Table of contents

An electron-beam tracking technique has been used to locate the sources of field-emitted electrons for subsequent in situ examination with a scanning electron microscope. The electrical characteristics of individual emitters were measured and used to estimate their physical size assuming simple geometric models. Micrographs of two of the locations are shown both before and after the application of an electric field sufficiently large to destroy the emission source.

A method is proposed for measuring the local current density in Tokamak-type devices by means of Thompson scattering of ruby laser light with the scattering k-vector parallel to the toroidal direction. The typically rather small shift of the profile ( approximately 10% of the width) may be determined by suitable summation of the signal observed in six to eight spectral channels. Questions of accuracy involving line-shape, photon statistics and channel calibration are discussed, and a null method is proposed for the last of these. A direct readout proportional to the current density may be obtained with accuracy better than about 10%.

Reports on the preliminary results obtained with a new method of microparticle detection using forward scattering from a focused 4 W argon ion laser beam. For the first time microparticles are shown to be present in the interelectrode gap in the few microseconds prior to vacuum breakdown. The frequency of occurrence is high and it is shown that many of the observations cannot simply be explained as being due to microparticles leaving the anode under uniform acceleration as is usually assumed for conventional microparticle breakdown theories.

The Boltzmann equation has been used to calculate the dissociative attachment coefficient in oxygen for E/N=10-200 Td and this has been compared with observations.

The decay of an electromagnetic wave incident on an inhomogeneous plasma into a background electromagnetic and a forward-going ion acoustic wave (stimulated Brillouin scattering) has been examined by various workers in the underdense region of a non-uniform plasma. It is possible to satisfy the matching conditions for this instability near the resonance layer where large electric fields are excited for a wave obliquely incident on the plasma. A calculation based on the fluid equations is presented and growth rates and thresholds for stimulated Brillouin scattering in this resonance region are calculated and compared with those of the parametric decay and oscillating two-stream instabilities.

The loss (dissociation) of CO in the positive column of a glow discharge has been measured in pure CO and in CO-He and CO-N₂-He mixtures similar to those used in efficient 5 mu m CO lasers. The loss coefficient is about an order of magnitude greater in the mixtures including helium than in the pure CO, but it is not necessary to postulate more than one dissociation mechanism. Positive ion, negative ion, C₃O₂ polymerization and direct electron impact dissociation have been proposed as the mechanism of the CO loss: only the direct electron impact dissociation channel is compatible with the results; whether an excited state is involved has not been determined.

Ionization growth measurements have been carried out in carbon monoxide for values of E/N (electric field/gas number density) between 109 and 182*10^-17 V cm² and N values between 3.3 and 235*10¹⁷ cm^-3. No attachment was detected and the ionization coefficients obtained agree with the 'apparent' ionization coefficients of Bhalla and Craggs (1961) but also show an increase with number density. The secondary ionization coefficients obtained differ from the results of Bhalla and Craggs but the breakdown measurements are in good agreement. Thomson's backscatter equation is shown to fit the current-voltage variation.

An experimental investigation has been made of laser-induced breakdown thresholds in the rare gases, using single picosecond pulses from a ruby laser with time durations of 18+or-4 ps. Breakdown threshold intensity measurements, carried out at pressures below approximately 7000 Torr, were always pressure-dependent. The results show important differences from those already obtained on picosecond timescales at the same wavelength of 0.694 mu m, but are in general agreement with similar measurements made at laser wavelengths of 1.064 and 0.53 mu m. They support the view that present multiphoton ionization theory is inadequate for describing ionization in high ( gamma approximately=1) radiation fields.

The transient build-up of the average electron energy due to the application of a step microwave high-power field to an ionized medium is investigated. Closed-form solutions are presented for three models of the collision cross-section. These are the constant-collision cross-section model characterizing weakly ionized media, the Rutherford scattering model for fully ionized media, and an approximate analytical expression which fits experimental data of the collision cross-section in a hydrogen plasma. The nature of the heating time necessary for the average electron energy to reach 95% of its steady-state value is investigated in each case. In particular, the dependence of the heating time and the steady-state electron energy on field amplitude and frequency is illustrated for the three models.

