Table of contents

CaF₂(EU) is known to be a poor thermoluminescent (TL) phosphor. This is attributed to the high stability of Eu²⁺ against oxidation/reduction upon irradiation/heating. This Letter demonstrates a new possibility of achieving considerable reduction of Eu²⁺ ions in CaF₂(Eu) upon irradiation, in the presence of other lanthanide ions such as Ce³⁺, Dy³⁺ and Ho³⁺, this also leads to a greater than tenfold increase in the TL sensitivity of otherwise pure CaF₂(Eu) with the TL emission still coming mainly from Eu²⁺ ions. The possible role played by the other trivalent lanthanide ions in the reduction of Eu²⁺ upon irradiation is discussed.

An ionic beam technique is suggested for observing the effects of surface plasmon dispersion at optical frequencies. Near a metallic surface the energy levels of a one-electron ion are Stark-split by the ionic image field and external electric fields. The latter can be adjusted to obtain fine tuning between the ion and the surface plasmon frequency at omega _s. The technique is suited for the scattering of light of frequency Omega approximately= omega _s, by the ions near the surface for which a double-peak structure is expected to reveal information about the theoretical form of the surface plasmon dispersion relation.

Half-order LEED beam intensities and half-widths produced by the low-temperature ( square root 2* square root 2)R45 degrees displacive phase on W(001) have been determined as a function of nitrogen adatom coverage. The clean surface structure is found to be strongly inhibited by very low coverages (1-2*10¹³ atoms/cm², or 1-2% of a monolayer), and the radius of influence of a N adatom pair is estimated as 20 AA. The results provide evidence for the influence of long-range effects, such as charge-density waves or surface-state coupling, on the formation of the ordered displacive phase.

The upward curvature of the conductivity or diffusion plots of ionic crystals in the frame of a single mechanism, is attributed to an excessive fall of the Gibbs free energy g per defect. It is shown that, contrary to published suggestions, thermodynamics demands a simultaneous increase of the enthalpy h and the entropy s of the defect with temperature. The excessive fall of g is due to an increasing difference between h and T_N.

The electron mobility in heavily doped GaAs is calculated by using an approach by Yussouf and Zittartz (1973) to account for the electron scattering in a Gaussian random potential. The cases of random and correlated impurity distribution are considered. Comparison with experimental data gives evidence that correlation is essential for the electron transport. It gives rise to a weak dependence of mobility on the electron concentration, which agrees very well with experiment.

The quadrupole coupling constant, asymmetry parameter and the direction of the principal axes of the EFG tensor of deuterons in CoSiF₆.6D₂O have been obtained at room temperature. A temperature variation study of ²D resonance reveals an abrupt change in the quadrupolar splitting frequency and in the linewidth of the individual deuteron line at the transition temperature T_c=262+or-2K during the cooling cycle. Furthermore, the coexistence of two phases over a temperature range of 4K below T_c confirms the first-order character of the phase transition.

EXAFS has been measured at the Nb K-edge in the layered compounds NbSe₂ and at the Nb and Rb K-edges in the same crystal after intercalation with the alkali metal. Changes in bond length were measured to an accuracy of 0.004 AA. These changes were compared with changes in lattice parameters observed by X-ray diffraction. It was found that whereas the Nb-Nb distance increased on intercalation, the Nb-Se bond length did not. This result is interpreted by reference to the band structure of NbSe₂.

Investigations of inelastic neutron scattering in cubic ammonium hexachlorides that have almost identical structures give the following energies E₁ for the first excited NH₄⁺ torsional state at 82K: 7.5 meV for (NH₄)₂PdCl₆, 8 meV for (NH₄)₂PtCl₆, 14.5 meV for (NH₄)₂SnCl₆, 15.5 meV for (NH₄)₂PbCl₆ and 18 meV for (NH₄)₂TeCl₆. The relation between E₁ and activation energies E_a for NH₄⁺ reorientation from proton relaxation differs from that of most other ammonium salts. The shape of the potential seen by NH₄⁺ is discussed E_a and E₁ are found to be very sensitive functions of the structure parameters.

