🔬 mesoscale physics

Interedge backscattering in quantum spin Hall-based NS and SNS junctions

This paper utilizes a microscopic Bernevig-Hughes-Zhang model to investigate how interedge backscattering in quantum spin Hall-based NS and SNS junctions, driven by factors like barrier geometry and disorder, deviates from quantized conductance and hybridizes Andreev branches to modify superconducting interference patterns.

Original authors: Cajetab Heinz, Patrik Recher, Fernando Dominguez

Published 2026-08-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Cajetab Heinz, Patrik Recher, Fernando Dominguez

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the hidden world of quantum materials, scientists are searching for a special kind of electricity that flows without resistance and without losing its way. This search focuses on a state of matter called the quantum spin Hall effect, where electrons are forced to travel along the very edges of a material, like cars stuck in a single-lane highway. In this unique highway, the direction an electron moves is locked to its spin, a tiny internal magnetic property. This locking creates a powerful protection: an electron cannot simply bounce backward and get stuck because doing so would require it to flip its spin, which nature forbids in this setting without outside interference. This protection is the key to a new kind of superconductivity, a state where electricity flows with zero energy loss. When scientists attach a standard superconductor to the edge of such a material, they hope to create exotic particles that could serve as the building blocks for future quantum computers. However, real-world devices are never perfect, and the interfaces where these materials meet are often messy, creating unexpected pathways that can ruin the delicate protection scientists rely on.

A team of researchers at technical universities in Germany has investigated exactly how these messy interfaces affect the flow of electricity in these hybrid devices. They focused on a specific problem: what happens when the two opposite edges of the quantum spin Hall material, which should be completely independent, accidentally talk to each other through a conducting barrier near the superconductor. In an ideal world, an electron traveling on the top edge would never reach the bottom edge, and vice versa. But in the actual devices being built, the materials used to connect the superconductor can form a bridge. The researchers used detailed computer simulations to model these bridges, treating the electrons as waves moving through a complex landscape of energy levels. They discovered that this bridge allows electrons to hop from one edge to the other, effectively creating a backdoor that lets them reverse direction without breaking the fundamental rules of the system.

The team found that this backdoor has a direct and measurable impact on how well electricity conducts through the device. In a perfect setup, the electrical conductance at zero voltage is a fixed, universal number, a standard value that acts as a fingerprint for the quantum state. The researchers showed that when the edges are coupled through the barrier, this perfect number drops. The amount it drops depends on how wide the barrier is, how strong the superconducting connection is, and how well the different materials align with each other. They also discovered that if the material contains impurities or disorder, the situation becomes even more complex. Surprisingly, adding a bit of disorder can actually make the conductance higher again, not because the system is working better, but because the disorder causes the electron waves to interfere with themselves in a way that cancels out the backscattering. This effect is fragile; applying a magnetic field or a small voltage pushes the system back to a state where the edges are coupled and the conductance drops.

When the researchers applied this understanding to a more complex device, a junction where a normal section is sandwiched between two superconductors, they saw how the coupling changes the rhythm of the supercurrent. In these junctions, the supercurrent usually oscillates in a predictable pattern as the phase difference between the two superconductors changes. The coupling between the edges disrupts this pattern, opening small gaps in the energy levels where the current should flow freely. These gaps appear at specific points in the cycle and their size depends on the chemical makeup of the device. Crucially, the researchers found that these gaps do not open and close at the same time; when one gap is large, the other is small, and they oscillate out of sync. This behavior leaves a distinct signature in the magnetic interference pattern of the device. Instead of a smooth, repeating wave, the pattern shows a "even-odd" effect where every other peak is suppressed. The researchers noted that the specific peaks that get suppressed depend on the relative size of the two gaps, offering a way to measure the hidden dynamics of the edge coupling.

The study concludes that the simple picture of independent edges is often an oversimplification for real devices. The coupling between edges is not a rare accident but a common feature controlled by the geometry and material properties of the interface. By measuring the conductance in a simple junction or the interference pattern in a more complex one, scientists can now tell if their device is behaving as an ideal quantum system or if it is being influenced by these hidden bridges. This work provides a practical guide for experimentalists, showing that the absence of a perfect signal does not necessarily mean the device is broken, but rather that it is revealing the complex, microscopic details of how its parts are connected. The findings suggest that to build reliable quantum devices, engineers must carefully control the barrier regions and material alignment to either minimize this coupling or account for it precisely, turning a potential obstacle into a measurable feature of the system.

Technical Summary: Interedge Backscattering in Quantum Spin Hall-based NS and SNS Junctions

Problem Statement
The identification of Majorana bound states (MBS) in quantum spin Hall (QSH)–superconductor hybrid junctions remains experimentally challenging. A primary obstacle is the lack of microscopic control over the superconductor–QSH interface. In realistic devices, band alignment and work-function mismatches often create a conducting barrier region at the interface. While idealized models assume well-separated helical edge states where time-reversal symmetry prevents backscattering within a single edge, a conducting barrier provides a nonlocal path for particles to propagate from one edge to the opposite edge. This interedge coupling can qualitatively alter transport signatures, potentially obscuring the fractional Josephson effect and other topological features. The specific microscopic conditions governing this coupling, and how it depends on interface geometry, band alignment, and disorder, were not fully understood.

