Search

An electron is usually considered to have only one type of kinetic energy, but could it have more, for both its spin and charge, or by exciting other electrons? In one dimension (1D), the physics of interacting electrons is captured well at low energies by the linear Tomonaga- Luttinger Liquid (TLL) model, with hallmark predictions, such as spin-charge separation, having already been observed. Recent theoretical work has focused on extending the theory to deal with more realistic curved dispersions. However, experimental realisations have remained elusive until recently. Here, we report on measurements of many-body modes using a momentum-resolved tunnelling spectroscopy technique in gated 1D wires connected via air-bridges. We map the 1D dispersion in a variety of devices, both in and out of equilibrium, and observe the formation of two separate Fermi seas, associated with spin and charge excitations, at energies up to five times that of the Fermi energy, which cannot be accounted for by the noninteracting model. By reducing the length of the system, we also observe the emergence of higher-order 'replica' parabolic dispersions with higher momenta or negative effective mass, which is consistent with one of the leading nonlinear Luttinger theories. Determining the bare electron mass m_0 in crystals is also another problem often hindered by many-body effects. Here, Fermi-liquid physics renormalises the band mass, making the observed values density-dependent. In 1D geometries, however, the effect of interactions is strongly amplified, and they naturally decouple from the mass. By changing the level of confinement in the wires, we are able to tune the electron density down to about 18 electrons per micron or, equivalently, an interaction parameter of rs~4. This allows us to extract a constant bare mass value of m_0 = 0.0545m_e, about 20% lighter than observed in GaAs in geometries of higher dimensionality. Finally, by progressively occupying more 1D subbands, our system also allows us to change the amount of inter-subband screening by over 50%, consequently varying the effective interaction strength, in situ, all the way from the weakly to the strongly interacting regime. This ability is the first of its kind for 1D solid-state system. Our spectroscopy technique therefore offers itself as an important and powerful tool for probing strongly correlated systems where new emerging phenomena are occurring. This knowledge is now being used to develop systems for improved energy efficiency and could prove useful in designing the next generation of sustainable materials.

Modern semiconductor physics research is often carried out at low temperatures where the carrier mobility is high and the thermal broad- ening of quantised energy levels is minimised. 4.2 K, the temperature of liquid helium, is thought of as relatively warm. Temperatures of 300 mK and below are readily accessible with sorption-pumped helium-3 and 3He/4He dilution refrigerators. At such low tempera- tures, the phonons that normally facilitate thermal equilibrium be- tween the charge carriers and the lattice are greatly suppressed. This can lead to the electron system thermally decoupling from the lattice. Conventional thermometers such as ceramic oxide or ruthenium ox- ide resistance thermometers thermally couple to the lattice, meaning that in this regime, they only effectively measure the lattice temper- ature, TL. The effect is further pronounced in low dimensional car- rier systems such as two-dimensional electron gases (2DEGs). This is problematic because the electron temperature, Te, is often of far greater importance than TL in such systems. Recent efforts have been made to develop thermometry techniques that are directly sensitive to Te. These techniques include using the thermopower of quantum point contacts and the Coulomb blockade of quantum dots. However, these techniques have issues with the ease and reliability of fabrication of the thermocouples, and the complexity of their subsequent oper- ation. This thesis presents a direct electron thermometry technique called bar-gate thermometry (BGT) that utilises the diffusion ther- mopower of the 2DEG as a thermocouple. BGT uses simple metal- lic bar-gates, fabricated directly onto the semiconductor wafer, and requires no particularly sophisticated electronic instruments. BGT proves to be a reliable and relatively simple thermometry technique. A device featuring a BGT is presented which allows a determination of the thermal relaxation length, l, of the 2DEG. This comes hand in hand with effective measurements of the inelastic scattering time, τi, and the inelastic scattering exponent, αi, of the system, which are not otherwise easily measured. Finally, an example of how BGTs can be used to measure the thermal conductivity of mesoscopic 2DEGs is given, which is just one possible use case of the technique. An overview of all the necessary and relevant theory needed to under- stand the techniques and results, as well as information on the design and fabrication of the devices used in the experiments is also included.

