Magnetic excitations in the Kitaev material Na$_2$IrO$_3$ studied by neutron scattering

Inelastic neutron scattering on Na$_2$IrO$_3$ reveals a 1.7 meV magnon gap and antiferromagnetic Heisenberg interactions, distinguishing its magnetic excitations from those in the sister compound $\alpha$-RuCl$_3$ and demonstrating that low-energy ferromagnetic fluctuations are not a universal fingerprint of Kitaev physics.

The Gist

Here is an explanation of the paper, translated from scientific jargon into a story about dancing magnets and a "magnetic dance floor."

The Big Picture: The Search for a "Quantum Magic Carpet"

Imagine you are looking for a special kind of material that behaves like a quantum magic carpet. In physics, this is called a "Quantum Spin Liquid" (QSL). In a normal magnet (like the one on your fridge), the tiny atomic magnets (spins) all line up neatly, like soldiers in a parade. But in a QSL, the spins are constantly dancing and swirling, never settling down, even when it's freezing cold.

This chaotic dance is special because it creates "anyons," particles that could be the building blocks for future, super-powerful quantum computers.

For years, scientists have been hunting for materials that act like this magic carpet. Two main suspects have been on the list:

1. $\alpha$-RuCl$_3$ (The "Star Student"): This material showed signs of the magic carpet, but it also had some "bad habits" (ferromagnetic fluctuations) that made it messy.
2. Na$_2$IrO$_3$ (The "Mysterious Twin"): This is the material studied in this paper. It looks very similar to the Star Student, but scientists weren't sure if it was truly a magic carpet or just a regular magnet in disguise.

The Experiment: Listening to the Atomic Dance

The authors of this paper, led by Alexandre Bertin, decided to listen to the "music" of the Na$_2$IrO$_3$ atoms. They used a giant machine called a neutron spectrometer (think of it as a super-sensitive microphone for atoms).

• The Problem: Na$_2$IrO$_3$ is made of Iridium, a heavy metal that "eats" neutrons (the particles used to listen). It's like trying to hear a whisper in a room full of heavy velvet curtains.
• The Solution: They couldn't use just one crystal. Instead, they glued 63 tiny crystals together onto a single plate, creating a "super-crystal" big enough to get a good signal.

The Discovery: The Gap and the Silence

When they listened to the vibrations (magnetic excitations) of this super-crystal, they found two major things:

1. The "Safety Net" (The Magnon Gap)

They found that the atoms need a tiny bit of energy (1.7 meV) just to start dancing. It's like a safety net under a trapeze artist; the artist can't fall all the way to the ground, they have to jump a little bit first.

• Why it matters: This "gap" tells us that the magnetic structure is stable and organized in a specific "zigzag" pattern. It's not a chaotic mess; it's a structured dance.

2. The "Silence" (No Ferromagnetic Fluctuations)

This is the most important part. In the "Star Student" ($\alpha$-RuCl$_3$), scientists heard a loud, low-frequency hum. This hum represented ferromagnetic fluctuations—a type of magnetic behavior where atoms want to line up in the same direction (like a crowd cheering in unison).

• The Twist: In Na$_2$IrO$_3$, there was no hum. The room was silent.
• The Analogy: Imagine two bands playing the same song. The Star Student band has a loud, booming bass drum (ferromagnetic noise) that drowns out the melody. The Na$_2$IrO$_3$ band, however, plays the melody perfectly without that distracting bass drum.

What This Means: The "Rulebook" is Different

Scientists had a theory (the Kitaev Model) that predicted how these atoms should dance. They thought the "bass drum" (ferromagnetic behavior) was a necessary part of the Kitaev dance.

This paper proves that the bass drum is optional.

• The Difference: The two materials are made of slightly different ingredients. In $\alpha$-RuCl$_3$, the atoms have a "friendly" interaction that makes them want to line up together (ferromagnetic). In Na$_2$IrO$_3$, that same interaction is "rivalrous" (antiferromagnetic), so they refuse to line up.
• The Conclusion: You don't need the "bass drum" (ferromagnetic fluctuations) to have a Kitaev material. The "magic carpet" (Kitaev physics) exists in Na$_2$IrO$_3$ just fine, even without the noise.

