Band Theory of Itinerant f -Electron Compounds

Abstract

Recent progress of band theory for the Fermi surface in the Ce and the U compounds is reported. A symmetrized relativistic APW method has been completed and applied to some compounds in which the de Haas-van Alphen effect (dHvA) has recently been measured. The exchange and correlation interaction of electrons are taken into account within the framework of the local-density approximation. On the basis of the itinerant-electron model for the 4/ electrons, it has been revealed that the origins of the major dHvA frequency branches observed in the valence-fluctuating compound CeSns can be clarified reasonably well by band theory. The experimental results for another valence-fluctuating compound CeNi can be explained well also, though the agreement between theory and experiment is not so good as for CeSns, perhaps because of limitations of the muffin-tin potential shape approximation used. On the basis of the itinerant-electron model for the 5/ electrons, the origins of the major dHvA frequency branches observed in UC and UB12 have been clarified satisfactorily well. In contrast to the shape of the Fermi surface, the heavy cyclotron effective masses observed in all these compounds except UB1z cannot be explained by the band structure alone. Theoretical results for related materials such as LaSns, LaNi, Th and a-U are also reported. § 1. Introduction 27 In recent years, various /-electron compounds such as the cerium compounds and the uranium compounds have attracted interest and been studied intensively both experimentally and theoretically, because they show many interesting anomalous properties due to the f electrons such as the Pauli paramagnet, the spin fluctuation, the heavy fermion and the magnetic ordering.!) As a first step toward a unified understanding of the electronic states in the whole temperature range, it is important to clarify the ground state properties of these compounds, and the Fermi surface study is useful for it. On an experimental side, a measurement of the de Haas-van Alphen (dHvA) effect is a powerful tool to observe the Fermi surface directly, and has been carried out systematically for the /-electron compounds together with the magnetoresistance measurement. 2 > The Fermi surface of the /-electron compounds, however, is usually so complicated that its shape cannot be deduced from experimental results alone. It should be determined by investigating the origins of the experimental dHvA frequency branches by the Fermi surface which is predicted by band theory. On a theoretical side, a new method has been developed so as to carry out reliable quantitative calculations of the energy band structures for the /-electron compounds. 3 >