A theoretical analysis, based on Kubo's formalism of current correlation functions, is made for the ferromagnetic Hall effect in the case of transport limited by electron-phonon scattering. The antisymmetric, off-diagonal conductivity is, to first order in the magnetization, found to be of order zero in the electronphonon interaction (assumed to be weak) and, to this order, is equivalent to Karplus and Luttinger's results. Several corrections are analyzed and found to be of higher order in the scattering potential. The analysis is carried through for monocrystalline iron by using Wood's dispersion curves. It is found that R s = +0.55 X10 -12 £} cm/G at r=293°K, which has the correct sign and is smaller, by a factor of roughly J, than Dheer's experimental values for iron whiskers. Our temperature dependence of R s is almost entirely as p xx 2 , which fails to explain the temperature dependence of R s below 75°K in iron whiskers.

A systematic investigation is made of the consequences of the pseudopotential theory of metals on the thermodynamic properties of metallic liquids, including mixtures. We obtain expressions for the internal energy, the Gibbs free energy, and excess chemical potentials of general metallic mixtures with any number of constituents. We recommend on this basis optimum pseudopotentials for all liquid alkali metals. We obtain good agreement for the excess chemical potental of lithium in selected alloys. Finally, on the basis of metallic theory, we obtain a new scaling law for the critical constants of liquid metals which seems to be well satisfied by the metals of the groups I and IIIa, except perhaps for lithium and thallium.

No abstract

The structure of the liquid alkali metals is obtained by an optimized iterative solution of the Percus-Yevick theory and by using recent pseudopotential models and screening functions for these metals. This solution involves an optimized expression for the correction function to g(r), knowing one guess and one calculated result for g(r). This is obtained by setting a weighted residual corresponding to the Percus-Yevick equation equal to zero. The residual is evaluated to first-order in a change in this correction function. This is usually a good approximation after the first few iterations where a more restrictive expression for the correction function is proposed. Convergence of this optimum iteration procedure to an error of 0.1% in the norm of the correction function is accomplished within 100 iterations. Complete results are presented for the five liquid alkali metals near their melting point. The method gives a successful qualitative check of the effective Ashcroft-Singwi potentials used (initially obtained from a best fit to the force constants in the solid alkalis), particularly in K, Rb, and Cs.