This thesis describes techniques for the generation of ultrashort optical pulses using semiconductor lasers. Active modelocking through gain modulation is applied to 1.?m InGaAsP semiconductor lasers which utilise microlensed optical fibres as external-cavity components, and tunable optical pulses of ~5ps duration have been generated. Particular attention has been given to suppression of the noisy subpulse features associated with low-frequency modelocked semiconductor lasers. This has led to two suppression methods applicable to symmetric (or balanced) and asymmetric cavity configurations, allowing the generation of clean, single feature short pulses of 7-10ps duration with peak powers in excess of 300mW. The physical phenomena relating to such processes will be outlined. Further amplification of these pulses using an erbium- doped fibre amplifier has resulted in output peak powers in the 5W range. A study of the phase-noise characteristics of these modelocked laser systems was undertaken using a high-speed photodetector and wide-band spectrum analyser arrangement. This characterisation allowed improvements in the rms pulse timing jitter from >2ps to below 300fs in the frequency range 50-5000Hz. Novel modelocked laser configurations were also constructed where the cavity components included both a semiconductor amplifier and an erbium-doped fibre amplifier. Linear and ring hybrid lasers were investigated with resultant pulse durations as short as 3ps with high peak power and excellent stability. CW and mechanical Q-switched doped fibre lasers were configured in diffraction grating tuned cavities with high output coupling (R=4%). Unprecedented output powers of up to 700mW tunable over ~100nm around 1.55mum, and 800mW tunable around 1.08mum were obtained. Also the value of 800W peak power for the 80ns output pulses from the Q-switched erbium-doped fibre laser amounts to the most intense pulse created thus far from any fibre laser device. Optimisation of the total tuning range attainable from erbium-doped fibre lasers via length tuning resulted in a variety of nonlinear phenomena, namely self Q-switching and optical bistability at the long wavelength tail of the tunable range.

An X-ray preionised mercury bromide discharge laser (502-504 nm) has been designed, constructed and optimised. The double pulse forming line system built to drive the laser has proven to be both versatile and reliable. It utilises multiple-paralleled thyratrons and has demonstrated the characteristics required for long life operation of high average power gas discharge lasers. The supply produces a voltage pulse of 240 nsec duration and its output impedance is variable from 300 mO to 3.6O. With twelve thyratrons in parallel current risetimes of 83 kA musec-1 into a 300 mO matched load and 2 nsec jitter have been achieved. A voltage transient on the leading edge of the discharge pulse, to aid gas breakdown, up to 1½ times the line charge voltage is possible through a 'spike line' incorporated in the system. The X-ray preioniser, based on a cold cathode field emitter, produces an exposure of 5 mR in the discharge region which is adequate for effective preionisation of the laser. A full parametric study of the laser covering gas composition and pressure. X-ray source effects energy loading and other pulsed power supply effects has been carried out. The peak output energy recorded was 710 mJ with neon buffer gas at 4 atmospheres pressure. This represents an energy extraction of 0.5 JL-1. Pulse lengths of 92 nsec (FWHM) and efficiencies greater than 1.5 % have been achieved.

The optical perturbation of discharge current voltage characteristics (optogalvanic effect or OGE) has been investigated for CO2 and CO laser discharges. A quantitative power perturbation model is constructed and a series of experiments show, a close agreement for changes of all the major operating parameters of the CO2 laser. The theory consists of an evaluation of the microscopic kinetic relaxation processes leading to the changes in thermal balance of the discharge that occur due to the absorption and amplification of the resonant laser radiation. A generation of compact and efficient cw CO2 lasers has been developed which produces higher stable output powers per unit length than previously reported, and these have been actively stabilized by OGE to provide a high degree of frequency stability (<50 kHz) and amplitude stability (< 0.5%) which is a six orders and two orders of magnitude improvement over the passive resonator capability, respectively. New optogalvanic effects have been discovered both at high frequencies (up to 100 kHz) and for sequence (00°2) laser transitions in CO2 and also in the cw CO laser. Preliminary investigation of optogalvanic detection of standing wave saturation resonances in low pressure discharged CO2 have been carried out and some analysis of discharge noise has been necessary to evaluate the ability of such a narrow band detector for laser stabilization. The empirical evidence provided by the temporal response of OGE combined with the gas composition dependence (including N2 free mixtures) proves conclusively that no major ionization mechanism described so far can be responsible for the effect. A thermal explanation of the effect due to modified kinetic cooling of the laser gas has been developed from existing qualitative explanations. This proposed "gas temperature" power perturbation model provides for the first time an accurate (~20%) prediction of perturbations.(amplitude, phase, and frequency) due to the resonant interaction of a CO2 laser beam over a wide range (up to 4 orders of magnitude) of detailed parametric changes, with a CO2 or laser mixture discharge.

