Table of contents

This paper presents an experimental investigation of the effect of surface charge upon the impulse flash-over of metallic-particle-contaminated insulators, which were clamped between two parallel planar electrodes, in at 1.0 bar. The results show that the accumulated surface charge greatly affects the impulse flash-over strength of insulators. The influence of the surface charge is closely dependent upon its polarity and density, as well as upon the electrode configuration. For a PTFE cylindrical insulator with a metallic particle in contact with the anode, a 25% reduction in the 50% impulse flash-over voltage occurs when negative charge of peak density about is present. However, with positive charge, the impulse flash-over voltage is not decreased, but increased instead, the enhancement in impulse strength being 21% at . When metal inserts are used to shield the electrode triple junction, charge accumulation on the insulator surface will be predominantly determined by the field normal to the surface. In this case, the accumulated surface charge appears to have little influence upon the impulse flash-over voltage.

Hydrogenated amorphous silicon nitride films with a high nitrogen content were prepared by RF plasma-enhanced chemical vapour deposition from a mixture of and at . The influence of the main operating variables on the quality of the material was examined. Characteristics and properties of the films were investigated using optical absorption, infrared spectroscopy, x-ray photo-electron spectroscopy and capacitance - voltage measurements. From infrared transmission and reflection measurements, it was found most of the IR peaks obtained by direct transmission in the range , which have often been mentioned in the literature, are caused not by absorption but rather by a strong reflection due to the Reststrahlen effect. IR studies carried out with polarized light revealed an anisotropic beam - sample interaction probably due to short-range phonon vibration modes: a parallel `LO mode' (), a parallel and perpendicular `TO mode' (). For sufficiently high values of the N/Si ratio, a decreasing percentage of or an increasing RF power leads to a decrease in the trapped charge density which should be related simultaneously to an increase in the N/Si ratio and a decrease in the Si - H bond density in the deposited films. It is suggested that, under our experimental conditions, the charged particle bombardment due to the RF discharge does not play an important role in the charge trapping and that the trapping rate seems to be controlled only by the nature of the traps.

It has been proposed that the interaction of a substrate with a suitable elemental adsorbate can be used to monitor its surface chemical activity. If the extent of adsorbate - substrate interaction can be measured by a surface chemical probe such as x-ray photoelectron spectroscopy, one could compare the surface activities of a variety of solids by studying how the same adsorbate reacts with each surface. The scope of this technique has been investigated by using an electropositive metal (Ni) as an adsorbate on a range of ceramic and polymeric compounds. This adsorbate is expected to probe the electronegativity or anionic activity of the surface which, for simple compounds, is directly related to overall anionic concentrations. A model of Ni adsorption in single crystal oxides has been developed which links the extent of adsorbate - oxide interaction to anionic point defects. Experimental data on Mg-doped are in quantitative agreement with this model. It has been seen that even non-ideal surfaces such as polymers and ion-damaged materials show a definite qualitative correlation between the overall anionic concentration and the extent of Ni - substrate interaction. However, in the case of complex compounds, especially those involving transition elements, there may be several cationic defects influencing the anionic lattice which should be recognized in interpreting the overall anionic activity observed. This technique is a simple, yet unique way of comparing the surface chemistries of different substrates.

Electrodeposited multilayered nanowires grown within a polycarbonate membrane constitute a new medium in which giant magnetoresistance (GMR) perpendicular to the plane of the multilayers can be measured. These structures can exhibit a perpendicular GMR of at least 22% at ambient temperature. We performed detailed studies both of reversible magnetization and of irreversible remanent magnetization curves for CoNiCu/Cu/CoNiCu multilayered and CoNiCu pulse-deposited nanowire systems with Co:Ni ratios of 6:4 and 7:3 respectively in the range 10 - 290 K, allowing the magnetic phases of these structures to be identified. Shape anisotropy in the pulse-deposited nanowire and inter-layer coupling in the multilayered nanowire are shown to make important contributions to the magnetic properties. Dipolar-like interactions are found to predominate in both nanowire systems. Magnetic force microscope (MFM) images of individual multilayered nanowires exhibit a contrast consistent with there being a soft magnetization parallel to the layers. Switching of the magnetic layers in the multilayered structure into the direction of the MFM tip's stray field is observed.

