Table of contents

An electron microscope study of ZrTe₃ shows that this material supports two modulations. One is unstable and lacks long-range order, and may be linked to compositional modulation. The second is associated with a phase transition already observed in the electrical and mechanical properties and may be attributed to the formation of a charge-density-wave normal to the metal atom chains.

Within the Falicov-Kimball model with hybridisation the trial function for the intermediate valence state which is described as the exciton condensate is constructed. There are several insulator states with different energy gaps. The narrow-gap state observed in SmB₆ may be derived in the framework of the approach developed, both the hybridisation and the excitonic effects being essential. The optical properties of this state are calculated.

The sharp defect-related bound exciton lines originally reported by Kunzel and Ploog (1980,1981) in the low-temperature photoluminescence spectrum of the MBE-grown GaAs are interpreted in terms of excitonic recombination at a pair of shallow acceptors. It is proposed that the various lines in the spectrum correspond to different separations of the two defects in the pair with the lowest-energy line arising from nearest-neighbour pairs. A pronounced optical polarisation of the lines is also reported.

The ion-pair transfer in ruby was studied as a function of an applied magnetic field. The N1 and N2 transferred lines are found to behave quite differently. Some of the observed features are interpreted as being due to ion-pair level crossings, while other characteristics appear to be more difficult to explain.

Room-temperature optical transmission measurements in the photon energy range 0.5-6 eV have been made on evaporated thin films of the semiconducting layered structures PbI₂ and BiI₃ during intercalation with hydrazine (N₂H₄). After intercalation, the optical absorption edge shifts to higher energies by approximately 1 eV for PbI₂, while for BiI₃ the shift is greater than 3 eV. The process is reversible. Unlike the case of the semiconducting transition-metal dichalcogenides, intercalation does not appear to introduce free carriers to the host lattice.

The positional parameters of the spinel MgAl₂O₄ were calculated using as a basis the space group F43m. The Al³⁺ and O^2- ion positions were varied, and the lattice energy was found for each configuration. The minimum lattice energy, summing electrostatic and repulsive energy contributions, was found to be -0.248495*10^-15 J/unit cell. This minimum energy occurred for ionic positional parameters satisfying Fd3m symmetry requirements. From the variations of energy with position, as well as other experimental considerations, it still seems appropriate to consider the space group for MgAl₂O₄ to be F43m. The specific values for the positional parameters yielding the minimum energy thus correspond to the intersection of energy surfaces of the two groups in question.

The self-diffusivity of Cd in CdS, D(Cd), has been measured by a radiotracer technique between 500 and 850 degrees C at the minimum total pressure (p_Cd^min) and saturated Cd pressure (p_Cd^sat) and also as a function of p_Cd at 675, 750 and 850 degrees C. Above 700 degrees C D(Cd, p_Cd^sat)=(0.058/sub //^*1.44)exp(-(1.85+or-0.03) eV/kT) cm² s^-1 and D(Cd, p_Cd^min)=(1.53/sub //^*3.17) exp(-(2.47+or-0.11)eV/kT) cm² s^-1, whereas below 700 degrees C D(Cd,p_Cd^sat)=(672/sub //^*16.3) exp(-(2.67+or-0.22) eV/kT) cm² s^-1 and D(Cd, p_Cd^min)=(7.6+or-2.8)*10^-14 cm² s^-1. Above 700 degrees C the results are satisfactorily explained in terms of a Cd_i^.. or a doubly ionised Cd_i-related donor and a (Cd_i^x Cd_i^.)pair or a Cd_s defect diffusion mechanism. Below 700 degrees C the situation is much less clear although there are indications of a V_Cd or V_Cd-related mechanism.

Within the recursion method, the local density of states is estimated from a truncated continued fraction (Jacobi matrix). For simple one- and two-band models, the authors compare the two methods for calculating the density of states. The quadrature method assumes no information other than the truncated fraction. Given exact band edges in addition to the truncated fraction, an analytic terminator can do better than quadrature if the internal singularities are sufficiently weak.

Numerical studies of the vibronic eigenstates in the case when an orbital triplet state is coupled linearly to a t₂ vibrational mode in cubic symmetry have been made. The vibronic energy levels and eigenvectors, as a function of pure linear coupling in the strong-coupling region, have been computed again. A detailed analysis of the vibrational contributions in the vibronic eigenvectors has been made. The possibility of, and conditions under which, the symmetry is lowered from cubic to trigonal have been investigated from the properties of the vibronic eigenvectors. The effect of a local strain on a trigonal Jahn-Teller distortion is discussed from the vibronic model.

The intermetallic compound system CePd₃B_x, with x=0, 0.12, 0.25 and 0.5, has been studied by X-ray photoemission, Bremsstrahlung isochromat and electron energy loss spectroscopy. As a function of the boron concentration strong variations of the occupied and unoccupied parts of the valence bands as well as of the 3d core levels are observed. An analysis of the data within the Gunnarsson-Schonhammer (GS) model (1983) yields an f occupation varying from n_f approximately 1 in CePd₃B_0.5 to n_f approximately 0.9 in CePd₃ (90K), while the f conduction electron hybridisation remains about constant over the entire composition range with Delta =150+or-20 meV. Moreover a nonlinear dependence of the lattice contraction on the f occupancy is found which is well accounted for by a calculation performed within the GS model.

