Anderson Hamiltonian description of the experimental electronic structure and magnetic interactions of copper oxide superconductors

TLDR

The authors argue that for Δ≪U and T≫Δ, charge‑transfer degrees of freedom and lattice aspects of the Anderson lattice Hamiltonian must be included in high‑Tc superconductivity models. They employ an Anderson‑Hamiltonian impurity‑cluster configuration‑interaction model to analyze valence‑band and core‑level photoemission data and compute the Cu‑O‑Cu superexchange interaction J, finding it dominated by charge‑transfer fluctuations. The analysis yields Anderson‑Hamiltonian parameters (Δ≈0.4 eV, U≈6 eV, T≈2.4 eV), shows that J is much larger than typical Néel temperatures and is dominated by charge‑transfer fluctuations.

Abstract

We describe valence-band and core-level photoemission data for copper oxide superconductors using the Anderson Hamiltonian applied to an impurity-cluster configuration-interaction model. We obtain experimental values of the parameters of the model, the copper\ensuremath{\rightleftarrows}oxygen chargetransfer energy $\ensuremath{\Delta}\ensuremath{\sim}0.4$ eV, the $d\ensuremath{-}d$ Coulomb interaction $U\ensuremath{\sim}6$ eV, and the ligand-$d$ hybridization $T\ensuremath{\sim}2.4$ eV. Using these parameters, we evaluate the linear Cu-O-Cu superexchange interaction $J$ and find it is dominated by the charge-transfer fluctuations. The magnitude obtained for $J$ is much larger than typical N\'eel temperatures of these materials, and is somewhat larger than that estimated from applying the resonating-valence-bond picture to ${\mathrm{La}}_{2}$Cu${\mathrm{O}}_{4}$. We point out that for $\ensuremath{\Delta}\ensuremath{\ll}U$ and $T\ensuremath{\gg}\ensuremath{\Delta}$ the charge-transfer degrees of freedom, and the lattice aspects of the Anderson lattice Hamiltonian, should not be neglected in constructing models for the high-${T}_{c}$ superconductivity. We also emphasize our resonant-photoemission result that the very small density of states at or near the Fermi level in all these materials has a substantial contribution from Cu $3d$ states, suggesting their importance for the superconductivity. We report other details of the resonant-photoemission data involving La and Ba states in the materials containing these elements.