Quantum computers are on the verge of revolutionising modern technology by providing scientists with unparalleled computational resources. With quantum-mechanical phenomena such as the superposition principle and entanglement, these computers could solve certain computational problems that are otherwise impossible for even the most powerful classical supercomputers. One of the major challenges standing in the way of this computing revolution is the accurate control of quantum bits. Quantum systems are extremely fragile and, by their nature, cannot be measured without destroying their quantum state. I wrote a numerical program to solve the time-dependent Schrödinger equation, the differential equation that describes the evolution of wave functions. The advantage of my code over other solvers is its speed. I used graphics processing units (GPUs), a technology that has only recently matured, to accelerate high-performance computing. Hardware- acceleration allows me to solve complex time-evolution problems within days rather than years. Such an exceptional speedup has enabled me to calculate the behaviour of single electrons in semiconductor devices. Electrons are particularly interesting because they are ubiquitous in modern technology, as well as being fundamental quantum particles. Using the simulations produced by my code, I track the time evolution of an electron wave function as it propagates along quantum circuits. By animating the evolution of the wave function, I am able to visualise the wave function of electrons propagating in space and time. This is a remarkable tool for studying the behaviour of quantum particles in nanodevices. I focused my thesis on the realistic modelling of devices that are readily available in a laboratory or on designs that could be fabricated in the near future. I began by modelling single electrons as quantum bits. I provide a definition for an optimal qubit and lay out the set of operations required to manipulate the quantum information carried by the electron. In all my simulations, I aim to model experimentally realistic devices. I calculated the electrostatic potential of a real nanodevice and simulated the time-evolution of a single electron. I show that it is possible to create a single-electron beam splitter by tuning the voltages applied to various parts of the device and I calculate the range of voltages in which quantum information is preserved and manipulated accurately. These results were verified experimentally by collaborators at the Institut Néel and were published in Nature Communications 10, 4557 (2019). Using my code, I developed a framework for general measurements of electron qubits and provided a design for a semiconductor device capable of performing positive-operator valued measures (POVMs). A POVM is a powerful measurement technique in quantum mechanics that allows quantum information to be manipulated in interesting ways. The proposed setup is suggested as an implementation of entanglement distillation, which is a useful error correction tool that transforms an arbitrary entangled state into a pure Bell pair. Entanglement is one of the most fascinating aspects of quantum mechanics and it remains a challenge to generate perfectly entangled particle pairs. An experimentally viable method for distilling - or perfecting - entanglement is crucial for the design of quantum computers or quantum communication systems. Using this design, I introduced a protocol to use electrons, rather than photons, in quantum-optics-like systems. These results were published in Phys. Rev. A 96, 052305 (2017). Going beyond single-particle behaviour, I compare different methods for generating entanglement between electron-spin qubits using the power-of-SWAP operation. By using realistic experimental parameters in my simulations, I demonstrate that generating entan- glement via electron-electron collisions in a harmonic channel cannot be implemented for multidimensional systems. These findings go against what researchers thought was possible and put forward the need for new solutions to particle entanglement. I provide an alternative by demonstrating that a method based on the exchange energy is more viable than previously thought. I present a semiconductor device structure and an electrostatic potential that experi- mental groups can use in order to obtain the most efficient entangling quantum logic gates. These findings were published in Phys. Rev. A 101, 022329 (2020). The results presented in this thesis provide a comprehensive description of the control of single electrons in a surface-acoustic-wave-based quantum circuit. However, work in this field is far from over. I present various research paths for future projects. These include going beyond the time-dependent Schrödinger equation to capture more complicated dynamics, using different hardware solutions to further accelerate numerical problem solving, and studying new systems of interest to extend this project beyond semiconductor physics.In all my simulations, I aim to model experimentally realistic devices. I calculated the electrostatic potential of a real nanodevice and simulated the time-evolution of a single electron. I show that it is possible to create a single-electron beam splitter by tuning the voltages applied to various parts of the device and I calculate the range of voltages in which quantum information is preserved and manipulated accurately. These results were verified experimentally by collaborators at the Institut Néel and were published in Nature Communications 10, 4557 (2019). Using my code, I developed a framework for general measurements of electron qubits and provided a design for a semiconductor device capable of performing positive-operator valued measures (POVMs). A POVM is a powerful measurement technique in quantum mechanics that allows quantum information to be manipulated in interesting ways. The proposed setup is suggested as an implementation of entanglement distillation, which is a useful error correction tool that transforms an arbitrary entangled state into a pure Bell pair. Entanglement is one of the most fascinating aspects of quantum mechanics and it remains a challenge to generate perfectly entangled particle pairs. An experimentally viable method for distilling - or perfecting - entanglement is crucial for the design of quantum computers or quantum communication systems. Using this design, I introduced a protocol to use electrons, rather than photons, in quantum-optics-like systems. These results were published in Phys. Rev. A 96, 052305 (2017). Going beyond single-particle behaviour, I compare different methods for generating entanglement between electron-spin qubits using the power-of-SWAP operation. By using realistic experimental parameters in my simulations, I demonstrate that generating entan- glement via electron-electron collisions in a harmonic channel cannot be implemented for multidimensional systems. These findings go against what researchers thought was possible and put forward the need for new solutions to particle entanglement. I provide an alternative by demonstrating that a method based on the exchange energy is more viable than previously thought. I present a semiconductor device structure and an electrostatic potential that experi- mental groups can use in order to obtain the most efficient entangling quantum logic gates. These findings were published in Phys. Rev. A 101, 022329 (2020). The results presented in this thesis provide a comprehensive description of the control of single electrons in a surface-acoustic-wave-based quantum circuit. However, work in this field is far from over. I present various research paths for future projects. These include going beyond the time-dependent Schrödinger equation to capture more complicated dynamics, using different hardware solutions to further accelerate numerical problem solving, and studying new systems of interest to extend this project beyond semiconductor physics.

