Table of contents

The variation of H_c2 with T_c in some structurally imperfect superconductors is examined. In none of these is the paramagnetic limit H_p(0)=1.84 T_c exceeded. For highly imperfect compounds with low T_c there is some indication that the Maki limit, H_c2(0)*=1.3 T_c, is approached.

The band structure of chromium dioxide, which is an important material because of its magnetic properties, has been calculated with the augmented spherical wave method by performing self-consistent spin-polarised calculations. The majority-spin electrons are metallic, whereas the minority-spin electrons are semiconducting. A new class of half-metallic ferromagnets is found in the rutile structure.

The non-linear Ginzburg-Landau equation is solved numerically to determine the distribution of magnetisation in a strongly enhanced paramagnetic layer in contact with a ferromagnetic substrate. The magnetisation profiles and the dependences of the total magnetisation on the thickness of the paramagnetic layer are computed for the Ni-Pd interface investigated experimentally by Gradmann and Bergholz. All four possible regimes on the free surface of the paramagnetic layer, i.e. strongly enhanced, 'live' suppressed and neutral surfaces are considered. For a strongly enhanced (but not 'live') free surface, the Ginzburg-Landau model predicts a maximum in the thickness dependence of the magnetisation. The model also predicts that a thin paramagnetic layer with two free strongly enhanced surfaces can become ferromagnetic below a critical thickness which depends on the surface enhancement.

The local structure around Mn atoms in icosahedral Al₈₄Mn₁₆ and in hexagonal Al₄Mn is compared by means of EXAFS and XANES at the Mn edge. The Al₄Mn crystalline phase has been selected as a possible reference structure since its composition corresponds closely to the actual composition of the icosahedral phase. The radial distribution functions are roughly similar in the two phases except for the skewness of the nearest-neighbour peak. A larger disorder beyond the first-neighbour shell causes the contribution from next nearest neighbours to be smeared in the quasicrystal, as compared with that in the hexagonal phase. The XANES spectra are identical, thus indicating similar point symmetry in the two phases.

The authors have explored the possibility of using H₂⁺ ion implantation to produce a Pd-H system for electron spectroscopic studies. This has been achieved by in situ implantation in a VG Scientific ESCA-3 MkII electron spectrometer using the AG-2 sputter ion gun. XPS and UPS studies were carried out on these samples. The results are in agreement with band-structure calculations and with XPS and UPS studies on the Pd-H system prepared by conventional methods.

The anelastic relaxation at intermediate temperatures (100-200K) in metal-hydrogen crystalline systems containing trapping centres presents anomalous features like a shift of the peak temperature and a variation of its apparent activation energy with hydrogen content. New results obtained in Nb-5 at.% Ti-H, Nb-2 at.% Ti-H and Nb-0.5 at.% Ti, and the existing data, have been interpreted by a comprehensive model which assumes a distribution of the interstitial site energies according to a density of states function. The process is thought to be originated from superposition of the elementary relaxations of Ti-H pairs whose population as a function of temperature and H concentration has been calculated using Fermi-Dirac statistics. The close similarity with the relaxation recently observed in amorphous materials and A15 superconductors doped with H indicates that the present behaviour is characteristic of the hydrogenated systems.

Heats of solution and impurity-induced lattice relaxations are calculated for hydrogen in all the transition metals. The hydrogen-metal interaction is described using the effective-medium theory and the metal-lattice relaxations are included in a harmonic-continuum model. The authors find that the trends in the heats of solution can be related to the magnitude of the electron density of the interstitial metal. The trends in the hydrogen trapping energies for vacancies are estimated from the difference between the calculated heats of solution and chemisorption energies. A very simple conceptual picture of the hydrogen-metal interaction results, which can easily be extended to treat for instance hydrogen in metal alloys.

The differential heat of absorption of hydrogen in zirconium has been measured by means of isoperibol calorimetry in the range of hydrogen concentration 0-0.5 H/Zr at temperatures 873-1123K. In the alpha phase the differential heat decreases with increasing hydrogen concentration except for an increase at very small concentrations but, in contrast, increases in the alpha + beta and beta phases, varying between 0.35 and 0.90 eV. Changes in the differential heat due to changes in elastic constants, lattice parameters and Fermi energy accompanying the absorption of hydrogen have been calculated and compared with the experimental result. The change in Fermi energy is a major contribution to the change in the differential heat.

