Table of contents

Three effects were observed. A Franck-Condon-type shift in the activation energy of the centres at 20°K was found to increase as the binding energy increases; a square law governed the relationship between the two. As the activation energy increases the phonon coupling of the hole bound to a centre increases. The strength of the coupling is accurately predicted for manganese- and copper-doped GaAs by applying to GaAs the theory which Hopfield applied to CdS. The coupling is reflected by the phonon satellites observed in the photoluminescence spectra at 20°K. An anomalous temperature change in the activation energy of one of the centres (V_Ga-Te) was also observed. This centre is analogous to the `self-activated' ZnS: Cl defect and a configurational coordinate model can be applied to explain the temperature dependence.

A correlation is established for zinc sulphide between a minimum in the temperature dependence of electroluminescence and a peak in the thermoluminescence emission. In most of the crystals examined which showed this minimum, it occurred between -30 and -100°c, depending on the frequency of the applied voltage. The associated thermoluminescence peak occurred at -130°c. From the measurements the depth of the traps concerned was deduced to be about 0.3 ev. A crystal from another source gave a minimum at lower temperatures associated with a trap depth of 0.12 ev showing that the correlation is not confined to traps 0.3 ev deep. An explanation is proposed for the minimum in terms of the impact ionization theory of electroluminescence and the effective depths of traps.

The theory of electron tunnelling through a thin dielectric film is modified by taking into account the penetration of the electrodes by the applied field. The numerical predictions differ only slightly from the existing calculations but some significant differences in the asymmetry of the current-voltage characteristics are produced. In particular, it is shown that the direction of easy current flow can reverse twice with increasing voltage. This depends on the barrier asymmetry and the relative amount of field penetration of the junction electrodes. Also the peak of the thermal (J, V) curve no longer occurs at a voltage equal to the metal-insulator barrier height, but at a slightly higher voltage. Some of the predictions of the modified theory are confirmed by experiment.

The sensitivities of some methods for measuring the optical constants and thicknesses of thin absorbing films by means of normally incident radiation have been investigated theoretically using an arbitrary standard. The ranges of values of n₁, k₁ and d/λ over which high sensitivity may be obtained have been determined from appropriate graphs for seven pairs of optical parameters. These ranges indicate that the three most suitable parameters for the simultaneous measurement of n₁, k₁ and d/λ are R, T and ϕ_T

A solution is obtained for Maxwell's field equations in the case of a cylindrical microwave cavity containing two concentric dielectrics, the outer being isotropic while the inner is a uniaxial crystal which has unequal transverse and longitudinal permittivities. The electric field along the crystal axis is then calculated in terms of the cavity parameters and microwave power, and related via the Pockels coefficient to the modulation produced in a light beam travelling along the crystal axis. Measurements of modulation depth for a given microwave input enable the Pockels coefficient at 9 Gc/s to be calculated, and the value so obtained in a crystal of potassium dihydrogen phosphate (KDP) is shown to agree reasonably well with previous measurements at low frequency and d.c.

Isothermal annealing of cold-worked nickel produces softening which initially occurs exponentially but which is followed by a sigmoidal stage. A theory of isothermal softening behaviour, based upon polygonization of subgrains, is presented.

An analytical and experimental analysis is presented for the pressure drop across a liquid nitrogen cold trap. The cold trap is positioned between a low-vacuum air-water-vapour mixture and a vacuum pressure gauge. The analytical study is used to calculate an upper bound for the pressure drop through the cold trap in terms of the geometry of the trap and the total and partial pressures of the original air-water-vapour mixture. These results are presented in tabular form, which allows the specification of a cold trap that would give a negligible total pressure drop. A limited experimental study was carried out, the results of which were in close agreement with the analytical analysis.

Determinations of primary α and secondary ω ionization coefficients in helium at pressures p₀ near to atmospheric and low values of E/p₀ (E is the electric field) show that α/p₀ = f(E/p₀) and ω/α = ϕ(E/p₀) for 30 < p₀d < 730 cm torr. The results obtained for ω/α and its dependence on E/p₀ are shown to be adequately accounted for on the basis of the assumption that the predominant secondary ionization process in helium at high pressures is the destruction of metastable states in the gas with the consequent production of non-resonance photons which liberate secondary electrons at the cathode.

An external probe with spatial resolution (half width)/(plasma radius) of 0·13 has been developed. It uses the magnetic field deflection method of Lin, Resler and Kantrowitz but quadrupole coils are used to improve the resolution and to minimize self-shielding by the plasma. The theory is discussed and the device is shown to be superior to the r.f. excited probes for moving plasmas. Results are presented for Mach 12 shock waves in argon.

A photoelectric pyrometer has been constructed to measure steady and transient temperatures of carbon and graphite arc electrodes. A description of the device and its calibration is given, together with some observations of electrode temperature decays in the range 4000-1300°K following arc interruption. The pyrometer has a response time of 100 μsec and for this application is therefore sufficient to follow most transients faithfully. Also, the spectral sensitivity peak is in the near infra-red at 1.55 μm, rendering it relatively insensitive to light from the plasma. At temperatures above 2500°K the pyrometer has considerable spatial selectivity in that the field of view includes only a very small area of the electrode surface (a disk of diameter 0.007 in.).

