The aim of the work in this thesis is to push the technology of solution-processed semiconductor lasers beyond the state-of-the-art and bring it closer to real-world implementation. An emphasis is put on the demonstration of mechanically-flexible lasers having low thresholds, high photostability and potential for cost-effectiveness and compact integration. Different gain materials, designs and pump sources are used to improve the performance and capabilities of these lasers. Distributed feedback resonators are chosen due to their planar fabrication and their potential for lower threshold than other cavities in the case of solution-processed lasers. Two types of gain materials are used: organic semiconductors and colloidal quantum dots. Encapsulation schemes compatible with the mechanical flexibility of the final devices, e.g. using transparent polymers or flexible glass membranes, are proposed and studied in order to extend the operational lifetime of the devices. One highlight of this work is the development of, to our knowledge, the first diode-pumped, mechanically flexible organic lasers encapsulated with thin-glass for high photostability. Other important outcomes include mechanical wavelength tuning of lasers, record performance for colloidal quantum dot lasers optically-pumped in the nanosecond regime and the demonstration of a red/green/blue laser. The capability for sensing applications of some reported formats of lasers are also shown.

The free-electron laser (FEL) is a powerful source of tuneable coherent radiation that currently demands kilometre-scale beam lines for high-energy vacuum ultraviolet (VUV) and X-ray output. Driving an FEL with a laser wakefield accelerator (LWFA) electron beam would radically reduce the size of such systems as well as delivering ultra-short duration radiation pulses. In this thesis, the production and optimal transport of high-quality electron beams in an LWFA and the feasibility of using such beams to drive a VUV FEL has been investigated. The ALPHA-X LWFA uses a 25 TW femtosecond laser pulse focused into a 2 mm gas jet to accelerate electrons. Using simulation codes [General Particle Tracer (GPT) and TRANSPORT], the initial ALPHA-X transport system has been analysed and an improved system using additional permanent magnet quadrupoles has been designed and installed. GPT has also been applied in analysis of the beam transport through a high resolution magnetic dipole electron spectrometer. It is shown that measurements of beam energy spreads of less than 1% imply a normalised transverse emittance of less than 1µ mm mrad, showing that these beams are suitable for driving an FEL. Efficient electron beam transport through an undulator (100 periods, period = 15 mm) is demonstrated implying an estimated source diameter of 300 µ close to the centre of the undulator (in agreement with simulation). Undulator radiation have been generated using electron beams of energy 83-131 MeV. Output radiation spans the range 180-500 nm and the scaling of photon yield with electron charge provides tentative evidence of coherent emission - the radiation flux is up to ~3 times greater than the expected spontaneous emission flux. The maximum number of photons peaks at 8 x 10p⁶ and, assuming a pulse duration of 100 fs, the maximum peak brilliance is 10¹⁸ photons/second/mm²/mrad²/0.1% bandwidth.

Free Electron Lasers can generate high power transversely coherent tunable radiation. X-ray radiation can be generated in the Self Amplified Spontaneous Emission regime, however this radiation has poor temporal coherence. The Echo-Enabled Harmonic Generation method can improve the radiation's temporal coherence in x-ray. In this thesis analysis of the Echo-Enabled Harmonic Generation technique reveals that electron pulse has a modal density profile. This density profile when matched to amplification profile of an undulatorchicane lattice generates a train of coherent radiation spikes. The interaction of multiple electron pulses is investigated in this thesis. Propagating multiple electron pulses through an undulator produce a train of radiation spikes. The temporal separation of the radiation spikes can be manipulated using magnetic chicanes. Two new techniques are then proposed to improve FEL performance when the electron pulse has a large energy spread, such as those produced in plasma accelerators. These techniques use seeded-undulators and chicanes to manipulate the electron phase space prior to injection through an undulator-chicane lattice.

