The strong confinement provided by plasmonic resonances has extended optics down to the nanoscale, allowing an unprecedented control over the interaction between light and matter. This could have far reaching applications in the development of ultra-compact and novel op- toelectronic devices. However, for commercial implementation of these plasmonic devices to become a reality there needs to be a shift toward designs which are compatible with the materi- als and processes of the established semiconductor industry. This is the overarching aim of the work presented in this thesis; here plasmonic devices are developed around the semiconductor on insulator architecture using a simple monolithic fabrication processes. Two waveguides are proposed and analysed, both produced through a single lithography step where a metallic slot or strip is formed on top of an SOI wafer. This process circumvents the etch step required to produce the waveguides used in silicon photonics. Despite this exceptionally simple fabrication procedure the designs support bound modes with areas as small as λ20/1000. Importantly, the mode size can be controlled through the width of the slot or strip and though careful design this can be used to effectively nanofocus light from larger low loss modes down to the nanoscale. The slot design is demonstrated experimentally with widths as narrow as 10nm. Following this, a similar design is implemented as a plasmonic laser. Here the SOI wafer is swapped for a suspended membrane of GaAs to provide the necessary gain. Lasers with slot widths as small as 100nm are demonstrated experimentally. A final device design is also discussed, where a highly effective cavity is formed through an array of metal resonators coupled to a semiconductor slab. The devices demonstrated in this thesis aim to provide a platform that allows the unique capabilities of plasmonics to be easily integrated with existing technologies.

Magnetised liner inertial fusion (MagLIF) represents a promising pathway to controlled thermonuclear fusion which would provide clean, plentiful energy. The concept uses a pulsed power machine to implode a metal cylinder or `liner' containing pre-magnetised and preheated fusion fuel; a critical limitation of such systems is the magneto-Rayleigh-Taylor (MRT) instability which primarily disrupts the outer surface of the liner. We carried out 3D simulations using Gorgon, an Eulerian resistive MHD code, to match experimental results showing large amplitude multi-mode MRT instability growth resulting from MagLIF-relevant liner implosions. These simulations under-estimated MRT amplitudes and wavelengths near stagnation due to a lack of azimuthal correlation, achieving good agreement only with the addition of an artificially azimuthally correlated initialisation. The experiment was repeated with an axial magnetic field, resulting in re-orientation of the MRT instability into a helical structure which has yet to be explained. We have shown that the missing azimuthal correlation could be provided by a combination of the electro-thermal instability (ETI) and an `electro-choric' instability (ECI); describing respectively the tendency of current to correlate azimuthally early in time due to temperature dependent Ohmic heating; and an amplification of the ETI driven by density dependent resistivity around vapourisation. We developed and implemented a material strength model to improve simulation of the solid phase of liner implosions and present test problems and benchmarking against the hydrodynamics code iSALE. Applied to simulations exhibiting the ETI and ECI, the inclusion of strength gave a significant increase in wavelength and amplitude of the ETI and ECI. Full circumference simulations of the multi-mode MRT instability provided a significant improvement on previous randomly initialised results and approached agreement with experiment. Simulations including an axial magnetic field reproduced helical structures associated with azimuthal currents induced by magnetic field compression, but did not reproduce experimental results.

The Northern Caspian lies along the southern-most boundary within the Northern Hemisphere where sea-ice forms. Although ice cover is typically observed between November and March, it experiences high spatio-temporal variabilities. This variability may be of interest in the long-term as it may act as an early indicator of large-scale climate change, as well as being an area of interest to industries and vulnerable species. Through empirical and model studies, this thesis carries out a sensitivity study of Caspian sea-ice for the first time. Caspian sea-ice concentration from satellite passive microwave data and surface daily air temperatures are analysed from 1978 to 2009. Relationships between mean winter air temperatures, cumulative freezing degree days (CFDD) and the sum of daily ice area are found for the first time. Mean monthly air temperatures of less than 5.5-9.5°C, and a minimum CFDD of 3.6-11.2°C, is required for ice formation in the Northern Caspian. Examination of climate projections from multi-model ensembles of monthly mean air temperatures suggest for the first time that the Northern Caspian may be largely ice-free by 2100 for the highest emission scenario. An ocean-ice-atmosphere model of the Caspian shows weak sensitivities of the minimum CFDD to varied sea-ice albedo and ice compressive strength. Sea level decline reduces the minimum CFDD and promotes formation of higher concentration ice. An atmosphere model of the Caspian is run with observed 2006 to 2009 sea-ice cover, with an additional run without ice cover, to quantify the sensitivity of the atmosphere to sea-ice for the first time. Ice cover removal results in up to 5-10% increase in precipitation, surface wind speeds and humidity, with up to 3°C increase in surface temperature, in the December-January-February climatology over the Northern Caspian. Additionally, extreme precipitation and extreme wind speeds intensify and extreme cold air events weaken with ice removal.