The dependence on temperature from 15 to 40K of the average tunnel frequency f_t in (NH₄)₂PbCl₆ is explained as resulting from very short lifetime approximately=10^-12 s in the first excited torsional state of NH₄⁺. A comment of f_t in (NH₄)₂SnCl₆ and NH₄ClO₄ is also given.

Self-consistent unrestricted Hartree-Fock calculations are performance for a CoO₆₀^10- cluster imbedded in a Madelung potential field simulating the rest of the solid. The O 2p-like and Co 3d-like stakes mix but are non-interacting, implying d electron localisation. The widths of the correlated p-like and d-like states are 3.94 eV and 5.38 eV respectively and these values agree well with experimental photoemission results.

A method of calculating electric-field gradients is given for the off-centred self-trapped exciton in KCl, first observed by Block, Wasiela and Merle d'Aubigne (1978). The method is based on the model for defect formation in alkali halides described previously and closely follows the treatment by Gourary and Adrian (1960) for F-centres. Using this expressions obtained for the electric-field gradients and the observed values of the quadrupole interaction parameters for the two central chlorine ions, the magnitude of the axial displacement is calculated to be 0.22 Bohr radii.

The author extends the treatment of Norgett, Stoneham and Pathak (1977) to the later alkaline earth oxides; CaO, SrO and BaO. Calculations show that the optical band observed for the V^- and allied defects in these compounds is to be ascribed to a small-polaron transition as in the model of Schirmer, Koidl and Reik (1974) for MgO. Ground state properties are calculated and compared with experiment where possible.

In a system of non-interacting electrons with disorder, such that there are both localised and mobile electron states, separated from each other in energy by a mobility edge, the contribution to the conductivity from the mobile electrons is proportional to their lifetimes, whereas that due to the hopping of localised electrons is proportional to their rate of hopping. Hence, as a scattering mechanism is switched on between localised and mobile electrons, the conductivity due to mobile electrons is decreased while that due to hopping electrons is enhanced. Which is the dominant mode of conduction would then depend on the strength of the scattering. In this article, the authors examine the Coulomb interaction between electrons as the scattering mechanism, and develop the results for the case of inversion layers. They find that the hopping mode of conduction contributes more and more as the Fermi level of the system approaches the mobility edge from below, and may finally take over. This contribution to the conductivity behaves as if it originated from an activated mobility, which is what experiments seem to indicate.

For pt.I see ibid., vol.12, p.2797 (1979). AC hopping conductivity calculations are performed using a method analogous to the one used in percolation theory for frequencies lower than those described by two-site models. An expression was obtained for sigma ( omega ) approximately omega f( omega tau _c), f being a slowly decreasing function of omega tau _c( tau _c is a few orders of magnitude larger than the characteristic time in two-site models, tau ₀ approximately 1/ nu _ph). Over a wide frequency range sigma ( omega ) approximately omega ^s, with s<1, and s may depend on omega , temperature and concentration of localisation centres. Agreement between the theory and the available experimental data is discussed.

The ground-state properties of the triangular antiferromagnetic Ising model with Ising constant J in a transverse field h are investigated by a real-space renormalisation group method. For h=0, the known properties (energy=-J, finite entropy, no order) are recovered. For h not=0, the ground state remains highly degenerate, no order exists and the correlation function exhibits power-law behaviour up to (h/J)_c approximately=1.41, where a phase transition takes place to a non-degenerate ground state. At h/J=(h/J)_c, power-law behaviour is still observed for the correlation function but with a different exponent. Just above this region a gap Delta opens ( Delta approximately ((h/J)-(h/J)_c)^s with s=0.8) and the coherence length of the correlation function xi diverges as xi varies as ((h/J)-(h/J)_c)^-v with v=0.8. This corresponds to a three-dimensional classical model where h is replaced by the temperature. In this type of system, where the ground state has a degeneracy proportional to N^2/3 it is predicted that a transition will occur at a finite T_c to a low-temperature phase without order but with power-law behaviour for the correlation function, as for the classical XY model in two dimensions.

