Large-scale superconductivity-induced conductance suppression in mesoscopic normal-superconducting structures

TL;DR

This paper presents a theory explaining large conductance suppression in mesoscopic normal-superconducting structures, highlighting an instability in 4-probe measurements and predicting sample-specific fluctuations beyond classical theory.

Contribution

It introduces a theoretical framework for understanding large conductance drops in hybrid structures, emphasizing the role of measurement configuration and sample fluctuations.

Findings

4-probe conductance measurements reveal an instability absent in 2-probe setups.

Ensemble average of conductance difference vanishes, aligning with quasi-classical theory.

Sample-specific fluctuations can cause significant negative conductance differences, beyond classical predictions.

Abstract

Experiments on hybrid superconducting normal-metal structures have revealed that even in the absence of tunnel junctions the onset of superconductivity can lead to a decrease in the electrical conductance by an amount many orders of magnitude greater than $e^2/h$. In this Letter we provide a theory of this phenomenon which shows that it originates from an instability in 4 - probe conductance measurements which is absent from 2-probe measurements. We compare the zero-bias,zero-temperature 4-probe conductances $G_{N}$ and $G_{S}$ of a normal diffusive metal in contact with a superconductor in both the normal (N) and superconducting (S) states respectively. In the absence of tunnel barriers, the ensemble average of the difference $\delta G=G_{S}-G_N$ vanishes, in agreement with quasi-classical theory. However we also predict that there exists macroscopic sample specific fluctuations in $\delta G$, which lie beyond quasi-classical theory and allow large negative values of $\delta G$ to occur.