Kondo peaks and dips in the differential conductance of a multi-lead quantum dot: Dependence on bias conditions

TL;DR

This paper investigates how bias asymmetry affects the differential conductance in a multi-lead quantum dot in the Kondo regime, revealing a threshold-dependent dip phenomenon useful for device characterization.

Contribution

It demonstrates that bias asymmetry can induce a conductance dip in the Kondo regime, providing a method to determine lead-dot coupling ratios experimentally.

Findings

Symmetric bias yields zero-bias conductance enhancement.

Asymmetric bias beyond a threshold causes a conductance dip.

The threshold depends on the coupling ratio r.

Abstract

We study the differential conductance in the Kondo regime of a quantum dot coupled to multiple leads. When the bias is applied symmetrically on two of the leads ($V$ and $-V$, as usual in experiments), while the others are grounded, the conductance through the biased leads always shows the expected enhancement at {\it zero} bias. However, under asymmetrically applied bias ($V$ and $\lambda V$, with $\lambda>0$), a suppression - dip - appears in the differential conductance if the asymmetry coefficient $\lambda$ is beyond a given threshold $\lambda_0= \sqrt[3]{1+r}$ determined by the ratio $r$ of the dot-leads couplings. This is a recipe to determine experimentally this ratio which is important for the quantum-dot devices. This finding is a direct result of the Keldysh transport formalism. For the illustration we use a many-lead Anderson Hamiltonian, the Green functions being calculated in the Lacroix approximation, which is generalized to the case of nonequilibrium.