Table of contents

The Dingle temperature (describing the scattering strength in a magnetic field) of edge dislocations in metals is calculated in the free-electron model by introducing an outer cut off radius equal to the cyclotron and orbit radius. No limiting scattering angle has to be considered. A more general approach involving magnetic-field-dependent wave-functions leads directly to a field-dependent Dingle temperature.

The phonon density of states for LaNi₅H₆ and LaNi₅D₆ has been measured by neutron inelastic scattering on a powdered sample. Within the host lattice frequencies an additional excitation has been observed at h(cross) omega =12 meV; this energy is independent of the mass of the light atom.

The structure factor of liquid nickel at 1873K was measured by neutron diffraction using an orientated single-crystalline sapphire sample container. The effective spin quantum number has been found to be 0<or=S<or=0.3. The result is compared with the PY hard-core structure factor and the agreement is improved by an extension corresponding to an attractive part in the pair potential.

The calculation of the impurity potential and its accuracy is discussed. It is shown how dilatation effects can be incorporated into the self-consistent method of calculation. It is argued that his method gives good impurity potentials in the noble metals: results of calculations for various impurities in copper are presented and are in reasonable agreement with experiment. It is shown that a muffin-tin potential for heterovalent impurities distorts the potentials and gives phase shifts which can differ from those of a self-consistent potential by significant amounts. Such inaccuracies can lead to incorrect results for calculations of physical effects like spin-flip relaxation times, etc., and this can lead to a misunderstanding of the physical situation.

The electric field gradients (EFGs) created at the nucleus of ¹¹¹Cd by Pd impurity atoms in the cubic Cu host are studied by the time differential perturbed angular correlation technique. The observed strong enhancement of near-neighbour (NN) impurity populations to the probe atom permitted the study of the onset of various NN configurations as the impurity concentration was increased from 0.5 to 4.5 at.%. The EFG attributed to one NN Pd impurity agrees very well with NMR data for the same alloy. The temperature dependence of this EFG follows a T^3/2 law in the range 80-973K. A measurement at 1173K indicates the presence of dynamical effects on the nuclear quadrupole interaction.

A systematic study has been made of the effects of Group III and rare earth impurities on the thermal oxidation of chromium. The elements were introduced by ion implantation. The oxidation properties and impurity behaviour were studied using alpha particle Rutherford backscattering, nuclear reaction analysis, CAMECA ion microprobe analysis, X-ray diffraction and scanning electron microscopy. Cr⁺ ion implantation was used to study radiation damage and annealing behaviour. For unannealed chromium the oxidation rate was found to be reduced by 60-90% by all the implanted species. Some evidence was found for a correlation with the mobility of the implanted species, X-ray measurements indicate the possible formation of a binary oxide.

Results of susceptibility and specific heat measurements on Ce_1-xTh_x alloys with 0.2<or=x<or=1 are presented. Susceptibilities were measured at temperatures between 1.6K and 1100K and specific heats between 1.3K and 20K. At the highest temperatures the cerium atoms are in the trivalent state for all compositions, but at low temperatures the magnetic moments are lost. The Stoner enhancement factor is found to be nearly independent of composition, although the density of states per cerium atom increases by about a factor of two with increasing x.

A canted structure has been previously discovered in GdMg equiatomic compounds. The authors study the specific heat of this compound and that of isomorphous YMg. The magnetic contribution in GdMg presents two anomalies at the Curie temperature and at the ferromagnetic canted structure transition. They discuss the equilibrium of the canted structure within different models involving high-order interactions, or a change of interactions in the ordered range. The theoretical predictions are compared to the actual data for the temperature variation of the magnetisation, canting angle and specific heat. The most probable origin of the canted structure seems to be the nonlinear behaviour of the conduction band magnetisation.

The general problem of anomalous diffusion is discussed in connection with the way in which the TDPAC (time differential perturbed angular correlation) technique may help determine the relative fraction of substitutional or interstitial positions of the diffusing atoms. The TDPAC technique referred to the Ag-In system is outlined. The present measurements give a fraction of 0.89+or-0.21 of Ag solutes in In substitutional positions. This suggests a dissociative mechanism for Ag diffusion in In. The value of 0.6+or-0.5 is obtained for the electric field gradient asymmetry parameter eta ₂ of the assumed interstitial positions. This value shows a low symmetry for the interstitial sites, indicating that these positions might be split interstitials of Ag solutes in In.

