Fate of diffusion under integrability breaking of classical integrable magnets

This paper investigates how integrability-breaking perturbations affect spin diffusion in a classical anisotropic Landau-Lifshitz magnet, revealing a sharp change in the diffusion constant and a crossover from non-Gaussian to Gaussian statistics in magnetization transfer.

The Gist

Imagine you are watching a massive, synchronized dance troupe performing in a giant stadium. This paper is essentially a study of what happens to that dance when you go from a perfectly choreographed routine to a chaotic, unorganized mosh pit.

Here is the breakdown of the science using everyday analogies.

1. The Setting: The "Perfectly Choreographed" Dance (Integrability)

In physics, an "integrable system" is like a dance troupe where every single dancer follows a strict, mathematical rule. If Dancer A moves left, Dancer B must move right to maintain balance. Because these rules are so perfect, the system is highly predictable.

In these "perfect" systems, things like heat or magnetism don't just spread out randomly (like a drop of ink in water). Instead, they move in a very specific, "anomalous" way. It’s like if you threw a ball into a crowd of dancers, and instead of bouncing around randomly, the ball followed a strange, wavy path because the dancers were so perfectly coordinated.

2. The Problem: The "Chaos" Factor (Integrability Breaking)

The researchers wanted to know: What happens if you introduce a little bit of chaos?

Imagine if, during the perfect dance, a few dancers started tripping, or a loud, unpredictable drumbeat started playing. This is what the scientists call "integrability breaking." They added tiny "perturbations" (little nudges of chaos) to their mathematical model to see how much "tripping" it would take to ruin the perfect choreography.

3. The Discovery: Two Big Surprises

The researchers found two major things happening when the chaos was introduced:

Surprise A: The "Cliff Edge" Effect (The Diffusion Constant)

Imagine you are driving a car on a perfectly smooth road. As you add tiny pebbles to the road, you barely feel them. But the researchers found that in these magnetic systems, there is a "sharp change."

It’s as if you are driving on a smooth road, and the moment you hit a certain amount of pebbles, the road suddenly turns into deep mud. The way "magnetism" spreads (the diffusion constant) doesn't just change slowly; it jumps abruptly. This suggests that the "perfect" way magnetism moves in a choreographed system is fundamentally different from the "messy" way it moves in a chaotic one.

Surprise B: From "Strange Patterns" to "Normal Mess" (Gaussian Statistics)

When the dancers are perfectly choreographed, the way they move creates strange, non-random patterns (the scientists call this non-Gaussian statistics). It’s like seeing a pattern in the way spilled milk flows because the table is perfectly tilted.

However, as soon as the "chaos" (the tripping dancers) is added, those strange patterns vanish. The movement becomes "normal" and random—what scientists call Gaussian. It’s like the difference between a beautiful, swirling pattern in a calm pond versus the random, messy splashing of a stormy ocean. The chaos "washes out" the special patterns of the perfect system.

4. Why does this matter? (The Big Picture)

You might ask, "Who cares about dancing magnets?"

The reason is that these mathematical models are the "blueprints" for how the universe works at a microscopic level. By studying these classical "dances," scientists are actually learning how to understand Quantum Mechanics—the strange, tiny rules that govern atoms.

The paper proves that even in a tiny, controlled mathematical world, there is a profound "bridge" between perfect order and total chaos. It shows that "normal" randomness is actually a very fragile state that only emerges once the perfect order of the universe is broken.

Technical Summary

Technical Summary: Fate of Diffusion under Integrability Breaking of Classical Integrable Magnets

Authors: Jiaozi Wang, Sourav Nandy, Markus Kraft, Tomaž Prosen, and Robin Steinigeweg
Date: April 28, 2026 (as per manuscript)

---

1. Problem Statement

A fundamental challenge in statistical mechanics is understanding how irreversible macroscopic phenomena, such as diffusion, emerge from time-reversible microscopic laws. In one-dimensional interacting systems, the nature of transport is highly sensitive to integrability (the existence of infinitely many conserved charges).

