No description available.

For many years industry has considered RF PA design to be a "black art". This perception has been held due to the lack of availability of meaningful information for analysis and design. Due to the emergence of large signal waveform based measurements and increased understanding in the literature, it is now possible to characterise devices and correlate the information for enhanced PA design in terms of efficiency, linearity and/or reliability. This has been well documented and demonstrated using on-wafer devices but where this thesis work begins, little work had been done in expanding this capability to higher more meaningful power levels using packaged devices. This work has successfully addressed both of these limitations and extended visibility of time domain waveform data to higher power levels. Thus, allowing for the uncovering of world record efficiency levels of 77 % (4W output power) for Si LDMOS devices at S band frequencies using waveform engineering based procedures, in this case Class F. A feat previously only reported at L-band frequencies. Other waveform based designs such as inverse Class F and Doherty modes of operation are also successfully demonstrated in this thesis. In both of these cases, voltage related issues affecting reliability were uncovered that merit further consideration in the design process. Waveform engineering was made possible by applying de-embedding the measured current and voltage waveforms to the current generator plane. That is the plane at which the device is free from any device and package parasitics with the current and voltage waveforms seen to be in good agreement with those typically found in literature. These were successfully applied at high power levels (110W) previously not reported. To further demonstrate the relevance of waveform de-embedding, a non-linear charge conservative model based on industry standard modelling techniques was compared against time-domain measurements conducted in several classes of operation. This form of model verification is often overlooked and provides a unique insight into the model's accuracy highlighting new areas of improvement. In most cases the model was shown to be in good agreement with measured data, providing a high level of confidence in the application of waveform engineering principles within the CAD domain. Thus providing the PA designer the facility to apply waveform engineering on both the test bench and within the CAD domain.

The ability to precisely analyse the composition of liquid mixtures by non-contact techniques in both static and flow situations is extremely desirable for a diversity of industrial, analytical and quality control procedures. Microwave resonators allow very accurate and sensitive characterisation of the dielectric properties of polar liquids due to the strong interaction of the latter with microwave electric fields. They have the useful dual role of both exact characterisation of the complex permittivity of a dielectric sample when it is inserted within a region of high electric field of the resonator, and effective volumetric heating of the same sample if its dielectric loss is large enough to permit heating. They offer tremendous potential for investigation of very small amounts of polar solvents in non-polar hosts. In this regard they are superior to other traditional composition analysis techniques such as liquid chromatography, gas chromatography and mass spectrometry in the speed of analysis ( 1 s), non destructive nature and scope for miniaturisation of the resonator size to suit the system under test. For minute sample volumes, the resonator perturbation technique is extensively used for dielectric measurements on polar liquids. In this project, it has been employed for highly sensitive compositional analysis of two-component dielectric mixtures contained in microcapillary segments. The primary evaluation system used here was mixtures of acetonitrile and toluene, chosen because of the large difference in their molecular electric dipole moments. The results obtained from this first system provided the inspiration to assess mixtures made of acetonitrile and water, which are much more closely matched in terms of their electric dipole moments. Three different types of resonators namely hairpin resonator, split ring resonator and sapphire dielectric resonator were used to analyse both the aforementioned solution systems. The results show very sensitive characterisation and are in close agreement with the theoretical predictions governing perturbation of resonators by dielectric samples. In the last phase of this research, a miniaturised sapphire dielectric resonator was designed and fabricated that provided added enhancement in measured sensitivity of both evaluation mixtures.

Reconfigurable antennas have been a subject of rapidly increased interest during the past decades. This has been prompted by the increased demand on new wireless communications technology in both civilian and military directions. Moreover, different types of reconfigurations have been identified and investigated to keep up with the demand for new technologies. In this research, the possibility of designing reconfigurable Dielectric Resonator Antennas (DRAs) have been explored with different types of reconfigurability directions, especially with the increased interest in the area of DRAs during the past three decades. These results have been satisfactory in general. The main aim of this research is to experiment with different reconfigurability designs, each purpose is to achieve one type of reconfigurability or more. This includes, polarisation reconfigurability in Chapter Three, frequency agility in Chapters Four and Five, beam steering and gain agility in Chapter Five. Furthermore, this research main aim has been to investigate new ways to exploit the advantages of the semiconductor plasma in reconfigurable antennas. However, research’s limited resources led to reduce the efforts in this area to only one experiment, which is presented in Chapter Six, based on a similar design presented in Chapter Four. Although the results have been conflicted for the last experiment, the results shown that the used reconfigurability medium (AlGaN/GaN HFETs) can be benefitted better from it in other application. Two models have been introduced for polarisation reconfigurability, a hemispherical DRA couple with reconfigurable annular slot excitation, and a notched rectangular DRA with reconfigurable parasitic strip(s). Both designs shown the possibility of achieving LP/CP radiations. In addition, rectangular DRAs that are excited with single, as well as multiple, slot have been studied. Prototypes have been built and measured with reasonable agreement between practical and simulated results. Furthermore, the work has been extended to study a reconfigurable DRA linear array where several designs have been investigated including single and dual-slot for two and four-element linear arrays. The single-slot model reconfiguration resulted in the expected beam steering alongside the array direction. On the other hand, both frequency tuning and beam steering have been achieved with the dual-slots models. Finally, the semiconductor plasma reconfigurable antennas have been considered with the investigation of AlGaN/GaN HFETs as a replacement for the well investigated and presented silicon SPIN diodes. The prototype has been measure and discrepancies between measurements and simulations have been discussed.