The Final Verdict

The paper concludes that Na$_2$IrO$_3$ is a valid Kitaev candidate, but it dances to a slightly different tune than its sister material.

• Old Belief: "If you see this specific low-energy noise, you know it's a Kitaev material."
• New Truth: "That noise isn't a fingerprint of the Kitaev magic; it's just a side effect of one specific ingredient. The real magic is in the bond-directional rules, which both materials follow, even if they dance differently."

In short: The scientists successfully listened to the "magnetic music" of a difficult material, proved it has a stable, structured dance, and realized that the "noise" we thought was essential for quantum computing materials isn't actually required. This clears the path to finding even better materials for the future of quantum technology.

Technical Summary

Here is a detailed technical summary of the paper "Magnetic excitations in the Kitaev material Na2IrO3 studied by neutron scattering" by Bertin et al.

1. Problem Statement

The search for materials realizing the Kitaev quantum spin liquid (QSL) state has focused heavily on honeycomb lattice compounds like $\alpha$-RuCl$_3$ and Na$_2$IrO$_3$. While $\alpha$-RuCl$_3$ has been extensively studied via inelastic neutron scattering (INS), revealing complex magnetic excitations including low-energy ferromagnetic (FM) fluctuations, Na$_2$IrO$_3$ has remained less understood in the low-energy regime.

• The Challenge: Na$_2$IrO$_3$ suffers from strong neutron absorption due to the Iridium (Ir) isotope and typically requires large sample volumes for INS, which are difficult to grow as high-quality single crystals.
• The Scientific Question: Does Na$_2$IrO$_3$ exhibit the same low-energy FM fluctuations seen in $\alpha$-RuCl$_3$? Are the microscopic interaction parameters (specifically the signs of the Heisenberg exchange $J_1$ and the Kitaev term $K$) fundamentally different between the two materials despite their structural similarities? Previous high-resolution Resonant Inelastic X-ray Scattering (RIXS) data covered high energies (>10 meV) but lacked the resolution to probe the low-energy magnon gap and dispersion crucial for validating microscopic models.

2. Methodology

The authors employed Inelastic Neutron Scattering (INS) on a large, co-aligned sample of Na$_2$IrO$_3$ to overcome absorption and size limitations.

• Sample Preparation: 63 thin single crystals of Na$_2$IrO$_3$ were grown via a self-flux method and co-aligned on an aluminum plate. This assembly provided a total mass of ~208 mg, sufficient to generate a detectable signal despite the strong absorption of Ir.
• Instrumentation: Experiments were conducted at the Institut Laue-Langevin (ILL) using two Triple-Axis Spectrometers (TAS):
  • Thales (Cold Neutrons): Used for high-resolution measurements at low energies ($E < 10$ meV) with an energy resolution of ~0.2 meV.
  • IN8 (Thermal Neutrons): Used to map higher energy excitations (up to 10 meV) with higher flux but broader resolution (0.8 meV).
• Measurement Strategy:
  • Constant-$Q$ scans to identify the magnon gap at the antiferromagnetic (AFM) zone center $(0, 1, 0.5)$.
  • Constant-$E$ scans to map the in-plane and out-of-plane dispersion of spin waves.
  • Temperature dependence studies from 1.5 K up to 100 K to distinguish between static order, critical fluctuations, and short-range correlations.
  • Background subtraction techniques involving sample rotation ($\pm 45^\circ$) to isolate magnetic signals from instrument and sample environment noise.
• Theoretical Analysis: The experimental data were compared against Linear Spin-Wave Theory (LSWT) calculations using the SpinW package. The calculations utilized the extended Kitaev-Heisenberg model ($H_{K\Gamma\Gamma'}$) with parameters derived from previous high-energy RIXS studies (specifically Model A3 from Ref. [33]).

