The phase behaviour of lattice polymers and peptides, under various conditions, is investigated using Monte Carlo simulation. Wang-Landau sampling is used so that, in principle, phase diagrams can be determined from a single simulation. It is demonstrated that the pseudophase diagram for polymer molecules, in several environments, can be plotted when sampling only from the internal degrees of freedom, by determining an appropriate density of states. Several improvements to the simulation methods used are detailed. A new prescription for setting the modification factor in the Wang-Landau algorithm is described, tested and found, for homopolymers, to result in near optimum convergence throughout the simulation. Different methods of selecting moves from the pull move set are detailed, and their relative efficiencies determined. Finally, it is shown that results for a polymer in a slit with one attractive surface can be determined by sampling only from the internal degrees of freedom of a lattice polymer. Adsorption of lattice polymers and peptides is investigated by determining pseudophase diagrams for individual molecules. The phase diagram for a homopolymer molecule, near a surface with a pattern of interaction, is determined, with a pseudophase identified where the polymer is commensurate with the pattern. For an example lattice peptide, the existence of the new pseudophase is found to depend on whether both hydrophobic and polar beads are attracted to the surface. The phase diagram for a ring polymer under applied force, with variable solvent quality, is determined for the first time. The effect, on the phase diagram, of topological knots in the ring polymer is investigated. In addition to eliminating pseudophases where the polymer is flattened into a single layer, it is found that non-trivial knots result in additional pseudophases for tensile force.

Temperatures below 0.1 kelvin can be nowadays routinely attained. The methods for achieving these temperatures rely on either mixing the rare and expensive isotope of helium with the more common isotope (dilution refrigerator) or on adiabatic demagnetisation of paramagnetic salt (ADR). Although both of these methods are mature, they still remain complicated enough to limit the usage only to specialized laboratories. The research done in this thesis revolves around a promising alternative to these techniques; using normal metal - insulator - superconductor (NIS) junctions. One of the defining properties of a superconductor is a gap in its electronic density of states. This gap enables it to act as an energy filter for electrons. Because of this property, when a proper bias voltage is applied over a NIS junction the normal metal part will cool down as current passes the junction. The cooling properties of NIS junctions were demonstrated almost two decades ago with cooling powers of the order of one picowatt. At present cooling powers of few hundreds of picowatts have been achieved. This thesis describes research on three areas related to NIS junctions. Firstly we use NIS junctions to cool low dimensional lattice systems, both 1D and 2D. The cooling of a 1D lattice (beam) is interesting for fundamental research. The 2D lattice cooling (membrane) is aimed at bringing NIS devices closer to more widespread use. An electronically cooled membrane would offer a platform on which applications, such as radiation detectors or superconducting electronics, could be integrated. Secondly we focus on the limitations of NIS cooling. In all cooling, one of the main problems is the dissipation of the extracted heat. As the other side of the junction (normal metal) is cooled, the other side (superconductor) is heated with many times larger power. This heat can then weaken the superconducting properties and heat up the phonon system around the junction. These effects act to counter the cooling effect and have been one of the main obstacles in scaling up the cooling power of NIS devices. We study these effects both numerically and experimentally. Thirdly, we study the cooling of silicon with superconducting tunnel junctions. In these superconductor - semiconductor structures the normal metal in a NIS structure is replaced with highly doped silicon. Specifically we study the effects of induced lattice strain to the electron-phonon coupling in silicon and hence to the cooling properties of these structures.

The free energy in the large temperature and density gradients in tokamaks can drive microinstabilities, which in turn drive turbulence. This turbulence is responsible for the transport of energy and particles over and above that predicted by neoclassical theory. Sheared toroidal rotation can suppress the turbulence and stabilise the underlying microinstabilities, thereby reducing the transport. This thesis investigates how variation of the equilibrium temperature and density profiles, over the same scales associated with the microinstabilities, affects how the ow shear stabilises the linear modes and suppresses the turbulence. A global gyrokinetic code is employed in this investigation, which retains the profile variation and simulates the full 3D domain of a tokamak plasma. How much ow shear is needed to stabilise the linear ion temperature gradient (ITG) mode is found to be dependent on its poloidal wavenumber, with longer wavelength modes needing more ow shear than the fastest growing mode. This dependence is present whether the ow shear is constant across the radius or if it has the variation typical in an experimental rotation profile. There is an asymmetry with respect to the sign of the ow shear in the effectiveness of the stabilisation, with the maximum linear growth rate occurring at finite negative shearing rates for the plasma studied here. This asymmetry arises from the profile variation, and is found to be significant in simulations of MAST L-mode plasmas, especially when the effects of kinetic trapped electrons are included in the simulations. Flow shear asymmetry is still present in nonlinear simulations, and the suppression of fully-developed turbulence depends on the sign of the shearing rate. With the experimental rotation profile, the heat ux arising from ITG turbulence is reduced by an amount comparable to the reduction in the linear growth rates. When the direction of the rotation profile is reversed, such that the sign of the ow shear is ipped while the magnitude remains the same, the turbulence is almost completely suppressed. A new diagnostic on MAST, beam emission spectroscopy (BES), is used to make a direct comparison between density fluctuations from simulation, and from experiment. Collisionless, electrostatic simulations with rotation are found to disagree significantly with experiment in the level of ITG turbulence activity and the correlation times and lengths of the turbulence. The inclusion of electron-electron and electron-ion collisions into static simulations is enough to bring the level of turbulent density uctuations down to within a factor two of the experimental levels, with the correlation lengths becoming comparable, while the correlation times remain an order of magnitude too large.

