Table of contents

The general form of the temperature dependence of resistivity in metallic glassy alloys is discussed in detail using the results of the scaling model of localisation and models of electron-electron correlation effects. It is shown that for a metallic glass of relatively low resistivity there would normally exist a resistance minimum at low temperatures, while for glasses of higher resistivity the conductivity below theta _D should be directly proportional to T, changing to a square root T dependence for higher temperatures. These predictions are then shown to be in accordance with new measurements on Cu₅₀Ti₅₀ and Ti₅₀Be₄₀Z₁₀ alloys.

The pressure dependence (0-7 kbar) of the magnetic susceptibility is reported for the intermediate valence system EuPd₂Si₂ in the temperature interval 77-300K. It is found that the thermally induced valence transition becomes more gradual on application of pressure The characteristic fluctuation temperature T_f, also seems to be pressure dependent.

High-resolution measurements of heavy-ion-induced X-ray satellite emission (HIXSE) spectra have been obtained for a series of Mo-Ti binary alloys and V(Fe_xCo_yNi_z)₃ long-range-ordered (LRO) alloys. Variations in the relative intensity distribution of the satellite lines of these alloys were observed and are attributed to the influence of the valence-electron distribution or chemical environment of the emitting target atom. Results are also presented which demonstrate the ability of this method to differentiate between different crystal structures of the same alloy. The sensitivity of HIXSE to changes in the valence-electron environment of alloys may provide a means of probing metal valence states in metals and alloys.

Several atom packings have been relaxed with a Johnson potential. This mere relaxation leads to states having significant differences. A procedure is given for obtaining configurations which can be considered as representative of the same amorphous structure. To give a precise geometric description of it, the authors work out a method of decomposing the whole volume into single units: tetrahedra, octahedra and more generally deltahedra.

Amorphous alloys of the types Zr_1-xNi_x, Zr_1-xCo_x, Zr_1-xFe_x and Zr_1-xCu_x were prepared by melt-spinning and studied by means of differential scanning calorimetry and X-ray diffraction. The crystallisation temperatures (T_x) and the activation energies for crystallisation ( Delta E) were determined. The former values were found to vary only slightly with concentration while the latter show a marked concentration dependence, which is particularly pronounced in Zr_1-xNi_x. The occurrence of compositional short-range ordering (CSRO) in amorphous systems is discussed with respect to its consequences for the hole model proposed earlier. It is shown that the predictions of the hole model are only slightly affected by the occurrence of CSRO, owing to a gross cancellation of the effects of a decreasing hole formation enthalpy and decreasing configuration entropy. The relatively strong temperature dependence of the configurational entropy expected for alloy compositions with strong CSRO leads to an increase in the experimental values of the apparent activation energies. From the concentration dependence of the experimental activation energies indications can be obtained about the concentration dependence of CSRO. It appears that the regions with a maximum CSRO do not, in general, correspond to the concentrations at which intermetallic compounds occur, nor do they correspond in all cases to regions located around the deepest eutectic composition.

A tight-binding model for studying the density of electronic states, N(E), in granular metals is proposed. The paper is concerned with the global states of the system and not with those of the individual grains. It is argued that such an approach is necessary in order to characterise metals composed of very small grains. Results of numerical calculations of N(E) for two- and three-dimensional systems are presented. The author investigates how N(E) is affected by (i) the presence of internal surfaces, (ii) randomness in the intergrain coupling strength, and (iii) connective properties of a percolating cluster of grains. In 3D systems all these effects result in an increased value of N(ER) at the Fermi level. This may lead to an enhancement of the critical temperature for the transition to the superconducting phase. The fractal dimensionality of backbones of 2D clusters at the percolation threshold, as determined from the band width, is consistent with that characterising certain self-similar fractal lattices.

K_beta emission and K absorption spectra of phosphorus in amorphous metallic Ni-P are presented and discussed. Modifications to the electronic distribution of p partial states of phosphorus between the pure element and alloy are deduced. An increase of the pure p states is evidenced. This modification involves a partial localisation of P p states in Ni-P. In contrast, among the strongly mixed s-p states lying in the upper part of the valence band, the proportion of s-character states increases.

The electronic energy bands of two group-IVB metal dihydrides ZrH₂ and HfH₂ have been calculated using the composite-wave variational version of the APW method. Comparisons are made between the results of the authors' calculations and the results of optical studies by Weaver et al. (1981). The presence of a hydrogen-induced valence band beneath the conduction band in the band structures has been supported by the XPS results of Veal et al. (1979) and the photoelectron spectroscopic results of Weaver et al.

