This thesis is concerned with structures and design techniques appropriate for the realisation of integrated optical receivers operating at multi-gigahertz frequencies. The development and practical proving of novel signal designs tailored specifically to very high bit-rate optical communication systems is reported. Timing imperfections and signal dependent noise - a result of the optical amplification deployed in all high-performance systems - are two major impairments that must be accommodated if optimum system performance is to be achieved. Here, a signal design that accommodates these impairments is developed and compared to established designs. The new signal designs are shown to provide improved performance, in particular, they exhibit tolerance to uncertainty in the exact level of the impairment. Following the derivation of the signal designs a range of practical realisations are described. A receiver amplifier GaAs MMIC for 4.8 Gbit/s operation with embedded signal shaping is described followed by the design and test of integrated post-detection filters for 10 and 15 Gbit/s systems. The susceptibility of the embedded signal shaping receiver to variations in photodiode capacitance leads to the development and test of a low inputimpedance common-gate 5 Gbit/s GaAs MMIC receiver. To effect signal shaping at very high data-rates a modified distributed amplifier structure is proposed which better utilises the capabilities of the available foundry processes. Two distributed amplifier based optical receivers with embedded signal shaping are devised and simulation results for 10 Gbit/s show the efficacy of this design approach. The implications of noise matching are investigated and a 2 GHz SCM receiver is used as a vehicle to illustrate the methods developed. The long term goal of receiver design is to fully integrate bnth optical and electrical components onto a single chip. A preliminary investigation of the feasibility of this goal is carried out on an experimental InP-based process. Two receiver designs for 10 Gbit/s were prepared as a precursor to a detailed design of an OEIC with embedded signal shaping that incorporates the novel topologies developed during this work.

No description available.

No description available.

No description available.

This research project explored the potential of forming an integrated optics technology based on silicon core waveguides suitable for application in sensors and communications in the wavelength range 1.2 to 1.6 mum. Integrated optics has evolved around the use of compounds such as lithium niobate and III-V semiconductors due to their available electro-optic properties. By contrast silicon has received relatively little attention as its indirect band gap has prevented the fabrication of light sources in the material and its centrosymmetric crystal structure means that it has no useful linear electro-optic effect. The lack of a demonstrated low loss integrated optical waveguide compatible with single mode optical fibres has been a further limitation. However, these major drawbacks in silicon waveguide technology may be more than offset by the potential advantages of forming silicon integrated optical devices using well established silicon microelectronics fabrication methods. The project focused research on waveguiding in silicon-on-insulator (SOI) structures with the aim of developing a practical low loss waveguide in these structures and understanding the various loss mechanisms. In principle the optical absorption of pure crystalline silicon over the wavelength range of interest allows waveguides with losses less than 0.1 dB/cm to be formed. SOI material formed by ion implantation has been developed for microelectronic applications and provides a commercial source of a silicon planar waveguide structure with high quality interfaces and low defect density. The project studied waveguides based on this material. Initially planar waveguides with silicon thickness from 0.57 to 7.3 microns and buried oxide thickness of 0.07 to 0.4 microns were studied. Fabrication methods and structures were identified which allowed multi-microns planar SOI waveguides to be formed with losses less than the benchmark of 1 dB/cm. For these structures a buried oxide thickness of 0.4 microns was found to be sufficient to prevent substrate leakage loss. It has been concluded that the predominate loss mechanism is scattering of light at the silicon to buried oxide interface. Rib waveguides were formed in SOI following the insight into loss mechanisms gained in the planar waveguide studies. Optical rib waveguides with widths from 2.73 to 7.73 microns were formed in SIMOX (Separation by IMplantation of OXygen) based SOI structures consisting of a 4.32 micron thick surface silicon layer and a 0.398 micron buried oxide layer. The effect of waveguide width, bend radius, Y-junction splitting and interface roughness on loss and mode characteristics were studied at wavelengths of 1.15 and 1.523 microns. The experimental results support the hypothesis that certain rib dimensions can lead to single mode waveguides even though planar SOI waveguides of similar multi-micron dimension are multimode. The propagation losses of waveguides 3.72 microns wide were found to be 0.0 dB/cm and 0.4 dB/cm for the TE and TM modes respectively when measured at 1.523 microns. The measurement uncertainty was estimated to be +/-0.5 dB/cm. These results are thought to be the lowest loss measurements for silicon integrated optical waveguides reported to date. During the course of the project other researchers have demonstrated useful electro-optic properties in silicon semiconductor junctions based on the free carrier plasma dispersion effect and room temperature electroluminescence in silicon based junctions. The combination of these developments with the practical waveguide structure demonstrated in this project now makes the possibility of developing a practical silicon based integrated optics technology a reality.

