Table of contents

We demonstrate that current many-body theory predicts that the high-density phase of the electron fluid is mechanically unstable for r_s greater, similar 5 and thermo-dynamically unstable for r_s greater, similar 4. The pressure-density phase diagram of the electron fluid is shown to consist of a high-density, liquid-like region and a low-density, gas-like region, separated by a two-phase region. We calculate the low-density segment of the coexistence curve in the limit of low temperature and pressure.

This paper presents the results of measurements of the growth of gold on rocksalt when the incident atoms, obtained by sputtering ejection, have energies of approximately 20 ev atom^-1. The results are compared with those obtained using thermal evaporation sources, when the incident atoms have energies of the order of 0·3 ev atom^-1.

The effect of the high arrival energy is manifested in an increased saturation island density (numbers of discrete groups of atoms per unit substrate area) and a change in the orientation of the deposit. The results are discussed in terms of incomplete thermal accommodation and production of atypical adsorption sites on the substrate.

A liquid gallium immersion dilatometer has been developed which is suitable for measuring thermal expansion coefficients of bulk non-metallic solids with high accuracy over the temperature range 50-700 °C. The dilatometer is constructed of vitreous silica and volume changes are measured by weighing the gallium expelled through a narrow-bore capillary tube. These experiments give thermal expansion coefficients relative to that of vitreous silica but the thermal expansion of silica is about 1% of that for the alkali halides, so that accurate absolute values may be obtained.

The apparatus was tested by measuring the expansion coefficient of mercury up to 300 °C, and the results obtained agree to within 0·1% with the very accurate data of Beattie et al. (1941). Thermal expansion coefficients have been determined in the temperature range 50-700 °C for liquid gallium with an estimated accuracy of ±0·2%, and for single crystals of sodium chloride and potassium chloride with estimated accuracies of about ±0·5%.

Contributions to the thermal expansion arising from the thermal creation of defects have been investigated in relation to the increasing positive curvature of the plot of thermal expansion coefficient against temperature observed for both salts above about 800 °K. Comparison with the results of ionic conductivity experiments shows that certainly for KCl and probably also for NaCl these effects are much too large to be accounted for in terms of vacancy formation and they are almost certainly higher-order anharmonic effects.

Gaseous mixtures of argon with krypton, and of argon with neon, were condensed on to substrates held at 7°K inside an electron diffraction camera. Diffraction studies of the deposits so formed were subsequently carried out with 50 kv electrons. The argon-krypton mixture crystallized into a single face-centred cubic phase whose lattice parameter depended in a non-linear manner on the atomic concentration of argon in the mixture and was intermediate between the lattice parameters of the two pure components. The argon-neon mixture crystallized into two separate phases, one consisting of crystals of pure neon, the other being very disordered and probably consisting principally of argon. These results are discussed with reference to a theoretical model of binary mixtures proposed by Prigogine and by Scott.

Neutron diffraction patterns for vitreous germania, synthetic vitreous silica and radiation-damaged synthetic silica have been obtained to 18 Å^-1. With the use of a special resolution function the results have been analysed to produce `radial distribution functions' free of truncation error ripples. The function for vitreous germania shows considerable improvement on previous x-ray measurements. Evidence for the concept of ordered groups of molecules in the glass is obtained from the radiation-damaged experiments.

The onset of long-range magnetic order in a heavily crushed powder of chromium takes place at 450 °K, below which a simple antiferromagnetic structure AF₀ exists. With reduction of temperature it gradually transforms into the structure AF₁ which is incommensurate with the unit cell. The subsequent spin-flip transition (AF₁ -> AF₂) occupies a temperature range and is incomplete even at 4·2 °K. When the sample is annealed the AF₀ structure becomes less prominent and the temperature ranges of both the spin-flip and commensurate-incommensurate transitions decrease, until the `normal' (single-crystal type) behaviour is reached. Study of a chromium-½% rhenium sample gives an indication of the nature of the AF₀-AF₁ transition.

The phase-shift method in solid state scattering theory is applied to the study of the scattering of long-wavelength phonons by point defects in the cubic lattice model of Montroll and Potts. Expressions for the mean free path and Debye temperature are derived for a solid containing a small concentration of point defects.

