Magnetic field effects in few-level quantum dots: theory, and application to experiment

TL;DR

This paper investigates the effects of magnetic fields on quantum dots using theoretical models and numerical methods, comparing results with experiments to understand the universal behaviors and limitations of current approaches.

Contribution

It introduces a method to analyze the scaling of Kondo resonance shifts over wide magnetic field ranges and compares NRG calculations with experimental data on quantum dots.

Findings

Good qualitative agreement between theory and experiment

Field-dependent Kondo conductance peak behavior captured

Deviations at high fields suggest limitations of equilibrium spectra for conductance

Abstract

We examine several effects of an applied magnetic field on Anderson-type models for both single- and two-level quantum dots, and make direct comparison between numerical renormalization group (NRG) calculations and recent conductance measurements. On the theoretical side the focus is on magnetization, single-particle dynamics and zero-bias conductance, with emphasis on the universality arising in strongly correlated regimes; including a method to obtain the scaling behavior of field-induced Kondo resonance shifts over a very wide field range. NRG is also used to interpret recent experiments on spin-1/2 and spin-1 quantum dots in a magnetic field, which we argue do not wholly probe universal regimes of behavior; and the calculations are shown to yield good qualitative agreement with essentially all features seen in experiment. The results capture in particular the observed field-dependence of the Kondo conductance peak in a spin-1/2 dot, with quantitative deviations from experiment occurring at fields in excess of $\sim$ 5 T, indicating the eventual inadequacy of using the equilibrium single-particle spectrum to calculate the conductance at finite bias.