VOLUME 73, NUMBER 10 PHYSICAL REVIEW LETTERS 5 SEPTEMBER 1994 Reflectance and Raman Spectra of Metallic Oxides, LaSrCoo and CaSrRuo: Resemblance to Superconducting Cuprates Ivan Bozovic E L G. in. zton Research Center, Varian Associates, Inc , Pa.lo Alto, California 94304-I025 J. H. Kim* and J. S. Harris, Jr. Department of Electrical Engineering, Stanford University, Stanford, California 94305 C. B. Eom and Julia M. Phillips AT&T Bell Laboratories, Murray Hill, New Jersey 07907 J. T. Cheung Science Center, Rockwell International Corp. , Thousand Oaks, California 91358 (Received 28 May 1993) %e have measured reflectance, ellipsometric, and Raman spectra on thin films of isotropic metal- lic oxides La05Sr05COO3 and Ca05Sr05Ru03. The spectra strongly resemble those of cuprate superconductors. In particular, reflectance is a nearly linear function of frequency throughout the infrared, and the electronic Raman scattering continuum is nearly independent of energy. Neither of the compounds is superconducting down to 4. 2 K, while their electronic excitations spectra are apparently like those of the cuprates. Hence, it seems unlikely that the corresponding "anomalous" dielectric response could be the sole root of high-temperature superconductivity. PACS numbers: 78. 30. Er, 78. 20. Ci, 78. 40.Kc So far, high-temperature superconductivity (HTSC) re- mains quite a rare chemical phenomenon. It has been un- equivocally observed only in layered cuprates [1], which may all be considered alike since HTSC is believed to originate in Cu02 layers; indeed, the "passive" lay- ers can be varied, but these may merely act as spac- ers and charge reservoirs [2]. Given their infrequent occurrence, one could suspect that HTSC compounds might possess some exceptional normal state characteris- tics, as well. Therefore, much effort has been devoted to identification of such properties, since clearly that could guide both the theory and the search for new HTSC materials. Extensive spectroscopic studies of cuprate supercon- ductors have indeed revealed the universal presence of some unusual features. First, throughout the infrared region — almost all the way up to the screened plasma frequency, which is typically found near 1 eV — the re- flectance decreases with frequency in an almost linear manner. Second, electronic Raman scattering takes place in the same spectral range, and its intensity is nearly in- dependent of frequency. These two features are illus- trated in insets to Figs. 1 and 2, where we have compiled broad-range reflectance and Raman spectra [3 — 5], respec- tively, for a few cuprate superconductors (single crystals or optical-quality epitaxial thin films). Similar spectra have indeed been reported by several other groups; over- all, there is a reasonable consensus, at least insofar as the raw experimental data are concerned. However, interpretation of these spectroscopic features is still a matter of great controversy. For example, ever 0.8 0. 6 U Q4 02 TBCCO BSCCO YBCO LSCO I I I I I 3 4 5 0 0 I I 2 3 4 ENERGY (eV) since 1987 spectroscopists have been debating [5, 6] whether the infrared absorption has only one electronic component (the free charge carrier response) or two (the Drude response plus a "midinfrared band"). Theorists, on their side, have proposed a vast number of HTSC theories, models, and scenarios (see, for example, Refs. [7 — 10]), and hence as many explanations for the optical spectra of cuprates. There may be one aspect, though, in which many of these theories seem to converge, and which may therefore deserve attention: The normal state of high-temperature FIG. 1. Broad-range specular reflectance spectra of CaosSrosRu03 (broken line) and LaosSrosCo03 (solid line). Inset: Reflectance spectra [3] of T12Ba, Ca2Cu, 0, 0, Bi2Sr2CaCu20s, YBa2Cu307, and La i.ssSro, »Cu04. 1436 0031-9007/94/73 (10)/1436(4) $06. 00 1994 The American Physical Society