Table of contents

Recent inelastic neutron scattering data in Cu (Larose and Brockhouse 1976) show an unexplained positive dispersion effect around the reduced wavevector xi =0.2 in the T₁A branch with a sound velocity proportional to (c₁₁-c₁₂)¹2/. This effect is explained in terms of a zero to first sound transition. The zero and first sound elastic constants, as well as the mean inverse lifetimes of the thermal phonons are derived for five temperatures from room temperature up to the melting point. Neutron scattering data in Ag and Au are also considered under these aspects.

Low-temperature specific heats have been measured between 1.5 and 4.2K in a Pd-H alpha phase sample of H/Pd ratio 0.005 in two different conditions: after deformation at room temperature, and after deformation at liquid nitrogen temperature and a short ageing at room temperature. In the latter case the value of the electronic specific heat coefficient is close to that of pure Pd, suggesting the formation of beta hydride.

The effect of including a finite ionic spin distribution on the preasymptotic form of the RKKY exchange interaction is investigated. For a simple model it is found that the short-range interactions in some Heusler alloys are all positive, in agreement with experiment.

Spin echo, X-ray and magnetization measurements are performed on Fe₃Si_1-xAl_x alloys for 0<x<0.25. The results are analysed by considering near neighbour environments for various sites.

The contention of Collins and Andrews (see ibid., vol.7, p.L43 (1976)) that the narrow peak in the soft X-ray emission spectrum of an 8% Fe-Al alloy is due to core L₁-L_2,3 transitions is refuted by the experimental observation that a less prominent peak is found for Fe₃Al. The arguments in favour of a localized state around an Al impurity in Fe are summarized.

The double ionization satellite of the MgL_2.3 to VV Auger band from solid Mg has been observed to be one third as intense as the main band when Al K alpha radiation is used for excitation. Arguments are presented to show that ionizations in the Mg K shell are responsible for this very high ratio. The arguments are supported by results obtained using Mg K alpha radiation with which no Mg K shell ionization is possible and the satellite is found to be much weaker. Simple calculations may be used to give qualitative estimates of these effects.

Higher-order partial structure factors are defined as the Fourier transforms of a class of correlation functions of the Ursell type. These partial structure factors are a generalization of those defined by Ballentine and Lakshmi (1975), but differ from those of Parrinello and Tosi (1975). The functional differentiation approach used in introducing the correlation functions allows the definition of their inverses, which are shown to be a class of direct correlation functions. Moreover, it is possible to construct a hierarchy of equations between partial structure factors of different orders which may be expressed in terms of either the number of particles of different species or the number-concentration formalism. The long-wavelength limit for the partial structure factors is shown to be related to well-defined thermodynamic properties of the system. Numerical results are presented for liquid alloys.

The electron density profiles, potentials and charge overlap forces between two like pieces of jellium have been computed as functions of the separation. Results are presented for the range of bulk metallic electron densities. Lattice effects are included via perturbation theory, and the total force acting between the halves is shown for Al (111) faces. An efficient scheme of iteration for self-consistent density functional calculations is described in the appendix.

A simple analytical model muffin-tin potential (MMTP) is introduced with three adjustable parameters. The energy-dependent logarithmic derivatives and scattering phase shifts required for bandstructure calculations are calculated as functions of these parameters. The s, p, and d levels of a transition metal bandstructure are strongly correlated with the corresponding parameter of the MMTP thereby providing a simple way to calculate the energy levels from the MMTP with the three parameters or, vice versa, the parameters from the energy band. This procedure is carried out in detail for copper, nickel, and iron.

Presents an extension of the theoretical model of Harris et al. (1974) to describe the magnetic properties of DyCo_3.4 alloys. The d band is treated in terms of a Hubbard Hamiltonian and the Dy moments, which are coupled with the d band through a local exchange interaction, are subject to a random anisotropy. The Hamiltonian is treated in the mean field approximation. A fit of the magnetization against temperature curve of DyCo_3.4 is presented and the magnetization curves of DyNi_3.4 and DyFe_3.4 are discussed in the light of the same theoretical model.