This thesis presents the milestone efficiency of 30.6% under 54x concentration achieved by a monolithic tandem solar cell containing strain balanced quantum well sub-cells. This is the first demonstration of tandem cells containing quantum well sub-cells. Characterisation and modelling of the quantum efficiency and dark currents of monolithic tandem cells containing quantum well sub-cells are also presented for the first time. Predictions performed by a simulation program indicate that 32.3% efficiency at 500x concentration can be achieved and in combination with an active Ge substrate efficiencies of 37.5% are possible. The model created in the simulation program is presented which makes possible the accurate fitting of quantum efficiency and prediction of dark current. The procedure developed to process tandem solar cells into concentrator devices is presented. A detailed account of the characterisation methods used for tandem solar cells are also described. Going further, a new optical concept for building integrated PV design was examined in this thesis. The design is based on a Buffon lens as an alternative to convex and Fresnel lens designs. The low cost design presented can be accommodated in the space of a double glazing window for glass facades.

Adaptive Optics (AO) is a technology which corrects for the effects of the atmosphere and so improves the optical quality of ground based astronomical observations. The bright “guide stars” required for correction are not available across the entire sky, so Laser Guide Stars (LGSs) are created. A Natural Guide Star (NGS) is still required to correct for tip-tilt as the LGS encounters turbulence on the uplink path resulting in unpredictable “jitter”, hence limiting corrected sky coverage. In this thesis an original method is proposed and investigated that promises to improve the correction performance for tomographic AO systems using only LGSs, and no NGS, by retrieving the LGS uplink tip-tilt. To investigate the viability of this method, two unique tools have been developed. A new AO simulation has been written in the Python programming language which has been designed to facilitate the rapid development of new AO concepts. It features realistic LGS simulation, ideal to test the method of LGS uplink tip-tilt retrieval. The Durham Real-Time Adaptive Optics Generalised Optical Nexus (DRAGON) is a laboratory AO test bench nearing completion, which features multiple LGS and NGS Wavefront Sensors (WFSs) intended to further improve tomographic AO. A novel method of LGS emulation has been designed, which re-creates focus anisoplanatism, elongation and uplink turbulence. Once complete, DRAGON will be the ideal test bench for further development of LGS uplink tip-tilt retrieval. Performance estimates from simulation of the LGS uplink tip-tilt retrieval method are presented. Performance is improved over tomographic LGS AO systems which do not correct for tip-tilt, giving a modest improvement in image quality over the entire night sky. Correction performance is found to be dependent on the atmospheric turbulence profile. If combined with ground layer adaptive optics, higher correction performance with a very high sky coverage may be achieved.

Blazars are numerous and often bright sources of cosmic γ-rays with several hundred such objects currently detected in this regime. However, there are several outstanding issues surrounding them. For example, it is uncertain exactly where in the kiloparsec- scale blazar jet γ-rays are produced and what the physical mechanisms responsible for the emission are. This thesis is an investigation in the physics and astronomy that can be learnt from studying the γ-ray energy spectra of blazars. Two studies are presented analysing bright blazars with the Fermi-Large Area Telescope (LAT). From the smooth spectral curvature seen in many of the objects and the lack of sharp spectral breaks it is concluded that the emission region is likely beyond the radius of the broad line region of 0.1 pc. It is seen that nearby confusing objects can lead to apparent breaks being ob- served in the spectra. However, a light curve analysis is performed of the bright blazar 3C 454.3 and it is seen that spectral breaks do sometimes occur when the object is in high flux states. From the shift in the peak of the γ-ray emission it is concluded that the high flux states are caused by an increase in the Lorentz factor of the emission region or by changes in the population of electrons in the emission region, but the origin of the breaks remains unexplained. Finally, a study is presented examining the effect that hypothetical axions or axion-like particles (ALPs) would have on the spectra if these particles should exist. It is found that under the right conditions a sudden flux boost could appear in the spectrum at a few TeV. The chances of the future Cherenkov Telescope Array (CTA) detecting the effects of photon-ALP mixing in AGN spectra is considered through simu- lated observations. These effects potentially present themselves in the spectra two ways: firstly the aforementioned flux boost and secondly due to the fact that the existence of ALPs could mitigate the pair absorption that γ-rays undergo when they traverse inter- galactic space. It is concluded that CTA will have good prospects of either detecting these effects or else, if no effect is seen, setting limits on the mass and coupling constant that ALPs could have.

