Table of contents

The current theories of the anomalous properties and degree of clustering in nonideal binary liquid alloys are reviewed. Measurements of the densities of these alloys are reported, and the results of incorporating these data into a calculation of Delta H are discussed in relation to possible implications for the properties of the alloys.

Measurements are reported of the electrical resistivity and thermoelectric ratio of potassium samples encased in polyethylene or Teflon tubes, or in contact with Kel-F. The samples encased in polyethylene display Kondo-like anomalies below a few tenths of a kelvin in both the resistivity and the thermoelectric ratio. These anomalies are not present in the samples in Teflon tubes or in contact with Kel-F. The observed anomalous behaviour is consistent with some, but not all, of the unusual behaviour reported by previous observers for the resistivity in the vicinity of 1K of potassium samples in polyethylene.

The effect of the sample thickness d (7 mu m<or approximately=d<or approximately=142 mu m) on the electrical resistivity rho (d,T) of single-crystal copper whiskers with residual electron mean free path l_e=242 mu m has been studied in the temperature range 4.2-40K. The observed deviations from Matthiessen's rule in the temperature-dependent part of the resistivity, Delta _s(d,T) Delta = rho (d,T)- Delta rho ( infinity ,T), reveal a T² variation in a narrow temperature range, 10-20K. The experimental data are in quantitative agreement with the size-effect calculations of Sambles et al. (1984) if the surface roughness is regarded as an adjustable parameter which decreases with increasing sample thickness.

The Hall resistivities of two amorphous Ni_100-xP_x alloys (x=20 and 24) and an Ni₈₀P₁₄B₆ alloy have been measured at 77K and at room temperature. The normal Hall coefficient R₀ is independent of temperature (77-300K), negative and its absolute value slowly decreases with increasing P content. The decrease in R₀ is probably due to an increase in the number of conducting electrons. For the alloys with the highest P content (d band completely full) R₀ is due to nearly free s-like electrons and the electron number density deduced from R₀ can be combined with the plasmon frequency results to obtain the effective mass ratio.

Self-consistent, relativistic band calculations for cobalt are presented. These have been performed as a function of axial ratio. A minimum in total energy is found close to the experimental value of c/a. Both spin and orbital contributions to the magnetic moment have been calculated and they show excellent agreement with experiment. The role of the energy bands in determining the axial ratio and magnetic properties is discussed.

The incommensurate-commensurate phase boundary of the magnetic phase diagram of a dilute Cr-Al system has been determined from measurements of the concentration dependence of the electrical resistivity at different fixed temperatures. The conflict existing in the literature between electrical resistivity and magnetic susceptibility measurements on the one hand and magnetoelastic measurements on the other is resolved by the present work. A triple point where the paramagnetic and the incommensurate and commensurate spin-density-wave states coexist is suggested to occur at about 2 at.% Al and 100K.

The electronic structure of d electrons in the cubic Laves phase of the intermetallic compound YCo₂ under an external magnetic field is calculated in the tight-binding approximation. The magnetic-field dependence of the induced moment is obtained at 0K and it is shown that the metamagnetic transition from the paramagnetic to ferromagnetic state occurs at about 90 T. It is also shown that the different hybridisations between the 3d band of Co and the 4d band of Y or the 5d band of R (where R is a rare earth) with majority and minority spins play an important role in the appearance of the local moment on Co in the magnetised state of the compounds YCo₂ or RCo₂.

A method which is based on the random phase approximation (RPA) and Monte Carlo (MC) computer simulation is applied to liquid Mg, Zn, Ca, Sr, Ba, Al, In and Pb to extract the effective pair potential nu _eff(r) using the experimental structure factor a(k) near the melting point. For Zn, Al, In and Pb the nu _eff(r) thus obtained have a markedly raised principal minimum, while for Mg, Ca, Sr and Ba they display a conventional attractive well at the first-neighbour distance. They present results, though not fully quantitative, give good support to the recent analysis by Jacucci and Taylor (1981) in which one expects the interatomic potential to be repulsive at the first-neighbour distance for simple metals with a high density of conduction electrons.

The temperature dependence of the resistivity of Fe_1-xB_x alloys implanted at room temperature is measured at different concentrations between x=0.18 and 0.31, and is compared with that of sputtered and quenched alloys. It is found that the implanted alloys are amorphous as soon as the eutectic composition is reached. The resistivity results obtained after irradiation at 6K with different ions of the Fe_0.76B_0.24 implanted film are presented in a companion paper (II).