Non-thermal plasma discharges, particularly dielectric barrier discharges (DBDs), are the most common method of ozone generation. The aim of this research was to optimize the micro-discharges in DBDs, to improve ozone generation efficiency. The electrical characteristics of DBDs were researched, and the effects of the physical and electrical parameters of DBDs on ozone generation were investigated. Four DBD based ozone reactors, including plate configuration and cylindrical configuration, were designed and developed. Three major energization modes, including transient (40 ms) AC power supply, continuous AC power supply and pulsed power supply, were used for the investigation of ozone generation performance. Under transient AC energisation in oxygen, it was found that the ozone generation efficiency at 2 bar absolute was 217 g/kWh, increased by 31% compared with that at 1 bar absolute. The ozone generation efficiency was found to increase with decreasing E/N in the range from 126 Td to 185 Td. Under continuous AC energisation, the ozone concentration was found to increase as the gas flow rate decreased (from 1 L/min to 0.4 L/min) or applied voltage was increased (from 3.5 kV to 6 kV). Under optimized conditions, the highest ozone concentration obtained was 271 g/Nm3, which is promising in comparison with previously-published data. The ozone generation efficiency was found to reduce as the ozone concentration increased above 30 g/Nm3. Furthermore, it was found that the AC energization frequency had no obvious effect on the behaviour of micro-discharges, or on the ozone generation efficiency. Pulsed DBDs for ozone generation was found to be less efficient than continuous AC energisation, for the conditions investigated herein. This research has achieved the desired combination of high ozone generation efficiency at high ozone concentration (>150 g/Nm3), based on DBDs The curve of ozone generation efficiency versus ozone concentration achieved shows more efficient performance than that in the literature: at the typical industrial ozone concentration of 150 g/Nm3 for waste water treatment, the ozone generation efficiency in this work was ~8.2 kWh/kg, ~20% higher than that in the literature.

The drive for miniaturisation of electronic circuits provides new materials challenges for the electronics industry. Indeed, the continued downscaling of transistor dimensions, described by Moore’s Law, has led to a race to find suitable replacements for current interconnect materials to replace copper. Carbon nanotubes have been studied as a suitable replacement for copper due to its superior electrical, thermal and mechanical properties. One of the advantages of using carbon nanotubes is their high current carrying capacity which has been demonstrated to be three orders of magnitude greater than that of copper. Most approaches in the implementation of carbon nanotubes have so far focused on the growth in vias which limits their application. In this work, a process is described for the transfer of carbon nanotubes to substrates allowing their use for more varied applications. Arrays of vertically aligned multiwalled carbon nanotubes were synthesised by photo-thermal chemical vapour deposition with high growth rates. Raman spectroscopy was used to show that the synthesised carbon nanotubes were of high quality. The carbon nanotubes were exposed to an oxygen plasma and the nature of the functional groups present was determined using X-ray photoelectron spectroscopy. Functional groups, such as carboxyl, carbonyl and hydroxyl groups, were found to be present on the surface of the multiwalled carbon nanotubes after the functionalisation process. The multiwalled carbon nanotubes were metallised after the functionalisation process using magnetron sputtering. Two materials, solder and sintered silver, were chosen to bind carbon nanotubes to substrates so as to enable their transfer and also to make electrical contact. The wettability of solder to carbon nanotubes was investigated and it was demonstrated that both functionalisation and metallisation were required in order for solder to bond with the carbon nanotubes. Similarly, functionalisation followed by metallisation was critical for bonding carbon nanotubes to sintered silver. A step by step process is described that allows the production of solder-carbon nanotubes and silver-carbon nanotubes interconnects. 4-point probe electrical characterisation of the interconnects was performed and the interconnects were shown to have a resistivity of 5.0 × 10-4 Ωcm for solder-carbon nanotubes and 5.2 × 10-4 Ωcm for silver-carbon nanotubes interconnects. Ramp to failure tests carried out on solder-carbon nanotubes interconnects showed current carrying capacity of 0.75 MA/cm2, only one order of magnitude lower than copper.

