Table of contents

A conducting `skin' is formed on chalcogenide glasses as a result of chemical interaction with O₂ or H₂O at elevated temperatures. Heating in high vacuum does not give a skin. Measurements of the resistivity and the Seebeck coefficient are seriously distorted by the presence of the conducting skin: this must lead to uncertainty about the usefulness of some reported data.

Electron spectroscopy of polycrystalline silver has been carried out in three different regions of the secondary electron spectrum: (i) characteristic energy loss region, showing surface and bulk losses, and interband and ionization losses; (ii) the Auger spectrum, showing the main M_4,5 N_4,5 N_4,5 Auger transition split into a triplet, confirming the high-resolution work of Aksela; (iii) characteristic energy gain region, where some 14 peaks have been observed superimposed on the `slow' secondary peak between 0 and 25 eV. Possible mechanisms suggested for the origin of these peaks include `plasmon gains' and direct transitions involving conduction band states above the vacuum level.

A quantitative theory of charge collection in electrostatic generators of the insulating carrier type is developed. First a parallel plate collector geometry is considered. An explanation for the possibility of perfect charge collection with any auxiliary voltage supply is postulated. Also, with this configuration, it is shown that double transport is possible, with a reversal of sign for the transported charge, although the magnitude of the charge is always less than the value arriving at the conductor.

Secondly a converging collector geometry is analysed. It is shown that with this geometry the magnitude of the charge leaving the collector can be greater than that of the charge arriving at the collector. This possibility could have useful applications in electrostatic generator design.

The thermoelectric potential of a molten salt thermocell is not only dependent on the molten salt but also on the types of electrodes used for measurement. The thermo-electric potential of a thermocell with reversible electrodes constructed of the metal of the salt is determined by the entropy changes of the cation and is termed the `cationic thermoelectric potential' by the author. Similarly, electrodes of the anionic component of the salt lead to an anionic thermoelectric potential. This paper is concerned with the thermoelectric potential of cationic thermocells containing pure salts, and binary mixtures of pure salts with a common anion or cation. Previous work has been reviewed and the theory of the thermoelectric potential and the entropies of transfer of ions in mixtures of salts has been extended. Experimental studies of cells containing argentous and cuprous halide mixtures between pure metal and alloy electrodes have generally confirmed this theory.

The possibility of using the thermoelectric potential of a molten salt thermocell for electrical generation was examined experimentally by the construction of a 13 V trial generator. This form of direct generation was shown to be possible but the expected efficiency of no more than 10% rather limited the field of application.

Optimization of a 337 μm HCN gas laser involving long-term stability, choice of beam splitter (polyethylene) and maximum power has been carried out. The gain determination leads to a value of 4% per metre. In addition, a liquid crystal camera has been designed to demonstrate visually the characteristics of several modes of the cavity.

One of the problems encountered when making matrix-addressed liquid crystal displays is the poor discrimination obtained. This arises because the optical effect used to form the display does not have an adequate `threshold' as a function of drive voltage, and elements which should remain OFF become partially ON. It is shown how these problems can be overcome by using AC signals to prevent the partial selection of display elements. Techniques are described which can be applied to static and scanned arrays based on the use of the dynamic scattering mode of a nematic liquid crystal.

A detailed appraisal of the collisionless sheath single cylindrical Langmuir probe technique applied to the study of electron densities and temperatures in afterglow plasmas is described. It is shown that for positive probe potentials (V) no significant disturbance (`pseudodepletion') of the plasma occurs until the electron current to the probe (i_e) approaches the natural diffusive positive ion current flowing to the reference electrodes. This is indicated by departures from linearity of i_e²-V plots obtained in the orbital-limited current régime. Departures from linearity have also been observed due to probe sheath expansion resulting in departures from cylindrical geometry above a critical value of V. It is shown that, for values of V below those critical values signifying the onset of pseudodepletion and sheath expansion, meaningful probe characteristics can be obtained in afterglows from which accurate electron densities can be obtained. Also described are the effects which arise from temporal changes of electron temperature, and the resulting changes in floating potential V_f. The latter has been measured as a function of the electron temperature (obtained from plots of the electron retarding region) and is shown to substantiate the original Langmuir theoretical expression for V_f.

The mobilities and ion-molecule reaction rates for negative oxygen ions have been measured in a drift-tube mass spectrometer as a function of E/N using a pulsed electron source which allows ion transients to be recorded. By curve-fitting these transients using a small digital computer to generate calculated waveforms, the mobilities and reaction rates are obtained by inspection. The rate coefficients for the three-body reaction producing O₃ from O⁻ were measured over the E/N range 30·0-63·9 Td and were found to agree well with earlier data. The two-body reaction producing O₂⁻ from O⁻ was investigated over the E/N range 86·5-252 Td and again good agreement with earlier work was found. The mobilities of O⁻, O₂⁻ and O₃⁻ ions were measured over as large a range of E/N as possible and were found to extrapolate back to zero-field mobilities of 9·2×10¹⁹, 5·6×10¹⁹ and 6·8×10¹⁹ V⁻¹s⁻¹cm⁻¹.

