M ar 2 00 3 From charge transfer type insulator to superconductor

Wei Guo,R. Han
2003-01-01
Abstract:The insulator-superconductor transition driven by impurity doping occurs uniquely in a class of copper oxides with a special magnetic structure. The magnetism of the high-Tc cuprate arises from unfilled d shell (3d ) of Cu ions in the CuO2 plane, the overlap of the atomic d wave function is weak because of large spacing between Cu ions(3.8Å). The strong correlation between local Cu moments(S = 1/2) comes from the indirect interaction mediated by p electrons at neighbor oxygen sites, i.e. the superexchange[1][2]. Even though the superfluid carriers in the high-Tc cuprates have the d wave character, however experimental observations confirm that the doped holes actually appear in the p band[3], which is consistent with the superexchange picture since d electrons are localized in the doping range 0 < x < 0.2. In this article we show that d wave superconductivity is not originated from hopping d electrons but p electrons. Undoped high-Tc cuprates are in fact charge transfer type insulators[4], where p band is above lower d band. p electrons being the mediation for magnetic ordering in the insulating phase became conductive after impurity doping, which causes the partial occupancy of the Wannier p state at a portion of oxygen sites and local d-level shift at neighbor Cu sites. If local d-level merge into p band, the hybridization between p and d electrons generates the Kondo exchange interaction between itinerant p electron spins and the local Cu spin. We show that at such a Cu site, the local Cu spin can be a mediation for electron pairing via Kondo exchange. The effective electron-electron interaction originated from the spin coupling is attractive and may yield a bound spin triplet with opposite spins χ+ = | ↑↓ 〉 + | ↓↑ 〉(S = 1, Sz = 0). The d wave symmetry of paired electrons comes from l = 2 partial wave channels of Bloch states forced by crystal and spin pair symmetry. Since electron pairs couple to the local spins in the CuO2 plane the effective pair-pair potential is strong and leads to high temperature superfluid transition.
What problem does this paper attempt to address?