Imaging short- and long-range magnetic order in a quantum anomalous Hall insulator

Using scanning superconducting quantum interference device microscopy, this study reveals that V-doped (Bi,Sb)$_2$Te$_3$ exhibiting the quantum anomalous Hall effect possesses a unique magnetic state characterized by the coexistence of local interactions within crystallographic grains and long-range ferromagnetic coupling between them, a behavior distinct from that of Cr-doped counterparts.

The Gist

Imagine a world where electricity flows without any resistance, like a superhighway with no traffic jams. This is the Quantum Anomalous Hall Effect (QAHE), a magical state of matter that scientists have been chasing for decades. It's so precise that it could one day be used as the ultimate ruler for measuring electricity, replacing the old, imperfect standards.

But here's the mystery: To make this "superhighway" work, scientists have to mix a special material (a topological insulator) with magnetic atoms. For a long time, they weren't sure how those magnetic atoms were behaving. Were they all marching in perfect lockstep (like a disciplined army), or were they just a chaotic crowd of individuals doing their own thing (like a mosh pit)?

This paper is like a high-tech detective story where the scientists use a super-sensitive camera to take a close-up look at the magnetic "crowd" inside a specific material made of Vanadium, Bismuth, and Antimony.

The Detective's Tool: The Magnetic Microscope

The scientists used a tool called a Scanning SQUID Microscope. Think of this as a tiny, ultra-sensitive "magnetic nose" on the tip of a needle. It hovers just above the material and sniffs out the tiny magnetic fields leaking out from the surface. It's so sensitive it can see magnetic patterns the size of a few hundred atoms.

The Big Discovery: A "Dual Nature"

The scientists found that the magnetic behavior in this material is a fascinating mix of two opposite personalities:

1. The "Village" Effect (Short-Range Order)
Imagine the material is made of tiny, distinct neighborhoods called grains. Inside each neighborhood, the magnetic atoms are very close-knit. They talk to their immediate neighbors and agree on a direction.

• The Analogy: It's like a small town where everyone knows each other. If the mayor (the magnetic field) tells the town to face North, everyone in that specific town turns North together.
• The Finding: The size of these magnetic "towns" (domains) matches the size of the physical crystal grains perfectly. This suggests that the physical boundaries of the crystal act like fences, keeping the magnetic influence local.

2. The "Chain Reaction" Effect (Long-Range Order)
Here is where it gets interesting. In some other materials (like the Chromium-doped ones studied previously), when you try to flip the magnetic direction, the towns flip randomly and independently. One town flips North, the next flips South, and the next flips East. It's chaotic.

• The Finding in This Paper: In this Vanadium material, the flipping is organized. When the scientists applied a magnetic field to flip the direction, they didn't see random flipping. Instead, they saw the "North" towns slowly expand, swallowing up the "South" towns.
• The Analogy: Imagine a game of "Tetris" or a spreading fire. Once a few blocks turn red, the red color spreads to the adjacent blocks, pushing the blue blocks back. The towns aren't acting alone; they are holding hands across the fences. If one town flips, it pulls its neighbor along with it.

Why Does This Matter?

This discovery solves a puzzle.

• The Puzzle: The material acts like a super-precise quantum ruler (which requires long-range order), but the magnetic domains are tiny and tied to crystal grains (which usually suggests short-range, chaotic order).
• The Solution: The material has both. The crystal grains create small, tight-knit magnetic groups, but these groups are strongly connected to each other, allowing them to act as one giant, coordinated unit when needed.

The Takeaway

Think of this material as a choir.

• In a bad choir, everyone sings their own song (superparamagnetism).
• In a perfectly disciplined choir, everyone is a single voice (perfect long-range order).
• In this material, you have small groups of singers (the grains) who are very tight-knit, but they are all listening to the same conductor and singing in harmony with the groups next to them.

This "best of both worlds" behavior explains why this material is so stable and precise for quantum measurements. It's not just a chaotic mess, and it's not a rigid robot; it's a flexible, cooperative system that knows how to hold its ground while still moving together. This understanding helps scientists build better quantum devices for the future.

Technical Summary

1. Problem Statement

The Quantum Anomalous Hall Effect (QAHE) in magnetically doped topological insulators (TIs) offers near-perfect quantization of Hall resistance, making it highly valuable for metrological applications. However, the microscopic magnetic origins of this effect remain ambiguous.

• The Conflict: Previous studies on materials like Cr-doped $(Bi,Sb)_2Te_3$ have reported conflicting magnetic behaviors, ranging from superparamagnetism (short-range interactions, random switching) to long-range ferromagnetic order.
• The Gap: While V-doped $(Bi,Sb)_2Te_3$ (VBST) has demonstrated precise QAHE quantization at ultra-low temperatures, the specific nature of its magnetic domain structure—specifically the interplay between crystallographic grain boundaries and magnetic exchange interactions—was unclear. It was unknown whether the magnetic order was dominated by local grain constraints or long-range coupling.

2. Methodology

The authors employed Scanning Superconducting Quantum Interference Device (SQUID) Microscopy (SSM) to image magnetic domains with high spatial resolution.