Sub-Poissonian entangled photon-pair sources based on semiconductor quantum dots are promising devices for the construction of a secure quantum network, thanks to their natural lack of multi photon-pair generation. In the previous decades, we have witnessed the achievements on a large variety of quantum dot light sources. However, they are still far from being used in an established standard fibre network. On the one hand, most of the sources are still operated in an environment that requires the installation of bulky laser setups, preventing their integration in a network. On the other hand, most of the experiments on entangled qubits generated from quantum dot sources are performed in the laboratory, resulting in a lack of work on the system supporting the remote operation of the device and high-fidelity transmission of entangled qubits over a long-distance in a large geographic scale. In this work, we demonstrate the integration of a quantum dot device in an optical network. To achieve this, both the source and the systems have been developed that favour the remote operation in a non-laboratory environment. For the source, a fully electrically operated telecom entangled on-chip pumping quantum dot device is designed and fabricated. It supports the operation at a low current of 1.6 mA and simultaneously achieves high wavelength tuneability of more than 25 nm by changing the applied bias voltage from 0 to -3.8 V. A in-laboratory entanglement fidelity of 97.3±0.2% is achieved with this on-chip pumping quantum dot light source. Several systems are developed in this work, supporting the integration of this developed source. We have proposed an innovative method for the precise alignment of the detection system to the quantum dot eigenbasis based on analysing time-resolved photon-pair correlations in a set of randomly oriented detection bases. It eliminates the restriction on the placement of the free-space polariser before the light is coupled from the quantum dot emitter into the first single-mode fibre. To support the classical traffic for sending the control signal to required components for source and system operation, we have developed a multiplexing system that allows the transmission of qubits at the telecom O-band with classical traffic at the telecom C-band over the same optical fibre without generating severe background at the quantum channel. To overcome changes of birefringence introduced by the deployed optical fibres, which causes the variation of the polarisation states during qubit transmission, we have developed a polarisation stabilisation system making use of the commercial polarisation controlling components. The system has an extremely low loss and a high duty cycle. The above development has enabled us to achieve a successful deployment and integration of the quantum dot device in the Cambridge Fibre Network. We demonstrate multiplexing of true single entangled photons tuned to 1310 nm from the on-chip pumping device with classical data traffic and achieve entanglement fidelity above 94% for over 40 hours.

With each and every passing year, weather patterns seem to become less predictable and more chaotic due to changes to worldwide climate patterns. Greenhouse gas-free energy production is a boon to the aim of preventing further rises in global temperature increases. This thesis concerns itself with one major aspect of green energy, namely photovoltaics, and some materials with great potential for application in photovoltaics, namely single-walled carbon nanotubes (SWNT), and perovskites. SWNT have been known for almost thirty years by now, but their intrinsic fondness of bundling up due to their 2pz-orbitals has to date prevented any large scale application of them. Functionalising of SWNT offers a solution to this problem, either by dispersing them with surfactants in water, or by polymer functionalisation. Here, the latter method was applied to the problem, and the efficacy of deposition from solvents more advantageous than water is shown, achieving values of transmission and sheet resistance close to the necessary values for application in solar cells. Next, an inorganic contender for solar cell applications, Cs2AgBiBr6, is characterised by magneto-optical studies. With these, a direct and indirect gap are identified at 2.85 eV and 1.31eV. Moreover, a strongly Stokes shift influenced AgBi anti-site defect colour centre at 2.25 eV is shown to be part of a self-trapped exciton at low temperatures. Thereafter, the optical properties between 1.4 K and 370 K in conjunction with room temperature x-ray diffraction measurements of MAPbI3 single crystals and thin films grown/annealed at different temperatures are used to explain the double emission structure visible in photo- luminescence and reflectance. A dynamic Rashba effect, likely due to spontaneous alignment of the methylammonium-ions, causes a direct and indirect gap, where the former dominates the surface and the latter dominates the bulk of the samples. The remaining chapter is a study of the polaronic influence on the transitions observed for MAPbBr3 and MAPbI3. In comparison between high field, high energy, and low field, low energy, an almost two-fold difference in exciton binding energy, and an increase in effective mass can be observed due to exciton/lattice interactions in comparison to the established high field literature values. Interestingly, for both perovskites studied, single crystals show more favourable values, indicating that device application improvements could be possible by making thin films more single crystal-like.