A new type of effective-field theory of the Ising model is presented. The differential-operator method is introduced into the exact spin correlation function identity obtained by Callen. The Curie temperatures are evaluated by using two different types of effective Hamiltonians. It is also shown how the Zernike and the Bethe-Peierls equations can be reproduced within one framework depending on the choice of effective fields in an effective Hamiltonian. The spin correlation function and the specific heat are presented.

The functions H/M=K(T,M²) and T(K,M²) are investigated near the critical point. It is shown that the coefficients in expansions of scaling functions and in corrections to scaling may be chosen in such a way that the functions k(T,M²) and T(K,M²) are analytic in the variable M² even at T=0 and K=0, respectively, contrary to the scaling law. On the basis of high-temperature expansion results, it is shown in three different ways that analyticities are fulfilled for the three-dimensional Heisenberg model. Postulating the two analyticities, it is concluded that beta =1/2 and delta =5 for that model. Postulates are generalised for other models. It is shown that all critical exponent estimate sets that are known exactly can be obtained on the basis of the two postulates which can be regarded as the basis of a generalised Landau theory.

The effects of the biquadratic exchange interaction and the uniaxial anisotropy on the Curie temperature of the ferromagnetic spin system of spin one with anisotropic exchange interaction have been calculated with the use of the pair model approximation. The dependence of the Curie temperature on the biquadratic exchange interaction and on the uniaxial anisotropy have been calculated for crystal lattices of different coordination numbers and for spin systems of various anisotropy parameters of the exchange interaction and discussed in comparative way. In a certain range of the anisotropy parameter of the exchange interaction, the biquadratic exchange interaction is shown to have no effect on the Curie temperature.

For pt.II see ibid., vol.9, p.3659 (1976). It is shown that basic features of the authors' previous theory of the dynamic properties of a one-dimensional Heisenberg magnet with an applied field are confirmed by computer simulation data for ferromagnetically coupled spins obtained by Loveluck and Balcar (1979).

Expressions for the magnetic susceptibility of a two-sublattice ferrimagnet are derived from semiclassical equations of motion. The dispersion equation of a bulk polariton is derived in general form, and the properties of bulk polaritons in ferrimagnets are discussed. The susceptibility is also used to calculate the magnetostatic correction to the frequency of long-wavelength magnons. The results are illustrated with numerical calculations for rare-earth iron garnets, and details of the mean field model and parameters used are given in an appendix.

The crystal-field parameter b₄ and transferred-hyperfine-interaction (THFI) parameters for ligands and second-neighbour ¹⁹F nuclei have been measured for Gd³⁺ at cubic sites in PbF₂. So also has b₄ for Eu²⁺ at cubic sites in CdF₂. The values of b₄ indicate, for Gd³⁺ that the ligand distance in PbF₂ is less than in SrF₂, and for Eu²⁺ that the ligand distance in CaF₂ is less than in CdF₂ in the opposite order to ligand distances in the undistorted host crystals. The ENDOR measurements of second neighbours show angular as well as radial displacements from the ionic positions in the undistorted host crystal. They also show that erroneous conclusions could be drawn from ENDOR measurements for only one direction of the external field, which may cast doubt on conclusions drawn from some earlier measurements of THFI parameters of second neighbours in fluorite crystals.

The available data on ligand hyperfine interactions for Gd³⁺ and Eu²⁺ in alkaline-earth fluorides is interpreted using a model in which the covalent contributions of s and p electrons on the ligands are assumed to follow a power-law dependence upon the ligand distance R. Electrostatic polarisation of the ligands, caused both by extra charge on the paramagnetic ion and by distortion of the lattice, produces a large contribution to the ligand hyperfine interaction parameters. A crude theoretical model is developed which indicates a similar magnitude of electrostatic contribution.

