Conditional statistics of electron transport in interacting nanoscale conductors

TL;DR

This paper develops a theoretical model and presents experimental evidence for informational backaction effects in electron transport through quantum dots, revealing how measurement influences transport statistics even with noninvasive detectors.

Contribution

It introduces a model predicting informational backaction in quantum dot transport and demonstrates experimental validation of these effects, including conditional noise enhancement.

Findings

Experimental evidence of informational backaction

Observation of conditional noise enhancement

Agreement between theory and experiment

Abstract

Interactions between nanoscale semiconductor structures form the basis for charge detectors in the solid state. Recent experimental advances have demonstrated the on-chip detection of single electron transport through a quantum dot (QD). The discreteness of charge in units of e leads to intrinsic fluctuations in the electrical current, known as shot noise. To measure these single-electron fluctuations a nearby coherent conductor, called a quantum point contact (QPC), interacts with the QD and acts as a detector. An important property of the QPC charge detector is noninvasiveness: the system physically affects the detector, not visa-versa. Here we predict that even for ideal noninvasive detectors such as the QPC, when a particular detector result is observed, the system suffers an informational backaction, radically altering the statistics of transport through the QD as compared to the unconditional shot noise. We develop a theoretical model to make predictions about the joint current probability distributions and conditional transport statistics. The experimental findings reported here demonstrate the reality of informational backaction in nanoscale systems as well as a variety of new effects, such as conditional noise enhancement, which are in essentially perfect agreement with our model calculations. This type of switching telegraph process occurs abundantly in nature, indicating that these results are applicable to a wide variety of systems.