Methodology
The authors employ a microscopic tight-binding formulation of the Bernevig–Hughes–Zhang (BHZ) model within the Bogoliubov–de Gennes (BdG) formalism. They model two configurations:

NS Junctions: A QSH edge adjacent to a superconductor via a conducting barrier.
SNS Junctions: A QSH region sandwiched between two superconductors, separated by a normal region and barriers.

The study neglects spin-nonconserving terms (Rashba and bulk-inversion-asymmetry contributions) to focus on the helical edge structure relevant to interedge coupling. The transport properties are calculated using a recursive Green's function technique to determine:

Differential conductance $G(V)$ for NS junctions.
Andreev bound state (ABS) spectra and equilibrium supercurrents for SNS junctions.
Superconducting quantum interference (SQI) patterns (critical current vs. magnetic flux).

The analysis systematically varies parameters including barrier width ( $W$ ), barrier length ( $L_b$ ), chemical potentials in the barrier and superconductor ( $C_b, C_s$ ), superconducting pairing amplitude ( $\Delta_0$ ), and disorder strength ( $\lambda$ ).

Key Contributions and Results

Mechanism of Interedge Coupling: The authors demonstrate that a conducting barrier adjacent to the superconductor allows quasiparticles to backscatter from one helical edge to the opposite edge without requiring magnetic perturbations or explicit time-reversal symmetry breaking. In NS junctions, this manifests as a deviation from the quantized zero-bias Andreev conductance ( $G = 4e^2/h$ ).
Dependence on Barrier Properties: The strength of the coupling is governed by an effective coherence length $\xi_{eff} = \hbar v_{F,barrier} / \Delta_{eff}$ , where $\Delta_{eff}$ is the induced gap in the barrier. The coupling decays exponentially with barrier width. The study identifies that orbital and Fermi-velocity mismatches between the barrier and superconductor significantly suppress the induced gap and conductance.
Role of Disorder: Contrary to intuition, disorder within the barrier can enhance the zero-bias conductance in the presence of time-reversal symmetry. This arises from the constructive interference between electron and time-reversed hole trajectories (analogous to reflectionless tunneling). This enhancement is suppressed by perpendicular magnetic fields or finite bias voltages, which introduce relative phase shifts.
Impact on SNS Junctions: In Josephson junctions, interedge coupling lifts the protection of level crossings at time-reversal-invariant phase differences ( $\phi = 0$ $ϕ = 0$ and $\phi = \pi$ $ϕ = π$ ), opening energy gaps denoted as $\delta_0$ $δ_{0}$ and $\delta_\pi$ $δ_{π}$ .
- These gaps oscillate as functions of the chemical potential and junction geometry, driven by the dynamical phase accumulated during propagation through the normal region.
- Crucially, $\delta_0$ and $\delta_\pi$ oscillate out of phase; one gap can close while the other remains finite. Thus, the observation of a single closed gap does not necessarily imply decoupled edges.
Superconducting Quantum Interference (SQI): The coupling modifies the SQI pattern. In the edge-dominated regime, it produces an "even–odd" modulation where every other lobe of the SQUID-like pattern is suppressed. The parity of the suppressed lobes correlates with the ratio $\delta_0/\delta_\pi$ . In the presence of reflection symmetry, a "sweet spot" exists where $\delta_\pi = 0$ while $\delta_0 \neq 0$ , a condition favorable for observing the fractional Josephson effect.
Bulk Band Contributions: When the chemical potential enters the bulk bands, the interference pattern evolves toward a Fraunhofer-like response. However, residual edge contributions remain visible as local minima and non-monotonic decay in the lobes. Disorder tends to restore the standard $\Phi_0$ periodicity, masking the even–odd effect.

Significance and Claims
The paper provides a microscopic explanation for how conducting interface regions modify the transport and spectral properties of QSH-superconductor hybrids. The authors claim that their investigation sheds light on the specific details controlling the coupling of helical edge states in actual devices.

The work establishes that:

Deviations from quantized conductance in NS junctions serve as a diagnostic for interedge coupling magnitude.
The Josephson spectrum and SQI patterns provide complementary information, being sensitive not only to the magnitude of coupling but also to the associated scattering phase.
The absence of an even–odd lobe modulation in SQI patterns should not be taken as definitive evidence of decoupled edges, as disorder can restore standard periodicity.

Ultimately, the study offers a practical framework for assessing departures from the idealized independent-edge description in quantum spin Hall–superconductor junctions, linking microscopic interface properties to observable macroscopic phenomena.

More like this