Table of contents

The authors argue that many phenomena associated with metal/nonmetal interfaces and similar situations with a large dielectric constant mismatch can be understood in terms of the image interactions due to charges in the nonmetal. The effects are additional to the traditional interactions, and are especially significant when no reactions between the phases occur. The image-charge concept allows one to rationalise much apparently unrelated information concerning: (a) the systematics of wetting and nonwetting of oxides by liquid metals; (b) the systematics of strong metal-support interaction in catalysis; (c) the spatial variation of stoichiometry in oxides grown on metals; (d) the dependence on thickness of the observed changes in the wetting by water of oxide grown on silicon; (e) some features of radiation-enhanced adhesion; and (f) a number of correlations of behaviour with nonmetal properties in which the precise choice of metal is not critical.

Alternative forms for the radiationless transition rate in the adiabatic coupling scheme are derived, with a minimum of special assumptions, for the case of coupling to several distinct promoting and accepting coordinates. Their relation to the static coupling scheme is discussed. The resulting expressions are suitable for use with state-of-the-art electronic structure calculations, i.e. not merely model systems. The appropriate application of pseudopotential theory in this context is considered.

The authors have carried out studies of the absorption of p-type polythiophene in the region 20-1600 cm^-1. The results show intense, polaron-induced, vibrational absorption and an impurity-pinned polaron translational mode.

The Raman spectrum of PbHPO₄ has been studied as a function of temperature between 10 and 320K (T_c approximately=310K). A soft-mode is observed whose frequency has a temperature dependence that is close to, but not the same as, the spontaneous polarisation which follows the proton ordering. Several of the Raman lines lose intensity continually with increasing temperature and are still discernible at 320K. A direct comparison with the structural data shows that the temperature-dependent behaviour of these Raman line intensities (and the soft mode frequency) is derived from the heavy-atom displacements, and is only indirectly related to the hydrogen ordering. The Raman data confirm the identical crystal symmetries now found above and below T_C, and support the conclusion that the transition involves simply an ordering of the protons, probably accompanied by some anomaly in the phosphate ion motions.

A survey of the recent progress on the properties of transition-metal impurities in the most common III-V compounds GaAs, InP and GaP is presented. A summary of experimental techniques available for identifying deep centers and/or locating deep levels in the band gap is given. The author concentrates on 3d isolated impurities, giving the known energy levels of Fe-group substitutional impurities in the three compounds. Complexes involving these impurities are also considered. In addition, 4d, 5d and rare earth impurities are briefly discussed.

The application of the random dipole model to weakly compensated classical impurity bands is reviewed, considering versions with or without a lattice. A numerical study of these systems was recently reported by Butcher, Cox and McInnes, (1984) who noted considerable disagreement between their results and predictions based on the random dipole model. To elucidate these differences the earlier theory of this model has been extended to dipoles distributed randomly on a lattice rather than on a continuum. This changes the distribution of potentials at neutral sites (identical to the density of states), particularly when the concentration of dipoles is large. The effect of allowing the dipoles to reorient in the local electric field so as to lower their energy is also considered. Using a simple pairing model of the correlation between the dipoles that this induces, it is found that the distribution of potentials is roughly halved in width, in good agreement with the numerical work. The gain in energy from the correction is also calculated; agreement here is less satisfactory, which may indicate that the distribution of potentials is a rather insensitive probe of the ordering of the dipoles.

The electronic structure of alpha -Al₂O₃ has been studied by means of cluster calculations in order to investigate the influence of chemical bonding on the quadrupole coupling of ²⁷Al and ¹⁷O. Charge transfer from oxygen into Al(3s, 3p, 3d) orbitals was found to account for the major part of the quadrupole coupling constant of ²⁷Al. The contribution of these orbitals to the valence states is in agreement with X-ray emission spectra and with previous band calculations in which, however, Al(3d) participation was neglected. The calculated charge density maps are comparable with those from refinements of X-ray diffraction data.

In order to clarify controversial discussions concerning the electronic excitation spectrum of ferromagnetic semiconductors such as EuO and EuS exact T=0 results for the spectral density and the quasiparticle density of states are presented and discussed within the framework of the sf model. The physics of the sf model turns out to be strikingly different in the strong-coupling region g/W>or approximately=0.5 (atomic-like behaviour, g is the sf exchange constant and W is the Bloch bandwidth), in the intermediate-coupling region 0.1>or approximately=g/W. It has been overlooked in the past that typical magnetic semiconductors (e.g. europium chalcogenides) are spread over all three regions. Mean-field treatments are reasonable in a certain sense for weakly coupled systems such as EuO but are, however, completely inappropriate for more strongly coupled systems such as EuS. Due to the sf exchange the conduction band consists of several quasiparticle sub-bands, overlapping in the weak-coupling but with real gaps between them in the intermediate- and strong-coupling regions.