Current-growth measurements have been made in mixtures of oxygen and hydrogen for a wide range of experimental parameters and mixture concentrations. These mixtures were chosen because hydrogen removes the primary O⁻ negative ions formed in oxygen, through a rapid associative detachment reaction, making the mixture behave as a non-attaching gas. This simplifies the analysis required to obtain ionization coefficients and should give more accurate values of the ionization coefficient for oxygen than can be obtained from measurement in pure oxygen. To check that at any value of E/N the mean energies of the O₂-H₂ mixtures were not significantly different from those of pure oxygen, the Boltzmann equation for the gas mixtures was solved and the transport properties of the mixture calculated. It was found that for O₂-H₂ mixtures containing less than 10% hydrogen, the mean energies and the other transport properties were not altered in any significant way. The values of the ionization coefficients obtained in these experiments are in general 30% below earlier published values but are in agreement with some recent current-growth analysis by Wagner (1971) which incorporates detachment and ion-molecule effects.

The intensities of certain spectral lines of glow discharge columns in hydrogen, helium and mercury vapour have been studied in a transverse magnetic field varying from 0 to 3000 G. It is observed that the line intensity rises with increasing magnetic field. After reaching a maximum at a value of the magnetic field which differs for different wavelengths, the intensity decreases. Beckman's expression for the radial electron density and the electron temperature in the positive column in a transverse magnetic field has been assumed, which leads to a quantitative explanation of the observed results. The magnetic field for maximum intensity is found to be a function of the energy of the upper level from which transitions take place for a particular spectral line. Quantitative comparison with experimental results is not possible, as the theory contains some constants whose values are too uncertain.

Formative time lags have been measured in mercury vapour for E_s/p₀=138-422 V cm⁻¹ Torr⁻¹, and have been compared with time lags calculated from models based on Davidson's theory of the temporal growth of a gas discharge. The observed times were compatible with a model having one slow process and one faster secondary ionization process: that is, a process in which delayed photons are produced by the destruction of metastable atoms by collisions in the vapour (coefficient δ₁/α), and a faster process such as positive ion bombardment (coefficient γ). The analysis could not distinguish between the various faster processes so all were included in the coefficient γ. Data are given for the appropriate values of both these secondary ionization coefficients as functions of E_s/p₀. The curve for δ₁/α as a function of E_s/p₀ resembles the excitation probability curve for the ³P₂ metastable state of mercury and the corresponding delayed photon process is predominant for E_s/p₀<220 V cm⁻¹ Torr⁻¹.

An apparatus has been developed to test under direct stress conditions metals such as cast iron which have high damping capacity and a markedly nonlinear stress-strain relationship. The apparatus consists of a maintained, single degree of freedom vibrating system in which the specimen is stressed uniformly. A wide range of damping and stress amplitudes can be covered depending on the specimen dimensions and the material damping. For example, stress levels in cast iron from 15 lbf in⁻² up to the fatigue limit of the material can be achieved. The apparatus is therefore suitable for investigating the variation of damping and dynamic modulus with stress amplitude from very low to high cyclic stress levels. Some typical results are given.

Measurements have been made in order to compare the scratch hardness with the corresponding indentation hardness of lead, copper and mild steel at room temperature. Both types of measurement were made using the same conical indenter, and the effects of lubrication, normal load and apical angle were investigated. It is shown that the scratch hardness values were most reproducible when measured with conical indenters of apical angles in the range 120°-170°. However, these values were rarely the same as the indentation hardness for a given cone. The hardness characteristics of lead, at room temperature, differed from those of the other metals in that the scratch hardness was generally 1·6 times greater than the indentation hardness, and both kinds of measurement were independent of cone angle. It was concluded that this was due to ideal plasticity in a metal deformed at, or above, half the melting point, and that this type of behaviour would be anticipated in other metals indented at sufficiently high homologous temperatures. Finally, the scratch hardness measurement was used to demonstrate the greater degree of anisotropy observed in single crystals, under these experimental conditions, compared with the identation technique.

A technique to measure the force-penetration behaviour of projectiles penetrating metal targets has been developed and used to examine the penetration characteristics of projectiles impacting soft 2S aluminium.

The maximum force measured during complete penetration of a plate of given thickness was found to be essentially independent of projectile geometry, but the geometry influenced the total energy required for the same process. The results have been discussed in terms of the yield behaviour of the material, the effect of the projectile geometry on the deformation pattern, and the energy loss due to friction.

