Vapour phase rubies were examined by several techniques. X-rays showed that the c-axis varied from place to place in each boule by up to 3º. Superimposed on this was a distribution of mosaic structure with the c-axis varying from grain to grain by up to 3º. Annealing did not improve these imperfections. Boules grown in the 90º orientation showed less mosaic than those grown in other orientations. Polygonisation, observed by Scheuplein and Gibbs (1960), provided an explanation. X-ray examination of corundum gave similar results. The random dislocation density revealed by etching was about l0(^5) dislocations/cm(^2) in corundum and ruby, therefore, adding chromium did not produce dislocations. The imperfections did not correlate with the boules, crystallography. The paramagnetic resonance linewidth increases with the imperfection in the sample and this increase is angular dependent. The basic linewidth depends on the chromium concentration. Sample 337c showed little angular broadening in the + (^1)(-2) to + (^3_)(_2) and the – (^3)(_2) to – (^1)(^2) transitions. In G2a the latter transition showed a maximum broadening of about 2(^1)(_2) times near the polar angle 45º. X-rays showed that each sample had similar c-axis misorientations but G2a contained much more mosaic than 337c. The misorientation predicted by the broadening G2a is 0º 25. The – (^1)(_2) to + (^3)(_2) transition in 337c was about 22 oersteds wide and did not vary much with angle. In G2a this transition broadened to 33 oersteds at polar angles 45º and 30º. The chromium concentration predicted from the basic linewidth at 55º in both sample is 0.43 weight % of chromium. The concentrations given by chemical analysis are 0.052 weight % for 337c and between 0.02 and 0.032 weight % for G2a.

Measurements have been made using a pulse saturation technique at 35 Gc/s of the relaxation times for transitions within the ions of the ‘irongroup' of transition metal elements placed in various lattices. A microwave spectrometer has been constructed in which the need for a separate local oscillator source has been overcome. A detailed study of the relaxation times and the texture in ruby single crystals has shown that the relaxation depends on the c-axis misorientation within the crystal, but is independent of the mosaic structure. A study of the temperature dependence of the relaxation times from 1.5 K to 120 K has shown that the chromium ions occupy two types of site within the lattice, one perfect the other distorted. Similar results have been obtained for samples of ruby grown by Verneuil and Gzochralski techniques. The effect of impurities in ruby has been considered and a 'figure of merit' has been empirically devised to describe the effects of cross-relaxation between chromic and ferric ions. X-irradiation of rubies has given indirect evidence for the existence of the Cr(^2+) ion. A brief examination of the relaxation times in chromium doped spinel and rutile has led to an explanation of the frequency dependence of the relaxation times for transitions across the lower Kramer's doublet in terms of an Orbach process. An examination of three alums has supported the suggestion of Kochelaev that the relaxation process in these materials is governed by the vibrations of the water complex surrounding the paramagnetic ion.

Measurements of thermally stimulated currents, photo conductivity and photoluminescence have been used to study the imperfection centres in photosensitive CdSe crystals. These crystals were grown from the elements using a flow technique. Seven closely spaced, discrete sets of traps, with concentrations between 10(^12) and 10(^14) cm(^-3), lie between 0.15 and 0.36 ev below the conduction band. They empty in the temperature range 90 to 210 K. Two other groups of traps are found after illumination at various temperatures with intense white light.(1) Illumination between 250 and 275 K creates four new sets of traps with associated T.S.C. peaks at 215, 230, 250 and 270 K. Also in this temperature range holes are thermally freed from sensitising centres which lie 0.6 ev above the valence band. The consequent increased recombination substantially modifies the T.S.C. peak shapes and no sensible values of the trapping parameters can be obtained from T.S.C. curve analysis. Illumination at higher temperatures destroys these four centres.(2) Three, further T.S.C. peaks appear at 295, 335 and 365 K as a result of illumination above 250 K. They increase in height with increasing temperature of illumination. Such behaviour can be attributed to either the photochemical creation of traps or the existence of traps surrounded by repulsive potential barriers. The traps have depths of 0.43, 0.52 and 0.63 ev. They have capture cross sections of about l0(^-20) cm(_2) and densities of up to 10(^19) cm(_2).Illumination at increasing temperatures in the range 90 to 400 K results in a progressive reduction in the free electron lifetime. This is due to the photochemical creation of class 1 fast recombination centres. Simultaneously the intensity of a 1.15 p luminescence emission band increases and that of a 0,95 p band decreases. There is no direct evidence to suggest that these effects are related.