Physically realising quantum computation is our long term goal. Currently characterising and verifying quantum states is a hard problem. Verification is necessary in order to understand and control quantum systems. There are also issues in the physical realisation of these systems, the apparatus used does not match the quantum scale and so produces errors in the measurements. There are many different candidates for physically realising quantum computation, we consider measurement based quantum computation using cluster state systems. We explore the different ways of verifying these systems in the presence of various experimental imperfections and non-idealities. We present a scheme to reduce the cross-talk found when verifying the state through stabilizer operator measurements. The cross-talk comes from physical constraints on the measurement apparatus. Our scheme reduces the cross-talk to almost 50% of the original value. We consider square, triangular and hexagonal connectivity lattices. We also use the Clauser-Horne-Shimony-Holt (CHSH) inequality as a way to verify atoms trapped in optical lattices through its entanglement. Imperfections arise in this system through finite entropy in the creation process that leads to vacant lattice sites. By optimising the conventional measurement settings we improve the tolerance of the system to incomplete measurement and vacancies. We find violations of the CHSH inequality for very large vacancy rates. We analyse further errors in the detectors and calculate the tolerance of the system in these cases. We study the effects of superselection rules and their connection to the single particle entanglement question. We review a system presented by Paterek et al. and verify it again using the CHSH inequality. We introduce errors into the measurement process. By optimising the measurement settings we increase the tolerance of the system for four different error models. We also explore how non-ideal state preparation affects the detectable violations.

This thesis reports on experimental and numerical investigations of ion acceleration and the underlying mechanisms of energy transfer in the interaction of intense laser pulses with ultra-thin foils undergoing relativistic induced transparency. The optimisation and optical control of the ion beam properties including the beam flux, maximum energy and energy spread is important for the development of applications of laser-driven ion beams. Multiple laser-ion acceleration mechanisms, driven by sheath fields, radiation pressure and transparency enhancement occur in intense laser pulse interactions with an ultra-thin foil. This is experimentally and numerically demonstrated in the work presented in this thesis. Results from an experimental investigation of ion acceleration from ultra-thin (nanometer-thick) foils using the Vulcan petawatt laser facility are presented. Spatially separating the multiple beam components arising from the differing acceleration mechanisms enables the underlyi ng physics of the individual mechanisms to be investigated. In the case of foils undergoing relativistic induced transparency, it is shown that an extended channel and resulting jet is formed in the expanding plasma at the rear of the target, resulting in higher laser energy absorption into electrons and enhanced ion acceleration in a localised region. This results from volumetric heating of electrons by the laser pulse propagating within the channel. The measured maximum energy of the protons in the enhanced region of the jet is found to be highly sensitive to the laser pulse contrast and rising edge intensity profile of the laser. It is shown, using a controlled pre-expansion of the target, that an increase in the maximum proton energy by a factor two is achievable. Numerical investigations of the interaction, using particle-in-cell (PIC) simulations, show that an idealised sharp rising edge Gaussian laser intensity profile produces the highest proton energy, though this condition could not be achieved experimentally. The simulations show that controlled pre-expansion of the target, by variation of the rising edge intensity profile, enables better conditions for channel formation and energy coupling to electrons and thus protons. A detailed numerical (PIC) investigation of the mechanisms of laser energy transfer to electrons and ions in thin foils undergoing relativistically induced transparency is also presented. The role of streaming instabilities in the transfer of energy between particle species is investigated. It is found that in addition to the relativistic Buneman instability, which arises from streaming of the volumetrically heated relativistic electrons with the background ions during transparency, ionion streaming in the expanding plasma also plays a role in enhancing the final ion energy. Enhancement of proton maximum energies via ion-ion streaming from shock-accelerated aluminium ions is observed in 1D PIC simulations and the energy exchange is demonstrated to be sensitive to the plasma density. Energy transfer between co-directional ion species is also observed in higher dimension 2D simulations. The simulations show that the greatest enhancement in proton energy is due to streaming of electrons in the region of the plasma jet formed in the expanding plasma.

This thesis presents a systematic study of the fabrication, efficiency and applications of micro-pixelated III-nitride light emitting diodes (micro-LEDs). Efficiency droop studies of micro-LEDs and the development of new-types of micro-LEDs have been investigated. The size dependence and temperature dependence of micro-LED efficiency droop have been analysed, providing a deep understanding of efficiency droop issue for general lighting and also demonstrating the advantages of micro-LEDs to alleviate such efficiency droop. Micro-LEDs on flexible substrates have been fabricated, which combine the flexibility of soft substrates and the high efficiency of inorganic LEDs for potential applications in flexible displays, biomedicine, etc. In addition, micro-LEDs on Si substrates were also fabricated to reduce the micro-LED device fabrication cost. The size-dependent efficiency droop study demonstrates that the smaller micro-LEDs have more uniform current spreading, which causes their higher efficiency and higher thermal saturation current density. In addition, the temperature-dependent efficiency droop study shows that both the radiative and Auger recombination coefficients decrease with increasing temperature, and the temperature dependence of the radiative and Auger recombination coefficients is weaker at a higher current density. So micro-LEDs possess stabler temperature dependent efficiency when operating at a high current density. Flexible micro-LEDs have been fabricated using metal bonding and laser lift off techniques with only one transfer step. Also, the micro-LED arrays on Si substrates were developed for the first time, which have lower fabrication cost compared with micro-LEDs on sapphire substrates. The applications in micro-display and visible light communication have been demonstrated for both the devices above. This work suggests that employing micro-LED techniques is a way to improve LED efficiency for general lighting. With development of flexible micro-LEDs and reduced fabrication cost of micro-LEDs on Si substrates, the future applications of micro-LEDs are expected to be greatly expanded.

