Conjectures for the microscopic theory of high temperature superconductivity

Abstract

Based on experimental results and our previous theoretical work, a microscopic theory of high temperature superconductivity is conjectured. In this conjecture, superconducting and antiferromagnetic long-range orders are driven by interlayer coupling. Strictly in two dimensions, the microscopic Hubbard model has an (resonating valence bond) insulator-to-metal transition at $x=x_{c}$ near optimal doping for zero temperature, leading to a quantum critical point, and one of the crossover lines is given by the pseudogap temperature $T^{*}$. We argue that various singular and non-Fermi liquid properties observed near optimal doping are due to the presence of this quantum critical point. In our conjecture, the crossover line $T^{*}$ also practically divides the superconducting region into two, depending on the doping level with respect to $x_{c}$. For $x \leq x_{c}$ the superconducting state has significant antiferromagnetic correlations, while for $x > x_{c}$ it has virtually no antiferromagnetic correlations, thus justifying the conventional BCS theory based on the noninteracting electrons. Inelastic neutron scattering resonance and systematically reduced superfluid density in the superconducting state below $x_{c}$ have their natural explanations in the present scenario. The present approach supports interlayer pair tunneling model where the superconducting condensation energy comes from the lowering of the c-axis kinetic energy in the superconducting state. Comparison of the present scenario with some of the leading theories based on the Hubbard and $t-J$ models is given. The generic features of both hole-doped and electron-doped cuprates as well as heavy-fermion superconductors may be understood in the {\em unified} framework within the present picture.