Pseudogap and Kinetic Pairing Under Critical Differentiation of Electrons In Cuprate Superconductors

Abstract

Superconducting mechanism of cuprates is discussed in the light of the proximity of the Mott insulator. The proximity accompanied by suppression of coherence takes place in an inhomogeneous way in the momentum space in finite-dimensional systems. Studies on instabilities of metals consisted of such differentiated electrons in the momentum space are reviewed from a general point of view. A typical example of the differentiation is found in the flattening of the quasiparticle dispersion discovered around momenta (π, 0) and (0, π) on 2D square lattices. This flattening even controls the criticality of the metal-insulator transition. Such differentiation and suppressed coherence subsequently cause an instability to the superconducting state in the second order of the strong coupling expansion. The d-wave pairing interaction is generated from such local but kinetic processes in the absence of disturbance from the coherent single- particle excitations. The superconducting mechanism emerges from a direct kinetic origin which is conceptually different from the pairing mechanism mediated by bosonic excitations as in magnetic, excitonic, and BCS mechanisms. Pseudogap phenomena widely observed in the underdoped cuprates are then naturally understood from the mode-mode coupling of d-wave superconducting (dSC) fluctuations repulsively coupled with antiferromagnetic (AFM) ones. When we assume the existence of a strong d-wave channel repulsively competing with AFM fluctuations under the formation of flat and damped single-particle dispersion, we reproduce basic properties of the pseudogap seen in the magnetic resonance, neutron scattering, angle resolved photoemission and tunneling measurements in the cuprates.KeywordsMott InsulatorSuperexchange InteractionDrude WeightUnderdoped RegionSuperconducting MechanismThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.