Electron energy distributions have been measured in and near the negative glow of an abnormal glow discharge in helium at pressures in the range 3-25 Torr. The experimental technique involved the use of a differentially pumped retarding-field analyser. The measured distributions are distinctly non-Maxwellian. In particular, the distribution in the negative glow close to the cathode dark space shows a prominent 'tail' of high-energy electrons and a 'beam' component at the cathode fall potential ( approximately 300 V). The significance of these results for hollow cathode helium/metal-vapour lasers is discussed.

The ratio of the radial diffusion coefficient to mobility D_r/ mu _exp has been measured for the first time in five gases at high E/N (ratio of electric field to gas number density). The range of E/N was as follows: helium 3<or=E/N<or=847 Td; argon 8<or=E/N<or=1271 Td; air 14<or=E/N<or=1412 Td; methane 14<or=E/N<or=5650 Td; nitric oxide 56<or=E/N<or=1412 Td. The Townsend ionization coefficient alpha _T/N has also been measured in each gas for the same E/N range.

The radiative energy loss from an argon plasma in a shock tube has been investigated, with emphasis on the role of reabsorption. Electron densities reached values of 1.5*10²⁴ electrons m^-3 behind the primary shock, and 2.5*10²⁵ electrons m^-3 after shock reflection. For the primary shock, emissivity effects outweighed reabsorption, whereas the axial gradients of gas properties caused by radiative cooling, which were observed by other investigators in experiments at lower electron densities, increased in these experiments. However, for the reflected shock region, the relative influence of reabsorption increased sufficiently to reduce the cooling rate due to radiation.

Tests were performed to examine the dielectric properties of perfluoropolyether fluids (Fomblin Y) and their transparence to ultraviolet radiation. The purpose was to evaluate the advantage of their use in high-power spark-gaps triggered by pulsed N₂ lasers (3371 AA). The results show that the perfluoropolyether fluids exhibit many advantages over the commonly used transformer oils.

Electron swarm motion has been simulated by using a Monte Carlo technique and swarm parameters have been evaluated for electrons in both hydrogen and carbon monoxide gases for a wide range of E/N varying from 30 to 3000 Td. Comparison has been made with the experimental results for the ionization coefficient, the drift velocity and both the radial and longitudinal diffusion coefficients. A set of electron/molecule collision cross-sections has been assembled for each gas which gave a good fit between the simulated and experimental values over the entire E/N range.

The emission of secondary electrons from atomically clean metal surfaces due to bombardment by singly and multiply charged ions is considered. Attention is restricted to ion kinetic energies of less than several hundred eV and molecular ions and alloy surfaces are both excluded. Previous measurements of the secondary electron emission coefficient gamma are used to derive an empirical relationship between gamma , the ionization energy of the ion and the work-function of the surface. This relationship is presented graphically and is more precise for multiply charged than for singly charged ions. It is proposed that the relationship is applicable to any combination of atomic ion and pure metal surface.

Studies of pulsed free-burning arcs in air on steel cathodes at kiloampere (I<23 kA) current levels have been undertaken using high-speed photography, total-voltage, voltage-distribution and radiation measurements. The results show a marked dependence of arc behaviour on cathode condition and in particular, on the total radiative power losses from the arc column. With current pulses of 10 ms and longer, the arc changes from being convection dominated at early times (t<3 ms), to being radiation dominated at later times (3 ms<t<6 ms) in the arc pulse. For t>6 ms, the arc becomes unstable, and this is believed to be associated with the diminished importance of cathode jet convection.

The cumulative effects of an axially inhomogeneous plasma column on the self-focusing and third harmonic generation of electromagnetic beams have been investigated when the scale length of inhomogeneity is larger than the wavelength. The nonlinearity for self-focusing is caused by the heating and subsequent redistribution of electrons by the electromagnetic beam, whereas for harmonic generation it arises through the energy-dependent collision frequency. A parametric study of these nonlinear phenomena has been carried out and the role of plasma halfwidth and peak electron density have been explicitly studied.

High-speed photographs have been taken of the luminous core of an AC arc in an orifice air flow. The arc area is normally directly proportional to the current. In this case the luminous core areas can be correlated by the parameter (r_c²P₀^0.5/I(1- alpha ), which is derived from an integral analysis of arc behaviour. The analysis includes the influence of enthalpy flow, kinetic energy transport and, in an empirical fashion, radiation. Under some conditions, however, regular oscillations in arc diameter are observed. These are found to reduce the interrupting ability of the test head.