Intrinsic (non)-Gilbert damping in magnetic insulators calculated from a minimal model and \textit{ab initio} spin Hamiltonians

This paper presents an analytically solvable minimal model and a corresponding *ab initio* numerical approach to demonstrate that while magnon-phonon coupling yields comparable Gilbert damping in both bulk and monolayer magnetic insulators, non-Gilbert damping from four-magnon scattering is strongly enhanced in two dimensions and becomes independent of spin-orbit coupling, a framework validated against materials like YIG and CrSBr.

The Gist

Imagine you are trying to send a message across a crowded room by whispering a secret from person to person. This "whisper" is like a magnon—a ripple of magnetic energy moving through a material. In the world of future electronics (called "magnonics"), scientists want to use these whispers instead of electricity to process information because they use less power and generate less heat.

However, there's a big problem: The whisper gets lost.

As the ripple moves, it loses energy and eventually fades away. This fading is called damping. The goal of this paper is to figure out exactly why these magnetic ripples fade, especially when we shrink our devices down to the size of a single sheet of paper (a "monolayer").

Here is a simple breakdown of what the researchers discovered:

1. The Two Main Culprits: "The Floor" and "The Crowd"

The paper identifies two main reasons why a magnetic ripple dies out:

• Magnon-Phonon Damping (The Floor): Imagine the magnetic ripple is a surfer riding a wave. The "phonons" are the vibrations of the floor (the atoms in the material) beneath them. If the floor is shaking, the surfer gets bumped off course and loses energy.

  • The Finding: Whether you are in a thick block of material (3D) or a single thin sheet (2D), the floor shakes enough to cause a similar amount of energy loss. It's a consistent, predictable drag.

• Magnon-Magnon Damping (The Crowd): Now imagine the surfer is surrounded by other surfers (other magnetic ripples). If they crash into each other, they scatter and lose their original direction.

  • The Finding: This is where things get wild. In a thick block (3D), the crowd is spread out, so collisions are rare. But in a thin sheet (2D), everyone is packed into a narrow hallway. The surfers crash into each other constantly. This causes a massive spike in energy loss that gets much worse as the material gets thinner.

2. The "Gilbert" vs. "Non-Gilbert" Mystery

Scientists have a standard rulebook (called the Gilbert Law) for predicting how fast these ripples fade. It's like a speed limit sign that says, "The faster you go, the more you slow down."

• The Floor (Phonons): Follows the rulebook perfectly. If you know the speed, you know the slowdown.
• The Crowd (Magnon-Magnon): Breaks the rulebook. In thin 2D sheets, the slowdown doesn't follow the standard math. It behaves unpredictably. The paper calls this "Non-Gilbert damping." It's like the surfers in the hallway suddenly deciding to stop based on a rule that has nothing to do with their speed.

3. The "Magic" of Weakness

Usually, scientists think that to stop these ripples from fading, you need materials with strong "magnetic glue" (Spin-Orbit Coupling). They thought weak glue meant weak magnets.

The Surprise: The paper shows that in 2D materials, the "crowd crash" (magnon-magnon damping) happens even if the magnetic glue is weak. You can't just rely on weak glue to save you; the geometry of being a thin sheet makes the crashes inevitable.

4. Testing with Real Materials

To prove this wasn't just a math game, the authors tested their theory on two real materials:

1. YIG (Yttrium Iron Garnet): The "gold standard" of magnetic materials, used in thick blocks.
2. CrSBr: A new, super-thin magnetic material (like a single layer of graphene).

The Results:

• In the thick YIG block, the "crowd crash" was present but manageable.
• In the thin CrSBr sheet, the "crowd crash" was huge. It was the dominant reason the signal faded, far outweighing the "floor shaking."

The Big Takeaway

If we want to build tiny, super-efficient magnetic computers using ultra-thin sheets of material, we have a new problem to solve.

We can't just look for materials with low friction (low damping). We have to figure out how to stop the "magnetic surfers" from crashing into each other in that crowded hallway. The paper suggests that applying a strong magnetic field might help clear the hallway and reduce the crashes, offering a potential solution for the future of miniaturized technology.

In short: Shrinking magnetic devices down to a single layer makes the internal "traffic jams" (collisions between ripples) much worse, breaking the old rules of how we expect them to behave.

Technical Summary

Here is a detailed technical summary of the paper "Intrinsic (non)-Gilbert damping in magnetic insulators calculated from a minimal model and ab initio spin Hamiltonians."

1. Problem Statement

The field of magnonics relies on the generation and control of spin waves (magnons). A critical bottleneck for practical applications (e.g., low-power computing, neural networks) is the relaxation (damping) of these spin waves.

• The Challenge: While the Landau-Lifshitz-Gilbert (LLG) equation is the standard framework for simulating spin dynamics, it assumes a linear relationship between the damping rate and the magnon frequency (Gilbert damping). However, experimental evidence increasingly shows deviations from this "Gilbert behavior," particularly in low-dimensional systems.
• The Gap: There is a lack of unified understanding regarding which intrinsic damping mechanisms (magnon-phonon vs. magnon-magnon) dominate in magnetic insulators, and how these mechanisms evolve from bulk (3D) systems to the monolayer (2D) limit. Specifically, it is unclear if the strong spin-orbit coupling (SOC) often associated with Gilbert damping is the primary driver in 2D materials, or if other mechanisms take over.