A method for estimating values of the effective constants of a heterogeneous medium, consisting of a continuous and a discontinuous phase, is proposed. This method incorporates a new physical model that reduces this problem of a two-phase system to an equivalent problem of a polycrystalline aggregate. It also takes the inclusion geometry into account in the lower order correlation functions - that is, those of order two and less. The upper and lower bounds and the improvements on them are derived for the thermal conductivity of a statistically isotropic heterogeneous system, and compared with the experimental results of other authors in order to check the effectiveness of this model. Good agreement between theoretical and experimental results is observed in spite of large differences between the conductivity of the two phases, that of the inclusions being about a thousand times that of the matrix. This method can therefore be expected to yield very good results when the differences are small.

Attention is drawn to attempts which have been made to calculate thermal and elastic properties of particle-, fibre- and laminar-reinforced composite materials from corresponding properties of their constituents. The results of measurements of the linear coefficients of thermal expansion of small (similar 0·5 mm) hollow hemispherical specimens of pyrolytic graphite, silicon carbide and triplet hemispherical composite specimens of these components over the approximate temperature range 300 to 800 K are then reported. In conclusion it is shown how dilatation in composite structures such as these is amenable to the application of classical elasticity theory, and how agreement between the results of such calculations and observation provides evidence for the absence of layer separation resulting from constrained expansion.

Hall-effect and resistance measurements on single-crystal alloys of Hg₃Te₃ with In₂Te₃ indicate the types of conduction and the related band structures for both atmospheric pressure and high-pressure phases of the system. The addition of In₂Te₃ to Hg₃Te₃ gives a zero bandgap at 7±2 mol% In₂Te₃. Electrons in the high-pressure cinnabar (HgS) form of Hg₃Te₃ occupy off-centre, heavy effective mass minima, with a mobility near 10 cm² V⁻¹ s⁻¹. The variation with pressure of the electrical properties of the ordered adamantine vacancy structure found at 37·5% In₂Te₃ can be fitted by a simple two-band model. The presence of an impurity level above the lowest minimum is inferred. A new rock-salt phase has been discovered above 42 kbar and 46 kbar for the 37·5% and 50% alloys respectively. Low-temperature measurements indicate that the phase is metallic: it is possibly superconducting at low temperatures.

A small-signal analysis for an infinitely extended semiconductor with negative differential mobility shows that perturbations of the total-current density always lead to an absolute instability. Therefore the experimentally observed switching behaviour of highly doped Gunn diodes cannot simply be interpreted as a consequence of the diffusion-induced absolute instability as was done by Guéret in 1971. This interpretation to apply rather requires the introduction of a zero-field boundary condition which is shown to cancel the total-current-induced instability.

Measurements of the dielectric properties of electron-beam-evaporated films of silicon monoxide are presented. Results for frequencies from 1 MHz to 10 GHz show these films to have a dielectric relaxation phenomenon around 3 GHz. It is concluded that the relaxation is that due to torsional vibration predicted by Hirose and Wada.

The Nernst coefficient has been measured for the Cd₃As₂-NiAs eutectic with the flow of heat both parallel to and perpendicular to the aligned metallic inclusions, in a magnetic field of up to about 5T. At low fields the Nernst coefficient is negative for both orientations, but it becomes positive for a field in excess of about 1·6 T when the heat flow is parallel to the inclusions. Such measurements suggest an improved performance for a thermal radiation detector based on the Nernst effect in the eutectic when the heat flow is perpendicular to the inclusions. This has been confirmed by a determination of the detector figure-of-merit for both directions of heat flow. All the experiments were performed at 300 K.

An anomaly is noted between reported low-temperature thermomagnetic measurements on bismuth crystals and the intrinsic theory developed by Korenblit to account for phonon drag effects. It is shown that an intrinsic description of bismuth is invalid at low temperatures, even for the purest available samples. Improved agreement between theory and experiment is obtained when an appropriate modification of the theory is made. Measurements on short bismuth crystals are reported which indicate that the large phonon drag effects observed in larger samples are quenched, possibly owing to boundary scattering of phonons in the presence of impurities.

Phototransient currents in polyethylene have been investigated in the visible and near-infrared region. Their polarities were found to be opposite to those observed by previous workers. No measurable steady-state photoconduction was observed. An explanation has been offered on the basis of redistribution of charge over a finite region at a barrier near the electrode-polymer interface.

The scattering of photons from perfectly diffusing surfaces and the scattering of molecules from surfaces covered with adsorbates follow similar laws, namely Lambert's and Knudsen's. Hence estimates of the conductance of channels for rarefied molecular flow may be simply obtained by optical analogue experiments.