Fermi surface changes due to localized–delocalized f-state transitions in Ce-115 and Pu-115 compounds

Abstract The current theoretical understanding of the electronic structures of the CeMIn 5 compounds (with M=Co, Rh, and Ir) and PuCoGa 5 and PuRhGa 5 is reviewed. Both the Ce-115 and Pu-115 compounds have in common that their f-electrons appear to be poised on the edge of a localization–delocalization boundary. We consider here the implications of this characteristic on the emergent electronic structure, addressing, in particular, the computed Fermi surfaces. In order to capture the electronic structures to the best as possible, we employ three different density-functional theory based computational approaches, viz. the local-spin density approximation (LSDA) for delocalized f-states, the open-core approach for localized f's, and the LSDA+ U approach for moderately localized f's. The Fermi surfaces of CeCoIn 5 and CeIrIn 5 are rather well described assuming delocalized 4f electrons. Conversely, the better description is obtained for CeRhIn 5 assuming a localized 4f state. For the Pu-115 compounds, LSDA calculations predict a rather two-dimensional Fermi surface. However, the Fermi surface of PuCoGa 5 computed with the around-mean field LSDA+ U scheme is also quite two-dimensional and, moreover, similar to the LSDA Fermi surface. Taking other available experimental data on PuCoGa 5 into account, it appears that the electronic structure of PuCoGa 5 is best described by the AMF-LSDA+ U approach with a Coulomb U of about 3 eV.