NMR evidence for an antisite-induced magnetic moment on Bi in a topological insulator heterostructure MnBi$_2$Te$_4$/(Bi$_2$Te$_3$)$_n$

This study utilizes NMR and magnetization measurements to demonstrate that Mn antisites in MnBi$_2$Te$_4$/(Bi$_2$Te$_3$)$_n$ heterostructures induce an antiparallel magnetic moment on Bi atoms, creating a new ferromagnetic component that is crucial for understanding and engineering the Quantum Anomalous Hall Effect in these topological insulator systems.

The Gist

Imagine you have a magical, super-thin sandwich. This isn't a normal sandwich with ham and cheese; it's made of layers of atoms. Some layers are "magnetic" (like tiny compass needles that want to point in a specific direction), and some are "non-magnetic" (like plain bread that doesn't care about direction).

Scientists call this a Topological Insulator. It's a material that acts like an insulator on the inside (blocking electricity) but conducts electricity perfectly on its surface. When you add magnetism to this, you get a "superpower" material that could revolutionize computers, making them faster and using way less energy.

The star of this show is a material called MnBi2Te4 (let's call it the "Magnetic Layer"). But pure Magnetic Layers are tricky. To make them useful, scientists stack them with layers of a non-magnetic material called Bi2Te3 (the "Plain Layer").

Here is the simple breakdown of what this paper discovered, using some everyday analogies:

1. The "Self-Organizing" Sandwich

Usually, to make these sandwiches, scientists try to stack the layers perfectly by hand (like building a Lego tower). But in this study, they used a method called "self-organization." Imagine pouring a hot soup of atoms and letting them cool down. As they cool, they naturally settle into layers, but not perfectly. Some spots have 1 plain layer, some have 2, some have 3.

The scientists found that this "messy" sandwich actually has some hidden secrets that a perfectly neat one doesn't.

2. The "Rebellious" Spies (Antisites)

Inside the Magnetic Layers, the atoms are supposed to be orderly. The "Manganese" (Mn) atoms are the soldiers, and they are supposed to stand in neat rows, pointing up and down in an alternating pattern (Up-Down-Up-Down). This is called Antiferromagnetism.

However, the study found "spies" or "defects." Sometimes, a Manganese soldier sneaks into a spot meant for a Bismuth (Bi) atom. We call these Antisites.

• The Analogy: Imagine a dance floor where everyone is supposed to face the wall. Suddenly, a few dancers (the Mn spies) decide to face the opposite direction.
• The Discovery: These spies don't just stand there; they mess with the neighbors. They force the nearby Bismuth atoms (who usually don't have a magnetic personality) to start acting magnetic too!

3. The "Ghost" Magnetism on Bismuth

This is the biggest "Aha!" moment of the paper.

• Before: Scientists thought Bismuth atoms were like neutral bystanders. They didn't have a magnetic "voice."
• Now: The paper proves that because of the Mn spies, the Bismuth atoms do get a magnetic voice. They start pointing in a direction, creating a tiny magnetic field.
• The Analogy: It's like a quiet person in a room (Bismuth) suddenly starting to shout because a loud neighbor (the Mn spy) is standing right next to them. The loud neighbor's energy is "inducing" the quiet person to make noise.

The scientists used a technique called NMR (Nuclear Magnetic Resonance). Think of NMR as a super-sensitive microphone that can "hear" the tiny magnetic whispers of individual atoms. They heard the Bismuth atoms whispering, proving they had become magnetic.

4. The "Tug-of-War" (Spin-Flop)

The scientists put this sandwich in a giant magnet to see how it reacted.

• Low Magnet: The magnetic layers are fighting each other (Up vs. Down), canceling out.
• Medium Magnet (The Flip): At a certain point (around 3.5 Tesla), the external magnet gets strong enough to win the tug-of-war. The layers suddenly "flip" or "cant" (tilt) to align with the magnet.
• The Result: The scientists measured exactly how much the layers tilted. They found that the "messy" parts of the sandwich (where the plain layers are mixed in) change how hard it is to flip the switch.