For pt.I see ibid., vol.15, p.1229 (1985). The electrical properties of amorphous Fe_1-xB_x alloys produced by ion implantation at room temperature have been studied in a companion paper (I). This paper reports the results of 6K irradiation (with 100 keV ¹H, 540 keV ⁸⁶Kr and ²⁰⁹Bi ions) and subsequent annealing of the Fe_0.76B_0.24 alloy. The electrical resistivity measurements are analysed in terms of irradiation-induced 'point defects' and are compared with those reported previously for Fe-B alloys quenched from the melt. Both the spontaneous recombination volume ( nu ₀=29+or-4 atomic volumes) and the intrinsic resistivity of induced defects ( rho _d=133+or-20 mu Omega cm) are found to be lower than those measured in Fe-B alloys quenched from the melt. The displacement cascade effects induced by heavy-ion irradiation are also studied: the irradiation efficiency is found to be lower for ⁸⁶Kr irradiation than for ¹H irradiation, as in crystalline metals. However, in contrast to the case of crystalline metals, the increase in saturation resistivity is the same whether or not displacement cascades occur during irradiation.

Measurements are reported of the elastic energy dissipation coefficient, Q^-1, following absorption of hydrogen into a series of palladium-silver alloys, which initially had been in both pre-annealed and mechanically deformed states. The measurements have enabled assessments to be made of the extents to which peaks in plots of Q^-1 against temperature over the range 85-300K can be attributed either to the hydrogen-dislocation interaction or to relaxation effects associated with reordering of hydrogen interstitials and vacant interstitial sites in the bulk lattice.

For pt.I see ibid., vol.15, p.1249 (1985). The elastic energy dissipation coefficient, Q^-1, has been measured over the temperature range 85-300K at vibration frequencies of about 3 kHz with Pd-Ag alloys containing hydrogen. In hydrogen-loaded pre-annealed specimens two relaxations, H₍₁₎ and H₍₂₎, have been observed, which are Zener effects due to stress-induced movements of hydrogen interstitials within regions of high and low hydrogen concentration, respectively. The evolution of the activation energy, the limiting relaxation time, and the relaxation time spectra and strengths have been investigated with increasing the hydrogen content. The degree of short-range order existing in the binary Pd-Ag alloys is believed to play an essential role in determining the main features of the two relaxations.

The concept of an elastic strain corresponding with the state of stress produced in thin-film structures containing one-dimensional composition modulation is considered as an explanation for the increased elastic moduli observed experimentally. The extent of modulus enhancement is predicted successfully using a pseudopotential energy calculation approach.

A model Hamiltonian has been formulated for Hg_{3- delta} AsF₆ embedded in the I4₁/amd space group based on eight Wannier functions constructed from gaussians. The authors centre the gaussian orbitals, and thus the Wannier functions, on special sites, which do not coincide with the mercury atoms but are related to the Wyckoff positions for the embedding I4₁/amd space group. Near the Fermi energy the model Hamiltonian, an 8*8 matrix, simulates accurately the bands resulting from an ASW calculation which used 160 basis functions.

The electronic structure of substoichiometric TiN_x is studied in terms of the KKR GF and KKR CPA methods. In accordance with recent XPS studies the cases of both a single vacancy and concentrated vacancies show peaks in the vicinity of the minimum in the density of states of the stoichiometric compound. The nature of the 'vacancy sub-band' is analysed in detail in terms of the resonance condition for virtual bound states. The Fermi energy for TiN_x, x=0.75, is lower than in the stoichiometric case, the density of states at E_F being about 10% higher in the non-stoichiometric system than in the pure system. It is argued that the results obtained are also characteristic of other substoichiometric refractory carbides and nitrides.

The authors report photoemission measurements of the core levels and valence bands of the metallic glasses Fe₈₂B₁₈ and Fe_80-xNi_xB₂₀ (for x=10, 30 and 40). In agreement with previous experiments they observe a broadening of the full width at half maximum of the Fe₈₂B₁₈ valence-band spectra in comparison with those of pure iron. Interaction between boron valence electrons and iron d electrons, which is presumably responsible for the broadening, involves shifts of iron and nickel d bands in the valence band of Ne-Ni-B alloys. Consequently, the respective contributions of iron and nickel are visible in the valence-band spectra. The shift of the boron 1s level to higher binding energies for high nickel content indicates that the chemical bonding interaction is stronger between boron and nickel than between boron and iron.

