The influence of device geometry on many-body effects in quantum point contacts: Signatures of the 0.7 anomaly, exchange and Kondo

TL;DR

This study investigates how the geometry of quantum point contacts influences many-body electron effects like the 0.7 anomaly, exchange interactions, and Kondo signatures, revealing correlations and dependencies through experimental and electrostatic modeling.

Contribution

It demonstrates the relationship between device geometry and many-body effects in QPCs, highlighting the role of exchange interactions and their correlation with the 0.7 anomaly.

Findings

Enhanced g-factor correlates with subband spacing.

Zero-field splitting relates to the 0.7 anomaly and exchange effects.

Kondo signatures show no clear geometric dependence.

Abstract

The conductance of a quantum point contact (QPC) shows several features that result from many-body electron interactions. The spin degeneracy in zero magnetic field appears to be spontaneously lifted due to the so-called 0.7 anomaly. Further, the g-factor for electrons in the QPC is enhanced, and a zero-bias peak in the conductance points to similarities with transport through a Kondo impurity. We report here how these many-body effects depend on QPC geometry. We find a clear relation between the enhanced g-factor and the subband spacing in our QPCs, and can relate this to the device geometry with electrostatic modeling of the QPC potential. We also measured the zero-field energy splitting related to the 0.7 anomaly, and studied how it evolves into a splitting that is the sum of the Zeeman effect and a field-independent exchange contribution when applying a magnetic field. While this exchange contribution shows sample-to-sample fluctuations and no clear dependence on QPC geometry, it is for all QPCs correlated with the zero-field splitting of the 0.7 anomaly. This provides evidence that the splitting of the 0.7 anomaly is dominated by this field-independent exchange splitting. Signatures of the Kondo effect also show no regular dependence on QPC geometry, but are possibly correlated with splitting of the 0.7 anomaly.