The ⁵⁹Co resonance in ferromagnetic cobalt-vanadium alloys containing 0·32-25·2 at.% of V has been studied at 4·2 °K. The continuous-wave linewidth of the main line is found to be unaffected by the addition of small quantities of vanadium, but spin echo measurements have disclosed satellite lines at 176 MHz and 212·2 MHz. The lines are shown to be due to cobalt nuclei with one vanadium atom in the first near-neighbour shell, and one vanadium atom in the third near-neighbour shell (and none in the first two shells) respectively.

The technique of nuclear magnetic pulsed double resonance has been used to study the quadrupole interactions of ¹⁴N in a single crystal of ammonium dihydrogen phosphate. The double resonance spectra are examined in some detail, and the advantages of this technique over conventional methods are discussed. For ¹⁴N in ammonium dihydrogen phosphate eQ_NV_ZZ/h = 24·0±2·0 kHz, and n = 0.

A theory is given of the spin density distribution of the ferromagnetic component in canted antiferromagnets having a molecular ligand of the type AX₃. It is found that for the well-known types, such as CO₃^2-, NO₃^-, no appreciable spin is covalently transferred to the central ion. However, a great exchange polarization of the AX₃ radical will occur and produce a negative spin density around the A ion. The theory is compared with experimental neutron scattering data on MnCO₃. The analysis shows that f, the spin transfer parameter for Mn²⁺ in this salt, has a value of 6%; this is to be contrasted with values of the order of 1% observed in salts of a less covalent character than MnCO₃. The analysis also shows that the magnetic moments on the oxygen and the carbon ions are of roughly equal magnitude, but of opposite sign; this is in qualitative agreement with the theoretical discussion of the exchange polarization in the CO₃^2- radical. The existence of the aspherical Kaplan spin distribution is discussed, and it is concluded that the experiments with this salt are not yet precise enough to give evidence of the effect. The analysis is based only on the neutron data and shows a fair agreement with the magnetization data for the total canted moment.

A non-perturbative, representation-independent method of studying the linear response of a disordered system to an external field is presented. For a homogeneous system, the momentum space representation of the basic equations reproduces Van Hove's results.

This method is applied to evaluate the static conductivity of a disordered system in the absence of a magnetic field. It is found that σ can be split into a portion proportional to the unit tensor, which is negligible in the weak coupling limit, and a second portion which exhibits the functional structure of the elementary expression one gets from the Boltzmann equation. In the strong coupling case, the first term may be important.

Measurements have been made of the current noise in gold films deposited on glass substrates with surface resistivities in the range 10 to 10⁴ Ω/sq. The experimental results have shown that the spectral noise power density may be expressed by

where I is the direct current flowing through a gold film of surface resistance R at a temperature T, f is the noise-channel frequency, F(T) is a function of T and G(S) is a function of film structure. Observations have also been made on the phenomenon of noise bursts.

The experimental data are interpreted in terms of a trap-assisted tunnelling conduction mechanism with a range of barrier widths. Further support is given to the proposed model from current-noise measurements made on cermet films.

The high-field transverse magnetoresistance of lead has been investigated experimentally and theoretically. A simple model of the third-zone sheet of the Fermi surface has been used to determine the ranges of magnetic field directions which give rise to open orbits on its surface. These field directions lie within a solid angle around <001> (B region) and, for a limited extent outside the B region, on planes of rational indices belonging to the zone <011> (Aprime spikes). The experimental results compare favourably with these predictions. The two principal Aprime spikes extend 16·5±0·25° from <011> in {100} and 15·1±0·25° in {01, macron1}, in substantial agreement with the values given by Schirber. The radius of the B region is 3-4°, comparable with the estimated value of 4°. There was evidence to suggest the existence of the shorter Aprime spikes predicted theoretically. No change in the lengths of the Aprime spikes could be detected at a pressure of 2·3 kilobars.

A sum-rule method is used to determine the response of an Ising ferromagnet to a time-dependent transverse magnetic field. A similar method yields the energy distribution of the inelastic neutron scattering cross section. The results are valid for arbitrary temperature including the critical region, and the effect of a static parallel magnetic field of finite magnitude is also considered.

