Critical local moment fluctuations and enhanced pairing correlations in a cluster Anderson model

Abstract

The appearance of unconventional superconductivity near many heavy-fermion quantum critical points (QCPs) motivates investigation of pairing correlations close to a "beyond Landau" Kondodestruction QCP. We focus on a two-Anderson-impurity cluster in which Kondo destruction is induced by a pseudogap in the conduction-electron density of states. Analysis via continuous-time quantum Monte-Carlo and the numerical renormalization group reveals a previously unstudied QCP that both displays the critical-local moment fluctuations characteristic of Kondo destruction and leads to a strongly enhanced singlet-pairing susceptibility. Our results shed light on the extent to which different kinds of magnetic interactions induce pairing correlations in non-Fermi liquid settings, thereby providing new insights into the mechanism for unconventional superconductivity in quantum critical metals.Heavy-fermion metals represent a prototype system in which unconventional superconductivity is driven by antiferromagnetic correlations [1,2]. Experiments on these strongly correlated systems have revealed quantum criticality in a wide range of compounds. Detailed theoretical and experimental studies have provided evidence for different classes of quantum critical point (QCP). One class follows the Landau theory, in which criticality is dictated by the fluctuations of an order parameter. Here, the broken-symmetry state is a spin density wave, and the theoretical description starts from the spatial and temporal fluctuations of the antiferomagnetic order parameter [3][4][5].Another class of QCP goes beyond the Landau framework, in that it involves new critical modes besides orderparameter fluctuations. The additional critical modes describe a critical destruction of the Kondo entanglement between the localized magnetic moments and conduction electrons [6,7]. An important example of a Kondodestruction QCP occurs in CeRhIn 5 , which has the highest T c among all the Ce-based heavy-fermion superconductors [8][9][10][11] and is generally believed to have a d x 2 −y 2 superconducting gap symmetry. A sudden change of the Fermi-surface topology at the antiferromagnetic QCP in CeRhIn 5 , accompanied by a diverging tendency of the carrier effective mass [12], defy explanation within the spin-density-wave scenario for quantum criticality. Instead, these properties provide evidence supporting the Kondo-destruction picture.How unconventional superconductivity arises near a Kondo-destruction QCP has yet to receive systematic theoretical study. The question is challenging because the normal state is a non-Fermi liquid with Landau quasiparticles turned critical. It has previously proved illumi-nating to study the notion of Kondo destruction within quantum impurity models [13][14][15][16][17]. Indeed, the QCP in the single-impurity pseudogap Kondo and Anderson models captures similar critical properties to the Kondodestruction QCP in the lattice problem (as solved within an extended dynamical mean field theory [6]), including a vanishing Kondo energy scale, ω/...

Quantum phases driven by strong correlations

“…This led to the important question whether a Kondo destruction QCP can also promote superconductivity. Theoretical calculations suggest that this is indeed the case 71,72 .…”

Superconductivity in an extreme strange metal

“…Very recently, in a two-impurity Anderson model that features Kondo destruction 20 , the singlet-pairing susceptibility was found to be strongly enhanced by critical local moment fluctuations 34 . Because singlet pairing may be subject to Pauli limiting, this phase—though stabilized by quantum critical fluctuations—might be suppressed by the applied magnetic field even well before the QCP is reached.…”

Unconventional Superconductivity from Fermi Surface Fluctuations in Strongly Correlated Metals

“…In the context of unconventional pairing and superconductivity, we expect the spin-flip part of the RKKY interaction to be essential for driving the formation of spin-singlet Cooper pairs 30,32,33 .…”