ELECTRONIC PROPERTIES OF NOBLE METALS, THEIR ALLOYS, AND FERROMAGNETIC IRON

Abstract

The work reported in this thesis can be divided into two categories (i) electronic properties of noble metals and their alloys, and (ii) Fermi surface (FS) and optical properties of iron. In the case of noble metals we have used the eight-cone model of Ziman to determine the FS of these metals. we find that this model with the local pseudopotential fails to give the correct FS anisotropy. Responding to the need of improvement in the model so that it may give the correct FS anisotropy we include a non-local term in the pseudopotential. This improves the FS considerably. We try to see if this FS model can give good results for pressure derivatives of FS. We obtain a reasonably good • agreement with the experimental results. Having obtained a good model for FS of noble metals, we try to see if this model is good enough to be used for evaluation of transport properties, as has been done by some of the workers of this field. Thus we determined Fermi velocities of elec-trons. We see that the eight-cone model does not give good results for the trends in the variation of Fermi velocities, either with the local or with the non-local pseudopotential. Hence we conclude that this model is reasonably good for the static properties, but not for the dynamics. Our earlier work on the noble metals has demonstrat-ed that using the interpolation scheme, it is possible to obtain a good fit to the FS as well as the optical data. In this thesis we extend our earlier work to the noble metal alloys, in particular to the Ag-Au system. One reason for choosing this is that excellent optical data as well as FS data is available. The aim here is to see whether one could be able to fit both these sets of data with the same band structure. In this way we hope to be able to see which parameters change on alloying. The second part of the thesis is devoted to iron, a transition metal which shows ferromagnetism. One reason for choosing iron is that in our earlier work on nickel we showed that a constant exchange splitting is sufficient to explain the FS data. We would like to see if the same could be said about iron. Unlike nickel, the FS of iron has been studied more in detail. Even so the FS geometry of ferromagnetic iron is not yet unambiguously established. There are some orbits which are still bone of contention and though recently some interference orbits have been observed experimentally yet one could not approach a unani-mous conclusion as to which model is appropriate to repre-sent the FS of ferromagnetic iron. This controversy, naturally, led us to make an attempt in this direction. We started with the paramagnetic iron in the course to study the ferromagnetic iron. The interpolation scheme, in which the basis set contains seven orthogonalized plane waves (OPWs) and five tight-binding functions, has recently been used to parametrize the energy bands of paramagnetic iron. Using these bands we have calculated the density of states (DOS), joint density of states (JDOS), the imaginary part of the dielectric function ( e2(w)),and the real part of optical conductivity (a1(w)) for paramagnetic iron. The Brillouin zone integrations are done using the Special Directions method with 66 Special Directions, taking 50 K points along each direction. Our results on e2(w) and a1(w) for pdramagnetic iron have been compared with experi-mental results on ferromagnetic iron for these quantities. We obtain a good agreement with the optical data.