This dissertation reports on several experimental projects studying electronic transport in thin-film electronic devices. Self-assembly methods and graphene were used to realise devices contacting films of self-assembled PbS quantum dots. The devices have exhibited single-electron tunnelling with a high yield. The electrical properties of the junctions are studied individually and collectively using statistical tools to extract correlations between device geometries and electrical data. The dissertation includes discussion of the theory of relevant electronic transport including numerical simulations. Several initiated projects deriving from this work are introduced. A second device reported in this thesis is a memristive switch. Contacting thin films of Al2O3 with graphene delivered junctions which exhibit memristive behaviour with an ultrahigh on-off conductance ratio. The conduction state of the junctions is correlated with morphological changes in the devices, whereby conductive flament formation in the junction is found to lead to electrically-controllable and reversible gas encapsulation in bubbles in the structure. The device is measured electrically and topographically, and the correlation between the two aspects is studied. A discussion of memristive conduction is included with numerical simulations. A third section reports on a project studying thermoelectricity in self-assembled molecular junctions, as they show potential for improved thermoelectric efficiency for energy harvesting; this is discussed in the dissertation. Strategies to benchmark the studies are presented with relevant devices fabricated and measured. These include the development of a measurement protocol to study thermoelectricity in devices, studies of electrical coupling between various molecular structures and graphene electrodes, molecular-structure dependence of electrical and thermal conductance of junctions. Preliminary results and on-going work are discussed.

This thesis presents an exploration of how electrochemical etching (ECE) can be used to create pores in nitride semiconductors in order to engineer their properties. ECE offers a simple, flexible approach to pore formation, driven by an electric field between the material surface and a suitable electrolyte. This means that the material must be conductive in order to form pores. This makes doping density and potential the two key parameters for controlling the process. A particular focus of this work has been to understand how pores can form in a subsurface doped layer, below a non-porous non-intentionally doped capping layer. It is found that this is enabled by nanopipes formed at dislocations, which allow the electrolyte to reach the doped layer with minimal damage to the capping layer. Large birefringence of 0.14 has been produced in porous GaN structures and this has been thoroughly studied through optical measurements, finite element modelling and 3D structural data obtained using focussed ion beam milling combined with scanning electron microscopy. Characterisation of structures with periodic layers of porous GaN has also been demonstrated, using x-ray diffraction measurements. This has been used to measure the layer thickness and porosity of these samples and predicts their optical reflectivity with higher accuracy than the standard method for such characterisation. Forming pores in nitride semiconductors other than GaN has also been demonstrated. Porous AlGaN UV reflectors have been created, with a peak reflectivity of 89% at 324 nm and ECE has been applied to InGaN quantum wells with the aim of enhancing the luminescence. Measurements indicate that ECE creates both enhancement and a blueshift in the emission peak. Controlling and characterising pore formation at this scale is a huge challenge and further study of the pore morphology is required to verify the mechanism causing this.