This thesis presents comparison between experimental measurements from the spherical tokamak MAST, two-fluid simulation data and theory of the Geodesic Acoustic Mode (GAM) in tight aspect ratio strongly shaped tokamak plasmas. The first identification of a strong ~10kHz mode detected in both potential and density fluctuations of the edge plasma in MAST using a reciprocating probe is given. The mode is radially localised, with outer limit ~ 2cm inside the separatrix, and is affected on application of resonant magnetic perturbations (RMP) generated by external coils. A shift in frequency with plasma rotation is found, and a suppression of the mode is observed above a certain threshold. Non-linear coupling to high wave number turbulence is evident, and an increase in power of turbulence fluctuations is seen after suppression. These observations are then interpreted in the context of known low frequency plasma modes present in the toroidal configuration. The supposition that the observed mode is a geodesic acoustic mode is considered and motivated by experimental observations and numerical simulations.

Studies of Overhauser DNP in liquids are presented in this thesis, where the polarisation is achieved in-situ using TEMPO-derived radicals at a magnetic field of 3.4 T (143 MHz/94 GHz 1H NMR/EPR frequency). The dielectric heating of lossy water solvent is unavoidable at high field, and so knowledge of temperature effects is important to properly compare enhancement results. It is shown that the temperature dependent DNP enhancement of water protons can be determined provided that the 1H NMR shift is sufficiently resolved and the nuclear relaxation T1I is sufficiently fast. Considerable sensitivity gains are made at modest temperatures, e.g. [E] ~ 40 at ~40 degrees C, and much greater enhancements are achievable at elevated temperatures, e.g. [E]~ 130 at ~ 100 degrees C. Since high radical concentrations (100 mM TEMPOL) are used, the leakage and saturation factors approach 1, enabling an experimental determination of the coupling factor from the enhancement. A value of E = 0:055+0:003 is found at 25 degrees C, which agrees well with values in the literature calculated from molecular dynamics simulations. The DNP enhancement is measured as a function of temperature for three organic compounds dissolved in water: glycine, L-proline and acrylic acid; with enhancements of -17, -16 and -11 at ~40 degrees C. To the author's knowledge, this is the first report of solute molecule enhancements for direct in-situ liquid DNP at this field. Significant enhancements are obtained, however, further analysis of the results reveals significantly weaker coupling of the electron spin to the solute molecule protons than to the solvent molecule protons. Discrepancies between experimental coupling factor ratios and those calculated from a force-free hard-sphere model suggest that the classical analytical models used to describe Overhauser DNP may require refinement. In addition to these temperature studies, simultaneous saturation of two EPR hyperfine lines is investigated and achieved, resulting in an increase in observed DNP enhancement.

Strongly localised concentrations or depressions of plasma density and magnetic field strength (\blobs") are ubiquitous in the edge region of tokamak fusion experiments. They contribute significantly to heating and transport in that region, and therefore to overall energy confinement. The existing fusion plasma literature in this area focuses primarily on blobs sufficiently large that a uid description is appropriate. However, the blob population may include some - not necessarily easily detectable - whose characteristic lengthscales are on the order of the ion gyro-scales. This implies that a description at the uid level is unlikely to capture the full dynamics. In this Thesis, therefore, we report hybrid (particle ions, uid electrons) particle-in-cell simulations of ion gyro-scale blobs, which enable us to examine the effects of finite Larmor radius on their dynamics, evolution, and their ability to heat the near-edge plasma. We find that ion gyro-scale blobs are advected with the background flow, and develop a twin-celled vortex structure. Asymmetry then arises from finite ion Larmor radius kinetics, manifesting in the size of the internal vortices, the shape of tails forming from blob ejecta, and the growth of a Kelvin-Helmholtz instability. Small scale blobs are also found to increase ion energies more than larger blobs as a result of ion pick-up at the upstream blob-background boundary, which may result in a significant increase in plasma energy caused by a blob population that is not yet directly observable. Finally, we examine the creation of ion gyro-scale blobs using hybrid simulations of kinetic interchange and Kelvin-Helmholtz instabilities, and present statistics of the sizes of blobs created by these instabilities, and power-laws for the resulting particle displacements.