Optical parametric oscillators have been known and used for a long time as efficient sources of non-classical states of light both below threshold of oscillation, where they generate squeezed vacuum states and bi-partite entangled states, and above threshold of oscillation, where they generate intensity correlated twin beams. The singly-resonant cavity, where only one of the three field involved in the parametric amplification process is resonated (signal), is in principle a simpler configuration to realize experimentally but, to the best of our knowledge, theoretical investigations of non-classical features of the light from a singly-resonant OPO (SROPO) are missing. One of the reasons is that SROPOs operate with strongly non-degenerate frequencies while much of the literature on squeezing focuses on the degenerate or close to degeneracy cases. Recent interest in non-classical correlations of the strongly non-degenerate regime of parametric down-conversion makes the study of entanglement in SROPO important for the optimization of coherent sources with fluctuations below the shot-noise level. There are clear technical advantages for SROPO configurations: only resonance of the signal field has to be maintained, continuous temperature tuning and suppression of mode-hopping. As a matter of fact even if the doubly resonant configuration, where both the signal and the idler fields are resonated, has a much lower threshold pump power, the tuning behavior is complicated and is massively affected by changes of the crystal temperature or pump wavelength, causing the signal and idler wavelengths undergoing jumps, and the tuning is generally non-monotonous. This is because the operation wavelengths are determined primarily by the requirement for simultaneous resonance for signal and idler, and not only by a phase-matching condition as in the case of singly resonant configuration. It is in this spirit that in Chapter 4 we apply the input-output theory of optical cavities to formulate a quantum treatment of a continuous wave singly-resonant optical parametric oscillator. This case is mainly relevant to largely non-degenerate signal and idler modes. We show that both intensity and quadrature squeezing are present and that the maximum noise reduction below the standard quantum limit is the same at the signal and idler frequencies in a way similar to the doubly resonant case. As the threshold of oscillation is approached, however, the intensity-difference and quadrature spectra display a progressive line-narrowing which is absent in the balanced doubly-resonant case. By using the separability criterion for continuous variables, the signal-idler state is found to be entangled over wide ranges of the parameters. We show that attainable levels of squeezing and entanglement make singly-resonant configurations ideal candidates for two-colour quantum information processes because of their ease of tuning in experimental realizations. Another very interesting feature of SROPOs which, this time, has no counterpart in the doubly-resonant regime is described in Chapter 5 where model equations for the evolution of signal and idler pulses in a synchronously pumped optical parametric oscillator are derived and numerically integrated. A novel regime of giant sub-threshold pulses driven by quantum fluctuations is described through the analysis of stability eigenvalues, growth factors and pseudospectra. Subthreshold pulses driven by quantum fluctuations are found at various mirror reflectivities in the non degenerate regime where signal and idler have different group velocities. Giant sub-threshold pulses open the possibility of observing macroscopic continuous variable entanglement with nonclassical features. This important feature is peculiar to the singly-resonant configuration and has no counterpart in the doubly-resonant regime. Very interesting classical features of SROPOs light are investigated in Chapter 6 where we show that spatio-temporal dynamics of singly resonant optical parametric oscillators with external seeding displays hexagonal, roll and honeycomb patterns, optical turbulence, rogue waves and cavity solitons. We derive appropriate mean-field equations with a sinc² non-linearity and demonstrate that off-resonance seeding is necessary and responsible for the formation of complex spatial structures via self-organization. We compare this model with those derived close to the threshold of signal generation and find that back-conversion of signal and idler photons is responsible for multiple regions of spatio-temporal self-organization when increasing the power of the pump field.