At elevated temperatures alumina insulators are a source of small electrical pulses of approximately 10(^-14) coulombs when subjected to an electrical stress. This phenomenon, termed pulse breakdown, has been studied on single crystal and polycrystalline alumina over the temperature range 0ºC to 900ºC. The study has been made over a range of ambient gas pressures and at voltages up to 3 kV. The d.c. conductivity of alumina has also been investigated over this range of conditions. The study has shown that pulse breakdown is dependent on the ambient gas, the electrodes and the alumina surface. The results have been interpreted by means of a surface discharge model, in which discharges occur at both electrodes. The surface discharge at the cathode has been studied in detail and is shown to be produced by a corona discharge which is triggered by the desorption of impurities from the alumina surface. The discharge did not occur once single crystal alumina had been heated to 900ºC. The possible explanations for this effect are considered. The d.c. conductivity of single crystal alumina has been shown to be affected by chemical treatment of the alumina surface. This is due to the surface conduction of the alumina being affected by chemisorption. The activation energy for the surface conduction of alumina has been shown to be 1.7 eV. The very much higher activation energy of 3.9 eV attributed to the bulk conduction shows that alumina is not an ionic conductor below 900ºC.

This thesis details three main studies. The first is an investigation of the effect of coating Fe nanoparticles in a gas to isolate the magnetic moments. An isolation or enhancement of the already increased magnetic moment of a Fe nanoparticle would have the potential for exploitation in high-moment materials. The two other investigations are of the behaviour of Fe nanoparticles in the rare earth matrices Ho and Dy. Transition metals and rare earth metals normally couple antiferromagnetically at their interface, however the intention of this work was to determine if this also happens when the Fe is incorporated as pre-formed clusters. The motivation for this is that if the interactions between the rare earth and the transition metal is switched to be ferromagnetic then the Curie temperature of the rare earth could be increased without a large decrease in its saturation magnetisation. Fe nanoparticles consisting of -200 atoms and -2nm in diameter were manufactured using a gas aggregation source then coated with H2(g). VSM, XMCD and TEM measurements were taken of these samples and the magnetic moment per atom of Fe was found to drop significantly compared to that of isolated clusters in Ag matrices. A comparative study using N2(g) was conducted yielding similar results. This is attributed to the gas permeating the whole cluster rather than forming a shell. Addition of atomic Fe to a rare earth matrix decreases the total magnetisation due to antiferromagnetic coupling. Fe nanoparticles deposited into rare earth matrices heavily quench the rare earth moment. Samples of 2-35% Fe by volume contain Fe nanoparticles large enough to disrupt the rare earth spin wave. The Fe nanoparticles couple ferrimagnetically to the rare earth producing the low overall magnetic moment. Several magnetic phase transitions were observed in all Fe/rare earth alloys. Structural measurements using EXAFS indicate that the Fe clusters may have changed to an expanded lattice within the Dy matrix.

The magnetisation dynamics of a single square nanomagnet, the interaction between a pair of nanodiscs, a partially built writer structure and a range of magnetic tunnel junction sensor heads were studied using Time Resolved Scanning Kerr Microscopy (TRSKM) and four probe contact DC electrical transport measurements. Large amplitude magnetisation dynamics of a single square nanomagnet have been studied by TRSKM. Experimental spectra revealed that only a single mode was excited for all bias field values. Micromagnetic simulations demonstrate that at larger pulsed field amplitudes the center mode dominates the dynamic response while the edge mode is almost completely suppressed. The magnetisation dynamics occurring in a system comprised of two laterally separated magnetic nano-discs were also investigated. The polar Magneto-Optical Kerr Effect was used to measure the dynamic response of each disc independently so as to demonstrate that dynamic dipolar interactions between non-uniform spin wave modes in the different discs may be identified from the difference in their phase of oscillation. Results show a stronger dynamic dipolar interaction than expected from micromagnetic simulations highlighting both the need for characterisation and control of magnetic properties at the deep nanoscale and the potential use of dynamic interactions for the realization of useful magnetic nanotechnologies. TRSKM measurements were made simultaneously of the three Cartesian components of the magnetisation vector, by means of a quadrant photodiode polarisation bridge detector, on partially built hard disk writer structures. The rise time, relaxation time, and amplitude of each component has been related to the magnetic ground state, the initial torque, and flux propagation through the yoke and pole piece. Dynamic images reveal “flux-beaming” in which the magnetisation component parallel to the symmetry axis of the yoke is largest along that axis. A comparison of the magnetisation dynamics excited with different pulsed excitation amplitudes was also made. The results shows that more effective flux beaming is observed for higher pulse amplitudes. Lastly the microwave emission of Tunnel Magnetoresistance (TMR) nanopillars has been measured using a four probe contact DC electrical transport measurement technique as a magnetic field is applied in the plane of the film at different angles (ϕ_H ) with respect to the easy axis. Experimental spectra revealed that a more complicated spectrum containing several modes is observed as ϕ_H is increased. The modes were identified as edge and higher order modes from the statistical distribution of modes from different devices and micromagnetic simulations. The in-plane and out-of-plane components of the Spin Transfer Torque (STT) were estimated by analytical fitting of experimental data for the lowest frequency edge mode for the value of ϕ_H where the amplitude of the said mode was a maximum and its frequency a minimum. The estimated values are larger than expected perhaps due to the macrospin approximation made in deriving the analytical model. The results presented in this thesis can contribute to the understanding of magnetisation dynamics in industrially relevant data storage devices as well as the realization of a dipolar field coupling mechanism for arrays of nanooscillators.

