Reinterpretation of Scanning Tunneling Microscopy on an Adsorbed Magnetic Atom

TL;DR

This paper reinterprets STM observations of the Kondo effect on magnetic atoms, showing that asymmetric line shapes are due to resonant tunneling levels rather than traditional Kondo peaks, challenging previous explanations.

Contribution

It introduces a microscopic theory explaining STM line shapes via resonant tunneling levels, clarifying the role of Kondo coupling and challenging the Fano line shape interpretation.

Findings

Asymmetric line shapes are mainly governed by resonant tunneling levels.

Kondo peaks only appear with combined substrate and tip coupling.

Most asymmetric line shapes do not involve the Kondo effect.

Abstract

The observation of the Kondo effect in mesoscopic systems under bias$^{1,2}$ has opened a new chapter in the physics of the Kondo phenomenon. Various types of $dI/dV$, where $I$ and $V$ denote current and source-drain (s-d) bias, respectively, line shapes have been measured by scanning tunneling microscopy (STM)$^{1,3-11}$. However, explanation by single Fano line shape$^{1,12-16}$ is not relevant and even misleading. Here, we provide consistent explanations for various asymmetric and symmetric line shapes in terms of a microscopic theory that shows the creation of two resonant tunneling levels (RTLs) when bias is applied$^{17}$. One side Kondo coupling between adatom and substrate does not create Kondo peak that appears only when the system has an overall Kondo coupling including both substrate and tip. The structure of an asymmetric line shape is mostly governed by the RTL peaks. Therefore, Kondo effect is negligible in most asymmetric line shapes.