Deuteron spin-lattice relaxation time T₁ was measured in ND₄ClO₄ and (ND₄)₂PtCl₆ at frequencies from 4 to 7 MHz and for the temperature range from 1.7 to 40K. Shallow level-crossing minima were revealed in the frequency dependence, allowing the tunnelling splittings between the A and E symmetry levels to be estimated as about 6.2 MHz and 10.8 MHz for ND₄ClO₄ and (ND₄)₂PtCl₆, respectively. The shape of the free induction decay was observed to depend on the resonance frequency and was found to be much shorter at the level-crossing minima for ND₄ClO₄. Activation energies of 3.2 and 3.6 kJ mol^-1 were obtained for ND₄ClO₄ and (ND₄)₂PtCl₆ respectively.

The Zeeman splitting of the D and C lines in the excitation spectra of residual Al acceptors in ultra-pure germanium has been measured using photo-thermal ionisation spectroscopy. The field dependence of the splittings has been studied over the range 0.3-2.5 T, with the magnetic field parallel to the (001), (111) and (110) crystallographic directions. The earlier assignment of a doubly degenerate final state for the C-transition is shown to be inconsistent with the present high-resolution results. The assignment of both the D and C lines to Gamma ₈⁺ to Gamma ₈^- transitions is proposed, whereby the ground-state splitting is very small compared with the excited-state splittings at the magnetic field strengths used. From the observed line splittings accurate g-factors for the corresponding excited states are calculated: D-line: ' Gamma ₈^-, r=-0.121 to 0.005, g'₁=+8.4 to 0.4; C-line: 3 Gamma ₈^-, r=-1/13 'exact', mod g'₁ mod =3.1+or-0.2. It is suggested that the observed quadratic Zeeman effect cannot be explained within a simple perturbation model, and should be compared with the results of an extensive numerical calculation allowing for many interacting states.

In rare-earth ions incorporated in the crystal, the absorption bands within the 4f^N+1 configuration are of free-ion type, except that they are slightly perturbed by the Stark effect caused by the crystal field. On the other hand, the transitions 4f^N+1 to 4f^N5d show broad absorption bands, since the 5d orbit is strongly affected by the crystal field. Furthermore, the transitions 4f^N+1 to 4f^N6s caused by the configuration interaction between 4f^N5d and 6s orbits. In this paper, the elements of the energy matrix of the 4f^N5d-6s system in the cubic field and the general formula for the shape of the absorption spectra 4f^N+1 to 4f^N5d-6s are calculated.

For pt.I see ibid., vol.12, p.4081 (1979). The excitation spectra of MgS:Eu²⁺ phosphors were measured at 300 and 80K. Two distinct bands assigned to the ⁸ Gamma ₄(t_2g) and ⁸ Gamma ₃(e_g) bands appear for lower concentrations of Eu²⁺ ions and in addition, two bands assigned to the ⁸ Gamma ₁(6s) and ⁶ Gamma ₄(t_2g) bands are observed for higher concentrations of Eu²⁺ ions. The shapes of the absorption spectra 4f⁷ to 4f⁶5d-6s of an Eu²⁺ ion in MgS are calculated by using the energy matrix for the 4f^N5d-6s configurations and the atomic spectrum data for a free Eu²⁺ ion. Comparison between the observed and calculated results confirms the band assignment and leads to the determination of some energy parameters.

A semiclassical analysis of the environmental effects on fluorescence quantum efficiencies and lifetimes is presented. It is shown, in particular, that the local absorbency of the medium surrounding the fluorescent molecule has a significant impact on the fluorescence quantum efficiency. The theory presented here may be useful in interpreting experimental results concerning fluorescent probe techniques in chemistry and biology.

Arsenic selenide-telluride glasses have been investigated by X-ray absorption and photoelectron spectroscopy. The core electron energy shifts and chemical shifts in K-absorption edge measurements associated with the glass-crystal transitions of pure As₂Se₃ and As₂Te₃ have been studied. The effect of composition on the core level energy and valence bands of As₂(Se,Te)₃ glasses, has been discussed. Mixed-composition glasses are found to be considerably ionic.