Transport across an Anderson quantum dot in the intermediate coupling regime

TL;DR

This paper introduces the dressed second order (DSO) approximation for modeling electron transport in an Anderson quantum dot with intermediate coupling, capturing key phenomena like conductance peak crossover and Kondo effects.

Contribution

The DSO approximation provides a new analytical approach to describe transport in quantum dots beyond sequential tunneling, including the crossover from Coulomb blockade to Kondo regimes.

Findings

DSO captures the crossover from thermally to tunneling broadened conductance peaks.

It qualitatively reproduces Kondo-like zero bias anomalies and conductance enhancement at low temperatures.

Results agree with experimental observations of Kondo to empty orbital regime transition.

Abstract

We describe linear and nonlinear transport across a single impurity Anderson model quantum dot with intermediate coupling to the leads, i.e., with tunnel coupling of the order of the thermal energy k_B T. The coupling is large enough that sequential tunneling processes alone do not suffice to properly describe the transport characteristics. Upon applying a density matrix approach, the current is expressed in terms of rates obtained by considering a very small class of diagrams which dress the sequential tunneling processes by charge fluctuations. We call this the "dressed second order" (DSO) approximation. One major achievement of the DSO is that, still in the Coulomb blockade regime, it can describe the crossover from thermally broadened to tunneling broadened conductance peaks. When the temperature is decreased even further, the DSO captures "Kondesque" behaviours of the Anderson quantum dot qualitatively: We find a zero bias anomaly of the differential conductance versus applied bias, an enhancement of the conductance with decreasing temperature as well as the onset of universality of the shape of the conductance as function of the temperature. We can address the case of a spin-degenerate level split energetically by a magnetic field and show that, if we assume in addition different capacitive couplings of the two spin-levels to the leads, one of the resonance peaks is vanishing. In case spin-dependent chemical potentials are introduced and only one of the four is varied, the DSO yields in principle only one resonance. This seems to be in agreement with experiments with pseudo-spin. Furthermore, we get qualitative agreement with experimental data showing a cross-over from the Kondo to the empty orbital regime.