A three dimensional parameterised model of an X-band 2nd harmonic gyro-travelling wave amplifier (gyro-TWA) with a helically corrugated interaction region has been created and optimised in the Particle-in-Cell code MAGIC-3D, to achieve an output power and saturated efficiency of ~1.0MW and ~27% respectively at 9.4GHz. This numerical model has been benchmarked to an experiment [Bratman, 2000] which demonstrated an output power and saturated efficiency of ~1.1MW and ~29% respectively at 9.4GHz, for similar input parameters. The numerical model has been coded in the Cartesian co-ordinate system which offers greater numerical stability over previous models, and has been shown to accurately and consistently reproduce results comparable to the experimental measurements. The good agreement between the simulation data and the experimental measurements naturally present the numerical model as a suitable benchmark tool to investigate potential efficiency and bandwidth enhancement of the amplifier, achieved through parameter profiling of the microwave circuit. The model predicts that a helical down taper of length 14cm to an output mean radius (r0) and corrugation amplitude (l) of ~11.3mm and ~1.8mm respectively i.e. ~80% of the original helical waveguide's r0 and l values, positioned 4cm before the end of the original uniform helical interaction region of the amplifier, could increase both the saturated efficiency of the amplifier by ~2.5% at 10.0GHz, from ~28.6% to ~31.1% and its bandwidth by 800MHz, from 1.8GHz to 2.6GHz. In addition, an X-band Marie-type mode converter has been simulated and fabricated which effectively converts from the fundamental mode in rectangular waveguide to the cylindrical TE01 mode with minimal reflections, over an optimised 2.0GHz bandwidth. This converter has been used to test a Penning cathode mesh with the experimental measurements confirming that the mesh transmitted an RF signal in the TE01 mode without reflection or mode conversion.

Quantum signals have intriguing properties and a characteristic feature of them is their intrinsic noise. This results in uncertainty relations restricting our ability to measure conjugate variables with absolute precision simultaneously. In the context of amplification, this noise forbids an unknown quantum signal to be amplified perfectly in a deterministic manner. In the first part of this thesis we propose a method to amplify coherent states probabilistically. Our method is based on coherent state comparison and photon subtraction. We found that for an input chosen at random from a binary set of states, under certain circumstances the fidelity can reach 100%. The probability of success is very high (~ 10 - 40%) and it increases with gain. We tested the experimental performance of our protocol for a gain of g² = 1:8 and verified that the experimental results were in line with the theoretical predictions. For an input state chosen from a binary set the fidelity was > 98% and the success rate of our amplifier was > 26000 ampli ed states per second. In the second part of the thesis we propose a new form of orbital angular momentum and angle states. These states consist of a sum of overlapping Gaussians in the angular position representation. We calculated both the uncertainty product and the entropic uncertainty relation for orbital angular momentum and angle. We found that in both cases our new states have a lower uncertainty than the intelligent states. Bringing all results together, our proposals have implications in quantum communications: as our amplification protocol gives a perfect fidelity while maintaining a high success probability it can find application as a quantum optical repeater, and as our overlapping Gaussian states are well-defined for any value of the angular uncertainty and have lower uncertainty relations than the intelligent states, they could find applications in protocols exploiting the high-dimensional basis of orbital angular momentum states.

Optical microscopy has long been an established tool in the biomedical sciences, being the preferred choice in the study of single cells and tissue sections. The realisation of the confocal laser scanning microscope in the 1980s led to major advances in the way optical microscopy is implemented, paving the way for the use of interference techniques such as 4Pi microscopy to increase the optical resolution, and for nonlinear microscopy techniques such as two-photon microscopy, which allows deeper penetration and the imaging of live specimens as a consequence of reduced photo-bleaching, and coherent anti-Stokes Raman scattering (CARS) microscopy, which produces high-contrast images without the need for fluorescent staining. In this work, I discuss advances in nonlinear and interference techniques available for biomedical imaging. I present a simultaneous near-field and far-field viewer for use in aligning the input beams in a CARS microscope and in a sum-frequency-generation- based two-photon microscope. I show 3D optical sectioning of whole mouse embryos using the Mesolens, a giant microscope objective capable of subcellular resolution in a 5 mm field of view, and present theoretical calculations on its use for two-photon microscopy. I present fast recording of synaptic events in neurones, with reduced photo-bleaching, using widefield two-photon microscopy. Finally, I show multiple super-resolved sections are obtained using a laser scanning standing wave microscope, generating precise contour maps of the surface membrane of red blood cells and revealing 3D information from a single image.