Universality and thermoelectric transport properties of quantum dot systems

TL;DR

This paper investigates the universal behavior of thermoelectric properties in quantum dot systems within the Kondo regime, providing theoretical calculations and fitting functions to predict experimental outcomes.

Contribution

It extends the understanding of universal thermoelectric transport coefficients to include thermopower and thermal conductance using renormalization-group methods.

Findings

Universal thermoelectric coefficients are derived.

Numerical results illustrate the universal behavior.

Analytical fitting functions for experimental predictions.

Abstract

We discuss the temperature-dependent thermoelectric transport properties of semiconductor nanostructures comprising a quantum dot coupled to quantum wires: the thermal dependence of the electrical conductance, thermal conductance, and thermopower. We explore the universality of the thermoelectric properties in the temperature range associated with the Kondo crossover. In this thermal range, general arguments indicate that any equilibrium property's temperature dependence should be a universal function of the ratio $T^{*}=T/T_{K}$, where $T_{K}$ is the Kondo temperature. Considering the particle-hole symmetric, spin-degenerate Anderson model, the zero-bias electrical conductance has already been shown to map linearly onto a universal conductance through a quantum dot embedded or side-coupled to a quantum wire. Employing rigorous renormalization-group arguments, we calculate universal thermoelectric transport coefficients that allow us to extend this result to the thermopower and the thermal conductance. We present numerical renormalization-group results to illustrate the physics in our findings. Applying the universal thermoelectric coefficients to recent experimental results of the electrical conductance and thermo-voltages versus $V_{gate}$, at different temperatures in the Kondo regime, we calculate all the thermoelectric properties and obtain simple analytical fitting functions that can be used to predict the experimental results of these properties. However, we cannot check all of them, due to the lack of available experimental results over a broad temperature range.