Why Does This Matter? (The "So What?")

You might ask, "Why do we care about a few atoms sneaking around?"

1. The Quantum Anomalous Hall Effect (QAHE): This is a fancy term for a super-efficient way to move electricity without losing energy as heat. It's the "Holy Grail" for making super-fast, cool computers.
2. The Problem: To get this effect to work perfectly, the magnetic layers need to be perfectly balanced.
3. The Solution: This paper shows that the "defects" (the Mn spies) actually create a new type of magnetism (the Bismuth voice) that acts like a Ferromagnet (a permanent magnet).
   • The Analogy: If you are trying to build a bridge, you usually want perfect bricks. But sometimes, a slightly different brick (the defect) can actually help lock the structure together in a new, useful way.

In Summary:
This paper is like a detective story. The scientists used a "magnetic microphone" (NMR) to listen to a messy, self-made atomic sandwich. They discovered that the "mistakes" in the sandwich (atoms in the wrong spots) are actually creating a new, useful magnetic force on the Bismuth atoms. Understanding this helps engineers design better materials for the next generation of ultra-efficient electronics.

Technical Summary

1. Problem Statement

MnBi2Te4 (MBT) is the first intrinsic magnetic topological insulator (MTI), combining topologically protected surface states with intrinsic magnetic order. It is a primary candidate for realizing the Quantum Anomalous Hall Effect (QAHE) in zero external magnetic fields.

• The Challenge: While ideal MBT is an A-type antiferromagnet (AFM), real-world samples often contain structural defects and form heterostructures with non-magnetic Bi2Te3 (BT). Specifically, self-organized growth methods produce samples with varying numbers of BT quintuple layers ($n$) separating MBT septuple layers, leading to a complex mix of magnetic phases.
• The Knowledge Gap: Structural defects, such as Mn antisites (Mn atoms substituting Bi atoms), are known to exist but their precise impact on local magnetic interactions and the induction of magnetic moments on otherwise non-magnetic atoms (like Bi) remains poorly understood. Understanding these local interactions is critical for engineering QAHE devices, as defects can degrade the QAHE state.

2. Methodology

The authors employed a multi-modal approach to investigate a self-organized single crystal MnBi2Te4/(Bi2Te3)n heterostructure grown from a Mn0.81Bi2.06Te4.13 melt.

• Sample Characterization:
  • Transmission Electron Microscopy (TEM): Confirmed the sample is a single crystal with an average of $n=2$ (two Bi2Te3 quintuple layers) separating MBT septuple layers, though a distribution of $n$ values exists.
  • SQUID Magnetometry: Performed macroscopic magnetization measurements ($M$ vs. $H$) at 4.2 K with the field applied along the $c$-axis ([0001]) up to 7 T.
• Nuclear Magnetic Resonance (NMR):
  • Technique: Zero-field and field-swept NMR using a phase-sensitive spin-echo spectrometer at 4.2 K.
  • Probes: $^{55}$Mn (magnetic probe) and $^{209}$Bi (non-magnetic probe).
  • Conditions: External magnetic fields ($B_{ext}$) applied along the $c$-axis, ranging from 0 to 6 T. This range covers the Spin-Flop Transition (SFT) from AFM to canted AFM (CAFM).
  • Analysis: The study analyzed frequency shifts, linewidths, and signal intensities to deduce local hyperfine fields, canting angles, and the orientation of magnetic moments relative to the external field.

3. Key Results

A. Macroscopic Magnetization and Spin-Flop Transition

• The $M(H)$ curve showed an uncompensated ferromagnetic (FM) contribution at low fields, attributed to the sample's inhomogeneity (mix of $n=0, 1, 2, 3$) and structural defects.
• A distinct inflection point was observed around 3.0–3.5 T, corresponding to the Spin-Flop Transition (SFT).
• Unlike stoichiometric MBT ($n=0$) which shows a sharp step, the transition here is broad due to the distribution of $n$ values and magnetic disorder.
• Above 3.5 T, magnetization increases linearly as Mn spins cant toward the external field, but saturation was not reached within 7 T (consistent with literature suggesting saturation requires ~60 T).

