Superconductor junctions blur the line between two quantum effects

Researchers have measured the full differential conductance matrix, defined as G_ij = ∂ I_i/∂ V_j, moving beyond standard linear response measurements in three-terminal normal-superconductor-normal devices. These devices are based on topological insulator nanowires, and the work reveals how crossed Andreev reflection and elastic co-tunneling, two competing quantum processes, respond to applied voltages. Specifically, the team observed that crossed Andreev reflection is enhanced when V₁ = V₂, while elastic co-tunneling is enhanced when V₁ = -V₂, offering a potential pathway to control these subtle quantum effects.

Bias Voltage Symmetry Explained by Self-Gating Effects

Researchers at the Osipyan Institute of Solid State Physics, Russian Academy of Sciences, and HSE University analyzed transport measurements of three-terminal normal-superconductor-normal (NSN) devices constructed from topological insulator nanowires, focusing on how bias voltages influence the flow of electrons and holes. Their analysis challenges interpretations of earlier work suggesting a straightforward control mechanism for quantum processes within these devices, revealing a more complex interplay of effects.

The initial interpretation proposed that this bias voltage symmetry offered a means to systematically favor one quantum process over the other; however, the new analysis suggests this symmetry arises from a “fine tuned self-gating effect” rather than intentional control. This self-gating, a bias-related change in the potential landscape within the device, complicates the relationship between applied voltage and observed conductance, introducing an additional layer of influence beyond the expected crossed Andreev reflection and elastic co-tunneling contributions.

The researchers state that determining the relative values of crossed Andreev reflection and elastic co-tunneling probabilities solely from transport measurements is fundamentally limited. They explain that the measured conductance matrix, G_ij, can remain constant even if the probabilities of local and non-local electron transfer change, particularly when moving beyond the linear response regime. This means that observing a specific conductance value does not definitively indicate the dominance of either crossed Andreev reflection or elastic co-tunneling, as previously suggested.

Their analysis, based on the scattering matrix formalism, reveals that the observed bias symmetry isn’t a direct consequence of controlling crossed Andreev reflection and elastic co-tunneling, but rather a result of the complex interplay of these processes and the self-gating effect. They state that the hypothesis about the origin of the bias voltages impact on the non-local conductances raised in a previous reference is not justified and an alternative explanation of the data is needed.

Differential Conductance Equations Relate Crossed Andreev Reflection and Elastic Co-tunneling

Recent advances in superconductivity research have focused on meticulously characterizing electron transport in nanoscale normal-superconductor-normal (NSN) devices, but distinguishing between competing quantum mechanical processes within these systems has proven remarkably difficult. Researchers are now challenging interpretations of experiments designed to differentiate between crossed Andreev reflection (CAR) and elastic co-tunneling (ECT), two mechanisms governing electron flow across superconducting junctions. A new analysis indicates that previously attributed bias-voltage control over these processes may stem from an unanticipated effect: self-gating within the device itself.

This detailed measurement is crucial because most conductance studies rely on simplifying assumptions that may obscure subtle effects. E.S. Tikhonov and V.S. Khrapai, in conversation with the author, observed indications that crossed Andreev reflection is enhanced for V₁ = V₂, whereas elastic co-tunneling is enhanced for V₁ = -V₂, in analogy with experiments on quantum dot devices, and used the former combination to investigate the gate voltage dependence of G_ij.

Based on the sign of the non-local conductance, Feng et al. made conclusions about the dominance of crossed Andreev reflection or elastic co-tunneling, and the case of the dominant crossed Andreev reflection was interpreted as evidence of unusually long-range crossed Andreev reflection. Here, the authors state that the interpretation of the experiment is misleading in two respects. First, the bias voltages impact the non-local differential conductance randomly, rather than systematically, and the bias symmetry of the non-local conductance can be explained by a fine tuned self-gating effect.

Second, complete knowledge of the G_ij is insufficient to make conclusions about the relative values of the crossed Andreev reflection and elastic co-tunneling probabilities. The researchers emphasize that correlating the sign of G₁₂ with the dominance of crossed Andreev reflection or elastic co-tunneling is ambiguous. The analysis highlights a fundamental limitation: determining the relative contributions of crossed Andreev reflection and elastic co-tunneling requires more than just transport measurements.

The team found that complete knowledge of the G_ij matrix is insufficient to conclude the dominance of either process. They suggest that future studies should incorporate additional measurements, such as shot noise analysis, to provide a more complete picture of electron transport in these complex nanoscale devices.

Non-Local Conductance Fails to Distinguish Crossed Andreev Reflection from Elastic Co-tunneling

E.S. Khrapai and others have challenged interpretations of recent experiments examining quantum effects in nanoscale devices. The team’s work scrutinizes claims made by Feng et al. regarding the ability to control these processes through applied bias voltages. Feng et al. Second, complete knowledge of the G_ij is insufficient to make conclusions about the relative values of the crossed Andreev reflection and elastic co-tunneling probabilities, particularly regarding the dominance of one of them.

They found that while Feng et al. The researchers emphasize that the sign of the conductance, G₁₂ > 0, does not necessarily imply a stronger crossed Andreev reflection process. The researchers emphasize that the sign of the conductance, G₁₂ > 0, does not necessarily imply a stronger crossed Andreev reflection process.