed in each sample with no threading dislocations (by Transmission Electron Microscopy (TEM)). The composition of the QRs is close to 100 % GaSb with high purity GaAs centres. The ring density per layer is approximately 1010 rings/cm2 with no significant variation in size or density in the separate layers. II Rapid thermal annealing (RTA) has been used to tailor the optical properties of 10-layer stacks of type-II GaSb self-assembled QDs and QRs embedded within GaAs grown by molecular beam epitaxy. An increase in PL emission intensity and a blue shift in peak energy in both types of QD stacks were observed, along with changes in the activation energy for PL quenching. These effects were attributed to Sb-As intermixing and size effects with corresponding changes in the band structure and an increase in the oscillator strength associated with the transformation towards type-I behaviour. It has been concluded that postgrowth rapid thermal annealing can be used to tune the spectral response and control carrier recombination and escape properties of stacked GaSb QD for more effective use in devices such as solar cells and lasers. The final part of the investigation examined the properties of multi-layer QDs/QRs single junction solar cells (SC) to obtain an understanding of the operation and characteristics of the devices. Three kinds of solar cells were fabricated; one is intrinsic layer with ? layers of QDs/QRs, another comprises 10 layers and the final one is control cells (without QDs/QRs).

The aim of this research was to investigate the reactive ion etching (RIE) of gallium arsenide (GaAs) in mixtures of methane and hydrogen (CH ₄H₂) plasma and to evaluate thier advantages over chlorinated plasmas. This was performed in order to find the optimum etching conditions for GaAs such as, the best etch rate with greatest degree of anisotropy, the finest smooth side walls and the lowest surface roughness. The induced damage to GaAs due to RIE was investigated by current-voltage (I-V) and capacitance-voltage (C-V) measurements. From a study of the behaviour of the DC self bias voltage and an analysis of electrical characterization, a possible model of the mechanism of etching GaAs in methane and hydrogen mixtures was proposed. The main contributions of this research are as follows: Etching of GaAs while maintaining the total flow rate of the gas mixture and its residence time in the chamber constant. Studying the effect of the physical component of the etching mechanism on GaAs by investigating the variation in the DC self bias voltage for all process parameters. Analyzing the effect of process parameters on the deposition rate of carbon polymers on the surface of the inorganic mask during GaAs etching. Investigating the effect of process parameters and carbon polymer deposition on SiO ₂mask erosion during GaAs etching. Examining the electrical characteristics of highly doped p-GaAs following RIE in CH ₄/H₂ plasma and comparing with those of H ₂ etched samples.

This work describes a comprehensive modelling and analysis hierarchy which pursues the convergence point between system-level behaviour and low-level physical design of 3-D integrated circuits. The complete modelling, design and analysis methodology can be applied to wider technological design areas in deep submicron very large scale integrated systems such as massively parallel computational systems and high-speed digital signalling. The electrical parasitic signature of Through Silicon Vias (TSVs) are exhaustively defined with novel closed-form equations for various coupled and isolated 3-D interconnect configurations. The trade-offs in electrical performance for the 3- D interconnect for industry-standard on-chip CMOS signalling circuits including low-voltage, differential , current mode and shielding practices are then presented to define signalling conventions for 3-D circuits and arc compared to the planar 00- and off-chip interconnect. Thermal compact models of 3-D packages are discussed and used to develop a stand-alone thermal simulation tool to provide fast analysis of state-of-the-art IC packages. Moving from the physical-level, the communication infrastructure of 3-D ICs is rigorously investigated with particular attention to the packet-switching Network-on-Chip framework. The model for average distance is derived and rigorously analysed with cycle-accurate simulations to optimise the partitioning of multifunctional dies. The scalability and performance of 2-D and 3-D networks is then assessed where a custom cycle accurate simulator is developed for several traffic patterns, switch architectures and network configurations. Finally, to bring the physical and communication-level models into perspective, a set of hierarchical models are presented which are used to assess the computational efficiency of 2-D and 3-D silicon-based processors for early-chip planning. The development of comprehensive hierarchical models from the physical to the circuit to the system level in this work contributes significantly towards understanding the promises and limitations of future IC package design based on TSV integrated 3-D die stacking.

No description available.

No description available.

This thesis focuses on theoretical modelling of dilute nitride semiconductor Quantum Well (QW) lasers and Semiconductor Optical Amplifies (SOAs) considering compositional fluctuations of N in GaInNxAs1-x material system. Incorporation of N into GaInAs results in a huge reduction of the material bandgap and this has been successfully modelled using a Band Anti-Crossing (BAC) model in which the N acts as a defect in the GaInAs conduction band mixing with it and pushing it downwards. However compositional fluctuations of N in GaInNxAsl_x result in Quantum Dot (QD)-like fluctuations in the Conduction Band ,Edge (CBE). The observed QD-like fluctuations cause a broadening of the gain much in the .way it does for an inhomogeneous array of QDs. Therefore, the influence of these compositional fluctuations on the performance of GaInNxAs l_x QW lasers and SOAs motivates the research work.

