Unfilled inner shells: Rare earths and their compounds

Abstract

In this chapter we review the present status of our understanding of the photoemission process from rare-earth (RE) metals, their compounds and alloys, and discuss the impact of this technique o11 various aspects of the electronic structure of RE solids. As is the case in other areas of electron spectroscopy (e.g., Auger, Leed, Electron Energy Loss)and, especially in those areas related to surfaces, the volume of experimental information is not matched by our theoretical understanding. A few of these aspects will be discussed in the following paragraphs. This problem results from the inherent many-electron nature of the 4f states and from our inability to treat correlated many-electron shells within the available schemes for calculating the electronic structure of solids. This aspect of the 4f electrons, together with the fact that they often have binding energies smaller than uncorrelated valence electrons, is responsible for the variety of electronic properties displayed by RE materials. The desire to elucidate the peculiar behavior of some RE solids has presented a challenging task to the electron-spectroscopists. The steady improvements of the experimental techniques in this field, especially in photoemission spectros-copy will, in turn, provide challenging information to test theories, not necessarily specific to RE solids. Although optical spectroscopy provided us with a quite detailed understanding of the absorption and emission spectra of the rare earths during the 1960's [4.1,2], it was toward the end of that period that questions like the one above were asked and answered. During that period the Eu chalcogenides and some other RE sulfides (especially GdS) received greater attention than even the metals themselves. The first photoemission studies, which showed unam-biguously that the 4f levels could indeed have lower photoionization energy than the valence electrons, were performed on such materials. They involved energy distribution curves (EDC) [4.3] and spin polarization [4.4] measurements. Although the high spin polarization (~ 30%) near photothreshold in EuS films was a clear indication of the low 4f-binding energy, E~ I, ill the Eu chalcogenides, the EDC measurement using photon energies up to 10.2eV on GdS gave ambiguous results regarding the position of the 4f levels in this metallic sulfide. However, in that work it was nevertheless established that in