Photoluminescence spectra are measured for GaP doped with tellurium in the density range from 2*10¹⁷ cm^-3 to 3*10¹⁹ cm^-3 at low temperatures. For doping levels higher than about 10¹⁸ cm^-3 an intense luminescence band is observed immediately below the well known (Te⁰, X) line originating from the radiative decay of excitons bound to isolated dopants. For donor concentrations ranging from 10¹⁸ cm^-3 to 5*10¹⁸ cm^-3 this band is characterised by a pronounced fine structure revealing about 19 distinct lines. From measurements of dependence on the doping, the pump level, the temperature, and the time after a short excitation pulse, the authors interpret this structured band as luminescence from impurity cluster bound excitons (ICE). Experimental findings are compared with detailed model calculations confirming his interpretation, where it is concluded that the most intense lines within the ICE band arise from the radiative recombination of excitons bound to donor pairs (Te₂⁰, X).

The effect of eliminating the dangling sp³ orbitals from the basis set in cluster calculations of the electronic structure of simple deep-level defects in Si is examined through the use of the extended Huckel theory (EHT) method. Using EHT parameters selected from those that allow a good reproduction of silicon bands, the authors investigate the electronic structure of the systems Si:N, Si:V⁰, Si:Si_int assuming a T_d symmetry. In all cases the ordering of the levels in the band gap is in agreement with the available results obtained by ab initio calculations. When the dangling orbitals are excluded from the orbital basis set the gap is free from fictitious dangling-bond states and also it is found that the band edges obtained for clusters of increasing sizes are centred around the corresponding gap of the infinite crystal. The locations of the defect levels remain almost fixed for clusters including more than 47 atoms while the valence band width and the positions of the band edges slowly tend to those of the infinite crystal. These two facts suggest that the deep-level positions can, at least qualitatively, be fixed with reference to the band edges of the host crystal through calculations on clusters of relatively small sizes.

A theory for the depolarisation dynamics of hot luminescence in the F centre in KCl is worked out on the basis of a vibronic model. The hot luminescence spectrum is calculated by using eigenvalues and eigenstates of a pseudo-Jahn-Teller system composed of 2p and 2s states coupled to a_1g and t_1u phonon modes, within the Markov approximation for the vibronic relaxation. Because of the pseudodegeneracy, the 2p to or from 2s nonradiative transition occurs at every step of the relaxation, resulting in the depolarisation of the hot luminescence in agreement with the observed spectrum obtained by a recent experiment.

The Kondo effect in a disordered medium is investigated. The existing perturbation theories for static impurities and for magnetic ones are unified and examined in leading logarithmic order. When determining T_K the competition of two effects are shown: the decrease of the density of states at E_F decreases T_K; diffusive processes increase it. T_K depends in a non-monotonic way on the disorder. The experimental situation is discussed for both dilute magnetic alloys and metallic glasses.

For pt.I see ibid., vol.18, p.3739 (1985). The scattering of electrons on two-level systems (TLS) in a disordered medium is investigated. The perturbative series obtained is summed up in leading logarithmic order. It is found, that localisation effects increase T_K, the crossover temperature. Furthermore T_K shows a maximum as a function of disorder. For a certain range of parameters a stronger than logarithmic dependence is obtained ( delta rho ^d=2 approximately T^-1, delta rho ^d=3 approximately T^-12/).

The transverse susceptibility of spin-one Ising systems with uniaxial single-ion anisotropies have been calculated using the linked-cluster expansion method. Series to the tenth order in the paramagnetic phase and to the eighth order in the ordered phase have been obtained for the FCC lattice with both easy-axis and hard-axis anisotropies of various magnitudes. It is found that the correlations of fluctuations play an essential role in predicting the true temperature behaviour of the transverse susceptibility.

A recent study of second-neighbour effects in uniform magnetic chains is extended to dimerised antiferromagnetic chains. This is the first time that the Jordan-Wigner transformation has been applied to the NNN dimerised chain. The resulting Hamiltonian is then treated in the thermal Hartree-Fock approximation, yielding a distortion of the mirror symmetric bands found in the NN case. While the formalism is developed for the XY-Ising model, a detailed application is made for the Heisenberg case with weak NNN exchange interactions. Elastic (phonon) contributions to the free energy of the dimerised system are made by following the Beni-Pincus approach in order to study the spin-Peierls transition. The authors find a limiting value of the spin-phonon coupling constant ( lambda ) such that for lambda <0.611 the transition temperature increases with G, the NNN exchange constant, while for lambda >0.611, it increases with -G. A comparison is made with the results of an alternate approach using the NNN XY uniform chain and XY spin-phonon interactions according to Pytte (1974).

