Table of contents

The reversible transition from the semimetallic (1T₂) to the metallic (1T₁) octahedral form of the layer crystal TaS₂ which takes place on heating at 325 K is investigated at pressures up to 60 kbar and in the temperature range 170-330 K. Resistivity data as a function of temperature reveal a semimetal-metal transition under pressure within the 1T₂ phase, and a model to account for this result is proposed. A new shear-induced transition under pressure in the trigonal prism metal 2H TaS₂ is described, and the resistivity values suggest that 1T₁ is the resultant phase at high pressure.

The second sound velocity decrease has been measured as a function of the temperature (1.3<or=T<or=2.1 K). The results are compared with two theories, both based on the Hall and Vinen moving vortex model (1956).

A simple extension of the effective medium theory for the conductivity of a branching medium is presented. new feature of the calculation is that a typical conductivity is assigned to the infinite clusters only. Excellent agreement is obtained with the results of an exact calculation by Stinchcombe (1974). In particular, the quadratic dependence on p-p_c near p_c is correctly reproduced.

The deposition of lead vapour on to sulphur- contaminated copper surfaces was followed by Auger electron spectroscopy. By detecting the formation of a complete binary monolayer (lead fills in the ad-layer vacancies) the spectrometer was calibrated for both adsorbates.

The order structure of CeD_2.29 at room temperature has been determined by profile analysis of powder neutron diffraction data. A non-stoichiometric tetragonal superstructure is adopted; space group I4₁md, lattice parameters a=5.538(1), c=11.046(2)AA. The ideal composition of the superstructure is CeD_2.5 but the fluorite phase is not ordered at this composition because of a high level of intrinsic disorder. It is presumed that the superstructure exists over a range of composition up to approximately CeD_2.4 and identical order structures have been determined in this range for LaD_2.30 and PrD_2.37.

Alkaline-earth fluorides have been irradiated at temperatures between 4 and 300 K using electrons of energy 70 keV, 300 keV and 1 MeV. The coloration produced has been investigated by both EPR and optical methods. At 77 K and lower temperatures, a rapid first-stage coloration occurs which saturates and is followed by a slower second-stage coloration. The first stage is characterized by perturbed F centres and compensating V_K centres. It seems that the perturbed F centres are associated with nearby interstitial fluorine ions but impurities may also be directly involved. The second stage is characterized by the production of isolated F centres and compensating H centres and appears to be the result of intrinsic radiolysis. Thermally induced annealing of the first- and second-stage coloration is investigated. The first stage coloration of BaF₂ is described in some detail in an Appendix.

It is shown by EPR methods that H centres are produced in alkaline earth fluorides by heavy irradiation with electrons at low temperatures and it is suggested that a very inefficient intrinsic process of radiolysis is involved. From structure observed in the EPR spectrum in SrF₂ it appears that H centres may be stabilized at other defect sites and that observed effects of intrinsic radiolysis may be controlled to some extent by impurities.

Electronic energy-loss calculations have been performed for incident ions with charges ranging from Z₁=1 to Z₁=50 channeling in Si, Ge, Al, Cu, Au and W crystals. The calculations are carried out within a model due to Cheshire (1969) which itself is an extension of a model employed by Briggs and Pathak (1974). Energy-loss S_e against Z₁ curves have been constructed with an electron-ion interaction potential of Thomas-Fermi form. The dependence of this S_e against Z₁ relation for a given target (fixed Z₂) on the incident-ion speed nu ₁ has also been found as well as the S_e against Z₂ dependence for fixed Z₁ and nu ₁. Predictions of the model for the S_e against Z₁ dependence (fixed Z₂) are in fairly good general agreement with experiment, the experimental oscillatory structure of the curves being confirmed in the calculations.

Ultrasonic attenuation measurements have been made in GaAs (n-type carrier density 1.3*10¹⁶ cm^-3) between 2K and room temperature at 40 MHz 80 MHz 160 MHz and 320 MHz. The ultrasonic attenuation is dominated by damping due to phonon-phonon interactions and is interpreted in terms of the Woodruff-Ehrenreich theory (1961). In GaAs the attenuation frequency exponent approaches 2 at the highest temperatures and falls towards 1 in the region omega tau =1 (about 30K) these results confirm the predictions of the Woodruff-Ehrenreich theory. Details are given of the Gruneisen parameter average which is operative in this model for the phonon damping.

The effect of disorder upon vibrational frequency spectra is investigated, using a simple two-dimensional glass-forming model system. It is found that the presence of topological disorder alone, in the glass, may cause its spectrum to differ considerably in appearance from those of related crystals. The introduction of an element of angular randomness increases the disparity between glass and crystal spectra further. Other types of imperfection, such as the breaking of interatomic bonds and distortion of individual molecular units, are also shown to produce characteristic spectral effects. The two-dimensional model system studied has strong structural similarities with a number of real three-dimensional glasses such as SiO₂ and B₂O₃, and the results obtained must call into serious question the validity of interpreting the observed vibrational spectra of such materials in terms of those of corresponding crystals.

The electronic band structure of FCC ortho-hydrogen is computed, using a nonselfconsistent KKR method and spherically averaged molecules. A comparison is made between three different local exchange and correlation potentials, due respectively to Slater (1963), Kohn and Sham (1965), and Lundqvist and Lundqvist (1972). The analysis shows that the latter (which includes correlation effects) is more appropriate to the present case. The valence band is found to be about 0.3eV narrower than the one recently obtained by Gomez et al (1973). A possible explanation for the discrepancy is given. The first conduction band shows a strong s-p hybridization, and the average band gap is of the order of the excitation energy to the first two excited electronic states B¹ Sigma _u⁺ and c³ Pi _u, of the free molecule.

