Understanding the magnetic structure and behaviour of domain walls has attracted recent attention in both academia and industry as it is of fundamental importance for the development of magnetic storage media. At the University of Glasgow we specialise in direct magnetic imaging through Lorentz electron microscopy. This PhD project presents an investigation into the development of magnetic imaging methods in the TEM and their application in imaging narrow domain walls in multilayer magnetic structures. Lorentz microscopy techniques are limited in quantitative magnetic imaging as this generally requires using scanning imaging modes which limits the capability of imaging dynamic processes. The first imaging method developed in this study is a phase gradient technique with the aim of producing quantitative magnetic contrast proportional to the magnetic induction of the sample whilst maintaining a live imaging mode. This method uses a specifically engineered, semi-electron-transparent graded wedge aperture to controllably perturb intensity in the back focal plane. The results of this study found that this method could produce magnetic contrast proportional to the sample induction, however the required gradient of the wedge aperture made this contrast close to the noise level with large associated errors. In the second part of this study we investigated the development of a technique aimed at gaining sub-microsecond temporal resolution within TEMs based on streak imaging. We are using ramped pulsed magnetic fields, applied across nanowire samples to both induce magnetic behaviour and detect the electron beam across the detector with respect to time. We are coupling this with a novel pixelated detector on the TEM in the form of a Medipix/Timepix chip capable of microsecond exposure times without adding noise. Running this detector in integral mode and allowing for practical limitations such as experiment time and aperture stability, the resultant streak images were taken in Fresnel, Foucault and low angle diffraction imaging modes. We found that while this method is theoretically viable, the limiting factor was the contrast of the magnetic signal in the streak and therefore the total image counts. Domain walls (DWs) in synthetic antiferromagnetically (SAF) coupled films patterned as nanowires offer exciting possibilities; the domain walls in these multilayers have narrower widths and reduced magnetostatic energy compared to those in single layer nanowires. In this study Co90Fe10/Ru/Co90Fe10 films were used to investigate the existence and structure of these walls. Nanowires were fabricated in these films and the DW structure was studied with respect to both wire width and varying magnetic layer thickness. It was found that while the DW structure does not appear to vary with the range of wire widths used, it changed significantly with varying thickness. The narrow DWs were observed to form only in samples with an unbalanced ratio of 1:25 : 1 or below; 1:75 : 1 and 2 : 1 samples were both observed to form transverse domain walls.

The desire to increase the amount of information that can be encoded onto a single photon has driven research in many areas of optics. One such area is the study of the orbital angular momentum (OAM) carried by a light beam. These beams have helical phase-fronts and carry an orbital angular momentum of l_hbar per photon, where the integer l is unbounded, giving a large state space in which to encode information. In the work that follows I discuss the development of new methods to measure the OAM carried by a light beam. An adaptation of a previously outlined interferometric technique is presented, resulting in a compact, robust measurement tool while dramatically reducing the number of degrees of freedom required for alignment. A new approach to sorting OAM is discussed, inspired by the simple example of the discrimination of plane waves focussed by a lens within direction space. This new approach is a telescopic system comprising two bespoke optical elements that transform OAM states into transverse momentum states; the various stages of development are outlined. Further to the development of this technique, investigations into the effects of misalignment and atmospheric turbulence on a communication link are presented. Outwith the area of optical communications, it is shown that by analysing the orbital angular momentum of light scattered from a spinning object we can observe a frequency shift many times greater than the rotation rate.

Strong coupling between a Fabry-Perot cavity mode and a quantum well exciton give rise to the new quasi-particles microcavity exciton-polaritons. Microcavity polaritons were for the first time demonstrated in 1992 in Ref [1] and since then, the field developed dramatically. Fundamental physics was explored in these systems including a quantum phase transitions of microcavity polaritons [2–4], demonstration of quantized vortices [5] and superfluidity [6]. Concerning applications [7], properties of microcavity polaritons are explored in optoelectronic devices e.g. low threshold electrically pumped polariton lasers, polarization sensitive optical bistable switches, spin memories and spin logic gates. Main research interest of this work concerned the field of quantum phase transitions and the many-body physics of microcavity polaritons which are relatively easily accessible experimentally as compared to similar physics in cold-atoms systems [8] from the point of view of equipment complexity. However, in polariton physics, samples with desirable properties play a crucial role. Microcavity samples commonly suffer from disorder, for both the exciton and the photon components of the polaritons, which strongly modifies polariton quantum effects and makes them difficult to interpret. This fact emphasizes the importance of the further development in the field of the sample design and growth, and this was one of the goals of this contribution. In particular, we worked to identify and suppress disorder in microcavity samples and to develop reproducible growth receipts providing samples with long photon lifetime. Photonic disorder was identified as cross-hatch dislocations and point-like-defects. A novel cross-hatch suppressing sample design was proposed and demonstrated, providing samples with long polariton propagation lengths in the order of millimeters in which genuine quantum fluid effects can be explored. Moreover, the origin behind the point-like-defects formation was identified as Ga nano-droplets deposited in microcavity during the molecular-beam epitaxial growth. These states were investigated using surface (differential-interference contrast microscopy, scanning-electron microscopy, chemical etching) and volume (focused-ion beam milling) techniques. Point-like-defect resulted in 0-dimensional polariton states exhibiting quantized energy levels which we have characterized in real and reciprocal space. The second part of this work was the investigation of quantum many-body effects in low disorder microcavities. In particular, we investigated polariton parametric scattering and demonstrated experimentally and theoretically scattering into ”ghost” branches which arises due to energy and momentum conservation of polaritons. Finally, we theoretically modeled quantum fluid effects of polaritons using Gross- Pitaievskii equation reproducing superfluid transition and bistability for spin independent polariton interactions.