Positive-column metal-vapour lasers are important because of the many visible and ultra-violet c.w. wavelengths attainable. In many such systems population inversion among the metal ion states is produced in charge-exchange reactions with helium ions. Although the power of such lasers should apparently increase linearly with the helium ion density and, hence, with the discharge current, several observers have noted a power saturation. This work describes a phenomenon which will account for this. The spontaneous emission from laser states in a helium-selenium discharge is observed across the tube, and the radial profiles of intensity show a central dip which deepens with discharge current and helium pressure. This axial depletion of laser states sets a limit to the achievable power. The dip effect is shown to be due to the removal of metal atoms from the centre of the tube by ionisation followed by a build-up of the vapour at the walls where the; metal is neutralised. States excited from the axially reduced neutral population will reflect the depletion, to some extent in their own distribution. The helium-cadmium discharge is selected as representative of metal-vapour systems for further study using the dip effect. By comparing the radial profiles of different excited states under the same discharge conditions, it is shown that electron collisions probably contribute to the excitation of levels which are normally assumed to result from charge-exchange and Penning collisions. Radial profiles of helium and cadmium metastables are examined. The behaviour of the helium profiles at high cadmium partial pressures is consistent with an electron temperature which depends upon the local cadmium concentration. The shape of the cadmium metastable profile indicates that this species is being destroyed in the discharge - either by electron collision or collisional mixing with the cadmium neutral resonance state. Observations of cadmium ion profiles at various cadmium partial pressures indicate that electron collisions are the principal source of cadmium ionisation. From the magnitude of the radial dip, the ionisation rate constant is calculated (1 - 5 x 10^-8 cm3 s-1). This varies with E/N, the reduced axial electric field. The ionisation rate is found another way, by modulating the discharge current and observing the phase lag of the profile dip. This experiment yields a value similar to the above.

Microwave excitation of pulsed metal vapour lasers (copper halide and strontium-ion recombination lasers) is investigated. Two waveguide coupling structures were designed and built, one based on a ridge waveguide; and the other based on a rectangular waveguide with a tapered narrow wall which was designed to produce a uniform, travelling microwave field. These coupling structures were designed to produce a transverse electric field at high pressure, with high electron densities, using pulsed microwaves with peak powers of up to 2.5 MW, compatible with the necessary requirements for copper and strontium based systems. The magnetron power supply was modified to produce double pulses of variable spacing (between 15 and 500 mus). Laser oscillation was observed on the cyclic transitions of neutral copper (=510.6 and 578.2 nm) and on the recombination transition of singly ionised strontium (=430.5 nm) for the first time in microwave excited systems. The performance of the tapered waveguide coupling structure was found to be superior to the ridge waveguide coupling structure. The efficiency of coupling of microwave power into the discharge was mainly dependent the buffer gas type and pressure and the electron density. At higher pressures, higher coupling efficiencies are observed (over 70 percent for pressures of over 500 mbar of helium). The performance of the lasers using both the coupling stractures was poor when compared to conventional copper based lasers. This was attributed to the interaction of the copper halide with the discharge and a nonuniform temperature distribution along the axis of the quartz tube. An average output power of 18 mW was achieved for both lines in a copper bromide laser. The strontium-ion recombination laser was operated at threshold average microwave input powers, no output powers were measured. The electric field and the electron density of the discharges in both laser systems were estimated and compared with those occurring in the respective conventional laser systems.

The work described in this thesis covers various aspects of pulsed copper and gold vapour lasers. The work is divided into four main parts : a computer model of the kinetics of the copper vapour laser discharge; construction and characterization of a copper vapour laser and a gold vapour laser system (to be used for photodynamic cancer treatment); analysis of the thermal processes occurring in the various forms of thermal insulation used in these lasers; and studies of the use of metal walls to confine a discharge plasma. The results of this work were combined in the design of the first copper vapour laser to use metal rather than an electrically insulating ceramic material for confinement of the discharge plasma. Laser action in copper vapour has been achieved in a number of metal-walled designs, with continuous lengths of metal ranging from 30 mm, in a segmented design, to 400 mm, where the discharge plasma was confined by two molybdenum tubes of this length. A theoretical explanation of the behaviour of plasmas in metal-walled discharge vessels is described.

Presently, a major factor restricting the achievement of higher power outputs from the CO2 laser is the occurrence of plasma instability. This instability may manifest itself in two forms causing either discharge striations (ionisation instability) or, more importantly, discharge arcing (thermal instability). Recent experimental and theoretical work, has identified these instabilities with the electron and ion kinetics of the CO2 laser discharge. The CO2 laser discharge, its development for high power output and the conditions leading to the two types of instability are reviewed. The requirement for full data on electron and ion processes for discharge modelling is clearly established. The techniques involved in mass spectrometric analysis of ions in gas discharge plasmas are reviewed and subsequently applied to the CO2 laser discharge. The major processes affecting the positive ion species in CO2 laser plasmas are determined in this way. The important negative ion species in the CO2 laser plasma are most easily and comprehensively determined by computational methods. These methods have allowed considerable insight into the dominant negative ion processes of the CO2 laser discharge. By application to specific laser situations correlation has been shown to exist between ion densities and the onset of experimentally observed discharge instability. The elucidation of the main ion processes, both by mass spectrometer and computer, has contributed detailed information for modelling of the CO2 laser discharge, and its instabilities, in a variety of configurations.

