arXiv subjects
Publications and source records attributed to Hyeong-Do Kim.
%auto-ignore This paper has been withdrawn by the authors because its results have been superceded by new experimental data and new theory presented in S.-K. Mo et al., cond-mat/0212110.
We have investigated 3d electronic states of doped transition metals in II-VI diluted magnetic semiconductors, Zn1-xMxSe (M = Mn, Fe, Co) and Zn1-xMnxY (Y = Se, Te), using the transition-metal L2,3-edge X-ray absorption spectroscopy (XAS) measurements. In order to explain the XAS spectra, we employed a tetragonal cluster model calculation, which includes not only the full ionic multiplet structure but also configuration interaction (CI). The results show that CI is essential to describe the experimental spectra adequately, indicating the strong hybridization between the transition metal 3d and the ligand p orbitals. In the study of Zn1-xMnxY (Y = Se, Te), we also found considerable spectral change in the Mn L2,3-edge XAS spectra for different ligands, confirming the importance of the hybridization effects in these materials.
Resonant photoemission (RPES) at the Ce 3d -> 4f threshold has been performed for alpha-like compound CeNi_2 with extremely high energy resolution (full width at half maximum < 0.2 eV) to obtain bulk-sensitive 4f spectral weight. The on-resonance spectrum shows a sharp resolution-limited peak near the Fermi energy which can be assigned to the tail of the Kondo resonance. However, the spin-orbit side band around 0.3 eV binding energy corresponding to the f_{7/2} peak is washed out, in contrast to the RPES spectrum at the Ce 3d -> 4f RPES threshold. This is interpreted as due to the different surface sensitivity, and the bulk-sensitive Ce 3d -> 4f RPES spectra are found to be consistent with other electron spectroscopy and low energy properties for alpha-like Ce-transition metal compounds, thus resolves controversy on the interpretation of Ce compound photoemission. The 4f spectral weight over the whole valence band can also be fitted fairly well with the Gunnarsson-Schoenhammer calculation of the single impurity Anderson model, although the detailed features show some dependence on the hybridization band shape and (possibly) Ce 5d emissions.