With its remarkable electro-thermal properties such as the highest known thermal conductivity (~22 W/cm∙ K at room temperature) of any material, high hole mobility (>2000 cm2/V∙s), high critical electric field (>10 MV/cm), and large bandgap (5.47 eV), Diamond has overwhelming advantages over Silicon and wide bandgap semiconductors (WBG) for ultra-high voltage and high temperature applications (>3 kV and >450 K, respectively). However, despite its tremendous potential, fabricated devices based on this material have not yet delivered the expected high-performance. This is due to three main reasons: (i) the lack of consistent physical models and design approaches specific to diamond-based devices that could significantly accelerate their development; (ii) the absence of shallow acceptor and donor dopant species which has resulted in poor room temperature performance; (iii) the technological issues of the manufacturing process. With the principal aim of modelling the next generation of diamond devices, this Ph.D dissertation endeavours to numerically model the main electro-thermal properties of diamond devices for power electronic applications. Optimized unipolar mode diamond field effect transistors have been designed by means of finite element simulations and their performance has been assessed against the state-of-the-art diamond FETs. Particular attention is given to the static and dynamic properties of deep dopant levels and their effects in WBG semiconductor-based devices. Moreover, by means of a more global comparison technique and through accurate theoretical analysis, diamond FETs and diodes' performance have been projected and compared with that of GaN and SiC devices. This work concludes with possible implementations of diamond devices in power converters and provides a roadmap of diamond devices for power electronics. These promising results give a new impetus to the rather small, but growing diamond community and enable future research in the field with the goal of bringing diamond to the commercial world.

Lead halide perovskites have shown remarkable device performance in both solar cells and LEDs. Whilst the research efforts so far have been mainly focussed on device optimisation, little is known about the photophysical properties. For example, the nature of the bandgap is still debated and an indirect bandgap due to a Rashba splitting has been suggested. In this thesis, we study the early-time carrier relaxation and its impact on photoluminescence emission. We first study ultrafast carrier thermalization processes using 2D electronic spectroscopy and extract characteristic carrier thermalization times from below 10 fs to 85 fs. We then investigate the early-time photoluminescence emission during carrier cooling. We observe that the luminescence signal shows a rise over 2 picoseconds in CH3NH3PbI3 while carriers cool to the band edge. This shows that luminescence of hot carriers is slower than that of cold carriers, as is found in direct gap semiconductors. We conclude that electrons and holes show strong overlap in momentum space, despite the potential presence of a small band offset arising from a Rashba effect. Recombination and device performance of perovskites are thus better described within a direct bandgap model. We finally study carrier recombination in perovskites and the impact of photon recycling. We show that, for an internal photoluminescence quantum yield of 70%, we measure external yields as low as 15% in planar films, where light out-coupling is inefficient, but observe values as high as 57% in films on textured substrates that enhance out-coupling. We study the photo-excited carrier dynamics and use a rate equation to relate radiative and non-radiative recombination events to measured photoluminescence efficiencies. We conclude that the use of textured active layers has the ability to improve power conversion efficiencies for both LEDs and solar cells.

In this thesis, we investigate how the optical and electronic properties of graphene may be modified in proximity to various other materials. We present several examples of how modification in this way can help make graphene better suited for specific device applications. We develop a method of up-scaling the fabrication of FeCl3-intercalated few-layer graphene from micron-sized flakes to macroscopic films so that it may be used as a transparent electrode in flexible light-emitting devices. We also find that photo-responsive junctions can be arbitrarily written into FeCl3-intercalated few-layer graphene by means of optical lithography. These junctions produce photocurrent signals that are directly proportional to incident optical power over an extended range compared to other graphene photodetectors. Through theoretical analysis of these junctions, we conclude that the enhanced cooling of hot carriers with lattice phonons is responsible for this behaviour. Finally, we trial rubrene single crystals as the light-absorbing layer in a graphene phototransistor. We find that rubrene single crystal-graphene interfaces exhibit enhanced charge transfer efficiencies under illumination with extremely weak light signals. Through a comparative study with similar devices, we conclude that the wide variation in sensitivity amongst graphene phototransistors is largely due to extraneous factors relating to device geometry and measurement conditions.

This thesis presents the development of a selective-etch fabrication process to create sub-10 nanometre metallic gaps and the subsequent use of the gaps to study the electronics of nanocrystals and molecules. A complete picture of the success of the process required both examination by scanning electron microscopy as well as probing the current response to an applied bias at low temperature. The empty gaps were fully characterised before self-assembling 7 nm CdSe nanocrystals onto the metal with the help of linker molecules. The I-V characteristics of the empty gaps showed a reduction of the tunnelling barrier height from the expected value (~5.1 eV, the work function of Au) when the results were fitted to the Simmons tunnelling model for a metal-insulator-metal system. Results indicate that after the barrier height is surpassed, a transition from direct to field-effect (Fowler-Nordheim) tunnelling occurs. After CdSe assembly, the collected I-V characteristics of the system at 77 K showed varied results. Many devices displayed conductance peaks at low voltages comparable to the results of the shadow evaporation process for 4.2 nm nanocrystals (also documented in this thesis). Several devices revealed switching between multiples of fundamental curves, suggesting conduction through multiples of nanocrystals.