Bandstructure calculations have been carried out for rhodium for the normal value of the lattice constant and one 5% larger, using Hubbard's method (1969). The results of the first calculation compare well with those of other workers and with X-ray photoemission measurements, while comparison with specific heat measurements suggest an electron-phonon interaction enhancement parameter lambda of 0.48. The approximate theory of Ratti and co-workers (1974) gives a much lower value of lambda =0.22 and comparison with experiment of the corresponding values predicted for the superconducting transition temperature T_c fails to decide between the two values. From the two band calculations values are calculated for dT_c/dV and it is predicted qualitatively that T_c will increase under pressure. With regard to the magnetic character of the groundstate it is confirmed that this will not be ferromagnetic according to the Stoner criterion, even in the expanded state.

The magnetic field dependence in the amplitude of de Haas-van Alphen or magnetoresistance oscillations introduces magnetic field dependent phase shifts which are not included in the standard formation of the Lifshitz-Kosevich theory. It is found that the magnetic field dependence of the oscillatory amplitude introduces an additional signal which is pi /2 out of phase in 1/H with respect to the signal which would be observed if the amplitude of the oscillation is strictly independent of magnetic field strength. A number of examples are presented which illustrate the magnitude of this magnetic field dependent amplitude effect when the large amplitude modulation field technique is employed to measure the frequency of the quantum oscillations.

The L_2,3 soft X-ray appearance potential spectra (SXAPS) of the 3d transition metals have been described as scattering of both the incident electron and the L shell electron into the conduction band (empty) state above the Fermi level (E_F) of these metals. The author uses the part of the conduction band that derives from the d localised states to present a parameter-dependence model to describe the appearance potential spectra of these materials (Sc-Ni). Choosing the appropriate parameters for Mn, a model calculation has been performed which produces the asymmetry apparent in the total X- ray yield of Mn.

The spherical solid model and the spin density functional formalism have been applied to calculate the screening of a positive point charge at different lattice sites in Al, Na and Cu. Results are obtained for the Knight shift, the electric field gradient, the heat of solution and the diffusion barrier. It is found essential to use the spin-polarised form to evaluate the Knight shift, especially at low metallic densities and for impurities with a high nuclear charge. Both the Knight shift and the electric field gradient are found to be markedly different for substitutional and interstitial positions. The calculated heat of solution of hydrogen is lowest for the octahedral position in FCC Al and for the tetrahedral position in BCC Na, indicating that no hydrogen trapping at vacancies occurs in these metals.

The screening effects of conduction electrons on the crystal field coefficients in metals are investigated and compared with an elementary (unscreened) point-charge model. The size of the interaction remains of the same order of magnitude but may be reversed in sign, both effects being in accordance with observations on metallic materials containing rare earth atoms.

The thermopower of Au-13 at.%Ag has been measured from 4.2 to 300K. Its near-zero value over this range and its independence of magnetic field make it a useful material for a thermopower standard.

The influence of thermal expansion of metals on the temperature dependence of the work function is considered. Both the temperature coefficient of the jellium work function and lattice corrections to the work function are calculated for (110), (100) and (111) planes of Al, and Pb. The calculation gives a negative temperature coefficient of the jellium work function for all metals and different signs of the total work function coefficient depending on the value of the core radius of the Ashcroft model potential and also on the kind of plane.

Lowde and Windsor (1970) have pointed out that good agreement between neutron scattering experiments on Ni and theory is possible if the effective Coulomb interaction is a mildly decreasing function of the momentum. The author presents a variational calculation of the wave vector-dependent susceptibility in the t matrix scheme for the short-ranged one-band model. The scheme satisfies the Ward-Takahashi identities for self-energy and the vertex part. Numerical results are presented assuming an idealised bandstructure for Ni. The computation yields results of the type suggested by experiment. The authors point out the reason why other theories give results which contradict experiment.

Measurements of the field dependence of susceptibility of a Cu-10 at.% Al-1 at.% Mn spin glass using a double AC technique are reported. The critical exponent for the field dependence of susceptibility delta chi / delta H is estimated to be -0.3+or-0.1. Some relaxation effects are also observed.