This work is a plasma diagnostic study of the hollow; cathode metal vapour laser. It can be divided into three sections. Section one concerns the study of the basic parameters (cathode fall potential electric field, cathode fall width; negative glow length and gas temperature) of the hollow cathode discharge. Section two concerns the use of absorption, linewidth and laser lieterodyne techniques to measure the number densities of He- (21S), (23S) states and He+ ions (electrons) since these species are involved in the pumping mechanisms (Penning and Duffendack) leading to population inversions. Section three concerns the emission, radial profiles of the excited states of He I and He II so the influence of the high energy electrons which enter the negative glow from, the cathode fall region can be studied as they cross the glow. Although the metastable number densities in the hollow cathode are comparable to those in the positive column, they behave differently. In the positive column both singlet and triplet metastables saturate with respect to the discharge current. In the hollow cathode the singlets saturate not only with respect to discharge current but also with respect to pressure. On the other hand the triplets increase linearly with the discharge current and follows' the cathode fall potential as a function of pressure. In the positive column the saturation of the metastable number densities with current combined with the cadmium ion drive-out to the walls lead to the laser power output limitation. In the hollow cathodo cadmium ion drive-out still occurs however this can be compensated by increasing the metal atom, number density since in this case the cathode fall potential, and hence electron energy, is not thereby decreased. Stark broadening and laser heterodyning were used to measure the ion number density and both techniques agree well. They show that the ion number density increases linearly with current, and follows the cathode fall potential as a function of pressure. Also they show that the electron (ion) number density in a hollow cathode is about two orders of magnitude larger than that in a positive column which may lead to an enhancement in the Duffendack reaction. The excitation radial profiles show that the negative glow receives a flux of electrons with energy derived directly from the cathode fall potential. These electrons play an important par in the collision processes in the negative glow. The correlation between these regions is shown clearly by studying the rate equations for different species. The behaviour of the species number densities in the negative glow with the discharge parameters can be related to the fundamental process occuring in the dark space.

The techniques of absorption and perturbation spectroscopy have been applied to the He-Cd laser discharge to determine species densities, excitation rates and de-excitation rates of relevance to laser oscillation on transitions in the cadmium ion. The first section, of this work describes the use of a line absorption technique to measure helium singlet and triplet metastable densities in a 3 mm bore capillary tube for both pure helium discharges (current range 10 - 200 mA., pressure range 0.5 - 15 torr) and for He-Cd discharges. For the pure helium discharge, with constant discharge current, the singlet and triplet densities show pronounced maxima of 2 X 10¹² cm^-3 and 9 x 10¹² cm^-3 respectively around 2 torr. At constant pressure the metastable densities saturate for currents above about 20 mA. The measured triplet densities are in fair agreement with values calculated using known cross-sections for production and loss processes. The addition of cadmium vapour has two principal effects: the metastable populations are almost halved when the optimum cadmium pressure for lasing is present while the current saturation is displaced towards higher currents (60 - 120 mA). The observed optimum performance of the He-Cd laser with respect to discharge current, helium filling pressure and cadmium partial pressure is related directly to the behaviour of the helium metastable densities as these parameters are varied. The second section describes the application of the technique of perturbation spectroscopy to the helium-cadmium laser. Chopping the radiation field inside the cavity of the later operating at 4416A induces perturbations in the populations of the 5s2 2D3/2,5/2 and 5p 2P3/2 levels of the cadmium ion. Analysis of the perturbations of the 5p 2P3/2 and 5s2 2P3/2 populations shows that the ratio of non-radiative to radiative de-excitation rates for the latter level (the upper level of the 4416A laser transition) is 0.9 +/- 0.2, with little dependence on discharge conditions. From the 5s2 2D3/2 population changes the cross-section for excitation of this level from the 5s2 2D5/2 level by electron collisions is found to be ~ 7 x 10^-16 cm2.

A high resolution study of magnetic-field-induced second harmonic generation in sodium vapour was performed using a single-frequency continuous-wave dye laser. Resonant enhancement was obtained by tuning the laser to the frequency of the 3S-40 two-photon transition of the sodium atoms. Coherent parametric emission of the second harmonic radiation ocurred in the presence of a symmetry-breaking magnetic field by means of a coherent electric-quadrupole emission at twice the laser frequency. A theoretical model of the second harmonic generation (SHG) was developed, and includes the roles played by atomic selection rules, Zeeman splitting and eigen function mixing in a magnetic field, and the effects of homogeneous and inhomogeneous broadening. The use of a single-frequency laser and an atomic nonlinear medium allowed an experimental investigation of a SHG process in unprecedented detail. This provided astringent test of the theoretical model developed to predict the properties of the nonlinear interaction in terms of fundamental atomic parameters. The excellent agreement obtained between theoretical and experimental results vindicated the theoretical methods used in the modelling.