While generic many-body systems are expected to exhibit "normal" diffusion (Gaussian statistics), integrable systems often exhibit "anomalous" signatures, such as non-Gaussian fluctuations in magnetization transfer. This paper seeks to answer:

• How does the diffusion constant ($D$) behave as a system transitions from an integrable to a chaotic (generic) regime via weak perturbations?
• Does the anomalous non-Gaussian signature of integrable transport survive weak integrability breaking, or is it restored to Gaussian statistics?
• What is the scaling relationship between the diffusion constant, the perturbation strength ($\epsilon$), and the system size ($L$)?

2. Methodology

To avoid the "curse of dimensionality" inherent in quantum many-body simulations, the authors utilize classical integrable magnets. This allows for large-scale numerical simulations on much larger system sizes, providing a more rigorous test of the thermodynamic limit.

• Model: The study focuses on the anisotropic Landau-Lifshitz magnet (LLM) in the easy-axis regime ($\delta > 0$). This model is the classical counterpart to the quantum spin-1/2 XXZ chain and shares its dynamical phases.
• Discretization: To maintain integrability in a discrete space-time lattice, the authors employ a specialized symplectic discretization using anisotropic stereographic variables ($\zeta$).
• Perturbations: Two types of integrability-breaking perturbations are introduced:
  • Model A: A space-time lattice perturbation involving alternating time steps ($\tau \pm \Delta\tau$).
  • Model B: A local rotation map ($R_\theta$) applied between steps.
• Observables:
  • Diffusion Constant ($D$): Calculated via the long-time limit of the current-current autocorrelation function.
  • Full Counting Statistics (FCS): Analysis of the time-integrated spin current $J(t)$ through its rescaled cumulants ($\kappa_n(t)$). Gaussian statistics are identified when $\kappa_{n>2}(t) \to 0$.

3. Key Results

A. Discontinuous Change in Diffusion Constant
The authors find that the diffusion constant $D$ does not change smoothly from the integrable value ($D_{int}$) to the chaotic value ($D_{chaos}$). Instead, in the thermodynamic limit ($L \to \infty$), there is a discontinuous jump at the integrable point ($\epsilon = 0$).

• Scaling Law: They establish a universal scaling form: $D_L(\epsilon) = f(\epsilon\sqrt{L})$.
• This implies that the derivative $\partial_\epsilon D_L |_{\epsilon=0}$ diverges as $\sqrt{L}$, confirming the non-analytic, discontinuous nature of the transition.

B. Restoration of Gaussian Statistics
The study investigates the "anomalous" non-Gaussian nature of integrable systems.

• Integrable Case ($\epsilon = 0$): Higher-order rescaled cumulants $\kappa_{n>2}(t)$ saturate to finite values, signifying non-Gaussian transport.
• Chaotic Case ($\epsilon > 0$): Beyond a characteristic crossover time $t^*$, the higher-order cumulants decay to zero ($\kappa_{n>2}(t) \to 0$). This indicates that integrability breaking restores normal Gaussian diffusion.
• Crossover Scaling: The crossover time scales as $t^* \sim \epsilon^{-2}$.

C. Classical-Quantum Correspondence
The results (specifically the discontinuity and the restoration of Gaussianity) are highly consistent with recent findings in the quantum spin-1/2 XXZ chain. The classical setting provides a clearer, more computationally controlled verification of these phenomena.

4. Significance

This work provides a rigorous numerical demonstration of how the "fine-tuned" property of integrability dictates transport properties. By showing that even infinitesimal perturbations can fundamentally alter the diffusion constant and the statistical nature of current fluctuations, the paper clarifies the boundary between integrable and generic many-body dynamics. It offers a compelling example of classical-quantum correspondence in transport phenomena and provides a framework for understanding how macroscopic irreversibility emerges from the breaking of microscopic conservation laws.