A modified reflection electron beam tomography for reconstructing the three-dimensional magnetic stray field on the surface of a magnetic head is presented. An algorithm for reconstructing the magnetic flux density in three dimensions from the deflection vectors of the electron beams has been developed and tested. An application to a one-turn coil model is presented. It is estimated that the magnetic vector field on the head surface can be reconstructed with a maximum error of about 20%. The results show the applicability of our method for reconstructing the three-dimensional magnetic vector field distribution on the head surface.

Degradation of gas mixtures and methods of its reduction in lasers were studied with reference to the DC-excited, transverse-flow CW laser of MLT 1200 type. The long-term output characteristics and their dependence on the mixture's composition determined in the experiment gave a hint concerning the thorough theoretical analysis of decomposition in laser mixtures. The observed optimal concentration of lies within the range and that of is about 2 - 3% for the laser under consideration. In the theoretical analyses of available data we focus our attention on the roles played by the gas pressure, the mixture composition and the growth of the afterglow region in the reduction of working gas degradation. The equilibrium conversion decreases with increasing gas pressure. A 50% reduction in is obtained when the pressure increases from 13.33 to 46.66 hPa. The observed optimal composition corresponds to low reactivity of the laser plasma with a small oxygen content. The optimal period for the laser gas mixture to reside in the afterglow region is larger by a factor of 1000 than the discharge residence time. The relatively small value of the equilibrium conversion in the MLT 1200 laser results from relatively high concentrations of atomic and electronically excited species. The formation of atomic oxygen is inhibited by reactions with electronically excited molecules, especially . Only very small concentrations of and , which are the most harmful reaction products, result according to our calculations for the MLT 1200 system.

The applicabilities of Miller's coefficient to the second- and third-order electronic susceptibilities are compared. To discuss the behaviour of Miller's generalized to the third-order susceptibility, the nonlinear refractive index is considered. The results obtained by employing the available set of 45 mainly centrosymmetric compounds and the set of about 130 non-centrosymmetric crystals show that the assumption of weak changes in Miller's should not be limited only to the second-order susceptibility. This assumption is even found to work better for the third-order electronic susceptibility than it does for the second-order one. The coefficient is employed to estimate the nonlinear refractive index. It is found that the estimation of its magnitude in partially ionic dielectrics may be improved by taking into account the ionicity of crystal bonds. An empirical formula to calculate is presented.

A diffusion model of sheaths with frozen ionization and recombination on emitting cathodes is considered. Governing equations are solved asymptotically and numerically. It is found, in particular, that the electric field in the sheath is directed from the cathode surface into the plasma at low current densities and in the opposite direction at high current densities. Special attention is paid to values of the electric field at the cathode surface, which can contribute to electron emission, and of the charged-particle density at the sheath edge, which provides a boundary condition for equations describing the distribution of the charged particles in the quasineutral region. On the basis of asymptotic and numerical results, approximate analytical formulae for these quantities are constructed.

The motion of melted metal in a droplet hanging at the tip of an arc electrode during arc welding is considered. The motion is induced by a surface tension gradient due to the non-uniformly heated surface of the droplet. It is shown that the melt flow is confined within a narrow boundary layer. The thickness of this layer and the melt velocity within it are estimated. The influence of the metal motion on heat transfer in the droplet is considered. A simple formula for effective thermal conductivity, which takes into account thermocapillary convection, is obtained. Estimates show that, for conditions typical for arc welding, the effective coefficient of thermal conduction exceeds the normal value by approximately tenfold. Calculated heat fluxes agree with those obtained from the observed electrode melting rates.

