Turning a Topological Material into an STM Tip: Seeing Altermagnetism at Atomic Scale

Forum topic · QianXun · 2026-10-15

Summary

A team at the University of Science and Technology of China, led by Chen Xianhui and Wang Zhenyu, has invented a new type of spin-sensitive scanning tunneling microscope (STM) tip made from bismuth bromide, a quantum spin Hall topological material. Because the helical edge states on such a tip lock spin to momentum, flipping the bias voltage polarity reverses the tunneling direction and the spin selectivity—no magnetic tip or external field needed. The tip must be sharpened by focused ion beam milling to atomic resolution (roughly the ten-thousandth of a human hair, ~10 nm scale) while preserving its topological edge-state electronic structure. After validation on iron telluride and the kagome magnet iron germanium, the team used the tip to image potassium vanadium selenoxide, a layered altermagnet, resolving two vanadium sublattices and observing a 90-degree rotation of quasiparticle interference signals when the bias polarity flipped—direct real-space evidence of d-wave, direction-alternating spin splitting. The results, published as an accelerated article preview in Nature, offer a new tool for studying altermagnets relevant to faster, more energy-efficient spintronic devices.

Using Helical Edge States as a Tool, Not Prey

In quantum spin Hall materials, the bulk is insulating but the edges host special conducting channels: electrons moving in opposite directions carry opposite spins ("spin-momentum locking"), known as helical edge states. Researchers traditionally *search* for such edge states — but a team from the University of Science and Technology of China (USTC), including Chen Xianhui and Wang Zhenyu, flipped the roles: they made the edge states part of the microscope itself. The work, with collaborators Xiang Bin (USTC) and Liu Junwei (HKUST), was published on October 13 as an accelerated article preview in *Nature*, titled "Atomic-scale spin sensing of a 2D d-wave altermagnet via helical tunneling."

Scanning tunneling microscopy relies on an atomically sharp tip near the sample surface; electrons tunnel across the vacuum gap, and the resulting current reveals atomic positions and electronic states. Conventional tips are non-magnetic metals and cannot distinguish electron spin. The team's bold idea: since helical edge states tie spin to direction of motion, switching the bias polarity between sample and tip reverses the tunneling electron direction — and thus the spin signature. No external magnetic field, no magnetic tip, just voltage polarity to read spin.

A Tip One Ten-Thousandth of a Hair Thick

Fabricating a tip from a bismuth bromide single crystal via focused ion beam milling was the technical crux: the apex must resolve single atoms (on the order of one ten-thousandth of a 0.1 mm human hair, ~10 nm scale) while preserving the topological edge-state electronic structure.

The tip then had to prove itself:

On iron telluride, stripe contrast reversed with the sample bias polarity.
On the kagome magnet iron germanium, interlayer contrast changed markedly with field-induced spin reorientation.

Finally the team targeted potassium vanadium selenoxide, a layered altermagnet. They first mapped the surface at single-atom resolution, distinguishing two types of vanadium sites. Then, extracting quasiparticle interference ripples scattered by defects, they found a huge contrast difference between two perpendicular directions — and when the sample bias flipped sign, the dominant signal direction rotated 90 degrees, matching the expectation of opposite spin characters along the two axes.

What Is Altermagnetic Order?

Think of atomic magnetic moments as a row of tiny compass needles. In ferromagnets they all align (macroscopic magnetism); in antiferromagnets they cancel pairwise. Altermagnets take a third path: moments cancel overall, but the two spin populations have different energies, and which spin is higher alternates with the direction of electron motion. This could enable faster, more energy-efficient devices, denser information storage, and reduced magnetic crosstalk.

The challenge is identification: bulk average magnetism cannot answer how moments are arranged, how the sublattices relate, or whether spin-energy splitting alternates with direction. The edge-state tip provides cross-validation: an atomic-scale real-space map, momentum-space d-wave anisotropy that flips sign with bias, interference contrast differences, and field-dependent spectroscopy.

Peer review was demanding: after first submission on December 23, 2025, *Nature* referees questioned why a helical edge state can sense spin at all. The team measured 5 different sample classes and designed 4 control experiments, ruling out lattice stress, tip-shape asymmetry, and orbital-selection biases as alternative explanations before acceptance in October 2026.

| Evidence layer | Measured | What it shows | |---|---|---| | Real space | Atomic map of two vanadium sublattices | Checkerboard oppositely arranged moments | | Momentum space | Interference contrast and 90° rotation | Opposite spin character along two axes | | Spectroscopy | Opposite energy shifts along two orthogonal axes | Direction-alternating spin splitting | | Controls | 5 sample classes, 4 control experiments | Excludes non-spin explanations |

Sources and Caveats

Details come from a *China Science Daily* report (Oct 14, via Sina), including team info, quotes, and submission/acceptance dates; the Nature paper appeared as an accelerated article preview (Oct 13).
The full paper was not independently verified; experimental details reflect the team's public descriptions.
Accelerated article previews are preliminary; final formatting and review record await the formal version.
The "one ten-thousandth of a hair" analogy corresponds to roughly the 10 nm scale.
Prospects (unconventional superconductivity, 2D magnets, moiré devices) are stated future directions without experimental support yet.