NSRRC Activity Report 2022

014 NSRRC ACTIVITY REPORT 2022 Next, the authors carried out Cu 2p core-level hard X-ray photoelectron spectroscopy (HAXPES) of CCRO and a reference material Li 2 CuO 2 at SP 12U1 , the Taiwan-contract beamline at SPring-8, Japan, as shown in Fig. 2 . CCRO and Li 2 CuO 2 comprise weakly coupled CuO 4 plaquettes. It can be seen from Fig. 2 that, while the gross features are similar, there is a very important difference between the main peaks of CCRO and Li 2 CuO 2 : the CCRO main peak consists of two peaks, while Li 2 CuO 2 exhibits a single peak. The Cu 2p spectrum of Li 2 CuO 2 can be explained by a full multiplet calculation for a single CuO 4 -cluster. On the other hand, the Cu 2p spectrum of CCRO cannot be explained in terms of a single CuO 4 -cluster model. Furthermore, it also cannot be explained by a nonlocal screening mechanism, as it cannot accommodate inter-cluster hopping because the CuO 4 plaquettes are isolated. The authors then showed that dynamical mean-field theory with local density approximation (LDA+DMFT) calculations could explain the Cu 2p spectra. Thus, the authors inferred that CCRO contains correlated Cu 2+ ions, which experience metallic screening. Finally, the authors discuss the experimental valence-band spectra of CCRO measured using incident photon energies ranging from 100 eV to 6.5 KeV ( Fig. 3(a) ) and LDA+DMFT calculations that identify the partial density of states of Cu 3d, Ru 4d, and O 2p states in the valence-band spectra ( Fig. 3(b) ). As can be observed in Fig. 3(a) , the valence band spectra reveal a systematic evolution with incident photon energy and, by taking into account photoionization cross- sections, features A and D are assigned to Ru 4d states, features B and E are assigned to Cu 3d states, and feature C can likely be assigned to the O 2p states. These assignments are confirmed by the LDA+DMFT results. Most interestingly, the authors also revealed that the Cu d xy orbitals cross the Fermi level and are related to a small experimental feature just 0.07–0.08 eV above the Fermi level, which are obtained after dividing the experimental spectrum of CCRO by the metallic gold spectrum. This provides evidence for the Kondo-type behavior of CCRO and, from the optimized LDA+DMFT calculations, the authors could estimate a high Kondo temperature in the range of 500–1000 K. 5 The authors concluded by stating that “the material class Fig. 1 : Valence-band resonant photoelectron spectroscopy of CaCu 3 Ru 4 O 12 with the experimental spectra recorded at the Cu 2p (L 3 ) resonance (hν = 931.2 eV) and at 10 eV below the resonance (hν = 921.2 eV). The inset displays the experimental Cu-L 2,3 X-ray absorption spectrum. [Reproduced from Ref. 5] Fig. 2 : Experimental Cu 2p core-level X-ray photoelectron spectrum of Li 2 CuO 2 reproduced from Ref. [1]. (b) Theoretical spectrum from the CuO 4 cluster model. (c) Experimental Cu 2p core-level HAXPES spectrum of CaCu 3 Ru 4 O 12 . (d) Theoretical spectrum from the LDA+DMFT method. [Reproduced from Ref. 5] Fig. 3 : Valence band spectra of CaCu 3 Ru 4 O 12 : (a) experimental results measured at different photon energies and (b) LDA+DMFT spectral intensities for the Cu 3d, Ru 4d, and O 2p states. Spectral broadening is taken into account using a 200-meV Gaussian to simulate the experimental resolution. (c) LDA+DMFT spectral intensities of the Cu 3d orbitals in CaCu 3 Ru 4 O 12 . The Cu orbitals are defined in the local axis of the CuO 4 plane, as shown in the inset. [Reproduced from Ref. 5]