The electronic structure of the relaxed exciton is investigated using a Hartree-Fock method. The model concentrates on a cluster of two sodium and two chlorine ions surrounded by an array of point charges. The excited states of the exciton and the associated transitions are discussed. The results are compared with experiment and with the phenomenological models proposed earlier workers. A number of features are confirmed, giving a general picture close to that of Kabler and his co-workers. One problem emerges, since it proves possible to understand the sigma luminescence or the pi luminescence, but not both at present. Possible reasons are discussed. In other respects, there is good agreement with experiment, especially for the optical transitions starting from the lowest triplet state of the exciton.

A two-component induced-moment system, made up of ions with singlet crystal-field ground state and triplet lowest excited state, is examined using a mean-field approximation for the magnetic interactions and treating the mixing of the components exactly in the limit of complete randomness. The results for the magnetization differ appreciably from those of the simple mean-field approximation that merely employs an average exchange interaction, for comparable concentrations of the two components and away from the critical temperature. At zero temperature the magnetization in this singlet-triplet model is the same as that found for the singlet-singlet model in an earlier calculation: at finite temperatures the former displays a smaller magnetization as well as a lower critical temperature. Crystal parameters for the numerical work are taken from experimental data on the crystal Tb_mY_1-mSb.

The relation between the principal axes of the magnetic susceptibility and those of the nuclear electric quadrupolar interaction is examined for the case of ferrous ions in surroundings of lower than rhombic, though nearly octahedral, symmetry. Formulae are obtained which make it possible, given the axes of the octahedral part of the field and the direction along which the ion has minimum susceptibility, to estimate the axial directions and anisotropy of the nuclear electric quadrupolar interaction. The formulae are applied to Fe(NH₄)₂(SO₄)₂.6H₂O, by making use of recent data on the magnetic axes and assuming that the octahedral axes correspond closely with the directions of the Fe-O bonds. It is found that they account quite well for the observed anisotropy of the Mossbauer absorption spectrum, but that the predicted sign of the quadrupolar splitting is opposite to that which appears in the literature.

The magnetization process of the Ising model with first- and second-neighbour interactions, J₁ and J₂ respectively, was studied for all combinations of signs and magnitudes of J₁ and J₂ by the four-sublattice molecular-field approximation. For the square and BCC lattices, ferro- (F) or paramagnetic phase (P), antiferromagnetic phase (AF), superantiferromagnetic phase (SAF) and type IV phase (IV) appear and the transitions among them are classified as: (i) When j₁>0, -¹/₂< gamma < infinity , then F. (ii) When J₁<0, J₂>0, then AF to P. (iii) When J₁<0, 0< gamma <¹/₂, then AF to IV to AF to P. (iv) When J₁<0, ¹/₂< gamma < infinity , then SAF to IV to P. (v) When J₁>0, - infinity < gamma <-¹/₂, then SAF to P, where gamma =z₂J₂/z₁J₁, and z₁ and z₂ are the numbers of the first- and second-neighbour interactions. The critical field, the magnetization curve, and the location of the critical point and the tricritical point were calculated for typical values of J₂/J₁.

Inelastic neutron scattering measurements on single-domain orthorhombic KNbO₃ at room temperature show no evidence of overdamped scattering, in contrast with previous neutron results in the cubic phase. As in the case of KTaO₃, the observed diffuse x-ray planes are shown to correspond to a highly anisotropic transverse acoustic branch with an unusually flat dispersion in the 'soft' plane. In addition, a well defined optic branch with a polarization related to the orthorhombic-rhombohedral phase transition is seen to become soft as q to 0 in the soft plane. A mode-eigenvector determination confirms the x-ray result that essentially only one atom (niobium) moves in the flat transverse acoustic mode. The connection with other experimental results and with the independent chain model is discussed.

The theoretical low-temperature field and temperature dependence of the forced magnetostriction of ferro-, ferri- and antiferromagnets has been explicitly derived in the spin-wave approximation. The magnetoelastic hamiltonian is cast in terms of operator equivalents for the crystal-field and exchange interactions, and the problem essentially reduces to finding the field dependence of thermal averages of the operator equivalents. In the case of ferromagnets, if the anisotropy is small compared with the applied field, the forced magnetostriction may be expanses as lambda = lambda ₀(T)+ lambda ₁(T)H¹2/+ lambda ₂(T)H+ lambda ₃(T)H² + lambda ₄(T)H³2/+..., whereas if the anisotropy is large, lambda = lambda ₀(T)+ lambda ₁(T)H+... In ferri- and antiferromagnets the additivity of the sublattice magnetostrictions is confirmed. For ferrimagnets, in a broad class of approximations, the field dependence is of the same form as for ferromagnets. For antiferromagnets the low-temperature analysis shows that lambda = lambda ₀(T)+ lambda ₁(T)H²+ lambda ₂(T)H⁴+ lambda ₃(T)H⁶+..., similar to the Callens' high-temperature result.

Measurements of the infrared absorption and emission lines of Cu²⁺ in cubic ZnS in high magnetic field are reported. The g factors are equal in both emission and absorption. The angular dependence of the relative intensities of the observed transitions and their dependence on field strength and temperature have been studied. Expressions describing the observed angular dependence of the intensities are derived from a static crystal-field model.