Self-consistent LMTO band-structure calculations for the paramagnetic ground states of UN, UC and UO are reported. Analysis of the radial charge densities reveals that the radial bonding charge density is very similar in the three compounds. The lattice parameters and bulk moduli, evaluated from the calculated equations of state, are within 3% and 11% of experiment, respectively. UO and UN satisfy the Stoner criterion for ferromagnetism, whereas UC fails to do so.

For pt.I see ibid., vol.14, no.3, p.639 (1984). The principal trends in the calculated paramagnetic ground-state charge densities, lattice parameters and bulk moduli of the uranium chalcogenides and pnictides are analysed in an approximation that neglects spin-orbit coupling. The increase in the size of the anion across a pnictide or chalcogenide series is responsible for a regular increase in the lattice parameter and a corresponding decrease in the bulk modulus. The decrease in the fractional volume occupied by the uranium atom leads to a narrowing of the uranium d-f band structure and a concomitant increase in the density of states at the Fermi level.

The band structure and the electron-phonon coupling constant are calculated for pi bands using a tight-binding parameter fit for V₃Si. Following this and previous calculations for the delta ₂ band and the sigma band hybridised with the delta ₁ band, an approximate relationship for the contribution to lambda from the various sub-bands of the 3 d band is developed. The relationship depends upon the individual width of the sub-band and its contribution to the density of states and shows that there is some compensation between N(0) (the electronic density of states at the Fermi level) and I² (the electron-phonon coupling constant). A narrow band gives rise to a large value of N(0) but a small value of I²; thus the product N(0)I² does not follow the behaviour of N(0).

Open orbits in the linear-chain mercury compound Hg_{3- delta} AsF₆ were investigated by the induced torque method at a temperature of 1.3K with magnetic fields up to 8.5 T. There is a primary open orbit in the (001) direction for all field directions in the (001) plane. It runs along electron and hole Fermi surface cylinders directed parallel to the direction k_c. There are also open orbits in (01n) directions where n>or=2. They result from magnetic breakthrough with a breakdown field of approximately 6.6 T between electron and hole cylinders. The results confirm the Fermi surface model formed from one-dimensional bands and show that there is a variation in the separation of electron and hole cylinders along the direction k_c.

For pt.I see ibid., vol.13, p.2313 (1983). The electronic structures and cohesive energies of hexagonal monolayers of the alkaline-earth metals Be to Ba are studied in terms of self-consistent all-electron local density functional calculations using the full-potential linearised augmented plane-wave method for thin films. The energy band structures of the 'light' atoms Be and Mg show wide s bands overlapping with p_z bands without any pronounced spectral features in their densities of states. In contrast, the 'heavy' alkaline-earth metals Ca, Sr and Ba are characterized by much narrower s bands and by increasingly sharper features in their densities of states originating from d-like states. Thus, the densities of states at the Fermi energy for the Be and Mg monolayers are low and increase dramatically for Ca, Sr and particularly Ba. In contrast to recent Hartree-Fock results, the author's study shows that the Be monolayer is metallic. The work functions of the monolayers are found to be about 0.2 eV higher than the experimental values for the clean metal surface. The high cohesive energy of Be is explained by the absence of p-core electrons and the admixture of pi bonding. Magnesium is found to form the most weakly bound alkaline-earth monolayer whereas those of Ca, Sr and Ba are stabilised by the contribution of d states.

Normal-state electrical resistivity data from 12-300K on Chevrel-phase superconductors of the type Cu_1.8Mo₆S_8-ySe_y, 0<or=y<or=8, are reported. The temperature dependence of the normal-state resistivity has been analysed on the basis of models proposed for A15 superconductors. It is found that the 'parallel resistor' model based on s-d scattering accounts for the observed variation satisfactorily in most of the compounds, provided the lower limit of the transport theory integral is taken to have a non-zero value obtained by a fit to the experimental data. Physically, this limit corresponds to the minimum energy required for a phonon to scatter an electron from the s band into the d band. The electrical resistivity below 40K could also be interpreted on the basis of interband electron-electron scattering.

The authors present a model calculation which not only indicates the modification of the region of validity of the scaling-theory predictions by a broadened effective field distribution but which also reproduces substantially the systematics of such behaviour as reported experimentally for PdMn.

AC susceptibility measurements on a series of (Pd_0.985Ni_0.015)_1-cMn_c alloys between 4.2 and 25K in various static biasing fields up to 250 Oe are presented for c up to 3 at.%. These data are compared and contrasted with those from other giant moment systems with regard to the dependence of T_c on c and the ease with which critical behaviour can be resolved.

