Diffusion and relaxation of topological excitations in layered spin liquids

This paper demonstrates that pump-probe experiments can uniquely identify 2D topological excitations in 3D layered spin liquids by revealing distinct subdiffusive spreading and anomalous decay laws, such as logarithmic propagation and a $n(t) \sim (\log^2 t)/t$ density decay, which arise from the topological constraint that only pairs of fractionalized excitations can hop between layers.

The Gist

Imagine you have a giant, multi-layered cake made of a very special kind of jelly. This isn't just any jelly; it's a "quantum spin liquid," a state of matter where tiny particles (let's call them "ghosts") behave in a very strange, rule-bound way.

The paper by Joy, Lange, and Rosch explores what happens when you poke this cake and watch how the ghosts move. Here is the story in simple terms:

The Rules of the Game

In this special jelly cake, the ghosts have a strict rule: They are trapped in their own layer.

• Inside a layer: A single ghost can wander around freely, like a person walking in a large room.
• Between layers: A single ghost cannot jump to the layer above or below. It's like a ghost that can walk on the floor but cannot climb stairs.
• The Loophole: The only way to move between layers is if two ghosts hold hands (form a pair). Only a pair can jump up or down together.

The Experiment: The "Poke"

The researchers propose a way to test this using a "pump-probe" experiment. Imagine shining a bright laser (the "pump") on the very top surface of the cake. This laser instantly creates a bunch of these ghosts right at the top.

Then, they use a second laser (the "probe") to watch how the ghosts spread out over time. They are asking: How fast do the ghosts sink into the deep layers of the cake?

The Surprising Results

1. The "Crowd Size" Rule
The most important discovery is about speed.

• In normal physics, if you drop a dye in water, it spreads at a speed that doesn't really care how much dye you used.
• In this quantum cake, the speed depends entirely on how many ghosts you created.
• The Analogy: Imagine a crowded hallway. If there are only a few people, they can move fast. If the hallway is packed, everyone moves slower because they keep bumping into each other.
• The Finding: The more ghosts the laser creates (the brighter the "pump"), the slower the whole process happens. The time it takes for the ghosts to reach the bottom is inversely proportional to the number of ghosts. If you double the number of ghosts, the process takes half as long. This is a unique "fingerprint" that proves the ghosts are following these special topological rules.

2. The "Slow Sink" (When Ghosts Don't Disappear)
If the ghosts are just wandering and never disappearing (no "annihilation"), they don't sink into the cake like a stone in water. They sink very slowly, like a drop of thick honey.

• The Math: The depth they reach grows with the cube root of time ($t^{1/3}$). This is called "sub-diffusion." It's much slower than normal spreading.

3. The "Logarithmic Crawl" (When Ghosts Disappear)
In reality, ghosts can bump into each other and vanish (annihilate), turning into harmless heat (phonons).

• When this happens, the spreading slows down even more. Instead of a power law, the depth grows only as the logarithm of time ($\log t$).
• The Analogy: Imagine trying to walk through a forest where every time you take a step, there's a chance you and a friend vanish. You end up moving incredibly slowly, barely making progress. The paper shows that even a tiny bit of this "vanishing" stops the ghosts from spreading deep into the cake quickly.

4. The Top vs. Bottom Mystery
If you have a finite cake (a slab with a top and a bottom), the researchers found that the density of ghosts on the top and the bottom takes a very long time to equalize.

• Even after a long time, the top layer might still be crowded while the bottom is empty. They approach each other at a rate that is "stretched exponential," meaning it gets slower and slower, almost like it's stuck.

Why Does This Matter?

Detecting "topological order" (this special rule-bound state) is notoriously hard. Normal tools (like looking at how the material reflects light) usually can't see these ghosts because they are "invisible" to local measurements.

This paper suggests a new way to catch them: Watch how they spread.
If you shine a laser and see that the spreading speed changes specifically based on how bright the laser is (following that inverse relationship), you have found proof that the material is a topological spin liquid. It's like identifying a specific type of bird not by its color, but by the unique way it flies when the wind changes.

Summary

• The Setup: Ghosts trapped in 2D layers, only pairs can jump between layers.
• The Test: Create ghosts on top and watch them sink.
• The Signature: The process slows down if you create more ghosts.
• The Movement: They sink very slowly (sub-diffusively), and if they can vanish, they sink even slower (logarithmically).

This provides a clear, measurable "smoking gun" to prove that a material is a quantum spin liquid, without needing to see the invisible particles directly.

Technical Summary

Technical Summary: Diffusion and Relaxation of Topological Excitations in Layered Spin Liquids

Problem Statement
Identifying topological phases of matter, particularly quantum spin liquids, remains a central challenge in condensed matter physics because their defining features—fractionalized excitations and non-trivial exchange statistics—are not accessible via local order parameters. Conventional probes like neutron spectroscopy couple to local operators, often exciting a broad continuum that obscures direct signatures of topological order. While recent nonlinear spectroscopy has shown promise in resolving fractionalized continua, there is a need for robust experimental strategies that can unambiguously detect the "emergent dimensionality" of excitations in three-dimensional (3D) materials composed of weakly coupled two-dimensional (2D) topological layers. Specifically, the authors address how to experimentally distinguish the dynamics of topological quasiparticles, which are confined to move within 2D planes, from ordinary quasiparticles (like magnons or phonons) that can diffuse freely in 3D.