The use of Insulated Gate Bipolar Transistors (IGBT) have enabled better switching performance than the Metal Oxide Semiconductor Field effect Transistor (MOSFET) in medium to high power applications due to their lower on-state power loss and higher current densities. The power ratings of IGBTs are slowly increasing and are envisaged to replace thyristors in medium power applications such as High Voltage Direct Current (HVDC) inverter systems and traction drive controls. Devices such as the MOS Controlled Thyristor (MCT) and Emitter Switched Thyristor (EST) were developed in an effort to further simplify drive requirements of thyristors by incorporating a voltage controlled MOS gate into the thyristor structure. However, the MCT is unable to achieve controlled current saturation which is a desirable characteristic of power switching devices while the EST has only limited control. The IGBT can achieve current saturation, however, due to the transistor based structure it exhibits a larger on-state voltage in high power applications compared with thyristor based devices. MOS Gated Thyristor (MGT) devices are a promising alternative to transistor based devices as they exhibit a lower forward voltage drop and improved current densities. This current research focuses on the Clustered Insulated Gate Bipolar Transistor (CIGBT) whilst being operated under soft-switching regimes. The CIGBT is a MOS gated thyristor device that exhibits a unique self-clamping feature that protects cathode cells from high anode voltages under all operating conditions. The self-clamping feature also enables current saturation at high gate biases and provides low switching losses. Its low on-state voltage and high voltage blocking capabilities make the CIGBT suitable as a contender to the IGBT in medium to high power switching applications. For the first time, the CIGBT has been operated under soft-switching regimes and transient over-voltages at turn-on have been witnessed which have been found to be associated with a number of factors. The internal dynamics of the CIGBT have been analysed using 2D numerical simulations and it has been shown that a major influence on the peak voltage is the P well spacing within the CIGBT structure. For example, Small adjacent P well spacings within the device results in an inability for the CIGBT to switch iv on correctly. Further to this, implant concentrations of the n well region during device fabrication can also affect the turn-on transients. Despite this, the CIGBT has been experimental analysed under soft-switching conditions and found to outperform the IGBT by 12% and 27% for on-state voltage drop and total energy losses respectively. Turn off current bumps have been seen whilst switching the device in zero voltage and zero current switching mode of operation and the internal dynamics have been analysed to show the influence upon the current at turn off. Preliminary results on the Trench CIGBT (TCIGBT) under soft switching conditions has also been analysed for the first time and was found to have a reduced peak over-voltage and better switching performance than the planer CIGBT. Through optimisation of the CIGBT structure and fabrication process, it is seen that the device will become a suitable replacement to IGBT in medium power application.

This thesis describes electrical measurements performed on low dimensional p-type devices, fabricated from GaAs/AlGaAs heterostructures. The Coulomb interaction between holes is similar to that between electrons. However, the kinetic energy is suppressed, which makes interaction effects particularly important. Holes may also be used to study band structure effects which arise from spin-orbit coupling in the valence band. The effects of Coulomb interactions in low dimensional electron systems are currently being studied extensively. Experiments presented in this thesis indicate the possible importance of Coulomb exchange interactions in both one and two dimensional hole systems (1DHSs,2DHSs). Tilted magnetic field studies of 2DHSs in the quantum Hall regime indicate that Landau levels at even filling factors will not cross. For high filling factor, this is attributed to a spin-orbit mixing effect which arises from the low symmetry ofthe system. At lower filling factor, activation-energy measurements verify that the energy gaps decrease and then increase as the field is tilted. However, the energy gap versus field dependences do not exhibit the curvature that might be expected from a perturbative anticrossing. It is speculated that the origin of this effect is a phase transition driven by the exchange interaction. Balanced arguments contrasting the relative strengths of the mixing and interactions theories are provided. The second part of this thesis describes a new method for the fabrication ofballistic 1DHSs, which exhibit clear conductance quantization. The quantization changes from even to odd multiples of e2/h as a function of the magnetic field in the plane of the heterostructure, as 'spin splitting' causes the 1D subbands to cross. Measurements of the 1D subband energy spacings are used together with the magnetic fields at which the crossings occur to calculate the in-plane g factors of the 1D subbands. These are found to increase as the number of occupied 1D subbands decreases. This enhancement of the g factor is attributed to exchange interactions; possible mixing explanations are also discussed. At higher magnetic fields, the pattern of quantization features shows that the subbands have crossed many times, and that the 1DHS can be strongly magnetized.

