Raman Spectroscopy of High-Temperature Cuprate Superconductors

Abstract

A variety of high-temperature cuprate superconductors were studied using Raman spectroscopy. Scattering from both electrons and phonons allowed the magnitude of the superconducting gap to be determined under different doping conditions and the presence of the pseudogap to be established in underdoped materials. The phonon spectra of Y1-xCaxBa2Cu3O7-δ (Y123), Y1-xCaxBa2Cu4O8 (Y124), and Bi2Sr2CaCu2O8+δ (Bi2212) superconductors were measured as function of temperature. In 10% calcium substituted Y123 the 340cm-l phonon mode showed clear evidence of superconductivity-induced changes, softening by 1.1 ± 0.8cm-1 at optimal doping and 4 ± 2cm-1 when overdoped. Similarly, in 10% calcium substituted Yl24 the 340cm-1 mode was seen to change below Tc, softening by 1.2 ± 0.6cm-1. The softening of the 340cm-1 mode in both materials is interpreted as evidence that the maximum superconducting gap energy is slightly greater than 340cm-1 in these samples. No changes that could be attributed to the onset of superconductivity were observed in the phonons of an optimally doped sample of.Bi2212, which is attributed to a strong dependence of such changes on doping in this material. Measurements performed on isotopically exchanged samples of both pure and substituted Y124 were used to evaluate the concentration and distribution of l8O within the oxygen sites of the material. Electronic Raman spectra of both overdoped and underdoped single crystals of Bi2212 were studied as a function of temperature. A substantial depletion of spectral weight at low energies is observed in the underdoped samples at temperatures well above Tc and a peak centred near 600cm-1 is observed in the normal state spectra of the optimally and underdoped crystals. This peak shows little dependence on doping or temperature and there is no apparent relationship between the doping dependence of this peak at 600cm-l and the pair-breaking peak observed below Tc. In contrast, the onset temperature of the normal state depletion, the depth of depletion, and both the position and intensity of the pair breaking peak are all dependent on sample doping. The depletion of normal state spectral weight in the underdoped materials is attributed to the opening of the pseudogap but the origin of the 600cm-l peak remains unclear.