The Hubbard model is investigated within the Hartree-Fock approximation for the linear chain where nearest and second-nearest-neighbour hopping is taken into account.

The order-dependent electrical resistivity, including the effects of static atomic displacements as might arise from vacancies, interstitials and dislocations and also thermally induced displacements is calculated. It is found that the structure factor for the static terms is modified by a term (1-q²( delta ²)+q²( delta _l)) where ( delta ²) are the mean square thermal displacements and ( delta _l) describes the correlated motion of atoms. In the systems studied (Cu₃Au and Al-Zn) an increase in the measuring temperature results in a definite though small decrease in the 'residual' resistivity through the coupling between the static and phonon terms.

The electrical resistivity and thermoelectric power of the liquid 3d transition metals (Ti, V, Cr, Mn, Fe, Co and Ni), Pd, Pt, Zr, Cu, Ag, Ge and Sn have been calculated using the t matrix approach and experimental structural data. Also the electrical resistivity of liquid alloys Sn-Ag, Sn-Co, Mn-Ge, Co-Ge and Mn-Cu, have been estimated in the same formulation. All the results are in good agreement with available experimental data showing that the electron transport properties of liquid transition metals and their alloys can be understood in terms of the t matrix approach.

Measurements of the absorption of longitudinal sound in aluminium at a microwave frequency of 35 GHz are reported. The phase problems associated with the very short sound wavelength have been partly overcome by using a semiconductor avalanche bolometer which detects scattered phonons rather than unscattered coherent ultrasound. A simple theory is developed to relate the bolometer signal to the ultrasonic attenuation coefficient and to its variation with temperature. The results for the pair breaking region of the superconducting state would appear to be in agreement with the BCS Bobetic theory, and the broadening of the transition to the region where quasi-particle scattering dominates is interpreted in terms of energy gap anisotropy.

The dipole-dipole interaction provides a mechanism for anisotropy energy affecting the rotation of magnetic clusters in a spin glass. The authors have calculated the distribution of the lowest energy barrier Delta between the two stable directions of the net magnetic moment of a cluster of n spins in the limit of large n. In the dilute or spin glass limit, i.e. for npz₁<<1<<n where p is concentration and z₁ is the number of nearest neighbours (NN) on the lattice, the distribution is a universal function of Delta /np and independent of the crystal lattice. For npz₁>>1, Delta is dominated by nearest-neighbour contributions and is normally distributed about a mean value Delta _n=(2npz₁)^1/2 sigma with a standard deviation sigma approximately= dipolar energy between two NN spins. Thus as n to infinity , the quantity Delta /n^1/2 tends to a fixed value. Within the framework of the magnetic cluster model, the above results are used to calculate the susceptibility measured in an AC magnetic field near the glass temperature T_g.

The paramagnetic resonance of Fe in hydrogenated and unhydrogenated Pd has been observed. A suitable explanation for the g-factor is found if one assumes the ionic ground state to be 3+ rather than 2+. Measurements of the demagnetization fields in the paramagnetic regime suggest that giant moments are present in the demagnetization field.

Complete expressions for elements of the paramagnetostriction tensor, including contributions from both the orbital motion and the spin of electrons, are derived for metals with an HCP structure. Anisotropic orbital contributions to the paramagnetostriction and isotropic ones due to spin are obtained. Changes due to the strain of both the orbital magnetic susceptibility tensor and spin paramagnetic susceptibility are anisotropic. If the linear compressibilities parallel and perpendicular to the c axis of the HCP metal are equal, the logarithmic volume dependence of the total magnetic susceptibility of a polycrystalline sample is expressed by the volume paramagnetostriction of the sample. The expansion coefficients of the total magnetic susceptibility with respect to strain are related to the elements of the paramagnetostriction tensor which can be directly measured.