The time-dependent Schrodinger equation is solved numerically for the quantum Hall effect system. A novel algorithm is employed. Wavefunctions and phase shifts are obtained for a number of scattering potentials. These calculations serve to verify and illustrate various theorems and expectations about the system. They also provide useful pedagogical examples and give checks on analytic approximation.

Static and dynamic spin correlations in the random quasi-one-dimensional Ising-like antiferromagnet CsCo_0.83Mg_0.17Cl₃ have been investigated by neutron scattering. The three-dimensional magnetic ordering temperature is strongly suppressed by the presence of impurities. The one-dimensional correlation length tends to a geometric limit at low temperatures in analogy with percolation in higher-dimensional magnetic systems. The quasi-elastic scattering is well described by exact calculations of S^zz(Q) for a dilute chain. In addition to the usual spin wave band of the host crystal, the spectrum contains localised modes that arise from the oscillations of spins at the ends of magnetic chains. Their frequency is approximately half that of the host band. Calculations have been made of the weighted contributions to S^xx (Q, omega ) from each randomly occurring length of chain and good agreement with experiment is obtained.

Low-temperature (0.05K<or=T<or=4.20K) DC susceptibility measurements have been performed on the antiferromagnetic frustrated system ZnCr_2xGa_2-2xO₄. Paramagnetic behaviour has been observed below T=1K, providing evidence for the existence of loose spins, resulting from the lattice frustration, far below the Neel temperature T_N and the freezing temperature T_f. The evolution with the concentration x of the contribution of these spins to the susceptibility is discussed.

A quantum statistical mechanical theory of ESR lineshapes is formulated for spin systems with exchange splitting. The theory is valid even for systems under the influence of strong magnetic fields and at low temperatures. A simple model Hamiltonian is deduced from a microscopic point of view. The magnetic susceptibility for these systems is also reduced to a tractable form which gives a firm basis for the analyses of related experiments.

For pt.I see ibid., vol.17, no.27, p.4843 (1984). On the basis of the previous formulation, theoretical analyses are given of the experiments on Cu(NH₄)₂Cl₄.2H₂O and Cu(NH₃)₄SO₄.H₂O (CTS). Certain prominent characteristics are explained: for instance, the asymmetry of the observed lineshapes and the systematic variations of the resonance positions of the field. These are inherent to the relaxation phenomena found in a spin system at high magnetic fields and low temperatures (HLS effect). Detailed discussions are also given to the relative intensities of the resonance lines and the nature of the allowed transitions.

It is shown how bilocal exciton-polariton theory in configuration space does explain the deviations in the exciton reflection spectra from the classical oscillator lineshape. Apart from a numerical matrix inversion the calculations are performed by analytical methods. Comparison with the experimental spectra for CdS and GaAs leads to very good agreement.

In radiation-damaged In- and Ga-doped KCl crystals many In- and Ga-related defects are formed. Their optical absorption bands strongly overlap, making it impossible to determine the optical transitions of a particular defect directly. With measurements of the magnetic circular dichroism tagged by electron spin resonance absorption bands of Ga⁰(1) and In⁰(1) centres, the analogues of the laser-active Tl⁰(1) centres, could be identified. Ga⁰(1) centres have absorption bands peaking at 930, 890 and 620 nm and In⁰(1) centres at 965, 860 and 610 nm. Up to 2.7 mu m no emission was found for Ga⁰(1) centres, whereas upon excitation at 965 nm In⁰(1) centres probably have a very weak emission around 2.2. mu m. In a model calculation the occurrence of radiative and non-radiative transitions upon excitation of the lowest excited state in Ga⁰(1), In⁰(1) and Tl⁰(1) centres is discussed.

The consequences of Auger recombination processes in a graded n-n⁺ contact in cadmium mercury telluride are computed and compared with a simplified analytic model to establish the equivalent surface recombination velocity of the contact. The major contribution is shown to derive from the relatively lightly doped regions of the contact, as the built-in barrier dominates the increasing recombination rate constant as the material becomes degenerate. The resulting recombination velocity is low (<60 cm s^-1). To act as an efficient electrically reflecting contact, the material must become degenerate within 1 mu m of the beginning of the contact graded region.

The initial steps of ultrahigh vacuum deposition of Al on cleaned Si (111) kept at room temperature were studied using low-energy electron diffraction (LEED), Auger electron spectroscopy (AES) and photoemission yield spectroscopy (PYS). It is shown that the new crystallographic order, the decrease of both the work function and the ionisation energy, and the change of the density of states at the interface can be explained by the covalent bonding between Al atoms and Si surface atoms and its distortions with changes in coverage.

Monte Carlo calculations of the herring-bone ordering phase transition of N₂ adsorbed on graphite have been performed for a variety of boundary conditions. The model consists of a rigid lattice of two-dimensional quadrupoles and for a small infinitely periodic system (PBC), a first-order phase transition is found, in agreement with previous work. For weak boundary conditions (WBC) the transition temperature is reduced and the transition remains first order. For strong boundary conditions (SBC) the transition is continuous. There is little residual ordering above the phase transition for the SBC and WBC transitions. The presence of weak and strong boundaries causes a film of adsorbed phase of thickness theta against the one-dimensional wall. ln theta is linear with temperature. A simple mean-field theory accounts for the qualitative behaviour of the boundary effects.