As we enter the era of burning plasmas in next step devices such as ITER, the confinement of fusion born a-particles for sufficient duration that they impart their energy to the bulk fuel ions in order to maintain the thermonuclear burn is an important challenge in magnetically confined fusion. Fast ion driven plasma instabilities can cause significant redistribution and loss of the suprathermal energetic particle (EP) population, degrading performance. With dimensionless parameters such as the ratio of fast ion to thermal ion beta (Bfi/Bth ~50%) and the relative fast ion velocity to the Alfvén velocity (vfi/vA ~2) similar to those anticipated in ITER, the Mega Ampere Spherical Tokamak (MAST) provides the ideal place to study such instabilities. During periods of Neutral Beam Injection (NBI) heating, 'fishbone' instabilities are observed that coincide with a reduction to the fusion rate measured by drops in the neutron emission. Via experimental observations, fishbones are identified to be low frequency internal kink modes that burst in amplitude and chirp downwards in frequency and are synonymous with high power tokamak discharges on a wide range of devices around the world. This thesis provides a detailed analysis of what occurs during a single fishbone event. Experiments have been performed on MAST that have been interpreted using fast ion plasma physics codes. Modelling of the instability shows a resulting flux of fast ions away from the core, providing evidence at a fundamental level that they drive sufficient levels of anomalous fast ion transport to explain experimental observations. The diffusivity is shown to scale with mode amplitude, and the effect of altering other fishbone parameters within the scope of the experimental observations have been explained by identifying the extent of the fast ion population that is resonant with the mode. Resonant surfaces that sweep through phase space during the chirp are presented that coincide with populous domains of the EP distribution function; it is the gradients in this distribution function that define the drive and or damping of the instability. Via the use of synthetic diagnostics, changes to the radial profiles of neutron emissivity caused by a fishbone are shown to match those measured experimentally.

Nano-deposits of Ni and Co, supported on porous oxide materials, serve as heterogeneous catalysts within Johnson-Matthey plc. in the steam reforming and Fischer-Tropsch processes, with the size, shape and dispersity of the metal crystallites linked to the catalytic profile. Here we study the magnetic properties of nickel systems synthesised on the nano-scale, with the aim of developing an industrially viable technique by which the diameter of the nickel species can be evaluated. A series of nickel nanoparticles, synthesised via the thermal decomposition of Ni(acac)2, are studied as a model for the catalytic systems. The nanoparticles were studied via magnetometry and microscopy to identify the super-paramagnetic and nuclear volume of the particles, respectively. The magnetisation studies demonstrate that the widely used Langevin function based method of particle sizing does not reflect the total nuclear volume, and a surface correction term is introduced based on the low temperature, high applied field magnetisation. To demonstrate the applicability of the proposed analysis, the study of a series of industrially-viable precipitation catalysts are reported. The catalysts are studied via x-ray diffraction (XRD) and gas adsorption to establish comparable values of crystallite diameter. The values of crystallite diameter determined from the magnetic analysis are demonstrated to be consistent with the range of sizes determined from the XRD and gas adsorption studies, with additional sensitivity to the polydispersity of the crystallites. During the study of the precipitation catalysts, the magnetic volume was demonstrated to be reduced from the nuclear volume. This behaviour was also confirmed via small angle neutron scattering experiments, which demonstrated a magnetic scattering volume reduced from the nuclear by = 1 nm, on the order previously reported for nano-ferromagnetic materials. Through these studies we have established the methods for determining the surface correction term to magnetic granulometry studies. We have demonstrated that the corrected values are in agreement with the nuclear volumes determined via TEM, gas adsorption and XRD and that our proposed technique for the study of catalyst crystallites requires a short time scale, is insensitive to the catalytic support and is sensitive to the distribution of crystallite diameters.

In this thesis a heuristic expression for the current through a GaAs/GaAlAs heterostructure is derived. This expression is shown to give rise to agreement between experiment and theory. The expression itself is derived within the effective mass formalism, which is discussed to show that its use will not generate large errors. This conclusion is contrary to previous work which will be shown to be in error due to misunderstandings concerning effective mass theory. To justify the approach used to obtain the tunnel current expression the behaviour of a wavepacket incident upon a square potential barrier is studied. The study shows that the wavepacket traverses the potential sufficiently rapidly to allow scattering to be neglected, and that the total transmission probability can be calculated from the solution of the time independent Schrodinger equation. The current expression is reduced to a one dimensional integral by assuming parabolic conduction bands, position independent mass and a thermalised electron distribution. The resulting expression is different from the usual Tsu-Esaki formula, a difference which can be seen to arise because the Tsu-Esaki formula does not account for the different velocities on each side of the barrier. The final stage, before any comparison is made to experimental results, is to show that the numerical technique of Vigneron and Lambin is more accurate than the WKB technique. A comparison of experimental results and the results of the numerical integration of the current density expression shows that they can only be reconciled if a resistance or diode is assumed in series with the tunnel barrier. This fitting parameter is then shown to be sufficient for good fits to be obtained between experiment and theory for the first time.

No description available.