Optical transmission and reflection spectra (mid ir through uv) and Raman spectra of superior-quality 90-, 180-, 400-, and 1000-nm-thick superconducting Y-Ba-Cu-O films are reported. Characteristic excitonic bands, and in particular the absorption band at -0.37 eV reported earlier, are not observed. It therefore seems unlikely that the high-TV superconductivity in cuprates could arise from excitonmediated electron pairing.PACS numbers: 74.70.Ya, 78.30.Er The discovery of high-TV superconductivity in cuprates 1 has stimulated unprecedented research activity. Much of the effort is focused on understanding the underlying mechanism of the superconductivity, and a number of models have already been proposed. 2 " 4 Some of them are specific enough to be directly tested experimentally. However, what is usually the most powerful technique for identifying the mechanism, tunneling spectroscopy, has been greatly hampered so far by the generally bad surface quality of the presently available oxygen-annealed cuprate pellets and films. Infrared spectroscopy probes much deeper into the bulk of the sample and thus could play an important role in the clarification of the mechanism of superconductivity in these materials.The optical reflectivity spectra of La-Sr-Cu-O and YBa-Cu-O have been measured 5 " 7 over a broad frequency range (far ir through uv). However, since reflectivity depends on both the real and the imaginary parts of the dielectric function, such spectra cannot be interpreted directly in general. Orenstein et al. 5 tried to fit their data by the simple classical Drude model, e(co) = e(oo) -ca^/coico + ir); their optimal parameter values were ^=2 and 3 eV for La-Sr-Cu-O and Y-Ba-Cu-O, respectively, and r=0.5 eV and e(°°)=4.5 for both compounds. The calculated reflectivity turned out to be much too large in the mid-ir and near-ir region (0.1 to 1 eV). Orenstein et al. 5 concluded that an optically allowed transition across a gap occurs at -0.5 eV in LaSr-Cu-O; a similar result was obtained for Y-Ba-Cu-O. Therefore, they inferred that the corresponding strong absorption band (with the peak absorption coefficient of -10 5 cm -1 ) should be directly observable in transmission experiments on thin films. Finally, they found this ir feature to be absent in undoped La2Cu04, thus tying it to superconductivity. Etemad et al. 5 also found the correlation between the occurrence of the 0.5-eV peak and superconductivity in La2-^Sr x Cu04. Kamaras et al., 6 by an analogous methodology, concluded that two strong electronic transitions occur at 0.44 and 1.3 eV in La-Ba-Cu-O, and at 0.37 and 2.5 eV in Y-Ba-Cu-O. In oxygen-depleted (nonsuperconducting) YBa2Cu306.2 samples, the 0.37-eV peak was absent. They interpreted these spectral features as the charge-transfer-exciton bands, in support of the excitonic superconductivity model of Varma et al. 2 Finally, Schlesinger et al. 7 also recorded similar spectra but derived just the opposite conclusion. Inspired by the recent report 8 of singlecrystal La2Ni04 reflectivity spectra, which ...

Measurements of the ac penetration depth of a-MoGe films in the presence of a perpendicular magnetic field reveal an anomaly in the ac response of the vortex lattice at a characteristic temperature below the H c i(T) line. The field and frequency dependence of this anomaly is found to be consistent with a Kosterlitz-Thouless-type melting of the two-dimensional vortex lattice. Moreover, we observe a crossover in the frequency dependence which suggests that the vortex lattice remains disordered on long length scales below the melting temperature.PACS numbers: 74.60.Ec, 74.60.GeThe question of the existence of a finite-temperature phase transition within the vortex state of type-II superconductors has received considerable attention since the discovery of the high-T^ superconductors. One possible phase transition in these superconductors, as well as conventional superconductors with reduced dimensions, is the dislocation-mediated Kosterlitz-Thouless (KT) [1] melting transition in the two-dimensional vortex lattice (VL). In conventional superconductors, the possibility of this phase transition was first suggested by Huberman and Doniach [2] and by Fisher [3]. At the transition, the solid vortex lattice phase which contains thermally created bound pairs of dislocations, melts as these pairs dissociate at large distances to create free dislocations. However, as Larkin [4] showed long ago, in the solid phase, any amount of disorder will disturb the lattice on long length scales, hence limiting positional correlations to a length scale R c > The consideration of disorder raises two important questions: (1) Can the disordered VL "melt locally" as each correlated volume (vortex bundle) melts internally via a KT-type transition with increasing temperature? (2) Do these correlated volumes freeze into a glasslike state at a lower temperature [5]?To date, there are two reports of experiments that have been interpreted as evidence for KT-type melting of the VL. The vibrating-reed experiments of Gammel, Hebard, and Bishop [6] on amorphous composite In/InO x films showed a large dissipation peak at around the predicted melting temperature. However, recently Brandt [7] has argued that these experimental results are likely related to a size-dependent electromagnetic crossover rather than the melting of the VL. The other case is the resistivity and I-V measurements of Berghuis, van der Slot, and Kes [8] on films of a-NbGe. These experiments were interpreted as a crossover of a pinned VL with a finite critical current at low fields to a melted vortex fluid with flux flow resistance at high fields. The crossover was found to occur close to the expected phase boundary.In this Letter, we present evidence for the existence of a melting transition on short length scales in the disordered VL by studying the behavior of the ac penetration depth of amorphous Mo77Ge23 thin films in the presence of a perpendicular magnetic field. Specifically, we interpret observed abrupt changes in the ac penetration depth as evidence for melting of the two-d...

The successful synthesis of high-Tc YBa2Cu3O7 films by means of electron-beam codeposition are reported. Several important growth parameters have been surveyed in a preliminary way. The substrates investigated include Al2O3, ZrO2, MgO, and SrTiO3, The films were characterized by resistivity measurements, x-ray diffraction, microprobe, and Rutherford backscattering analysis. Some TEM and critical current density studies were also carried out. The best results to date have been obtained on SrTiO3 substrates with which polycrystalline epitaxial growth has been achieved. Resistive superconducting transitions with zero resistance at 89.5 K and a 2 K width have been observed in these films.