The first part of this thesis deals with the study of mesoscopic fluctuations of the Coulomb drag resistance in double-layer GaAs/AlGaAs heterostructures, both in weak magnetic fields and strong magnetic fields. In the second part, measurements are made in a monolayer graphene structure, specifically of the quantum lifetime, and the mesoscopic resistance fluctuations at quantising magnetic fields.

This thesis reports on the electrochemical behaviour of Si in aqueous KOH solutions, and its implications on 3-D micromachining for transducer applications. First, wet chemical etches of Si are summarised, with etching reactions and suitable etch masks. Several ways for 3-D Si micromachining are reviewed, with special emphasis on electrochemical passivation techniques. Analysis of the electrochemical behaviour of Si requires the combination of the concepts of solid state physics and electrochemistry. The equilibration processes and the kinetics of the charge transfer at the semiconductor/solution interface are described on the energy band diagrams of the solid state physics. This is accomplished by defining an effective Fermi level for the solution, and by describing the movement of the ions and the surrounding water molecules on an electronic energy scale utilising a fluctuating energy level model. Arguments for chemical and electrochemical mechanisms for the dissolution of Si in aqueous KOH solutions are reviewed and compared. It is shown that an electrochemical approach is more consistent and accounts for the aniostropic nature and concentration dependence of the etching process. The analysis of the electrochemical behaviour of Si is based on a series of voltammetry (I-V) measurements, where the current of a Si electrode is plotted while its potential (with respect to the solution) is linearly increased with time. Different regions of the I-V characteristics and bias dependence of the etching mechanisms are analysed utilising the energy band diagram representation of the Si/KOH interface. I-V characteristics, obtained by ramping the electrode potential of a Si electrode in the same voltage interval with increasing rates, are used to examine the dynamics and the bias dependence of the passivation processes and the anodic oxide growth. Relying on the results of the voltammetry measurements, 3-D microstructures are successfully fabricated that use implantation/diffusion steps employed in a standard CMOS process. A novel approach to biasing the electrodes in an electrochemical cell is reported. With this technique, the solution is biased via a feedback loop which virtually grounds the reference electrode. This reduces the noise susceptibility of the apparatus and allows the use of samples incorporating multiple electrodes.

Optical and electron beam lithography has been used to fabricate high mobility GaAs:AlGaAs heterojunction FETs in which the current is controlled by Schottky barrier gates with novel geometries. The two dimensional electron gas (2DEG) at the heterojunction interface had a low temperature mobility of ~250,000 cm^{2V-1s-1 and a carrier concentration of 4.3x1011cm-2. Narrow channels of the 2DEG were defined by means of a split gate which consisted of two gold pads 15μm long, separated by ~1μm. A negative voltage applied to the gate removes carriers from beneath the gate resulting in a narrow channel in the gap. The channel width can be reduced to zero by further decreasing the gate voltage. At low temperatures (T ≤ 4.2K) the electron phase coherence length, Lφ, is greater than the width, W, and the transport is quasi one dimensional. Analysis of the low temperature magnetoconductance showed that for a channel of width ~450AA the phase coherence length varied as Lφsim 0.16μ m(T/K)^-0.35±0.06. A similar result was obtained from an analysis of the universal conductance fluctuations in channels of width ~ 1800AA. This suggests that the dominant electron scattering mechanism was due to electromagnetic fluctuations in the 2DEG for which Lφ would be expected to vary as T^-1/3. For high magnetic fields (O < B ≤ 8T) the magnetoconductance showed oscillations which were explained in terms of the magnetic depopulation of one dimensional subbands. A number of fine gate FETs were made with gate lengths of ~ 1000AA. The I-V characteristics of a strongly depleted channel were measured at 4.2K and it was found that I ∝V3/2 so that the current flow was dominated by space charge effects. For larger source drain biases I ∝ V and this was explained as being due to velocity saturation. The second voltage differential δ2V/δI2 showed structure at ~ 40meV and ~ 80meV and this was attributed to optic phonon emission by hot electrons.}