A model is developed in which the moment on a nickel atom is a function of atomic environment through its susceptibility and of magnetic environment through exchange interactions. Solution of this model leads to a fit of change of moment with concentration and, with the assumption of first-neighbour interactions, a further parameter-free fit to both polarized and unpolarized diffuse neutron scattering data. A detailed analysis of the Ni-Cu and Ni-Cr system is presented.

The charge and spin density distributions in antiferromagnetic Cr-Al solid solutions have been studied by means of X-ray and neutron diffraction experiments on powdered samples. The results show that the valence electrons of aluminium in the alloys must be localized so that they behave rather like 3d electrons having T_2g type symmetry. The departure from spherical symmetry of the spin distribution around the magnetic atoms is not large, but shows a tendency to become more E_g like with increasing aluminium content. These results suggest that the increase in magnetic moment with increasing aluminium content cannot be simply due to aluminium electrons filling the chromium d band. It is suggested that the change in magnetic properties is mainly due to a decrease in the chromium bandwidth relative to the exchange energy as the amount of d shell overlap is reduced by adding aluminium.

Nuclear quadrupolar couplings measured by the field cycling technique are reported for the first neighbouring shells of various impurities in aluminium (Mg, Si, Zn, Ga, Ge, Ag, Cd, In, Sn, Au, Cu). The quadrupolar solid effect lines due to dipolar couplings between nearby nuclei have been used for identification of the shells. The results indicate a strong anisotropy of the screening charge density with a minimum along the (100) directions. The distortions of the Fermi surface appearing near its intersections with the first Brillouin zone boundaries may explain the observed anisotropy.

Using the field cycling technique, nuclear quadrupole couplings are measured on the first neighbouring shells of titanium, vanadium and iron in aluminium. The results confirm previous measurements in AlSc, AlMn and AlCr: they show a progressive attenuation of the quadrupolar couplings on the nuclei close to the impurity along the 3d series.

The ⁶³Cu NMR on CuNi alloys in the critical concentration range for the transition to the paramagnetic state (45 at.%<or=x<or=54 at.%) has been studied between 4.2K and 300K by the spin-echo technique. The magnetic properties of such alloys are known to be strongly dependent on local environment effects and chemical clustering. Two types of CuNi alloys which are characterized by a different degree of chemical homogeneity have been examined. The observed inhomogeneously broadened ⁶³Cu line exhibits features which are characteristic of the magnetic polarization due to Ni atoms in different magnetic states (magnetic and nearly magnetic) resulting from local environment effects. From the analysis of the NMR parameters a model is proposed to interpret the magnetic polarization of the observed Cu atoms showing the fundamental role of nearly magnetic atoms in the vicinity of these Cu sites. At the same time it is shown that chemical inhomogeneities resulting from clustering of Cu atoms strongly reduce the magnetic polarization of these atoms.

The sign and magnitude of the magnetic hyperfine fields at Au and Sb sites in the ferromagnetic Heusler alloy AuMnSb have been measured, at 4.2K, by means of the Mossbauer effect. The values found were -240+or-20 kOe and +410+or-10 kOe for the gold and antimony fields respectively. The experimental results have been compared with the predictions of three theoretical models for hyperfine fields at nonmagnetic sites in the Heusler alloys.

The photoemission spectra of the d bands and of the Ni 2p_3/2 and Pt 4f core levels of Ni-Pt alloys have been measured with monochromatized X-rays. Within the 0.6 eV experimental resolution, the d band photoemission spectra appear to be described well by linear superpositions of the densities of states of the pure elements. The core level asymmetries of both Pt and Ni atoms change by no more than 20% with alloying, suggesting that they are not very sensitive to the magnetic state of the emitting atom nor to its local density of states near E_f, as determined by specific heat and susceptibility measurements. These results suggest that the core level asymmetries depend upon the local number of d holes, rather than the local density of states at the Fermi level. The core level asymmetries might provide a means to determine the number of d holes of certain atoms in alloys and compounds. It is suggested that the shape of the Ni 2p_3/2 core level spectra might be determined by the occupied Ni 3d bands.