As integrated circuit density increases, digital circuits characterized by high operating frequencies and low voltage levels will be increasingly susceptible to faults. Furthermore, it has recently been shown that for many digital signature and identification schemes an attacker can inject faults into the hardware and the resulting incorrect outputs may completely expose their secrets. On-chip error masking techniques such as error correction could be one of the options to mitigate the above problems. To this end, this thesis presents a framework of techniques to design error circuits.

Conventionally-designed digital circuits are encountering speed limits which cannot be overcome due to constraints imposed by the basic laws of physics. Synchronous designs are relatively easy to produce, but require the system clock to run at the speed of the slowest part of the entire circuit. Self-timed (asynchronous) circuits offer the promise of being able to use resources more effectively, however they require much more design effort while not always delivering the expected gains. This thesis presents a method of combining the advantages of both synchronous and asynchronous techniques by introducing and developing, for the first time, a clocked version of the well-known Muller C-Element as part of this research. Results of experiments demonstrating its effectiveness are given along with recommendations for further research and development.

Mutually coupled lasers have received much attention in recent years as a source of rich non linear dynamics, enabling optical configurations for new ranges of applications, particularly chaos-based encrypted communications. However the overall dynamics of the mutually coupled configuration has not received so much attention, particularly with regard to the practical range of parameters for the lasers. This thesis investigates the dynamics of the mutually coupled lasers system in respect of all the significant causes and influences on dynamics. Furthermore, the lasers here are widely tuneable, representing the more complex tunable devices found in modern communications systems. The system investigated is both self and mutually-coupled, which represents the mechanism which may arise in modern photonic integrated circuits (PICs), and therefore has relevance to the design and operation of such devices. However the configuration here has parallels with, and great relevance to the non self-coupled configuration, and the conditions for this (such as phase) are highlighted. The dynamics are investigated experimentally in an integrated device using some of the latest technology to obtain time series of a high bandwidth, in addition to optical spectra and high resolution radio-frequency spectra. These enable a range of dynamic analysis tools such as the correlation dimension (for which a unique algorithm is described) to be applied, and this provides insight into the dynamics of the system. The PIC is also modelled using a thorough and realistic travelling wave method, which is required for the complex system of multiple coupling sources, with varying delays and optical paths which may be found in modem PICs. The PIC parameters such as coupling magnitude are first investigated experimentally and then various methods are used to calibrate these relations within the travelling wave program. A number of tuning properties of the OBR laser(s) within the PIC are then given in both the model and experiment, with good correlation between the two found. The principal investigations of dynamics in the PIC are then described. This commences with details of the overall dynamics of the PIC, and the dynamic analysis tools required to represent and quantify these dynamics. The effects of varying coupling between the lasers over a range of frequency detuning values are then studied. The effects of coupling magnitude are quantified with respect to an overall transition to chaos in the system, and overall dynamic trends. Correlation is found between the experimental and modelled results. The system of phases in the PIC are then investigated, and this has impact on the permutations of phase between oscillators in respect of conditions such as spatial symmetry and index dispersion. The control of the system of phases as a single ensemble is then described, and this is found to produce a rich variety of dynamic behaviour from this subtle parameter. Phase controlled dynamics are demonstrated experimentally, and correlated dynamics in the model are demonstrated. Finally four wave mixing, which is present in many optical configurations is shown to have a profound effect on dynamics in the model, and experimental results are presented supporting these results. These parameters of coupling magnitude, detuning, phase and four wave mixing level represent the major contributing causes of dynamics in multi-laser PICs. The results presented here have implications for the design and use of PICs, either to exploit or avoid these dynamics.

Methods arc developed for modelling the effect of excitons (electron-hole bound states) on the optical properties of carbon nanotubes. The exciton effect is modelled using the methodology of the Bethe-Salpeter Equation, utilising a novel tight-binding formulation that is able to account for the effect of nanotube curvature. The exciton spectrum and fine-structure are calculated using a number of approximations to this model. The screening of excitons at high exciton density it; considered by both a static and dynamic screening model; the dielectric function being modelled by the random phase approximation and the plasmon pole approximation respectively. It is shown that the static screening model significantly underestimates the stability of the exciton at high exciton densities, in comparison to experimental results, but that the results of the dynamic screening model have better agreement with experiment, and predict stability of the exciton binding energy even at high exciton densities.