A calculation of the Néel temperature in KNiF₃ has been made. The superexchange interaction has been evaluated in two ways, using wave functions in which the covalency is represented in terms of either the molecular orbitals or configuration interaction schemes. The covalency parameters have been chosen to fit the observed metal-ligand spin transfer, but correspond closely to theoretical estimates. Although the two approaches give rise to expressions which appear to be quite different in form, they both predict an exchange splitting of about 250 °K, compared with an observed value of 180 °K. This is chiefly because they handle kinetic exchange, which is the dominant effect present, in ways which give identical results. Potential exchange produces large individual terms, but there is considerable cancellation amongst them, and the net effect is small.

This result, taken in conjunction with recent calculations on similar salts, strongly suggests that the strength of the antiferromagnetic coupling is due mainly to covalent components in the crystal wave function.

The change of magnetization with temperature of gadolinium in a field of 20 kOe has been investigated up to 50 °K using a sensitive technique. A departure previously found from the T^3/2 variation which is followed at higher temperatures has been confirmed. This can be understood in terms of the disappearance of higher-order effects which contribute to the higher-temperature T^3/2 region, and the effect of the applied field. At the lower temperatures the data come close to following a field-modified spin-wave T^3/2 law. A comparison with the measured exchange interactions of terbium has also been made.

The energy levels of the magnetic impurity modes are determined for a nearest-neighbour pair defect in a face-centred cubic ferromagnet. The coupling between the host atoms and the impurity atoms is assumed to be ferromagnetic. The calculations are performed by means of a combination of Green functions and group-theoretical methods. According to their transformation properties, eight different types of impurity mode may be distinguished. Only the energy of the modes with B_2u symmetry depends on the exchange integral between the two impurity atoms. The numerical computations are restricted to the resonant modes. Particular attention is given to the resonant A_g and B_2u modes, which are the only modes with energies in the lower part of the magnon band.

The intermetallic compound Au₅Mn₂ has been investigated by neutron diffraction and x-ray techniques and the temperature variations of the magnetic susceptibility, electrical resistivity and Young's modulus have been measured.

The compound is shown to be antiferromagnetic with a Néel temperature of 353 °K, and to have identical chemical and magnetic unit cells. Each contains four manganese atoms with ionic moments of spin S = fraction five-twos arranged in a co-linear antiferromagnetic array.

Above the Néel temperature the magnetic susceptibility showed Curie-Weiss behaviour, with a Curie temperature of +100 °K. Characteristic antiferromagnetic anomalies were also observed in the temperature variation of resistivity and Young's modulus.

A phenomenological model of non-radiative transitions for certain nickel-doped II-VI compounds is developed. The predictions of the model are shown to account for the apparent irrationality of the observed luminescence properties of these particular phosphors. Further experiments are suggested, the information from which should supply sufficient data to provide a rigorous check on any proposed theory of non-radiative transitions between impurity energy levels. The analysis also concerns the validity of conclusions about the nature of the luminescence transitions based on the configurational co-ordinate diagram.

Hall effect measurements on n-type indium antimonide have shown a magnetic field dependence of donor ionization energy which can be explained by quasi-free electrons in the conduction band screening the donor-ion potential. Electron freeze-out has been observed in the absence of a magnetic field.

Experiments on type II superconductors of low defect density reveal that the current-carrying capacity in the mixed state is profoundly influenced by any plane surface parallel to the magnetic field direction.

Similar surface effects are found to be present when the force on the flux lines is due to a temperature gradient rather than to a transport current. A temperature gradient localized near to the surface influences the capacity to support transport current in a manner which varies with the polarity of field and current.

It is concluded that critical current is controlled by interactions between the fluxoids and the sample surface, and not by a `resistanceless surface layer'.

Crude calculations based on the interaction model provide qualitative explanations of polarity effects in the critical current and of the presence of a strong minimum in the critical current of one polarity.

A thin film of beryllium evaporated in an electron diffraction camera, at liquid-helium temperatures, was examined as to its structure and for superconductivity - the former by reflexion diffraction, the latter by the `electron shadow' method. It was found (i) that the film was superconducting below about 7 °K (ii) that the structure was `amorphous'. The same film after annealing at room temperature gave the normal hexagonal diffraction pattern; in this state it was not superconducting at the lowest temperature used (4·2 °K).