🔬 mesoscale physics

Vortex-mediated spin current injection into two-dimensional superconductor NbSe2

This study provides the first direct experimental evidence of vortex-mediated spin current injection in an NbSe2/LAFO heterostructure, demonstrating that vortices in the liquid phase act as efficient carriers of spin angular momentum through a sign-reversed thermoelectric signal, thereby establishing a new pathway for superconducting spintronic devices.

Original authors: Meng Yang, Xiaolong Yin, Jingjing Liu, Yifeng Chen, Qinwu Gao, Hongxing Zhu, Danni Huang, Lang Chen, Shuo-Ying Yang, Junxue Li

Published 2026-09-18

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Meng Yang, Xiaolong Yin, Jingjing Liu, Yifeng Chen, Qinwu Gao, Hongxing Zhu, Danni Huang, Lang Chen, Shuo-Ying Yang, Junxue Li

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 world of modern electronics, information is often carried by the flow of electric charge, but there is a parallel universe of data transmission that relies on a different property: spin. Spin is an intrinsic form of angular momentum possessed by electrons, acting like a tiny internal compass needle that can point up or down. Scientists have long dreamed of building devices that use this spin to carry information, a field known as spintronics, because it promises faster speeds and far less energy waste than current technology. However, a major obstacle has always been the material used to conduct electricity. In ordinary metals, spin information travels only a short distance before it gets scrambled. In superconductors, materials that conduct electricity with zero resistance, the problem is even more severe. Inside a superconductor, electrons pair up to form what are called Cooper pairs. These pairs lock their spins in opposite directions, canceling each other out so that the pair carries no net spin at all. For decades, this meant that superconductors were dead ends for spin-based information, unless researchers could find a way to break this rule or find a different carrier entirely.

A team of researchers at the Southern University of Science and Technology in Shenzhen has now identified a new way to move spin information through a superconductor, bypassing the limitations of the electron pairs themselves. They focused on a specific type of defect that appears in superconductors when a magnetic field is applied, known as a vortex. Imagine a superconductor as a calm, frozen lake where the water cannot flow. When a magnetic field pushes through, it creates tiny, swirling whirlpools in the ice. These whirlpools are the vortices. While previous studies focused on how to stop these whirlpools from moving to keep the superconductor efficient, this team realized that if they could make the whirlpools move in a specific way, they could carry spin information with them. By creating a sandwich of two different materials—a superconducting crystal and a magnetic insulator—they managed to inject spin into the superconductor and watch it travel via these moving vortices.

The experiment involved a thin sheet of a superconductor called niobium diselenide, which was placed directly on top of a magnetic insulator known as lithium aluminum iron oxide. The researchers heated one side of this stack and cooled the other, creating a temperature difference that ran vertically through the layers. In a normal superconductor without the magnetic layer, this temperature difference causes the vortices to drift from the hot side to the cold side, generating a small electrical voltage. This is a known effect called the vortex Nernst effect. However, in the new setup with the magnetic insulator, something unexpected happened. The magnetic layer injected a stream of spin into the superconductor. The researchers found that this injected spin pushed the vortices to move in the exact opposite direction of the heat-driven flow. As a result, the electrical voltage generated by the moving vortices flipped its sign, becoming the opposite of what was seen in the control samples. This sign reversal was the key signature that spin was being carried by the vortices, proving that the magnetic layer was successfully driving the motion of these defects.

To be certain that this effect was truly coming from the vortices and not from some other source, the team carefully mapped out the conditions under which the signal appeared. They discovered that the effect only existed when the magnetic field was strong enough to melt the orderly arrangement of vortices into a chaotic, liquid-like state where they could move freely. If the magnetic field was too weak, the vortices were stuck in place, and no signal appeared. If the temperature was raised above the point where the material became a superconductor, the signal vanished completely. This last observation was crucial because it ruled out the possibility that the signal was caused by loose, unpaired electrons, known as quasiparticles, which can carry spin but only exist when the material is not fully superconducting. By showing that the effect disappeared exactly when the superconducting state ended, the researchers confirmed that the vortices themselves were the sole carriers of the spin information.

The findings suggest that vortices are not just obstacles to be suppressed, but active participants in spin transport. The study indicates that these swirling defects can act as highly efficient messengers, capable of carrying spin angular momentum over relatively long distances within a superconductor. This discovery opens a new path for designing superconducting spintronic devices, where information could be moved with minimal energy loss. The researchers demonstrated that by simply controlling the temperature and magnetic field, they could direct the flow of spin through the material. While the work is still in the early stages of proving the concept, it establishes a clear mechanism for how spin can be injected and transported in a superconductor, turning a previously overlooked feature of these materials into a potential tool for the next generation of computing technology.