2. Methodology

The authors employ a dual approach combining analytical modeling and ab initio numerical simulations:

A. Minimal Toy Model (Analytical)

• System: A square lattice (2D) or cubic lattice (3D) with ferromagnetic exchange interactions ($J$) and uniaxial anisotropy ($\kappa$).
• Formalism: The Boltzmann equation is used to calculate the relaxation rate ($\gamma_0$) of a $k=0$ acoustic magnon.
• Interactions Modeled:
  1. Magnon-Phonon (m-ph): Scattering where a magnon absorbs a phonon.
  2. Magnon-Magnon (m-m): Four-magnon scattering processes (one $k=0$ magnon interacts with a thermal magnon to produce two new magnons).
• Key Assumption: The analysis focuses on the low-temperature limit where the magnon decay rate is small compared to its frequency ($\alpha \ll 1$).

B. Ab Initio Numerical Approach

• Materials:
  1. Bulk Yttrium Iron Garnet (YIG): The prototypical low-damping ferrimagnetic insulator.
  2. Monolayer CrSBr: A van der Waals (vdW) ferromagnetic semiconductor.
• Workflow:
  1. Electronic Structure: Calculated using VASP and SIESTA codes.
  2. Spin Hamiltonian: Derived via perturbation theory using the TB2J software package.
  3. Phonons: Calculated using the finite-displacement method (phonopy).
  4. Damping Calculation: Performed using the authors' custom code, MagnoFallas, which implements the Boltzmann formalism to compute damping coefficients from the derived Hamiltonians.
• 2D vs. 3D Phonons: For the CrSBr monolayer, simulations were run with both intrinsic 2D phonons (flexural modes) and 3D phonons (simulating a substrate-supported scenario) to isolate dimensional effects.

3. Key Contributions and Results

A. Distinction Between Gilbert and Non-Gilbert Damping

The paper establishes a clear theoretical distinction:

• Gilbert Damping: The damping parameter $\alpha$ is independent of the external magnetic field (and thus the magnon gap $\Delta$). This is characteristic of magnon-phonon scattering.
• Non-Gilbert Damping: $\alpha$ depends explicitly on $\Delta$ (often decreasing as $\Delta$ increases). This is characteristic of magnon-magnon scattering.

B. Dimensionality Effects (2D vs. 3D)

• Magnon-Phonon Coupling:
  • Produces Gilbert damping in both 2D and 3D.
  • The magnitude is comparable in both dimensions.
  • Crucial Difference: In free-standing 2D monolayers, flexural phonons (out-of-plane modes) have very low energies and high occupation numbers. This leads to a significantly enhanced damping coefficient (approx. 10x stronger) compared to 3D phonons or substrate-supported layers.
• Magnon-Magnon Scattering:
  • 3D: Damping is weak ($\alpha \sim 10^{-5}$) and scales with anisotropy ($\kappa$).
  • 2D: Damping is strongly enhanced (reaching $\alpha \sim 0.03$ at 50 K in the toy model).
  • Independence from SOC: In 2D, the magnon-magnon damping becomes independent of spin-orbit coupling (anisotropy $\kappa$) at low fields. This is because long-range magnetic order in 2D Heisenberg magnets is forbidden at finite temperatures (Mermin-Wagner theorem); thus, the stability of the $k=0$ mode and its scattering are controlled by anisotropy in a way that renders the scattering rate finite even as $\kappa \to 0$.

C. Benchmarking with Real Materials

• Bulk YIG:
  • Magnon-magnon damping exceeds magnon-phonon damping over most relevant temperature ranges.
  • The calculated m-m damping is non-Gilbert and follows the theoretical prediction (Eq. 7).
  • The absolute value ($\sim 7 \times 10^{-5}$ at room temperature) matches experimental observations in high-quality YIG, suggesting that intrinsic m-m scattering is a major contributor to YIG's total damping, not just phonons.
• Monolayer CrSBr:
  • Confirms the toy model predictions: m-m damping is dominant and non-Gilbert.
  • m-ph damping is Gilbert-like but significantly enhanced when using 2D flexural phonons compared to 3D phonons.

4. Significance and Implications

1. Redefining Damping Limits: The study reveals that the theoretical lower bound for damping in perfect 2D magnetic crystals is likely set by intrinsic magnon-magnon scattering, not magnon-phonon scattering or spin-orbit coupling. This challenges the assumption that weak SOC materials (like vdW magnets) will automatically have ultra-low damping.
2. Non-Gilbert Behavior in 2D: The strong non-Gilbert behavior in 2D implies that standard LLG simulations using a constant $\alpha$ parameter may fail to accurately describe spin-wave dynamics in monolayers, especially at low magnetic fields.
3. Design Guidelines for Magnonics:
   • Substrate Engineering: For 2D devices, the choice of substrate is critical. Using a substrate that suppresses flexural phonons (effectively making the phonon spectrum 3D-like) could reduce damping by an order of magnitude.
   • Field Control: Since non-Gilbert damping decreases with increasing magnetic field (as the magnon gap $\Delta$ opens), applying moderate external fields could be a strategy to mitigate the high intrinsic damping in 2D magnonic devices.
4. Methodological Advance: The introduction of the MagnoFallas code provides a robust, open-source tool for the community to compute intrinsic damping from ab initio spin Hamiltonians, bridging the gap between first-principles calculations and macroscopic spin dynamics.

Conclusion

The paper demonstrates that while magnon-phonon interactions provide a baseline Gilbert damping, magnon-magnon scattering is the dominant intrinsic damping mechanism in 2D magnetic insulators, leading to strong non-Gilbert behavior that is independent of spin-orbit coupling. This finding necessitates a re-evaluation of material selection and device design for next-generation 2D magnonic technologies.