Using ultra-high vacuum techniques, sticking probabilities as a function of surface coverage for chemisorption of nitrogen on molybdenum have been measured at room temperature and above. At 298°K the initial sticking probability was 0.27 with saturation at about 4.4 × 10¹⁴ atoms/cm², these values decreasing with increase in temperature. Flash desorption experiments revealed only one dissociative binding state with a heat of desorption of 60.5 ± 2 kcal mole^-1. Carbon monoxide chemisorbed on molybdenum cannot be removed by heating in vacuo to the evaporation temperature of the metal and inhibits further nitrogen chemisorption almost completely. The adsorptive capacity may be restored by heating in hydrogen.

The effect of viscosity on the formation of spray drops of Newtonian liquids from fan-jet nozzles has been examined both photographically and quantitatively. The break-up of the liquid is governed by the flow properties of the liquid through the nozzle orifice and can be described in terms of four ranges of Reynolds numbers. Range 1 corresponds to turbulent flow in the nozzle: viscosity. has no significant effect on drop size. Range 2 is the critical region of Reynolds numbers corresponding to the transition between turbulent and laminar flow; in this region there is a tendency for the drop size to decrease with increasing viscosity. Laminar flow occurs in ranges 3 and 4 and drop size increases with increasing viscosity. In range 3 ligaments are formed at the edges of the spray sheet and they become more pronounced as the viscosity increases until, in range 4, sheet formation is suppressed and drop formation is due wholly to the break-up of ligaments.

Theoretical expressions have been developed for use with static or slow-moving nozzles, to describe both the mechanism of break-up of the spray sheet in ranges 1, 2 and 3 and break-up of the edge ligaments in ranges 3 and 4. The way in which these expressions may be modified to cover rapidly moving nozzles has been discussed.

This paper describes an experimental and mathematical approach to the evaluation of the variation of temperatures with distance and time downwind of a fire in a ventilated passage through a medium of low thermal conductivity. The experimental results obtained in a 30 ft long, 2½ in. square passage enabled a useful simplification to be made in the analysis. The method of analysis was applied to the similar problem of the air-cooling of a circular passage through a medium of low thermal conductivity and the results of this application agreed quite well with a previous analysis by Van Heerden.

A simple charge compensation mechanism for the substitution of mixed valency ions in oxide crystals requires a fixed ratio of the substituents in the melt to obtain equal concentrations of the substituents in the solid, or balanced substitution.

A direct relationship between the melt concentration ratios and the deviation from balanced substitution has been deduced from the theory and shown to hold in practice for the substitution of chromium and lithium in zinc tungstate. This relationship permits the required ratio for balanced substitution to be obtained directly from limited experimental data.

This note gives the derivation of an equation for the normal freezing of binary alloy systems which have parabolic liquidus and solidus curves. Conditions are given under which this modified equation reduces to the normal freezing equation for constant segregation coefficients.

(I, V) characteristics of aluminium-cadmium-sulphide-aluminium (or indium) thin film structures have been studied. After an irreversible breakdown, the (I, V) curve shows an interesting dual negative resistance phenomenon. When the upper electrode is positive the curve first follows a low-conductivity path from which it switches over to a high-conductivity state through a current-controlled negative resistance. On decreasing the voltage, the high-conductivity path is maintained until a critical voltage in the opposite direction is reached when the structure switches back to the low-conductivity state through a voltage-controlled negative resistance region.

Apparatus is described which is capable of producing thin films of bismuth oxide with refractive index 2.50 and index of absorption 0.03 (both measured at a wavelength of 5500 Å). The results described show the necessity for close control of the argon: oxygen ratio in the sputtering discharge, and a simple flowmeter is described that can provide such control.

The variation of damping with temperature of gadolinium wire has been measured between 130 and 300°K. A peak has been observed whose height depends on the previous annealing treatment. Its cause is considered to be micro-eddy currents.

Analysis of the electron diffraction pattern obtained from a permalloy crystal taken at different wavelengths has shown that the direction of easy magnetization is the [111] axis. Analysis of the pattern was performed vectorially. The magnetization at the specimen was found to be about 15 kG from the diffraction pattern obtained.

Electron drift velocities have been measured in nitrogen and hydrogen over the E/p range 4-40 v cm^-1 torr^-1 using a drift tube apparatus incorporating fast shutters. A brief description of the apparatus and technique is included.

The model for heat flow proposed in a paper by Alberts, Bohlmann and Meiring fails since the conductivity of amorphous silica is used for that of crystalline quartz. The experimental data show the fallacy.

A correction to, and extension of, an earlier paper by Sodha, Kaw and Srivastava is given.

The total hemispherical emissivity predicted from the thermal conductivity by an equation suggested in the literature is within 22% of the measured value for polished niobium in the temperature range 500-1000°C. For alloys an unknown constant may be introduced into the equation. This constant may be set to a value such that the disparity between prediction and observation for 20% Cr: 25% Ni: Nbstabilized steel in the temperature range 600-1200°C is no more than 4%.

A simple power law is shown to relate the tensile strength in brittle materials to the volume of material subjected to 95% or more of the maximum stress.