In this thesis the results of magneto-optical experiments will be presented. The experiments were performed on micro-arrays of square nanomagnets in order to characterise the static and time-dependent behaviour of the nanomagnets. The static behaviour was investigated in vector-resolved scanning Kerr microscopy experiments, while the time-dependent behaviour was investigated in time-resolved scanning Kerr microscopy experiments. In the latter so-called pump-probe experiments, magnetisation dynamics were induced by exciting the sample magnetisation with a pulsed magnetic field (pump). The magnetisation dynamics were then detected using the magnetooptical polar Kerr effect (probe). The longitudinal Kerr effect was utilised in the vectorresolved scanning Kerr microscope in order to measure the in-plane components of the static magnetisation. The experimental set-up and methodology of the vector- and timeresolved scanning Kerr microscopy experiments will be discussed in detail, in particular, the detection technique that allows three components of the vector magnetisation to be measured simultaneously. Since the spatial resolution of the magneto-optical probe was insufficient to resolve the spatial character of the magnetisation dynamics within individual nanomagnets, micromagnetic simulations were used to gain insight into the character of the excited modes. Extensive testing of different micromagnetic models was carried out in order to investigate the effect of the different models on the simulated dynamics. The results of measurements carried out on the arrays of square nanomagnets revealed that the static and time-dependent behaviour of the magnetisation became more complicated as the size of the nanomagnets was reduced. In particular, similar hysteresis loops were acquired when the elements were magnetised along the uniaxial anisotropy easy and hard axes, while fast Fourier transform spectra of time-resolved signals revealed that the character of the magnetisation dynamics changed significantly as the element size and/or applied magnetic field were reduced. Interpretation of the experimental results using micromagnetic simulations revealed that the elements had a non-uniform single domain ground state magnetisation. When the field was applied along either edge of the square elements and reversed, the magnetisation was found to switch via a series of metastable non-uniform single domain states. Furthermore, the increasing non-uniformity of the single domain ground state as the element size and/or applied field were reduced lead to significant changes in the mode character excited within the elements. Comparison of 2 experimental spectra with simulated spectra and Fourier images of the dynamic magnetisation revealed that as the element size and/or applied field were reduced, the mode character changed from one that occupied the majority of the volume of the element, to several modes that were localised near to the edges of the element that were perpendicular to the applied field. Furthermore, deviation of the direction of the wavevector of the dynamic magnetisation from the direction of the static magnetisation was found to lead to a dynamic configurational anisotropy within nanomagnets. Following the presentation of the experimental results, the recent developments for future experimental work are presented with the aim to study precessional switching in an isolated nanomagnet. The results obtained in the experiments presented in this thesis are expected to lead to a better understanding of the non-uniform magnetisation dynamics in square nanomagnets, which have application in future magnetic data storage technologies.

In this thesis the properties of graphene are studied via the various aspects of the quantum transport: doping of the graphene surface with organic molecules, flicker noise and transport in the quantum Hall regime. First, it was shown that certain molecules (toluene, aniline and water), which possess such common properties as non zero dipole moment and ability to undergo the electrochemical reaction, have a peculiar doping effect on graphene. The effect of toluene doping was studied in detail and is explained by the electrochemical reaction, which takes place in the vicinity of the graphene and results in a gate voltage dependent doping. Second, the flicker noise in graphene and its relation to the scattering mechanisms were studied. The flicker noise as a function of the carrier concentration was demonstrated to be sensitive to the scattering potential determining the resistance of the graphene. Therefore, as it was suggested, the flicker noise can be used as a tool for determining the dominant scattering mechanism in graphene, although it was found that the resistance and noise can originate from different scattering potentials. Also, the flicker noise spectrum was shown to decompose into individual lorentzians at low temperatures (below ∼ 25 K), where the fluctuations of the resistance is supposedly coming from the random jumps of electrons between the conductive channel in the graphene flake and the nearby impurity states. Third, the transport properties of the bilayer/trilayer graphene structure were studied at different temperatures and different magnetic fields including the quantum Hall regime. Bilayer and trilayer parts of the sample revealed the signatures of the quantum Hall effect predicted theoretically. The transport through the interface between bilayer and trilayer parts was also investigated. Signatures of the interface resistance were seen, although the observed behaviour is not explained. Under high magnetic fields the properties of the interface longitudinal resistance were described qualitatively by the classic transport equations.