No description available.

The heat transfer coefficients in even relatively simple components varies significantly at different points on the surface. Although this has been known for some time the amount of work which has been done is limited to the relationship between these variable thermal conditions and the generation of stress and strain during the quench. The object of the work was to investigate the relationship between the variables using standard quenchants. The variation in the surface heat transfer coefficients has been determined at points along the horizontal and vertical axes of a stainless steel plate and upper and lower surface of a plate held vertically and horizontally in water. Photographic studies of the appearance of the quenchant in contact with the surface of the plate have been carried out in conjunction with the horizontal and vertical quenches which are mentioned above. The data obtained have been used in the determination of the variation in stress and strain generated in the plate during the quench by involving a 2-D finite element analysis using the PAFEC package.

Spintronics is the field concerned with the control of electron spin. In logic devices electron charge is manipulated, in computer data storage the magnetisation of a domain is altered; spintronics offers a hybrid between the two. This could be exploited in non-volatile random access memory cells for low power data storage. All-metal Lateral spin-valve devices were fabricated by electron beam lithography to investigate spin transport phenomena. The fabrication and measurement processes were optimised and lateral spin-valve devices were successfully fabricated with spin diffusion lengths of (200±25) nm and (310±30) nm in 100 nm and 200 nm wide Copper wires respectively. Spin filtering was previously observed by patterning nano-scale wires to be laterally asymmetric. Here, nano-scale wires were patterned to have a laterally symmetric spin diffusion path. No increase in signal due to the filtering effect was observed, thus confirming the phenomenological model put forward. Also, the spin diffusion path in a lateral spin valve was split into a ring geometry. By applying a field gradient across the ring, the operational efficiency was improved by 30%. The observation of a mechanically induced spin current has been achieved for the first time. The design of an optical measurement system that rotates a sample at up to 200~Hz is presented here. Deviation in the moment on the surface of a paramagnetic Tungsten foil from the moment induced by the Barnett effect confirms that a spin current may be induced by mechanical rotation. In summary, design and development of magneto-electrical and mechano-optical measurement systems has been achieved. The improvement in the operational efficiency in lateral spin-valves could be used alongside materials such as Heusler alloys to provide cheaper efficient logic devices. The observation of a mechanically induced spin current in Tungsten precedes the future study of the effect in other paramagnetic materials, such as Platinum or Palladium.

The merging fields of photonics and organic electronics into organic optoelectronics has created a surge of enthusiasm over the possibility of developing low-cost and large-area advanced optoelectronic systems. These applications can combine the best functionalities of both fields, such as tailoring the organic semiconductors by chemical means, engineering the structure in which organic materials are embedded in, are to name a few. These advances have stimulated the excitement over the next generation of optoelectronic systems with enhanced capabilities and low-cost manufacturing processes compared to their inorganic counterparts. Such technology direction is mainly reflected by the high investments towards the aim of developing flexible, and roll-to-roll organic light-emitting diodes and organic solar cells. Interestingly, more sophisticated applications require a deeper understanding of the underlying mechanisms at play that merge concepts from the fields of photonics and organic electronics. Particularly, organic light-emitting diodes (OLEDs) under certain constraints (such as cavity light confinement, strong exciton-photon interaction) exhibit modified spectral emission compared to OLED devices that are not bounded by the same conditions. The introduction of the polariton concept as a quasi-particle, which is part-light and part-matter, has emerged to describe such new physical phenomena caused by this photon-exciton intricate interaction. Polariton physics is well established in inorganic semiconductors were a plethora of physical phenomena have been demonstrated, such as the appearance of Bose-Einstein Condensation or low-threshold laser devices. The later is what has as yet to be demonstrated from the field of solid state physics utilising organic semiconductors. This thesis is focused on the study of the physics and the engineering of organic light-emitting diodes that will aid in the realization of efficient organic polariton LEDs. The main body of work examines various organic semiconductor materials in their ability to reach the strong light and matter interaction regime and, subsequently, to be used in OLEDs as the emissive component. Furthermore, a degradation investigation highlights the issues that affect small-molecule based OLEDs, and finally, the possible pathways for achieving efficient polariton OLEDs are discussed.