Visible light emission from silicon nanostructures formed by Si+ ion implantation into a SiO2 matrix and subsequently annealed at high temperatures (mainly 1300°C and 900°C) in various annealing atmospheres has been investigated. Various analyses techniques, such as Rutherford Backscattering Spectroscopy (RBS), Secondary Ion Mass Spectroscopy (SIMS), Photoluminescence (PL), and Transmission Electron Microscopy (TEM) were employed to characterize the structures in terms of their composition and optical properties. RBS and SIMS analyses revealed nitrogen and carbon impurities in the samples which we conclude originate from contamination, such as N2 and CO+ in the ion beam. PL analysis with a 488 nm Ar laser at 300 K showed that there was no visible PL from the samples before Si+ implantation or from the samples after Si+ implantation but before annealing. Also, N+ implantation gave rise to no PL. Si+ implanted samples with 2 x 1017 Si+ cm-2 and 6 x 10 17 Si+ cm-2 exhibited, after annealing at 1300°C for 30 minutes in a nitrogen ambient, strong visible PL, with a broad spectrum at peak wavelengths of 580 nm and 760 nm, respectively. There was a weak dependence of the PL peaks at 580 nm and 760 nm on annealing tune and annealing temperature. However, there was no PL from N+ implanted and annealed samples. Annealing the Si+ implanted samples in forming gas (FG) at lower temperatures (up to 1000°C) increased the PL peak intensity up to a factor of two, however, the PL peak wavelengths were the same. It is concluded that hydrogen annihilates the non-radiative recombination pathways. This effect provides evidence for surface states playing an important role in light emission. From PL analysis, using a short wavelength laser (325 nm), it was found that silica samples showed one broad PL spectrum at a peak wavelength of 440 nm, whereas samples consisting of a 1 mum thick SiO2 film exhibited several peaks which we found to be due to optical interference. Detailed observations of the fine structure in the PL spectra at low temperature (18 K), from the silica samples which were annealed in FG at 900°C revealed strong evidence for interaction between excitons and Si-O vibrations localized in a very small region. TEM analysis showed that there were precipitates after annealing at 1300°C in N2 in a sample implanted at a dose of 2 x 10 17 Si+ cm2 with an energy of 400 keV whilst HTEM analysis showed that the microcrystallites varied in size from 2.5 nm to 7.5 nm. However, TEM failed to show any precipitates in samples which were implanted with the same dose at an energy of 200 keV and have strong PL, and this is also another strong indicator that the PL emission is not simply due to quantum confinement. In this thesis, we propose a luminescent model to describe the mechanism for light emission. Physically the emitting structures are envisaged to have three regions, namely, a Si core (Si precipitate), an interfacial transition region whose composition varies from Si to SiO2, and the surrounding SiO2 matrix. Light emission occurs by a two step process involving generation and confinement of excitons in the Si core, whose band structure is modified from that of bulk crystalline silicon, and radiative recombination through the interaction of excitons and Si-O vibrations within the interfacial transition region. The justification for this model is discussed in this thesis.

"Smart Structures" or "Smart Skins" will require structurally integrated sensing systems that can operate in practical situations. Optical sensing techniques are receiving considerable attention for the monitoring of such systems. Single ended polarimetric sensors were utilized with a large dynamic range for strain measurements as surface mounted and embedded strain sensors in composite materials (glass fibre and carbon fibre reinforced polymers). They were also used to monitor the strain and the formation of microcracks in the glue line of carbon fibre reinforced polymer (CFRP) concrete beams. The intrinsic Fabry-Perot was also used as a surface mounted sensor to monitor axial strain of GFRP coupons. Finite Element (FE) modelling was used in order to investigate the stress/strain distributions within the composite material and the embedded optical fibre. The modelling results show excellent agreement with the experimental results and suggest that the soft acrylate coating is debonding, thus reducing the sensor's dynamic range. Actuators and/or Sensors embedded into a host material will disrupt the physical properties of the host. Finite element analysis was used to determine and to minimise the stress concentrations which arise in a "Smart" material system due to the embedded optical fibre sensor. A parametric study was undertaken to determine the theoretical mechanical and thermal properties of the interface coating that minimises the disruption of the polymer composite host material properties due to the optical fibre inclusion. The effects of transverse tensile and thermal loading were studied, and also the residual thermal stress concentrations due to the manufacturing process were taken into consideration. The stress concentrations in the composite host are affected by the dimensions, mechanical and thermal properties of the interface coating. The results show that with careful selection of the interface coating properties die stress concentrations in the host material caused by the optical fibre inclusion can be reduced and be similar to those of the pure host material.

Surface characterisation and failure mechanisms of adhesively bonded glass butt joints were studied. Materials of relevance to the adhesive bonding of fibre optic components were employed: a fine annealed Grade A glass (Schott(TM) BK - 7), and a fast curing epoxy based optical adhesive. The joint behaviour and their durability under adverse environmental conditions were investigated, and the subsequent, failed joint fracture surfaces were examined using XPS, ToF-SIMS and SEM. Surface analysis techniques have been employed to characterise components of the adhesive systems and to interrogate the surface of failed joints with a view to establishing the locus of failure. Joints were immersed in pure water for period of 0 - 270 days at 35°C. Substantial reduction in the bond strength within a few days of water immersion was observed. Also for the joints left in for a prolonged exposure periods (> 100 days) the strength values reached a minimum value. The surface analysis of the joints has indicated a cochoidal brittle fracture through the adhesive in dry conditions, and an interfacial failure for wet joint fractures. Calculations of the polymer thickness on the interfacial surfaces indicated a decrease in overlayer thickness. This reduction reached a constant value for joints left in water for more than 100 days. During this investigation surfaces segregation of minor components of the adhesive such as amine and diluent was also identified by ToF-SIMS and XPS. In order to study this phenomenon further, reformulation of the adhesive has been carried out. Six adhesives were formulated, of which three resins were based on the change in diluent content of the system, three were based on the change in amine concentrations. This work has clearly identified amine segregation only at the interfacially failed joints. It is expected that when the joints cures very fast, even if the formulation contains very high proportion of amine, the segregation is minimal. Also, some evidence of diluent migration at the interfaces was observed during the reformulation studies.

No description available.

No description available.