The resistivity of lithium and three LiMg alloys with magnesium concentrations of 1.0, 5.0 and 10 at.% has been measured from 2 to 60K. Below 10K a term proportional to T² and to the residual resistivity is identified and associated with inelastic impurity scattering of electrons. At higher temperatures the deviations from Matthiessen's rule are found to become essentially independent of concentration.

On the basis of a spin-polarised density functional description of the electrons, the authors develop a 'mean-field' theory of magnetic phase transitions in metals. The one-electron-like finite-temperature Schrodinger equation is solved, formally, for random orientations of local moments and the corresponding grand potential is used in a statistical mechanics of the spin configurations. This latter, in the mean-field approximation, requires the knowledge of the electronic grand potential averaged over various ensembles of such 'spin' configurations. These averages are carried out with the help of the Korringa-Kohn-Rostoker coherent-potential-approximation (KKR CPA) method for dealing with electrons in random potential fields. Then, the whole procedure is made self-consistent on the average. The theory determines the local moment mu , the Curie temperature T_C, and the susceptibility chi (q,T) in addition to the electronic structure at finite temperatures without adjustable parameters. The authors illustrate the explicit calculations for iron. The local moment is found to be 1.9 mu _B above T_C, and the preliminary estimate of T_c is 1250K.

The authors describe the electronic structure of Fe and Ni above their Curie temperatures in their disordered local moment (DLM) states as a function of wavevector k and energy epsilon . In particular, they calculate the Bloch spectral function, A(k, epsilon ), averaged over the orientational configurations of the local moments, at selected points in the Brillouin zone and determine the shape and smearing of the 'Fermi surface'. They find that BCC Fe, with a local moment of 1.9 mu _B, can show an exchange splitting at some points whilst in other regions of the Brillouin zone no such splitting occurs. For comparison they also study FCC Fe, which also supports a substantial local moment. They find that it has similar features but the smearing of the 'bands' is more pronounced. On the other hand, the electronic structure of Ni is quite different, shows no such local exchange splitting, but is able to support a small local moment of 0.2 mu _B. The resulting picture of the electronic structure of Ni is that of a paramagnetic smeared 'Stoner-Wohlfarth' model.

Alloys of Fe_1-xAg_x obtained by vapour quenching have been investigated by X-ray diffraction and magnetisation measurements. The single BCC phase extends up to about x=0.14, while the single FCC phase exists above x=0.6. For x=0.14-0.6, there is a mixed BCC and FCC phase. The average magnetic moment decreases with increasing x. It deviates upwards from the simple dilution law in the BCC and FCC alloys with x=0.6-0.7. The Curie temperature, T_C, is about 900K for the BCC alloy with x=0.1, while T_C is about 550K for the FCC alloy with x=0.6 and decreases with increasing x to zero at about x=0.95.

An analysis of the isomer shift (IS) in intermetallic compounds with Gd is given in terms of the cellular model of Miedema and van der Woude (1980). Deviations from the predicted linear behaviour of the IS as a function of the coordination fraction are attributed to volume contractions upon compound formation. The volume dependence of the IS, as extracted from the analysis, is found to be in agreement with a value obtained from other experimental data. The Gd compounds were investigated with the 86 keV transition of ¹⁵⁵Gd.

Doppler-broadened lineshapes of positron annihilation in deformed aluminium samples were measured in terms of the peak parameter S in the temperature range between 4.2 and 273K. A remarkable increase in S was seen at temperatures below about 100K as the temperature decreased. The results were interpreted in terms of positron trapping in a dislocation as a shallow line trap accompanied by point-like defects (e.g., jogs) as a deep trap. It was found that the binding energy between a positron and a dislocation line is about 0.005 eV.

Based on the work of M. Kaveh et al. (see ibid., vol.13, p.953, 1983) and Dugdale et al. (see Phys. Rev., vol.157, p.552, 1967). Recently Kaveh and Wiser proposed that the apparently high values of the specific dislocation resistivity in bent copper crystals may be due to a deviation from Matthiessen's rule associated with the relatively high impurity content. It is pointed out in this Comment that if a small error in their work on the dislocation-impurity system is corrected, it leads to a result which is identical in form to that predicted by the earlier theory of Dugdale and Basinski on the impurity-phonon and dislocation-phonon systems, despite the apparently quite different mathematical procedures used.