• Sample: A 9 nm thick layer of $V_{0.1}(Bi_{0.2}Sb_{0.8})_{1.9}Te_3$ grown via molecular beam epitaxy on Si(111), capped with Te to prevent degradation.
• Instrumentation: A nanoscale SQUID sensor (80 nm effective loop diameter) patterned at the tip of an AFM cantilever.
  • Measurement Mode: The SQUID measured the normal-to-plane stray field ($B_z$) and its gradient ($B^a_c \propto dB_z/dz$) via sinusoidal modulation of the sample height (178 Hz) and lock-in detection.
  • Conditions: Measurements were conducted at 5 K with SQUID-sample separations of 150–260 nm.
• Data Analysis & Reconstruction:
  • Stray Field Mapping: Images of $B^a_c$ were taken across a hysteresis loop (from saturation to reversal).
  • Magnetization Reconstruction: An iterative inverse propagation algorithm was used to reconstruct the magnetization map ($M_z$) from the measured stray fields. The algorithm assumed a binary magnetization state ($\pm M_{sat}$) and optimized parameters (distance $d$, filter cutoff $\lambda_M$, and saturation magnetization $M_{sat}$) to minimize the Mean Squared Error (MSE) between measured and calculated fields.
  • Comparative Modeling: The reconstructed magnetic domains were compared against Atomic Force Microscopy (AFM) topography data of the crystal grains. A random magnetization model was generated based on grain boundaries to test the correlation between grain structure and magnetic domains.
  • Micromagnetic Simulations: MuMax3 simulations were performed to model domain evolution under varying intergrain exchange stiffness ($A_{ex,inter}$).

3. Key Results

A. Magnetic Reversal Dynamics

• Domain Expansion: Unlike the random nucleation of reversed domains seen in Cr-doped samples (indicative of superparamagnetism), the VBST sample exhibited domain expansion. Reversal occurred through the gradual growth of existing domains along their edges, a hallmark of long-range ferromagnetic coupling.
• Hysteresis: The stray field hysteresis loop showed a coercive field of $\sim 200$ mT, consistent with transport measurements on similar samples (though slightly lower than the 250 mT observed at 4.2 K, likely due to the 5 K measurement temperature).

B. Correlation with Crystallographic Grains

• Size Similarity: The reconstructed magnetic domains were found to be of similar size (peaking at 85–100 nm) to the crystallographic rotational twin grains identified in AFM images.
• Dual Nature of Order:
  • Local Constraint: The domain size distribution matches the grain size, suggesting that grain boundaries act as pinning sites and that intra-grain exchange is strong.
  • Long-Range Coupling: Despite the grain-size limitation, the mechanism of reversal (cooperative expansion rather than independent switching) indicates significant inter-grain ferromagnetic coupling.
• Quantitative Validation: The reconstructed saturation magnetization was $M_{sat} \approx 2.3 \, \mu_B/\text{nm}^2$, corresponding to a magnetic moment of 1.4–1.8 $\mu_B$ per V dopant ion, aligning with theoretical and experimental values for V-dopants.

C. Simulation Insights

• Micromagnetic simulations confirmed that the observed "edge-expansion" reversal behavior corresponds to an intermediate regime of intergrain exchange stiffness ($A_{ex,inter} \approx 10^{-15} - 10^{-14} \, \text{J m}^{-1}$).
• This value is lower than the intra-grain stiffness but high enough to prevent independent superparamagnetic switching, facilitating cooperative domain growth.

4. Key Contributions

1. Direct Imaging: Provided the first direct visualization of magnetic domain formation and evolution in a V-doped QAHE insulator with precise quantization.
2. Resolution of Magnetic Nature: Resolved the debate on the magnetic state of VBST by demonstrating a coexistence of short-range and long-range order. The system is not purely superparamagnetic (as seen in some Cr-doped variants) nor a perfect single-domain ferromagnet; it is a granular ferromagnet where grain boundaries influence domain size but do not sever long-range coupling.
3. Methodological Advance: Successfully combined high-resolution SQUID microscopy with inverse reconstruction and micromagnetic modeling to extract quantitative magnetic parameters ($M_{sat}$, exchange stiffness regimes) from stray field data.

5. Significance

• Metrology: Understanding the magnetic stability and domain structure is crucial for the robustness of QAHE-based resistance standards. The findings suggest that while grain boundaries exist, the long-range coupling ensures a stable global magnetic alignment necessary for the QAHE.
• Material Design: The distinction between V-doped and Cr-doped systems highlights that the choice of magnetic dopant significantly alters the magnetic exchange landscape. V-doping appears to support stronger inter-grain coupling, which may be a key factor in achieving the high-precision quantization observed in VBST.
• Fundamental Physics: The study provides a concrete example of how topological transport properties emerge from a complex magnetic landscape where local crystallographic constraints and global ferromagnetic interactions compete and coexist.

In conclusion, the paper establishes that the V-doped $(Bi,Sb)_2Te_3$ system possesses a dual-nature magnetic ordering: magnetic domains are pinned by crystallographic grains (short-range), yet they are coupled via long-range ferromagnetic interactions that drive cooperative reversal, distinguishing it from the superparamagnetic behavior of Cr-doped counterparts.