Spectroscopic investigation of 4f correlations in CeAg; comparison between LDA and many-body descriptions

High resolution photoemission and bremsstrahlung isochromat spectroscopy experiments have been carried out on CeAg and LaAg. From a comparison between these two compounds, we extract the cerium 4f spectral function, we compare it with band structure calculations and with many-body model simulations based on the single-impurity Anderson Hamiltonian. We show that the localized 4f charge estimated from previous band structure calculations cannot be identified with the 4f occupation number deduced from the Anderson model.     

Giant spin splitting through surface alloying

The long-range ordered surface alloy Bi/Ag(111) is found to exhibit a giant spin splitting of its surface electronic structure due to spin-orbit coupling, as is determined by angle-resolved photoelectron spectroscopy. First-principles electronic structure calculations fully confirm the experimental findings. The effect is brought about by a strong in-plane gradient of the crystal potential in the surface layer, in interplay with the structural asymmetry due to the surface-potential barrier. As a result, the spin polarization of the surface states is considerably rotated out of the surface plane.     

New spectroscopy solves an old puzzle: the Kondo scale in heavy fermions

Resonant inelastic x-ray scattering (RIXS) yields clear evidence of spectroscopic Kondo scales in heavy fermions. In YbInCu4 and YbAgCu4 RIXS probes the Yb2+ component of the hybrid ground state and the temperature dependence of the Yb 4f occupation. We report a sudden valence change at a phase transition in YbInCu4, but a continuous temperature dependence in YbAgCu4, consistent with the predictions of the Anderson impurity model, for a Kondo temperature T(K) = 70 K. These results solve a long-standing controversy and establish RIXS as a quantitative probe of the electronic structure of strongly correlated electron systems.     

Optical and photoemission evidence for a Tomonaga-Luttinger liquid in the Bechgaard salts

Combined optical and photoemission experiments on the quasi-one dimensional Bechgaard salts reveal the non-Fermi liquid character of these prototype quasi-one dimensional interacting electron systems. We show that various aspects of the exotic normal state properties along the chains are consistent with the predictions of the Tomonaga-Luttinger liquid theory. We also discuss the effect of interchain coupling on the insulator-metal transition, associated with the electron confinement-deconfinement crossover. Received 17 May 1999 and Received in final form 13 July 1999     

Crystal field and low energy excitations measured by high resolution RIXS at the L3 edge of Cu, Ni and Mn

Resonant inelastic x-ray scattering in the soft x-ray regime has been profiting much from technical advances that have lowered considerably the instru- mental linewidth. At the ADRESS beam line of the Swiss Light Source the SAXES spectrometer can be used to measure RIXS spectra at the L edges of the 3d transition metals with unprecedented energy resolution, of the order of 100 meV for Mn, Ni and Cu. We present here some preliminary spectra on CuO, malachite, NiO, NiCl₂, MnO and LaMnO₃. The dd excitations are very well resolved allowing accurate experimental evaluation of 3d state energy splitting. The low energy scale becomes accessible opening the way to the study of collective excitations in strongly correlated electron systems, like magnons and orbitons.     

Temperature dependence of the Kondo peak in photoemission spectra of YbAgCu4

Temperature dependent photoelectron spectra of the heavy fermion YbAgCu₄ show that the Kondo resonance is strongly depressed with increasing temperature. The observed reduction is fully consistent with the predictions of the Anderson impurity model, and reveals the transition from a Fermi liquid at low temperature to localized moments above the characteristic Kondo temperature. Despite this important experimental confirmation of the model, discrepancies exist between the experimental and calculated linewidth of the resonance. It is argued that a more realistic treatment of the photoemission process would improve the agreement between theory and experiment.