3. Key Contributions

• First Low-Energy INS Map: This study provides the first comprehensive mapping of the low-energy magnon dispersion in single-crystal Na$_2$IrO$_3$, filling the gap left by powder studies and high-energy RIXS.
• Resolution of the "Gap" Origin: The authors clarify that the observed low-energy gap is not a standard transverse magnon at the AFM zone center but arises from zone-boundary modes folded to the Bragg point due to the multi-domain structure ($C_3$ symmetry) of the honeycomb layers.
• Absence of FM Fluctuations: Crucially, the study definitively rules out the presence of strong low-energy ferromagnetic fluctuations in Na$_2$IrO$_3$, contrasting sharply with $\alpha$-RuCl$_3$.
• Validation of Microscopic Models: The experimental dispersion matches the LSWT predictions based on the RIXS-derived parameters without further tuning, validating the specific interaction model for Na$_2$IrO$_3$.

4. Key Results

Magnetic Order and Gap

• Ordering: Na$_2$IrO$_3$ undergoes a transition to a zigzag AFM order at $T_N \approx 15$ K with a propagation vector $\mathbf{q} = (0, 1, 0.5)$.
• Magnon Gap: A finite energy gap of $\Delta = 1.7(1)$ meV was observed at the AFM zone center.
• Nature of the Gap: The signal at the zone center is identified as a "soft" excitation originating from the zone boundaries $(0.5, \pm 0.5, 0)$ of the rotated domains, folded into the AFM point. This is distinct from the transverse magnon gap usually associated with anisotropy.
• Dimensionality: The dispersion along the out-of-plane direction ($l$) is negligible, confirming the highly 2D nature of the magnetic interactions.

Dispersion and High-Energy Behavior

• Dispersion Shape: The magnons exhibit a parabolic-like dispersion near the zone center, steepening at higher energies.
• Temperature Dependence: Magnetic correlations persist up to $\sim 6 \times T_N$ (approx. 100 K), indicating strong short-range correlations even above the ordering temperature.

Comparison with $\alpha$-RuCl$_3$

• Ferromagnetic Instability: Unlike $\alpha$-RuCl$_3$, which shows sharp, intense low-energy FM modes at the FM zone center, Na$_2$IrO$_3$ shows no evidence of FM scattering in the 2–9 meV range.
• Microscopic Origin: The difference is attributed to the sign of the nearest-neighbor Heisenberg exchange ($J_1$).
  • In $\alpha$-RuCl$_3$, $J_1$ is Ferromagnetic, competing with the FM Kitaev term to create FM instabilities.
  • In Na$_2$IrO$_3$, $J_1$ is Antiferromagnetic, which suppresses the FM instability despite the presence of a strong FM Kitaev term ($K$).

5. Significance

• Refining the Kitaev Paradigm: The results demonstrate that while both materials are "Kitaev candidates" with strong bond-directional interactions, they deviate from the ideal Kitaev model in fundamentally different ways due to the sign of the Heisenberg exchange.
• Fingerprinting Interactions: The study establishes that low-energy FM fluctuations are not a universal fingerprint of FM Kitaev interactions; rather, they depend on the competition with other exchange terms (specifically $J_1$).
• Model Validation: The agreement between the INS data and the LSWT calculations using parameters from high-energy RIXS confirms that the extended $H_{K\Gamma\Gamma'}$ model accurately describes the magnetic physics of Na$_2$IrO$_3$ across the entire energy spectrum.
• Experimental Benchmark: By successfully overcoming the neutron absorption challenges, this work sets a benchmark for studying other heavy-element Kitaev candidates where single-crystal growth is difficult.

In conclusion, the paper provides a definitive characterization of the magnetic excitations in Na$_2$IrO$_3$, resolving discrepancies with previous models and highlighting the critical role of the Heisenberg exchange sign in distinguishing the magnetic behavior of Kitaev materials.