COEXISTENCE OF LOCALIZED AND DELOCALIZED STATES IN BI-2212 Y/CA - AN INVESTIGATION OF THE INFRARED AND OPTICAL-ABSORPTION

We have studied the absorption of the system Bi-2212 Y/Ca (Bi2Sr2Ca1-xYxCu2O8+delta with 0 less-than-or-equal-to x less-than-or-equal-to 0.9) in the spectral range 600 to 20 000 cm-1. We observe systematical changes of the spectra as a function of Y-content. The temperature dependence of the NIR absorption of the samples possessing superconductivity (x < 0.5) shows a change at T(c), indicating its coupling to super-conductivity. We decompose the absorption spectra into at most three different contributions and observe that each changes as a function of Y content. First, a structureless contribution, modelled by alpha approximately (HomegaBAR-E(d))1.5/HomegaBAR is attributed to transitions between delocalized and non-occupied localized states with excitation energy E(d)(x). Second, a prominent peak at 13 500 cm-1 (almost-equal-to 1.7 eV) for all insulating samples, indicating the charge transfer gap and finally a near infrared (NIR) peak in the range from 1500 cm-1 to 7000 cm-1. The shape of the NIR peak suggests that the absorption is caused by photon assisted hopping of small polarons in disordered systems. We propose a schematic density of states including localized and delocalized states which is in accordance with absorption properties and the obtained fit parameters. In particular the Fermi energy cuts the bandedge of delocalized states at x almost-equal-to 0.5 as a result of hole doping (decreasing Y content). This can explain the metal (superconductor) /insulator transition in accordance with electrical transport properties and indicates the coexistence of delocalized and localized states at the Fermi energy for x < 0.5.