In rare-earth compounds, with 4f-localised states and itinerant electrons, the authors consider the interplay between charge ordering and electronic transitions. In the framework of a band model, they study the possible existence of a long-range charge ordering between occupied 4f sites and its influence on the stability of an intermediate valence state. The intraatomic Coulomb interaction between itinerant and localised 4f electrons can lower the electronic energy in an ordered intermediate valence state. The resulting charge ordering of the conduction electrons is also determined as a function of the Coulomb interaction. At T=0, as a function of the position of the localised states relative to the conduction band, two transitions can be obtained, firstly from a semiconductor to the ordered intermediate valence and secondly to a metallic state. The relevance of such a model is briefly discussed in the light of recent suggestive Mossbauer experiments in EuPd₃B.

A generalised coherent potential approximation is used to describe a chemically disordered material. It is shown that the method gives not only the density of states for such a system but also reliable values of the total electronic energy. The model is applied to a binary metal alloy where it is seen that a pseudo gap develops prior to the development of an ordered phase and in both cases the conductivity is anticipated to be lowered.

The local densities of states of TL_xZr_1-x(x approximately 0.3) metallic glasses, where TL is Co, Ni, Rh or Pd, have been investigated by soft X-ray emission spectroscopy. Unlike the local Rh and Pd 4d states which overlap the part of the Zr 4d band with high binding energy, the Co and Ni 3d states are centred on the Zr 4d states. The contribution to the density of states at the Fermi level from the TL elements is more pronounced for the 3d than for the 4d elements. The results are discussed in relation to density of states calculations on disordered and ordered alloys. Particularly good agreement is obtained with density of states calculations performed for ordered structures, indicating the dominant effect of chemical short-range order.

The authors have performed self-consistent linearised augmented plane wave calculations for the charge density of metallic beryllium, with and without warping of the interstitial potential, using the local density approximation for the exchange and correlation energy functional. Overall agreement with the newest X-ray form factors is good, and it is argued that the remaining discrepancies have their source in the adequacy of the local density approximation and, for some reflections, in neglecting the non-sphericity of the core charge density.

The data of Rowlands and Woods (1975, 1978) for the low-temperature electrical resistivity of strained samples of Ag show that the sign of the deviations from Matthiessen's rule changes from negative to positive as the temperature increases above 12K. An explanation is presented for this change in sign.

Measured pressure coefficients of resistance and the corresponding pressure coefficients of resistivity, estimated using 'law of mixtures' isothermal compressibility values, are reported for a series of Mg-based amorphous alloys. The pressure coefficients correlate with the thermopower and the resistivity. When combined with results for non-magnetic amorphous alloys containing transition metals, there are suggestions of universal correlations between the pressure coefficients and the electrical resistivity. Data are discussed in the framework of the diffraction model.

The authors report measurements of resistance as a function of magnetic field and temperature in quasi-two-dimensional Mg films condensed at room temperature. The anomalous contributions due to weak localisation and interaction effects have been separated and consistency was found between the temperature and magnetic-field dependence of the resistance. The spin-orbit scattering time tau _so depends on the film structure and preparation conditions, the inelastic lifetime tau _i(T) is compared with the values predicted by the theoretical models for the disorder-enhanced electron-electron theoretical scattering.

The authors present data on the effects of alpha irradiation and H implantation on the electrical resistivity, rho , and the superconducting T_c of Bi₃Ni. Two regimes of flux can be distinguished. At low doses, a large resistivity increase is due to atomic disorder irrespective of the impinging ion ( alpha or H); at the same time, only a slight T_c decrease is measured, in contrast with what is observed in A15 compounds. At high doses, specific effects due to H implantation are seen, probably implying significant modification of the electronic structure. There are indications that H enters the lattice as an interstitial, but that it also forms microbubbles.