In the first two chapters we introduce the concept of constant energy surfaces in semiconductors and show how the conduction electrons behave in electric and magnetic fields. We summarize the low field galvanomagnetic theory but detail the work of Toyozawa in his explanation of negative magnetoresistance. The third chapter deals with the experimental details involved in the selection of the PbS samples and the electrical measuring apparatus. The fourth chapter presents the experimental results obtained. From these results we conclude that the negative magnetoresistance in n-type PbS observed at temperatures below 30K is due to an interaction between the conduction electrons and a number of localized magnetic moments isolated at impurity sites. The origin of this localized magnetic moment was discussed and also a quantitative explanation of the concentration dependence of the negative magnetoresistance was given. The true saturating negative magnetoresistance component was separated from the observed and the concentration dependence of this saturating com- ponent was obtained and found to vary as n-0.75. The literature pertaining to negative magnetoresistance in semiconductors was reviewed and other effects giving rise to negative magnetoresistance were discussed. The second section of this thesis deals with quantum effects in semiconductors. The fifth chapter deals with the theory of these effects and in particular details the theory of the Shubnikov-de Haas Effect as developed by Argyres. Expressions are obtained for the number of conduction electrons per valley as a function of the period of the Shubnikov-de Haas magnetoresistance oscillations. With a knowledge of the total carrier concentration obtained from Hall Effect measurements we show how to arrive at the appropriate multi-valley model. The remainder of the chapter describes the cryostat, crystal holders, superconducting magnet and the electrical measuring apparatus. In the sixth chapter we describe a check on the working of the apparatus using a bismuth sample. The results for thirteen samples of n-type PbS are presented and from one sample with a large amplitude of oscillation a value of 4.5K for the non- thermal broadening temperature T' was obtained. The theoretical calculations and other experimental results of the band structure of PbS were reviewed. The experimental results appeared to conflict in that measurements on low concentration samples yielded either a (100) model or a nearly spherical surface at k = 0, whereas the results obtained from high concentration crystals indicated that a (111) zone boundary model was appropriate. From our results we concluded that the number of electrons per valley we obtained from the oscillatory period was in excess of that required to fill eight ellipsoids at (111) zone boundaries. This indicated that there was more than one section of the Fermi surface although only one oscillatory period had been observed. On the basis of these results we proposed the following two interpretations:- (a) The observed oscillatory period is due to carriers in (ill) valleys at the zone boundary and there exists elsewhere a second band to hold the excess electrons and whose cross section was not observed. (b) The observed single period is due to a spherical surface at k = 0 whose cross seotion is equivalent to an nv of between xi n and in addition there are several nearly spherical ellipsoids whose cross seotion were again not observed.

A systematic investigation of three transition metal oxide (Mo, W, V) alkali phosphate-based glasses was undertaken with a view to studying the ability of these to sustain mixed ionic and electronic conductivity. The glasses were characterised by a.c. analysis and uv/vis spectroscopy. In the molybdate and tungstate glass systems large and contrary trends in ionic and electronic conductivity were highlighted. The changes were identified as being predominantly related to the alkali oxide content, for glasses containing about 50 mol% transition metal oxide. As alkali oxide was added to binary phosphate glasses a deep minimum in the isothermal conductivity plots, reminiscent of the mixed alkali effect observed in mixed ionically conducting glasses, was noted. This effect was most dramatic in the tungstate glasses, but slightly less so for the molybdate glasses. In concert with these conductivity variations the nature of the charge carrier was simultaneously determined from the shapes of impedance and spectroscopic modulus plots. This highlighted that on the low alkali side of the minimum an electronic conduction (polaronic hopping) mechanism predominated, whereas on the high alkali side of the minimum an ionic conductivity mechanism prevailed. All glasses were characterised by uv/vis spectroscopy, using either transmission or diffuse reflection techniques. Most spectra, particularly those of high transition metal content, exhibit classical "bronze-type" spectra. Spectroscopically it was found that, concurrent with the variations in electrical conductivity, dramatic changes in the relative intensities of absorption peaks in the uv/vis region of the spectrum were occurring.