In this thesis a new computer code is developed to perform non-equilibrium Green’s function based calculations of electronic transmission, using a Hamiltonian computed from self consistent extended Hückle theory as input. Individual elements of this code are tested to ensure correctness. To evaluate its usefulness, the code is tested on porphyrin based systems against the more traditional density functional theory methods of generating the required Hamiltonian. It is then used on more complex porphyrin systems, and comments are made on the use of porphyrin in molecular electronics. Finally it is used on DNA based systems too large to be dealt with efficiently via density functional theory to provide predictions of the effects of DNA structure on its conductance.

There is an industrial drive for the improved detection of sub-mm sized sur¬face breaking defects using non-destructive evaluation (NDE) methods [1]. Electromagnetic acoustic transducers (EMATs) are a non-contact NDE technique that utilise the generation and detection of Ultrasound using primarily Lorentz force mechanisms [2]. They are relatively safe and inexpensive, however, they suffer from low generation efficiency. The precise industrial drive for this work is improved ultrasonic crack detection of surface defects hidden by a thin metallic paint coating. The majority of standard ultrasonic techniques are not applicable as they require direct contact to the sample surface. Laser techniques, while non-contact, are still impeded by the coating, and eddy current techniques are difficult to implement due to interference from the metallic coating. EMATs are applicable, however their low generation efficiency limits the minimum defect that can be detected. This work presents improved resolution surface wave EMATs using geometric focusing for the detection of sub-mm sized surface breaking defects. Three main design types have been presented: a pseudo-pulse-echo focused meander-line EMAT, a pitch-catch focused racetrack EMAT and a pitch-catch focused linear EMAT. The first two designs have been fully characterised, finding the relations between coil geometry, focal point location and size, and the optimum operation frequencies [3, 4, 5]. Both designs have been used to size the lengths of a set of drilled calibration defects to accuracies of 10.5 and 10.4 mm respectively, and the pitch-catch design has been used to create a calibration curve for defect depth measurements. In addition, both designs have been used to map a pair of real surface breaking cracks in an aluminium billet sample to sub-mm resolution. The pitch-catch design has been used to detect a set of mm-size real thermal fatigue cracks in steel through a 40 - 60 ktm thick metallic paint coating. A four-coil EMAT design based on the pitch-catch focused racetrack EMAT has been built and demonstrated to detect surface breaking defects regardless of their surface orientation. Finally, the meander-line, racetrack, and linear coil design types have been compared based on their signal strength and their performance at lift-off from a sample surface. The meander-line designs have the strongest signal to noise ratios (SNR), with over 40 dB found when in contact with the sample, but the largest SNR loss with increased lift-off, reducing to 0 dB by 0.3 mm lift-off. The linear designs have the weakest SNRs, under 30 dB when in direct contact, but the smallest SNR loss with increased lift-off, dropping to 0 dB by around 1 mm, depending on the frequency of operation. This makes the linear coil designs optimal for situations requiring higher lift-off. Lower frequency designs are shown to perform better with increased lift-off regardless of the coil design, however, lower frequencies have less spatial resolution capabilities. A proposed linear-meander-line phased EMAT design is presented to generated 1 MHz signals but with the improved lift-off capabilities of the linear designs. This proves that surface wave EMATs can be optimised for surface wave detection of sub-mm defects through a metallic paint coating. While pseudo-pulse-echo focused meander-line EMATs are already in exsitence, there was previously no published work on their capabilities and full charaterisation. The other focused designs presented here are new designs in the field.

In this Work, the potential and feasibility of using structured silicon membranes as an active electro-optical light modulators in the Mid-Wave Infrared MWIR range are evaluated and compared. The near IR pump-Mid IR probe used to investigate spectral, fluence dependence and time dependence optical properties of nano and microporous silicon membranes. The results indicated that the spectral features of both samples is structure dependent parameter including the skeleton dimensions , pore and porosity . Both membranes demonstrated ground state transmission of about 90 % with 60% achievable modulation when pumped at fluences less than 5mJ/cm2. The nanoporous membrane demonstrated response time as fast as few tens of picoseconds allowing its use in the GHz regime, while microporous silicon demonstrated three order of magnitude slower allowing its use in the MHz regime. The time resolved results indicated that the carrier recombination of the microporous silicon follows Shockly Read Hall (SRH) with recombination coefficient of about 1.65 s-1. For nanoporous silicon, the SRH type dominates the recombination at low excitation fluencies, while at higher excitations, the Auger type competes with SRH and even dominates the recombination process. The longer SRH recombination observed at 3.5 µm attributed to vibrational-electronic state coupling.