Passive spectroscopy diagnostics play a key role in advancing the physics of plasma exhaust in the edge and divertor regions of tokamaks. Information obtained from spectral line intensities and profile shapes is crucial for estimating plasma parameters, particularly for studies of cold and dense detached plasmas. This work aims to characterise the visible and near-infrared spectral regions with emphasis on the Balmer and Paschen hydrogen series lines, their diagnostic applications and interpretation techniques. Whereas the Balmer series lines are measured routinely, few observations of the Paschen series lines have been carried out in the fusion plasma context. Extending observations to the near-infrared region for more detailed studies of the Paschen series is addressed through diagnostic development with the aim of providing coverage of the visible to near-infrared spectral range (350-1900 nm) along the same optical line-of-sight. An initial spectral survey on MAST using a purpose built diagnostic provides new insight into the spectral features in the near-infrared and confirms the viability of Paschen line observations. Following the proof of concept measurements on MAST, diagnostic development on the JET ITER-like wall mirror-linked divertor spectroscopy system facilitated more refined measurements. The main outputs include first of its kind measurements of the Pa-alpha line and spatially resolved spectral line profile measurements of the Pa-beta line. In the visible range ELM-resolved Balmer and impurity emission profile measurements at high spatial resolution were obtained using a new filtered camera system. A detailed assessment of the diagnostic scope for parameter estimation from both high-n and low-n Balmer and Paschen series lines is presented, underpinned by a parametrised line profile model which captures the relevant broadening and splitting mechanisms, including the Zeeman, Stark and Doppler processes. Interpretation of JET divertor plasma measurements, in combination with synthetic data simulation results, highlights the importance of complementary Balmer and Paschen series measurements for refining parameter estimates in divertor plasmas.

This thesis presents the work towards the realisation of an interferometer using bright solitary waves in a ring. The splitting of a bright solitary wave, which is created from a 85Rb Bose-Einstein condensate in an optical waveguide, is realised through scattering from a narrow potential well formed from a tightly focussed red detuned laser beam. We observe reflection of up to 25% of the atoms, along with the trapping of atoms at the position of the potential. Such a reflected fraction is much larger than the theoretical predictions for a single, narrow Gaussian potential. A more detailed model, which accounts for the diffraction pattern of the laser beam, suggests that the presence of these small subsidiary intensity maxima is the cause of the enhancement in quantum reflection. An upgrade of the apparatus sees a new set of magnetic coils, a compact coil mount, and a crossed optical dipole trap with independently controllable beams implemented. This enables the control of magnetic curvature and dipole trap position, and maximises the optical access to the science cell. To generate a ring trap for the interferometry scheme, a spatial light modulator (SLM) is incorporated into the experiment. Through underfilling the SLM panel with the laser beam, and the use of the analytical first phase guess prior to the error minimising Mixed Region Amplitude Freedom (MRAF) algorithm, we are able to generate speckle-free, high quality holograms of arbitrary shapes. Furthermore, we demonstrate atom trapping in a ring potential, which is formed at the intersection of the SLM beam and a red detuned horizontal light sheet.

No description available.

II-VI semiconductor material ZnCdMgSe has the potential to enable optical devices emitting throughout the visible spectrum. While difficulties in doping of this material have hindered its development for conventional electrically-injected semiconductor lasers, the recent availability of efficient, high power InGaN-based laser diodes has created the opportunity for optically-pumped devices, and this work primarily focusses on the progression towards realising vertical external-cavity surface-emitting lasers (VECSELs) based on this material system. Challenges in the growth of a ZnCdMgSe distributed Bragg reflector (DBR), such as low refractive index contrast and limited growth thickness for maintaining material quality, lead to the design of novel thin-film VECSEL structures, and the development of epitaxial transfer techniques to overcome the absorptive InP growth substrate and buffer. Transfer of thin-film (few microns thick) multi-quantum well heterostructures is demonstrated for samples with areas of a few mm², successfully transferring and liquid-capillary-bonding the films to diamond heat-spreaders for thermal management. Continued challenging growth, namely heterostructure layer inaccuracies, mean that laser threshold is not yet reached, however extensive characterisation and analysis is carried out to inform future progress in realising the ZnCdMgSe thin-film VECSEL. The thin-film VECSEL architecture offers advantages beyond allowing for the use of novel materials, opening the potential for novel laser cavities and optical pumping schemes. The thin-film transfer method developed for the II-VI VECSEL is adapted for the transfer of III-V AlGalnP epitaxial structures from GaAs growth substrates, and AlGalnP thin-film VECSELs are demonstrated operating continuous wave at red wavelengths at room temperature. Laser operation is currently limited by pump-induced de-bonding from the diamond, with attempts made to counter this through the refinement of structre design (including strain balancing) and transfer method. Until thermal rollover occurs, performance is relatively comparable with the 'conventional'; gain-mirror AlGaInP VECSELs, with a maximum output power of 21 mW recorded at both 682 nm and 670 nm for low output coupling. Using the transfer method developed for the II-VI material, ZnCdMgSe multi-quantum well structures are used as colour conversion films for micron-size LED arrays. The resulting hybrid devices are demonstrated to have high modulation bandwidths, limited only by the LED modulation bandwidth, suitable for application in visible-light communication.