Neutron diffraction measurements on manganese-rich alloys are presented which provide information within the areas of magnetism and metallurgy. Three aspects of the alloy behaviour are considered in the present paper in separate but related sections. In the first of these, the decomposition of the metastable gamma phase is discussed. A detailed analysis of the diffuse scattering from a gamma Mn-Ni alloy, for which the diffraction pattern shows the characteristic cusp-like peak, is made. Diffraction measurements are also presented for Mn-Pd alloys. In the second section the contributions to the diffuse scattering from gamma Mn alloys which arise from short-range order are discussed. The temperature variation of the scattering is described in relation to the well defined temperature variation of order parameters for Cu₃Au. In the third section, the variation of magnetic moment with composition, for gamma Mn alloys as derived from neutron diffraction measurements at low temperature is presented.

The Co and Tb contributions to the magnetic density of TbCo₂ have been separated by means of diffraction experiments with polarised neutrons performed at various temperatures. The 4f Tb form factor measured at 200K is in good agreement with that calculated using the tensor operator method. As in other rare earth (R)-Co₂ compounds previously studied, the Co magnetic behaviour arises from a metamagnetic transition on the 3d band with the total field acting on it. The occurrence of such a transition is correlated to a maximum in the thermal variation of the susceptibility. In the RCo₂ compounds, if the 3d band susceptibility at the Curie temperature is below the maximum value, the metamagnetic transition is then observed. The transition is of first order; it is the case of HoCo₂. If the susceptibility at the Curie temperature is close to the maximum value, the Co moment is induced without any discontinuity. The transition is of second order; it is the case of TbCo₂.

Amorphous ErCo₂ exhibits at low temperature a sperimagnetic structure: the cobalt moments are parallel, while the erbium moments are opposed but strongly connected to their local easy-magnetisation axes. The present study with polarised neutrons and polarisation analysis shows that there exists a correlation between the erbium magnetic components perpendicular to the cobalt magnetisation. Their arrangement is ferro- or antiferromagnetic, depending on the angle between the cobalt moments and the easy axes. There is evidence for correlation between the local easy-magnetisation axes which are not distributed at random. As a result, the existence of magnetic subdomains may be deduced. The correlation between erbium moments vanishes when the temperature increases and the anisotropy disappears.

For pt.I see ibid., vol.9, no.7, p.1421 (1979). At low temperature the sperimagnetic amorphous alloy ErCo₂ gives rise to a very large small-angle neutron scattering of magnetic origin. Evidence has been given for a crossover in the diffusion pattern, with two different regimes: a k^-2 law below the crossover and a k^-3 law above it. These two regimes are explained by the nature of the magnetic correlations between the erbium moments in such a sperimagnetic alloy.

The temperature dependence of the low field AC susceptibility of a number of PtFe alloys has been measured. For ferromagnetic alloys the Curie temperature varies linearly with the solute concentration (up to approximately 15% Fe) and extrapolates to zero at a critical concentration of 0.76% Fe. Below this concentration a spin glass regime probably exists. The susceptibility-temperature curves for the weakly ferromagnetic alloys (1-4% Fe) are similar to those hitherto commonly associated only with spin glasses, while for the most concentrated alloys investigated (15 and 17% Fe) the curves bear some resemblance to those of AuFe alloys of similar compositions.

The zero temperature probability distribution P(h) of internal magnetic fields is studied, both in the Sherrington-Kirkpatrick random Ising model of a spin glass and in its natural extension to classical vector spins. Theoretical predictions and computer simulations agree that P(h) is linear for small h in the Ising case, and has a hole-P(h)=0 for h< eta -in the vector spin case.

The incoherent phase transitions of H and D in Nb have been studied by X-ray scattering. For all investigations Nb was loaded with hydrogen in situ in order to prevent precipitation of the new phase, before starting the experiment, which irreversibly changes the crystal lattice and gives misleading results. The incoherent phase boundary was determined from the variation of the lattice parameter with hydrogen concentration. The lattice parameter varies linearly with concentration at least up to 0.34 H/Nb. The thermal expansion coefficient increases with hydrogen concentration and shows an anomalous behaviour just above the alpha - alpha ' phase transition. For the alpha - alpha ' transition of H and D in Nb the critical temperature is T_c=171+or-1 degrees C and the critical concentration c_c=0.31+or-0.01 H/Nb. The triple point temperature of the alpha + beta to alpha + alpha ' transition is 88 degrees C for H and 99 degrees C for D in Nb.