The evolution of a fast pin-to-pin spark discharge in nitrogen is studied during the first s of its development with a single-wavelength interferometry method. The growth of the plasma kernel, the particle number densities, the temperature distribution and shock-wave position are measured with this technique. The energy dissipated into the gap is measured with a calorimetric bomb. The voltage across the gap and the electrical current flowing through the gap are measured simultaneously. The measured waveform of the electrical current is used as an input to a two-dimensional simulation model based upon the equations from continuum mechanics. The model successfully describes the shock-wave position, the particle number densities and the temperature. The model can therefore be used to produce realistic initial conditions for studies of ignition of combustible gases.

The effect of the presence of ions near the metal surface on the electron field emission by an arc cathode is studied theoretically. An important electron emission process is shown to be mediated by the neutralization of the ions: one electron from the metal is captured by each ion to form an excited neutral atom which is then ionized by the macroscopic field. The neutralization - ionization rates are obtained for ions in front of a metal surface for different fields and then used to evaluate the electron emission. This ion-assisted process contributes significantly to the electron emission, well below the top of the Schottky barrier. The contribution of this resonant process to the cathode heating is also evaluated.

We present a set of expressions for the determination of the activation energy of a thermoluminescence peak recorded with a hyperbolic heating scheme. It is demonstrated that the order of kinetics can be estimated from the fractional intensities at the points of inflection of the peak.

The published fixed-wavelength nulling ellipsometric data on chemically cleaned silicon (Si) surfaces are re-interpreted in terms of error analyses using a three-phase model (air - layer - substrate). The effective thickness of the surface layer formed by the chemical treatment as well as the angle of incidence at which the surface ellipsometric data have been measured are evaluated from this model in which the surface layer is considered an equivalent dielectric film of known refractive index. The error analyses which accompany the computation procedure show that such ultra-thin surface layer thicknesses can be evaluated with sufficient accuracy in spite of the inaccurate knowledge of the layer index, from nulling ellipsometers producing angular errors of a few hundredths of a degree. The correlation between the angle of incidence and the substrate's real index value found allows the latter to be corrected, provided that the angle of incidence is measured directly with sufficient accuracy. The layer thickness values are deduced from the null-ellipsometric data measured for Si(100) surfaces subjected to different chemical treatments and are compared with those found by spectroscopic ellipsometry. Mainly layer thicknesses of about 6.5 and 12 Å are detected respectively on HF-treated and SC1 ()-oxidized Si wafers. It is thus shown that preliminary results on the effect of chemical treatment of substrates could be obtained more rapidly from fixed-wavelength nulling ellipsometric measurements than they could by spectroscopic ellipsometric methods.

A semi-infinite, fractally scaled, transmission line is defined in terms of the impedance along the line and the admittance to earth. The relevant Telegrapher's equation has been derived and general and particular solutions have been obtained. The input impedance of the line has been examined as a function of frequency and the voltage development along the lines determined for a range of the scaling parameters. As has been reported for fractal circuits based on discrete components, the input impedance for frequencies less than the inverse of the relaxation time for the line is of a negative fractional power law form in frequency, characteristic of a fractally scaled system, but - uniquely for this case - for frequencies greater than the characteristic value, the input impedance of the inverse of the square root of the frequency response of a non-scaled line is recovered. The conditions under which the fractal line propagates an input signal have been investigated and are reported.

In the paper `Dielectric properties of ceramics', which appeared in J. Phys. D: Appl. Phys. (Vol 29, No 5, pp 1321 - 5 (1996)), the coefficients of the Devonshire model of ferroelectric materials are reported on the basis of the dielectric properties of the materials. However, there is a rather inappropriate use of the Devonshire theory applied to all the above materials.

It has been suggested that diffuse transitions due to local inhomogeneities would invalidate the simple version of the Devonshire model used to describe the dielectric behaviour of ceramics. There is some evidence of diffuseness in our results but there was only a slight departure from the Curie - Weiss law. The model has been used successfully for optimizing the composition of ceramics for maximum energy storage density.