Alloys of the composition Al_100-xCo_x for x approximately=14 are found to be of the nonBravais icosahedral structure when quenched from the melt at a rate of >1*10⁶K s^-1. X-ray measurements show the presence of FCC Al as well as the icosahedral phase. Measurements on alloys of different compositions indicate that the icosahedral phase has a stoichiometry of approximately Al₄Co. The ⁵⁷Fe Mossbauer effect spectrum of a sample containing 1 at.% Fe shows a quadrupole split doublet. Mossbauer effect parameters are compared with those for other icosahedral Al₈₆T₁₃Fe (T=Cr,Mn,Fe) alloys. Magnetisation measurements show the icosahedral Al-Co alloy to be a weak paramagnet with a room-temperature susceptibility of chi =1.3*10^-6 EMU g^-1. The electrical resistivity shows a weak temperature dependence with a minimum at approximately 15K.

The longitudinal magnetoresistances ( Delta / rho / rho ) has been measured in AuFe (18 at.%) alloy in the temperature range 4.2-200K and in magnetic fields (H) of up to 10 kG. The data have been analysed in the framework of the theory based on the vector mean-field model leading to the coexistence of spin glass and ferromagnetic orders at low temperatures. In the coexistence regime and at low fields, it is found that Delta rho / rho varies as H⁶7/, thereby indirectly implying a quadratic dependence of Delta rho / rho on magnetisation according to the above theory. The measurements of real and imaginary parts of AC susceptibility are also reported.

The local and total densities of states for the arrangement of Co atoms on the 2b sites in Y₉Co₇ and the local density of states for excess Co atoms on Y sites are calculated by a recursion method. All the calculated values of the local density of states on all sites at the Fermi level are not large. When the positions of Co atoms on the 2b sites are shifted a little, the adjacent Co atoms on these sites, as well as the excess Co atoms on Y sites, have a larger value of the local density of states at the Fermi level than all other sites and are capable of sustaining a local magnetic moment, or the weak ferromagnetism of Y₉Co₇. The paramagnetic susceptibility for Y₉Co₇ is calculated with two models each of which assumes excess Co atoms on Y sites and a 2b sub-band with strongly enhanced spin susceptibility.

The authors compare the magnetic properties of a single crystal of erbium before and after charging with hydrogen to a concentration x=0.035 at.(H)/(Er). The decrease with x of the Neel temperatures T_N (along the c axis) and T_H (within the basal plane) can be explained by: (i) a change in the electronic structure of the metal, i.e. a decrease of the number of conduction electrons (or of the Fermi level density of states); (ii) a modification of the interaction between itinerant and localised spins. Interesting features are observed for the low temperature transitions towards ferromagnetism. Along the c axis and in zero field there is an increase of the transition temperature T_c with x; for T>T_C, the same transition is observed in a finite critical field the value of which is smaller for x not=0 than for x=0. Within the basal plane, the transition towards ferromagnetism is observed only in a finite field; contrary to what is observed along the c axis, higher fields are needed to induce the transition for the x not=0 case. The interplay between magnetostriction and hydrogen-induced dilatation is most probably at the origin of these results; anisotropies in the hydrogen distribution may also be invoked.

The nuclear magnetic resonance of ⁴⁷Ti and ⁴⁹Ti has been observed in hexagonal close-packed titanium metal. Measurements of the spin-echo profile were carried out at 4.2K at various frequencies around 10 MHz by means of the pulsed nuclear resonance technique between about 40 and 50 kG. Because both ⁴⁷Ti and ⁴⁹Ti possess a nuclear quadrupole moment, the spectra of both isotopes show a satellite structure, which is typical of quadrupole-disturbed nuclear magnetic resonance. Since the gyromagnetic ratios of ⁴⁷Ti and ⁴⁹Ti are nearly identical, the resonances of the two isotopes were superimposed, leading to rather complex experimental spectra. By computer simulation of the experimental spectra, however, the quadrupole interaction and Knight shift parameters could be determined quite accurately. The deduced quadrupole frequencies for ⁴⁷Ti and ⁴⁹Ti are in full accordance with the ratio of the quadrupole momenta of the two isotopes and lead to an electric field gradient of 1.62+or-0.08*10¹⁷ V cm^-2 for hexagonal close-packed titanium metal. The isotropic and axial Knight shifts are found to be 0.21+or-0.02% and 0.01+or-0.02%, respectively. Although these values could be reproduced quite reasonably by theoretical estimations based on recent band-structure calculations, these estimations indicate that the partitioning of the Van Vleck contribution to the Knight shift into a hyperfine field and Van Vleck susceptibility is only a poor approximation to the exact expression for the Van Vleck term.