Methodology
The authors propose a theoretical framework combining a microscopic particle-based stochastic model with a coarse-grained continuum description to analyze non-equilibrium dynamics following a "quench" (sudden excitation).

1. Microscopic Model: The system is modeled as a stack of 2D $Z_2$ spin liquids (e.g., Kitaev models). The elementary excitations (e.g., visons, anyons) are treated as classical random walkers subject to hard-core constraints. The dynamics are governed by three processes:

   • In-plane diffusion: Single particles diffuse freely within a layer with rate $\Gamma_\parallel$.
   • Inter-layer pair-hopping: Due to topological constraints, single particles cannot hop between layers. Only pairs of excitations can hop to adjacent layers with rate $\Gamma_\perp$.
   • Pair-annihilation: Pairs can annihilate into the vacuum (emitting phonons) with rate $\Gamma_\lambda$.
     The model is simulated using stochastic particle-based methods on a cubic lattice.

2. Continuum Limit: In the low-density limit, the system is described by a nonlinear diffusion equation for the particle density $\rho(z,t)$:
   $$\partial_t \rho = D_\parallel (\partial_x^2 + \partial_y^2)\rho + D_\perp \partial_z^2 \rho^2 - \lambda \rho^2 + \text{noise terms}$$
   Here, the inter-layer diffusion term is quadratic in density ($\partial_z^2 \rho^2$) because it requires pairs, while the annihilation term is also quadratic ($-\lambda \rho^2$). The authors analyze exact solutions of the noiseless equation and perform scaling analyses to understand the role of noise and finite-size effects.

3. Experimental Protocol: The proposed detection scheme involves a pump-probe experiment where a laser or THz pulse uniformly excites the top surface of a layered sample, creating an initial density $\rho_0$. The subsequent relaxation and spreading of excitations into the bulk are tracked.

Key Contributions and Results

• Scaling of Time Scales: The central theoretical result is that all characteristic time scales in the relaxation process are inversely proportional to the initial excitation density $\rho_0$ (and thus the pump intensity $I_P$). This is derived from the scaling invariance of the nonlinear diffusion equation. Specifically, $\tau \propto 1/\rho_0 \propto 1/I_P$. The authors identify this intensity dependence as a "smoking-gun" signature of topological excitations, distinguishing them from trivial quasiparticles where time scales are typically independent of density.

• Subdiffusive Spreading (No Annihilation): In the absence of pair-annihilation ($\lambda = 0$), the excitation cloud spreads into the bulk subdiffusively. The mean depth $\bar{z}$ of the excitation cloud scales as $\bar{z} \sim t^{1/3}$. The density profile follows a universal parabolic form, and the top-layer density decays as $\rho(0,t) \sim t^{-1/3}$.

• Logarithmic Propagation (With Annihilation): When pair-annihilation is allowed ($\lambda \neq 0$), the propagation is significantly suppressed. The mean depth scales logarithmically, $\bar{z} \sim \log t$, rather than as a power law. Consequently, the total particle density decays as $n(t) \sim (\log t)/t$ in the mean-field approximation.

• Noise Corrections: The authors analyze the effect of stochastic noise.

  • For $\lambda = 0$, noise is an irrelevant perturbation, and the mean-field $t^{1/3}$ scaling holds accurately.
  • For $\lambda \neq 0$, noise is a marginal perturbation. In the long-time limit, noise induces logarithmic corrections, leading to a total density decay of $n(t) \sim (\log^2 t)/t$, which is slower than the mean-field prediction.

• Finite Slab Geometry: For a sample of finite width $w$, the authors derive the time $t_b$ required for excitations to reach the bottom layer.

  • If $w \ll \sqrt{D_\perp/\lambda}$, $t_b \propto w^3/\rho_0$.
  • If $w \gg \sqrt{D_\perp/\lambda}$, $t_b$ grows exponentially with $w$.
    Furthermore, the difference between the top and bottom layer densities decays as a stretched exponential in logarithmic time: $\Delta n \sim e^{-\alpha (\log t)^2}$. This extremely slow equilibration is a distinct signature of the constrained dynamics.

• Initial Conditions and Noise: The study confirms that the scaling laws are robust against different initial conditions (randomly placed particles vs. randomly placed pairs) and that the results hold even in the "ultraclean" limit where in-plane diffusion is driven by particle-particle collisions rather than disorder, provided the excitation is uniform in the plane.

Significance and Claims
The paper claims to provide a concrete, experimentally accessible protocol for detecting topological order in layered materials without relying on local order parameters. The significance lies in the proposal that the emergent dimensionality of quasiparticles (2D motion within layers, 3D motion only via pairs) dictates a unique, density-dependent relaxation dynamics.

The authors modestly state that their protocol offers a "relatively direct experimental proof" of the topological nature of excitations by verifying the inverse proportionality of time scales to pump intensity. They emphasize that this approach is applicable to a broad class of layered spin-liquid candidates (such as $\alpha$-RuCl$_3$) and fracton models. The work does not claim to solve the problem of detecting topological order in all materials but specifically targets systems where weak inter-layer coupling preserves the 2D topological character. The authors acknowledge that experimental feasibility depends on the ability to create surface-localized excitations and probe their density, suggesting that observables like dielectric function or Raman intensity could serve as proxies.