Technical Summary: Vortex-Mediated Spin Current Injection into Two-Dimensional Superconductor NbSe2

Problem Statement
The injection of pure spin current into superconductors remains a significant challenge in the field of superconducting spintronics. In conventional superconductors, Cooper pairs form spin singlets ( $S=0$ ) and cannot carry spin angular momentum. While spin-triplet supercurrents ( $S=1$ ) offer a potential solution, intrinsic triplet superconductors are rare, and inducing robust triplet pairs at interfaces is experimentally difficult. Consequently, previous experimental reports of spin injection have largely been attributed to thermally excited, spin-polarized quasiparticles. Although theoretical studies propose that vortices—topological defects in type-II superconductors—can act as carriers of spin angular momentum through mechanisms like spin-vorticity transmutation or the inverse vortex spin Hall effect (IVSHE), direct experimental evidence for vortex-mediated spin current injection has been lacking. A primary difficulty lies in distinguishing the thermoelectric signal generated by vortex-mediated spin injection from the conventional vortex Nernst effect (VNE), as both arise from vortex motion under thermal gradients.

Methodology
To address this, the authors fabricated a heterostructure consisting of a two-dimensional Ising superconductor, Niobium Diselenide (NbSe2), and a ferrimagnetic insulator, LiAl2Fe3O8 (LAFO). The use of a 2D superconductor and van der Waals assembly was chosen to ensure a clean interface and avoid the interlayer interdiffusion common in high-temperature growth methods. LAFO was selected for its low magnetic damping and insulating nature, which eliminates spurious contributions from the anomalous Nernst effect in the magnetic layer.

The experimental setup involved a longitudinal spin Seebeck geometry. An out-of-plane temperature gradient ( $\nabla T$ along the $z$ -axis) was applied using a top heater, while an in-plane magnetic field ( $B$ along the $y$ -axis) was swept. The resulting thermoelectric response was measured as a second-harmonic voltage ( $V_{2\omega}$ ) along the $x$ -axis. This configuration allows for the detection of out-of-plane vortex dynamics via the Josephson relation ( $E = B \times v$ ). Control devices using NbSe2 on nonmagnetic SiO2/Si substrates were fabricated to measure the conventional VNE. The study systematically mapped the field-temperature phase diagram and compared the thermoelectric signals between the NbSe2/LAFO bilayer and the single-layer NbSe2 reference.

Key Contributions and Results

Sign Reversal of Thermoelectric Signal: The most critical finding is that the NbSe2/LAFO bilayer exhibits a pronounced thermoelectric peak near the upper critical magnetic field ( $H_{c2}$ ) that has the opposite sign compared to the conventional VNE observed in the single-layer NbSe2. This sign reversal is robust across multiple devices.
Vortex Liquid Phase Exclusivity: By mapping the phase diagram, the authors demonstrate that the thermoelectric signal (both the VNE and the spin-injection signal) appears exclusively in the vortex liquid phase (between the vortex-lattice-melting field $H_{melt}$ and $H_{c2}$ ). The signal vanishes in the vortex solid phase (where pinning suppresses motion) and above the superconducting transition temperature ( $T_c$ ).
Exclusion of Quasiparticle Contributions: The observation that the thermoelectric signal vanishes above $T_c$ rules out the contribution of thermally excited quasiparticles, which typically generate Nernst signals above the superconducting transition in bulk materials.
Mechanism Identification: The sign reversal is interpreted as evidence that the vortex flow induced by spin current injection is opposite to the flow driven by the temperature gradient. This is consistent with theoretical mechanisms where spin accumulation gradients drive vortex motion (via spin-vorticity transmutation or IVSHE) in a direction opposing the thermal force.

Significance
The paper establishes vortices as efficient carriers of spin information within superconductors. By demonstrating that spin current can be injected from a magnetic insulator into a superconductor and transported via mobile vortices, the work provides direct experimental evidence of vortex-mediated spin current injection. This finding distinguishes vortex-based transport from quasiparticle-based transport and opens a new route for developing low-dissipation superconducting spintronic devices that utilize the topological stability and high effective spin Hall angle of vortices. The results suggest that superconducting vortices can serve as effective converters between spin and charge responses, potentially enabling new functionalities in spintronic applications.

Technical Summary: Vortex-Mediated Spin Current Injection into Two-Dimensional Superconductor NbSe2

More like this