The continuing refinement of crystal growth techniques has made possible the fabrication of semiconductor superlattices where the period can be as small as one lattice constant. Prediction of many of the properties of such systems requires a detailed description of their electronic structure. In this thesis, a self-consistent pseudopotential method which includes a parametrization scheme has been used to calculate the electronic properties of (GaAs)n(AlAs)n superlattices with n ranging from 1 to 4. The parametrization scheme is used to reproduce energy gaps at the principal symmetry points for the bulk constituents and the resulting parameter set is employed in all subsequent calculations. The n=l superlattice is found to be indirect with the conduction band minimum at R (equivalent to the zincblende L point) and all the thicker systems are pseudodirect in good agreement with experimental results. The lowest conduction band state at the zone centre for all systems is found to be mainly X-derived reflecting the importance of zone translating effects here. By analysing the states near to the band edges, the observed pattern of confinement in states of the n=l superlattice shows the band offsets to have at most a small role, in contrast to the thicker systems where a definite relationship was established. Moreover, the results suggest that Dingle's "15% rule" is consistently violated and that a valence band offset of about 30-40% is obtained which changes little with layer thickness. Attempts to study the effects of hydrostatic pressure on the n=3 superlattice were in part successful and predicted quite complex behaviour for the electronic states. Much of the discrepancy between the results obtained and the experimental data was attributed to the inadequacies of the empty-core pseudopotential to model the ions.

The main objective of this project was to measure the density of states in magnetron sputtered hydrogenated amorphous silicon and compare it with that obtained from material prepared by the conventional R. F. sputtering method and to try to decide on the critical factors which may influence it. It was also intended to compare the results to the well known density of states of glow discharge amorphous silicon. To this end, two types of experiments were used: the admittance-frequency measurements on Schottky barriers and space charge limited current measurements. In the case of the Schottky barrier analysis, the results were analysed using a trap release model and the temperature-frequency dependence of the A. C. response of Schottky barriers was investigated. The results obtained from several samples give densities of states around 10e16 eV-1 cm-3 , significantly lower than the values quoted for non-magnetron material reported in literature. For SCLC experiment, the results obtained indicate densities of states near the fermi level of 10e15-10e16 eV -1cm-3 Experimental data showed that the surface layers at front and back contacts were different, clearly indicating surface properties of the film. Therefore, it becomes clear that a complete description of the material requires the surface states to be included in the analysis and a computer program was developed for this. It was concluded that the true bulk density of states of the material could be substantially reduced if surface states were considered. The temperature dependence of SCLC characteristics was also investigated and it was found that tail states could be detected by the SCLC technique. The values of the density of states obtained suggest that, as far as bulk properties of of magnetron sputtered a-Si:H are concerned, these are not distinguishable from those of glow discharge material. However, significant evidence was found for the presence of surface layers and their related effects.

This thesis describes an experimental investigation of the photoluminescence emissions from firstly, bulk In0.53 Ga0.47 As and secondly, InGaAs\InP and AlAs\GaAs quantum well structures, as a function of hydrostatic pressure. Two high pressure systems have been developed and successfully used in the course of this work, an 8kbar piston and cylinder system and a miniature diamond anvil cell. From the high pressure measurements on the bulk InGaAs, both at room anti liquid nitrogen temperatures, it is shown that the pressure dependence of the direct band-edge luminescence is non-linear and independent of temperature. Using an empirical equation of state and making some assumptions as to the value of the bulk modulus, it is found that the band-edge luminescence is linearly dependent upon the lattice constant and may be described by a band-edge deformation potential of -8.25 +/- 0.1eV. Theory compares favourably with this value. From low temperature measurements of both quantum well systems it has been possible to deduce a quantitative description of how the conduction and valence band-edge discontinuities vary as a function of the applied pressure. It is shown that the band-offset ratio changes with pressure. This work constitutes the first observation of this phenomenon which had previously been thought either too small to be of consequence, or simply ignored. It is found that the conduction-band discontinuity in InGaAs\InP quantum wells decreases at -2.3 +/- 0.6meV/kbar while its valence-band discontinuity remains constant. The valence-band discontinuity in the AlAs\GaAs superlattice is directly measured to increase at +1.1 +/- 0. ImeV/kbar. An analysis of reported data for A1 Ga As/GaAs quantum wells shows that the pressure coefficient of the X l-x valence-band discontinuity is linearly dependent upon the alloy composition x. Theories of band-offset ratios and in particular the "model-solid" theory of Van de Walle and Martin (Van de Walle 1987), agree exceedingly well with these experimental findings. It is suggested that such agreement lends weight to the assumption that heterojunction band-edge discontinuities are intrinsic to the bulk properties of the host materials.