Unified description of correlations in double quantum dots

TL;DR

This paper provides an exact mapping of a double quantum dot model to a generalized Anderson model, revealing its connection to the anisotropic Kondo model and offering precise analytical and numerical insights into its charge and transmission properties.

Contribution

It introduces a comprehensive mapping of the double quantum dot model to a single impurity Anderson model, unifying various phenomena under a coherent theoretical framework.

Findings

Exact mapping to a generalized Anderson model

Accurate expressions for dot occupation and transmission

Validation against Bethe ansatz and renormalization-group results

Abstract

The two-level model for a double quantum dot coupled to two leads, which is ubiquitously used to describe charge oscillations, transmission-phase lapses and correlation-induced resonances, is considered in its general form. The model features arbitrary tunnelling matrix elements among the two levels and the leads and between the levels themselves (including the effect of Aharonov-Bohm fluxes), as well as inter-level repulsive interactions. We show that this model is exactly mapped onto a generalized Anderson model of a single impurity, where the electrons acquire a pseudo-spin degree of freedom, which is conserved by the tunnelling but not within the dot. Focusing on the local-moment regime where the dot is singly occupied, we show that the effective low-energy Hamiltonian is that of the anisotropic Kondo model in the presence of a tilted magnetic field. For moderate values of the (renormalized) field, the Bethe ansatz solution of the isotropic Kondo model allows us to derive accurate expressions for the dot occupation numbers, and henceforth its zero-temperature transmission. Our results are in excellent agreement with those obtained from the Bethe ansatz for the isotropic Anderson model, and with the functional and numerical renormalization-group calculations of Meden and Marquardt [Phys. Rev. Lett. 96, 146801 (2006)], which are valid for the general anisotropic case. In addition we present highly accurate estimates for the validity of the Schrieffer-Wolff transformation (which maps the Anderson Hamiltonian onto the low-energy Kondo model) at both the high- and low-magnetic field limits. Perhaps most importantly, we provide a single coherent picture for the host of phenomena to which this model has been applied.