Firstly, we investigated a novel method in which metal clusters are formed from the sputtering of a cryogenic solidified carbon dioxide gas matrix loaded with metal atoms, dubbed the Matrix Assembly Cluster Source (MACS). Using STEM images of the deposited silver clusters on TEM grids, we studied their size-dependence as a function of both the metal loading in the dry ice matrix and the energy ofthe sputtering ion beam used to extract them. Secondly, a MACS2 system was built, which enabled the deposition of metal clusters on a larger area. Silver, gold and binary gold-palladium clusters were produced with the MACS2 and deposited on presputtered graphite tape. Flakes of graphite loaded with metal clusters were produced and it is demonstrated that these samples show catalytic activity for the carbon monoxide oxidation reaction. Finally, gold nanoparticles were chemically embedded in porous silicon layers. Reflectometry measurements were used to determine the linear refractive index and the composition of the nanocomposites in near infrared. Ultrafast time-resolved pump-probe spectroscopy measurements are carried out on the samples, using an 800 om p-polarised pump and a 2.5 J.lm s-polarised probe. Carrier densities, linear and nonlinear optical constants are determined experimentally, supported by optical models.

In this thesis, from discrete spring-mass systems to continuous elastic solids, the possibility of achieving topological phases and elastic spin-Hall effect are analytically and numerically discussed. Originating from time-reversal symmetry breaking via applying external fields, a unidirectional and backscattering-immune edge state arises owing to the topological protection. Caused by the effective spin-orbit coupling, the elastic counterpart of spin-Hall effect arises at certain area of the momentum space. Also, the proposed arguments are verified by numerical calculation of practical mechanical crystals and elastic composites. We believe these studies pave the way for the future researches in topological elasticity. On the other hand, PT symmetry, which is a weaker restriction than Hermicity, allows real eigenvalues in a non-Hermitian Hamiltonian. However, it is challenging to introduce the PT condition into quantum mechanical systems. In this thesis, we consider an acoustic metamaterial made of periodically arranged spinning cylinders. By virtue of the rotational Doppler effects, the dispersion relation around the rotating speed of rods is significantly influenced by the rotation. The frequency shifts cause a PT symmetric Hamiltonian so that, at specific points, the spontaneous PT symmetry breaking emerges and exceptional points arise. Lastly a possible setup is discussed for the future experimental realisation.

It is an extremely exciting time for physics. In the last 100 years we moved from the formulation of Einstein's general relativity to the first direct observation of gravitational waves in late 2015 by the Laser Interferometer Gravitational-wave Observatory (LIGO). In that time we learned to use light to cool atoms to nearly absolute zero temperature, and to use atomic transitions in the microwave and optical regimes to devise the most accurate time and frequency references. We observed the wave-like behaviour of cold atoms in diffraction experiments using the periodic structure of light beams. Exploiting this wave-like behaviour, we constructed atom interferometers which allow us to test and measure gravity in a new scale. All of these experiments have one thing in common, from LIGO's giant 4 km arms, to the transportable atomic clocks sent to space, they all make use of a device that has become essential in many areas of science and technology: the Fabry-Perot optical cavity. This thesis delves deeply into the application of optical cavities at the forefront of experimental physics, and it is divided into three parts, each pertaining to a different field where optical cavities are a key technology.

The work presented in this thesis focusses on the development of a transportable atom-interferometry experiment and a compact fibre laser system towards precision measurements of gravitational acceleration. Interference fringes are shown with clouds of cold 87Rb atoms using co-propagating laser beams to drive stimulated Raman transitions. This is demonstrated both inside and outside of laboratory environments for which an integrated and transportable experiment is constructed. Further improvements are presented that enable the generation of clouds containing 1.7 · 108 atoms at a rate of 2.5 Hz and having a temperature of (7 ± 1) μK. This is largely due to the development of a compact laser system based on all-fibre coupled components. It is demonstrated that the laser system designed here can achieve fast frequency sweeps over 1.8 GHz within 2 ms, making it widely applicable in compact atom-interferometry experiments with rubidium atoms. This is shown by creating a Mach–Zehnder type interferometer with counter-propagating Raman beams, thus enabling measurements of gravitational acceleration. Since the laser system uses only two lasers and one fibre amplifier, a significant reduction in size is achieved, as well as a decrease in the total power consumption of the overall experiment by a third to (162 ± 7) W.