The Raman spectra of Nb₃(Al-Ge) with a superconducting-transition temperature of T_c approximately=20K were measured over the temperature range from 300K to 1.8K. For V₃Si, the unfound F_2g Raman-active phonon has been observed. In Nb₃(Al-Ge) the positions of the E_g and F_2g phonon peaks, as well as their temperature dependences, show similarities, but also some significant differences to the other high T_c compounds Nb₃Sn and V₃Si. This behaviour agrees well with a recent theory of the electron-phonon coupling in high T_cA15 materials.

A new approach to the analysis of positron annihilation angular correlation data based on the definition of the LCW folded k-space density in terms of real-space matrix elements and an orbital angular momentum l-based wavefunction expansion is presented and discussed. The computational convenience and interpretation utility of this approach is illustrated with an example calculation for niobium.

The angular correlation peak counts from positrons annihilating in single crystals of zinc have been measured from 95 to 660K in the (1010) and (0001) directions. A significant anisotropy has been observed between the counts in the two orientations which increases with temperature but in the opposite sense to a/c. The anisotropy increases near 290K, and then decreases to the nearly isotropic distribution of positrons trapped at vacancies. The vacancy formation energy, h_f, determined by the threshold temperature of 403K, is 0.52+or-0.01 eV for both orientations. A trapping-model determination of h_f is found to depend on the temperature dependence assumed in the saturation region; bet fits are obtained for h_f=0.59+or-0.03 eV from the (0001) data, and 0.60+or-0.03 eV from the (1010) data. Full angular correlation curves have been measured at 95, 340 and 600K for both orientations. The effects of the higher momentum components in the positron wavefunction are calculated and could account for the anisotropy observed beyond the Fermi momentum at 95 and 340K. It is suggested that the temperature dependence of the peak counts anisotropy may arise from a relatively faster decline of the Fermi cut-off in the (1010) direction relative to (0001), as expected from the band structure of zinc.

The vacancy formation enthalpy in copper, CuSb (0.1 at.% and 0.5 at.%) and CuIn (0.5 at.%) has been studied by measurements of the temperature dependence of the Doppler broadening of the positron annihilation photon line. Not only monovacancy formation enthalpies E_v^f have been deduced from the data but also the vacancy-solute-atom binding energy E_v^b. The extracted vacancy formation enthalpy for copper is E_v^f=1.29+or-0.03 eV. The value for the vacancy-indium binding energy is E_vin^b=0.29+or-0.06 eV. For a vacancy-antimony pair one finds E_vSb^b=0.32+or-0.06 eV.

For pt.I see ibid., vol.15, p.661 (1985). The authors have calculated the angle-resolved photoemission from nickel using a d-band-only model of the Ni bands with a local but spatially varying exchange splitting Delta _j on atom j. By the direction of Delta _j they mean the direction of the magnetic polarisation on atom j. They have extended the previous calculations for spiral magnetic configurations to random configurations with site-to-site correlation. They investigated trends as the strength of the exchange splitting and the range of order were varied. The calculations were made for different points in k space including points at the zone boundary. The results show that a three-peak structure can be observed only under certain circumstances. They fitted the results to the sum of three Lorentzians in order to clarify the trends. In all cases the apparent splitting of the outer peaks was less than the magnitude of the local exchange splitting used to calculate the photoemission. However, the decrease is significant only if the exchange splitting is small and the amount of order is small. Comparing the results with the experiments of Maetz et al. (1982) they obtain bounds on the range of order and the size of the first-neighbour correlation coefficient in nickel above the Curie temperature.