In order to correlate changes in magnetic properties in the Ag-Mn spin-glass alloys with changes in short-range order (SRO), EXAFS investigations of the Ag and the Mn K edges have been performed on Ag-Mn quenched alloys. These measurements follow X-ray diffuse scattering experiments on single crystals, which showed that SRO exists in Ag-Mn quenched alloys of composition greater than 6 at.% Mn. A knowledge of the chemical order which exists in these alloys is necessary to understand the variation of the magnetic susceptibility after annealing at moderate temperatures. As usual in EXAFS analysis, a check on the phase shifts has been achieved with elemental metals (Ag and alpha -Mn). The magnitude of the local variation of the lattice parameter about Mn atoms and the value of the Cowley SRO parameter ( alpha ₁) was measured for five alloys with Mn concentrations from 6-24 at.%. For each sample the Mn-Ag distance differed by only 2% from the Ag-Ag distance measured at the same temperature in pure silver. Thus only a very weak distortion effect exists in these alloys. The Cowley parameters ( alpha ₁) are in good agreement with the results of the analysis of X-ray diffuse scattering experiments. This confirms that the number of first-neighbour Mn atoms to an Mn atom is less than the statistical value.

Low-temperature specific heats for Fe-rich Fe_100-xZr_x amorphous alloys with compositions of x=8, 9, 10 and 11 were measured in the temperature range between 1.5 and 6K. The total specific heat is found to follow the linear relation C= gamma T+ beta with a negative coefficient beta , in sharp contrast with the ordinary relation C= gamma T+ alpha T³. The linearly temperature-dependent coefficient is extremely high and well exceeds 20 mJ mol^-1 K^-2. It increases further with decreasing Zr content. The large coefficient gamma can be ascribed to the sum of electronic and magnetic contributions. The magnetic specific heat can be expressed in the form A₀+A₁T+A₃T³ in the temperature range studied.

Measurements of thermal expansion, specific heat and forced magnetostriction have been performed on several Y₉Co₇ samples in order to study magnetism and superconductivity in this compound by means of the magnetovolume relation. An analysis of the zero-field thermal expansion and the forced magnetostriction data is hampered by large sample-dependent magnetic shape effects. The excess specific heat in the magnetic region is analysed in terms of a small number of antiferromagnetically coupled local magnetic moments. High-pressure experiments on resistivity, AC susceptibility and magnetisation curves show a depression of magnetism and an enhancement of superconductivity in all samples studied so far.

By rapidly quenching from the melt, metastable single-phase solid solutions of Ag_1-xPt_x (0<or=x<or=1) were obtained. Magnetic susceptibility and nuclear magnetic resonance (NMR) measurements were performed between 4.2 and 300K and at 4.2K, respectively. The concentration-dependent Knight shift of ¹⁹⁵Pt corresponds qualitatively to the variation of susceptibility with composition. From both variations with concentration a d hyperfine field of -107+or-10 T for pure Pt was deduced. The variation of the ¹⁰⁹Ag Knight shift is similar to that of ¹⁹⁵Pt and for Pt-rich alloys is dominated by the contribution due to conduction-electron polarisation. From these results and from the variation of the spin lattice relaxation time T₁ of ¹⁰⁹Ag with concentration a rigid-band behaviour can clearly be discarded. The measurements are in accordance with and can be interpreted by the latest results of KKK CPA calculations on the isoelectronic systems AgPd and AuPt, whose NMR and susceptibility properties are very similar to those reported for AgPt.

The authors have measured the two-dimensional angular correlation of positron annihilation radiation from single crystals of aluminium containing voids and other defects. No asymmetry was found between the (110) and the (100) directions for angles less than 10 mrad. This is contrary to what one expects if positrons are trapped in extended surface states on the void walls, which are primarily (111) planes. A tentative hypothesis involving positronium states in voids is offered.

A new triggering model of the martensitic transformation is presented. In this model martensitic transformation is thought to be triggered by coherent phonons, the ordering phonons emitted by a simulated emission process of phonons called the potential emission of coherent phonons attained by quenching. The thermodynamic basis of the present model is discussed briefly and the transformation process involving the stimulated emission of phonons and the formation of martensite is studied using a three-dimensional wave equation, Van der Pol's equation and Duffing's equation. The results are used to explain the main features of the martensitic transformation: the 'military' nature of the martensitic transformation (cooperative atomic movements), habit plane, mode softening, transformation plasticity, the reverse transformation on heating, etc. The emission ratio of coherent phonons in the quench process is given and is shown to be proportional to the cooling rate. The analogy between martensitic transformation and laser threshold is stressed. The analogy with the Landau theory of phase transformations and the difference between it and the present model are pointed out.