The MnAs spin system has been modelled as a three-dimensional Ising system under the assumption that the exchange interactions are independent of the external conditions (i.e. a rigid spin system). The magnetic behaviour of this system in magnetic fields of 20, 40, 60 and 90 kOe has been studied. The order parameter m=M/ mod M mod (where M is magnetisation) has been obtained by Monte Carlo simulation of an 8*8*6 spin system with hexagonal symmetry under periodic boundary conditions. The relaxation behaviour of m near T_C shows that the model system undergoes a first-order phase transition from a ferromagnetic into a paramagnetic state. The resulting magnetic phase diagram is in good agreement with the experimental one. The independent behaviour of the magnetic subsystem is shown to play an essential role in the transition mechanism.

AC susceptibility chi (T, H)= chi '-i chi " and static magnetisation measurements M(T, H) up to 5T were performed on theta -M_xV₂O₅ vanadium oxide bronzes containing Al and Fe donor atoms between 1.4 and 300K. The susceptibility of Fe/sub /xV₂O₅ (0.33<or=x<or=0.37) shows two magnetic transitions indicating that two magnetic subsystems are present. In the case of Al_xV₂O₅ (0.33<or=x<or=0.50) as well as in the mixed samples (Fe, Al)_0.33V₂O₅ only one transition is observable. For Al_xV₂O₅ and (Fe, Al)_0.33V₂O₅. The authors attribute the magnetic ordering to antiferromagnetically coupled spin pairs of V⁴⁺ ions. The results can be described by a simple model containing V⁴⁺ dimers with S=0 spin and paramagnetic V⁴⁺ ions with S=¹/₂. The second transition appearing only in Fe_xV₂O₅ is discussed in terms of a spin glass phase formed in the iron sublattice.

Results of ⁵⁷Fe Mossbauer studies of a layered cyanide Mg(Mg((CH₂)₆N₄)₂Fe(CN)₆)₂nH₂O as a function of dehydration are presented. The spectral features of the hydrated sample correspond to those of the low-spin Fe(II), whereas the spectrum of the dehydrated specimen is typical of the low-spin Fe(III). The variation of the spectrum as a function of water content has been discussed in the light of the nucleophilic attack of the water molecules in the Van der Waals gap on the Fe(CN)₆ group.

Nuclear magnetic resonance measurements of the proton resonance linewidth and of the proton spin-lattice relaxation time versus temperature in crystalline (PdAlCl₄(C₆H₆))₂ are reported. From the values of linewidth it is deduced that C₆H₆ groups perform reorientational motion around their sixfold symmetry axis both at room temperature and at liquid nitrogen temperature. The activation energy for such motion, deduced from spin-lattice relaxation times in the high temperature region, is found to be 1.4 kcal mol^-1. This relatively low activation energy is close to that for the reorientation of arene groups in several other organometallic compounds containing two arene groups.

A semi-rigid molecule model is used to explain both the temperature dependence of the ³⁵Cl nuclear quadrupole resonance (NQR) frequency nu _Q(T) and the spin-lattice relaxation time T₁(T) in p-bromobenzene sulphonyl chloride. This paper reports measurements of T₁(T) over the temperature range 80 to 285K which reflect the appearance of thermally activated semi-external modes (SEM) of the SO₂Cl group, giving a strong spin-lattice relaxation for temperatures greater than 160K. The measured temperature dependence of nu _Q(T) is explained by assuming the existence of two thermally activated SEM.

Debye relaxation behaviour is unique in the sense that all memory of excitation is instantaneously lost. It is shown that the presence of higher-frequency relaxations for components of the same dipole degrades this property by extending the memory into the timescale of relaxation, thus giving a nonDebye behaviour which can be expected to occur in all practical dielectric materials. In this case it is not possible, as has been suggested, that knowledge of the polarisation or of the depolarisation current at a single time is sufficient, alone, to define the relaxation behaviour of the system for all subsequent times. Furthermore the Guo and Guo experiment (1983) is shown to be incapable of discriminating between this picture of sequential relaxation and the alternative possibility of an ensemble of independent relaxations.