Positive current cross-correlations in a highly transparent normal-superconducting beam splitter due to synchronized Andreev and inverse Andreev reflections

TL;DR

This paper predicts positive current cross-correlations in a highly transparent NSN device due to synchronized Andreev and inverse Andreev reflections, challenging the usual association with crossed Andreev reflection.

Contribution

It introduces a new interpretation of positive current correlations as resulting from synchronized Andreev processes, not solely from crossed Andreev reflection, in highly transparent NSN devices.

Findings

Positive correlations occur at high transparency due to synchronized Andreev reflections.

At intermediate transparency, correlations become negative.

Analytic and microscopic models clarify the physical processes involved.

Abstract

Predictions are established for linear differential current-current cross-correlations dSab/dV in a symmetrically biased three-terminal normal metal-superconductor-normal metal (NSN) device. Highly transparent contacts turn out to be especially interesting because they feature positive dSab/dV. At high transparency, processes based on Crossed Andreev Reflection (CAR) contribute only negligibly to the current and to dSab/dV. Under these circumstances, current-current cross-correlations can be plausibly interpreted as a coherent coupling between the two NS interfaces in the form of synchronized Andreev and inverse Andreev reflections, corresponding to the process where a pair of electron-like quasi-particles and a pair of hole-like quasi-particles arrive from the normal electrodes and annihilate in the superconductor. Hence, positive dSab/dV does not automatically imply CAR. For tunnel contacts, dSab/dV is positive because of CAR. In between these two extremities, at intermediate transparencies, dSab/dV is negative because both processes which cause positive correlations, occur only with small amplitude. We use scattering theory to obtain analytic expressions for current and noise, and microscopic calculation using a tight binding model in order to obtain a clear interpretation of the physical processes.