Not available / text

In this paper, we report on an increase in emission intensity of up to 10 nW/mm2 that has been realized with a new novel two junction, diagonal avalanche control, and minority carrier injection silicon complementary metal–oxide–semiconductor (CMOS) light emitting device (LED). The device utilizes a four-terminal configuration with two embedded shallow nþp junctions in a p substrate. One junction is kept in deep-avalanche and light-emitting mode, while the other junction is forward biased and minority carrier electrons are injected into the avalanching junction. The device has been realized using standard 0.35 mm CMOS design rules and fabrication technology and operates at 9V in the current range 0.1– 3 mA. The optical output power is about one order of magnitude higher for previous single-junction nþp light-emitting devices while the emission intensity is about two orders of magnitude higher than for single-junction devices. The optical output is about three orders of magnitude higher than the low-frequency detectivity limit of silicon p–i–n detectors of comparable dimensions. The realized characteristics may enable diverse optoelectronic applications in standard-CMOS-silicon-technology-based integrated circuitry.

Not available / text

Not available / text

published_or_final_version / abstract / toc / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

No description available.

Non-destructive testing is an important field of research in many areas of industry and science. This field covers methods that test some property of an object while not impacting its future usefulness. This project is specifically interested in a method of NDE called Laser Doppler Vibrometry (LDV), which uses light to probe the motion of an object's surface. This thesis presents research into the possibility of developing a full-field LDV camera capable of measuring the vibration at a number of points across an object simultaneously. The approach chosen was to develop a single processing element which contained a light detector as well as the processing required to produce an output signal proportional to the vibration of the surface imaged by the element. This processing element, or pixel, could then be used to produce a full-field LDV camera capable of imaging an object's surface to provide vibration amplitude and frequency for a number of points across the whole surface simultaneously Some early work was carried out measuring the vibration of a piezoelectric target with a CMOS camera and the results were compared with theory. The method of processing chosen is called zero-crossing demodulation and a circuit was designed and simulated using this method. This circuit was then laid-out and a CMOS chip was fabricated with the pixel on it. The circuit was designed such that the different parts of it could be tested separately using electrical signals, and a number of electrical tests were performed to check how well the operation of the fabricated pixel compared with simulation. Using a laser interferometer the pixel was finally tested with real vibrating objects and compared with a commercial LDV.

Modulated Light Cameras represent a breed of cameras designed specifically to capture intensity modulated light. This is because using coherent detection it is possible to lift a signal of interest out of the background noise and thus increase the precision of measurements. This work presents a camera designed to detect the phase of amplitude modulated light. By implementing an in-pixel demodulation, wide-field detection of the phase of light is possible. The camera provides 32 by 32 pixels, each with a pitch of 115 μm with a fill factor of 16 %. This pixel used in the camera introduces a novel tuning mechanism that matches the camera to the frequency of operation and light conditions. This enables the camera to work at high modulation depths, and increases the detection frequency to 50 MHz. The camera also provides an improved linear response without compromising on dynamic range and pixel size. The noise response of the camera is also improved as compared with previous work performed. The camera has been demonstrated in wide-field range measurements of a scene (Imaging LIDAR). It has also been applied to wide-field heterodyne interforemetry and in ultra-stable interferometry.

Optical microresonators can form the basis of all-optical switching and control devices. The presented study is an exploration of the Transmission Line Modelling (TLM) method as a suitable candidate for designing optical microresonators. Chalcogenide glasses were identified as promising materials, with which to fabricate optical microresonators. The study presents the formulation of TLM in two dimensions to model nonmagnetic dielectric materials and a suitable computationally efficient yet flexible software design. Some methods for extracting spectral properties of resonators are compared and the modified difference Prony method was identified as a suitable tool to extract resonant frequencies and Q factors from a limited time signal. When applying TLM to microresonators of sub-wavelength dimensions it was understood that the method of discretisation plays an important role in accurately modelling microresonators. Two novel methods of discretisations -the same area method and the anti-aliasing method- were used to improve the accuracy significantly compared to existing mesh refinement techniques. Perfect matched layers (PMLs) were implemented to improve reflections from domain truncation using several methods. A Convolutional PML(CPML) was identified as the best, but it does not reach the efficiency of PMLs in the Finite Difference Time Domain (FDTD) method. Several frequency dependent refractive index models were proposed and implemented in TLM. A Tauc-Lorentz model was identified as the best fit to the experimental refractive index of three chalcogenide glasses, but a Sellmeier model with one term and a coefficient was efficient for TLM implementation. The main concern in the use of these models within TLM was shown to be the error arising due to mesh dispersion. Kerr nonlinear models were formulated and implemented in TLM and the models applied to the study of a waveguide junction. Compared to an equivalent implementation in a time domain beam propagation method, TLM models better represent the waveguide junction reflections.

No description available.