X-ray magnetic circular dichroism at rare-earth L23 absorption edges in various compounds and alloys

Theoretical interpretations of X-ray magnetic circular dichroism (XMCD) at rare-earth (called R hereafter) L23 absorption edges are reviewed using differing models, depending on the material under investigation. In the first chapter, we present an overview of recent developments for XMCD in XAS with a few general remarks, especially at R atom absorption edges. In Section 2, we first describe basic mechanism of XMCD at L23 edges of R systems, and then we essentially discuss two examples of XMCD spectra in: (i) R2Fe14B metallic compounds, with the help of a cluster model, and (ii) RFe2 Laves-phase compounds, using a tight-binding approximation for R 5d and Fe 3d conducting states. A good agreement between theory and experiment for R2Fe14B suggests that a cluster model provides a valuable method to quantitatively calculate XMCD spectra of R systems, even with quite complicated atomic arrangements. For RFe2 systems the XMCD spectral shape, especially for the L2 edge of heavy R elements, is more complicated than those of R2Fe14B systems, and this is explained by the competition of some different XMCD mechanisms. In Section 3, we focus on special series of Ce systems, related to XAS and XMCD studies at the Ce L23 edges. Two clearly differing cases are interpreted: (i) A well localized 4f1 system, i.e. CeRu2Ge2; (ii) A less localized 4f1 system, i.e. CeFe2, with a 3d partner. Then, from a more experimental point of view, we investigate the influence of substitution on the low temperature properties of CePd3 compounds, i.e. Ce ( Pd 1 - x Mn x ) 3 alloys where x is about 0.03, giving rise to ( CePd 3 ) 8 Mn ordered structure. We give another example: Ce ( Pd 1 - x Ni x ) 3 alloys with x taken up to about 0.25. Also the Ce L23 XMCD signal measured in pure CePd3 demonstrates that in the Ce based dense Kondo materials, only the 4f1 channel yields a magnetic response.