B. $^{55}$Mn NMR: Identifying Magnetic Phases

The $^{55}$Mn spectra revealed two distinct components:

1. "Black" Component (Main Mn Layers):
   • Below 1.5 T, this signal was weak (likely due to fast relaxation).
   • Between 3.0 and 3.5 T (SFT region), the frequency slope changed dramatically, indicating a transition from AFM to CAFM.
   • Canting Angle ($\theta$): By analyzing the frequency shift ($\omega \propto B_{hf} - B_{ext}\cos\theta$), the authors estimated the canting angle. At 4 T, $\theta \approx 53^\circ$; at 6 T, $\theta \approx 28^\circ$. Extrapolation suggests full alignment occurs at $\sim 8.5$ T.
2. "Red" Component (Mn Antisites):
   • Observed as an up-shifting signal at low fields (< 1.5 T).
   • This indicates Mn atoms at Mn$_{Bi}$ antisite positions are antiparallel to the external field (and thus antiparallel to the main Mn layers).
   • As the field increases, these moments align with the field, merging into the main signal.

C. $^{209}$Bi NMR: Discovery of Induced Moments (Key Finding)

• Observation: A strong NMR signal was detected around 110 MHz at zero field, attributed to $^{209}$Bi (spin $I=9/2$).
• Evidence of Magnetism: The effective magnetic field at the Bi nucleus was calculated to be ~16 T. Since Bi is non-magnetic in the parent BT compound, this massive internal field implies a magnetic moment induced on the Bi atoms.
• Mechanism: The induced moment arises from the hybridization of Mn 3d orbitals with Bi valence 6s/6p orbitals.
• Field Dependence: The Bi signal frequency behavior closely mirrored the Mn antisite signal:
  • Below 2 T: Frequency shifts down (indicating the induced Bi moment is antiparallel to the external field).
  • Above 2 T: Frequency shifts up.
• Conclusion: The Mn$_{Bi}$ antisite moments polarize their six nearest-neighbor Bi atoms, inducing a magnetic moment on Bi that is antiparallel to the Mn antisite itself. Consequently, the induced Bi moment is parallel to the main Mn layers, contributing a new ferromagnetic (FM) component to the system.

4. Key Contributions

1. First Direct Evidence: This study provides the first direct experimental proof (via NMR) that Bi atoms acquire an induced magnetic moment in MnBi2Te4/(Bi2Te3)n heterostructures.
2. Defect Characterization: It definitively characterizes the magnetic role of Mn$_{Bi}$ antisites, confirming they are AFM-coupled to the main Mn layers and act as the source of the induced Bi moment.
3. Quantitative Analysis: The authors successfully mapped the canting angle of the Mn sublattice as a function of the external field, providing a benchmark for theoretical models of the CAFM state.
4. Methodological Advance: Demonstrated the power of combining bulk magnetization with site-specific NMR to disentangle complex magnetic phases in inhomogeneous topological insulators.

5. Significance

• Understanding QAHE: The presence of these induced FM moments (from Bi polarization via antisites) adds a ferromagnetic component to the system. Since the QAHE relies on precise magnetic ordering, these defect-induced moments are critical parameters. They may either disrupt the ideal AFM state or, if engineered correctly, be utilized to optimize the QAHE.
• Material Engineering: The findings suggest that controlling the density of Mn antisites is a viable strategy for tuning the magnetic ground state of MTI heterostructures.
• Fundamental Physics: The discovery of a magnetic moment on Bi driven by Mn-Bi hybridization expands the understanding of magnetic interactions in topological materials, suggesting that non-magnetic elements in these lattices can play an active role in magnetism.

In summary, the paper moves beyond the idealized model of MBT, revealing a complex magnetic landscape where structural defects (antisites) actively induce magnetism on non-magnetic Bi atoms, a factor that must be accounted for in the design of future topological quantum devices.