🔬 mesoscale physics

Fractional diffusion without disorder in two dimensions

This paper demonstrates that simple local constraints in two-dimensional ice- and dimer-type models induce robust, tunable subdiffusion of defects through dynamical caging and an emergent entropic logarithmic potential, resulting in a fractal frontier dimension of 5/4 characteristic of loop-erased random walks.

Original authors: Nilotpal Chakraborty, Markus Heyl, Roderich Moessner

Published 2026-07-29

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Nilotpal Chakraborty, Markus Heyl, Roderich Moessner

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a crowded dance floor where everyone is holding hands, forming a giant, shifting web. In the world of physics, this is a bit like how particles move in a fluid: they drift, bump, and spread out randomly, a process we call "diffusion." It's the reason a drop of ink spreads evenly in a glass of water or why a smell fills a room. Usually, this spreading follows a predictable rhythm: the further you go, the more time it takes, but it happens at a steady, reliable pace. Scientists have spent centuries studying this, assuming that if you remove the obstacles (like dirt or disorder) and the system is simple, the particles will just diffuse normally.

But what if the rules of the dance floor themselves changed? What if the dancers were forced to follow a strict, invisible code that made it incredibly hard to move forward, even though the floor was empty? This is the puzzle physicists are tackling in the realm of "frustrated magnets" and "gauge theories." Think of these as special materials where the tiny magnetic spins (like little compass needles) are stuck in a game of "rock-paper-scissors" where no single arrangement makes everyone happy. In these systems, the movement of a single "defect" (a glitch in the pattern) isn't just about bumping into things; it's about navigating a maze created by the collective rules of the entire crowd. Understanding this is crucial because it helps us build new kinds of quantum computers and simulate complex physics in the lab, revealing how order and chaos dance together in the quantum world.

The Mystery of the Stuck Defect

In this paper, the authors Nilotpal Chakraborty, Markus Heyl, and Roderich Moessner investigate a very specific, counter-intuitive scenario: how a single defect moves in a two-dimensional grid of "ice" or "dimers" (models of magnetic materials) that has absolutely no disorder, no random messiness, and isn't tuned to a special critical point. You might expect that in such a clean, simple system, the defect would just wander around normally. Instead, they found something wild: the defect gets stuck in a kind of "dynamical cage," moving much slower than expected in a way that is "subdiffusive."

Think of it like a teenager trying to walk through a hallway. In a normal hallway, they walk at a steady pace. But in this "ice" hallway, every time they try to take a step forward, they are forced to retrace their steps, undo their previous moves, and try again, over and over. They aren't blocked by a wall; they are blocked by the rules of the hallway itself. The authors discovered that this defect doesn't just wander; it gets trapped in long, looping retractions. It's as if the floor itself is telling the walker, "You can't go that way because you just came from there, and the rules say you have to undo that step first."

The "Loop-Erased" Dance

To understand why this happens, the authors looked at the path the defect takes. They found that the defect's journey is mathematically identical to something called a "loop-erased random walk." Imagine you are drawing a line on a piece of paper. Every time your pen accidentally circles back and touches a line you've already drawn, you erase that loop and continue from the point where the loop started. The path you end up with is "loop-erased."

In their simulations, the defect behaves exactly like this. It tries to hop around, but the strict rules of the system (known as "Gauss' law" in this context) prevent it from making closed loops in a single direction. If it tries to circle back, the system forces it to undo the move. This results in the defect spending a huge amount of time "waiting" or retracing its steps. The authors measured this waiting time and found it follows a specific pattern: the longer the wait, the more likely it is to happen, following a power law where the exponent is roughly 1.84. Because the average waiting time is effectively infinite, the defect moves incredibly slowly.

A Frictionless Friction

The most surprising part is that this "friction" isn't caused by physical obstacles or dirt. It's purely "entropic," meaning it comes from the sheer number of ways the system can arrange itself. The authors mapped the movement of the defect to a "height field," which is like imagining the grid as a hilly landscape. The defect is like a ball rolling on this landscape. Usually, a ball rolls down a hill. But here, the landscape has a special "logarithmic potential"—a weird, curved hill that gets steeper the further you go, but in a way that depends on the "stiffness" of the material.

This potential acts like a gentle but persistent hand pushing the defect back. The strength of this push is tunable. By changing the rules of the system (like adding static charges or changing the spin of the particles), the authors showed they could change how slow the defect moves. In their simulations on a 400 × 400 grid, they measured the defect's movement and found that the distance it traveled squared ( $\langle r^2 \rangle$ ) grew with time ( $t$ ) according to the formula $\langle r^2 \rangle \sim t^{0.86}$ . In normal diffusion, the exponent would be 1.0. Getting 0.86 means the defect is significantly slower, a phenomenon they call "robust subdiffusion."

The Shape of the Path

The authors also looked at the shape of the area the defect explores over time. If you draw the boundary of the territory the defect has visited, it looks like a fractal—a shape that is rough and jagged at every scale. They found that the "roughness" of this boundary has a fractal dimension of 1.25 (or 5/4). This is a very specific number that matches the "loop-erased random walk" perfectly. It is different from a simple random walk (which would be 4/3) or a self-avoiding walk. This suggests that the defect's path is not just a random stroll; it is a highly structured, non-equilibrium growth process that belongs to a unique class of mathematical curves known as SLE(2).

When the Crowd Gets Too Big

Finally, the team asked: what happens if you have more than one defect? If you have a few, they still get stuck in their own cages. But as you add more defects, they start to interfere with each other. The authors found that the subdiffusion only lasts until the defect has traveled a distance roughly equal to the average space between the defects. Once it crosses that threshold, the "cage" breaks, and the defects start moving normally again, like a standard diffusion process. It's like a crowded party: if you are the only one dancing, you might get stuck in a weird rhythm, but if the room is full, everyone just bumps into each other and moves in a standard, chaotic flow.

Why It Matters

This work is a big deal because it shows that you don't need a messy, disordered system to get strange, slow movement. You just need the right set of rules. This "constraint-induced subdiffusion" could be the key to understanding how defects move in artificial spin ice (materials made of tiny magnets) and quantum simulators. It suggests that by tuning the rules of these systems, scientists could control how fast information or energy moves, potentially leading to new ways of storing data or simulating complex physics. The authors emphasize that this is a simulation-based finding, and while the math is solid, the real-world application in quantum platforms is an exciting possibility for the future.

Technical Summary: Fractional Diffusion Without Disorder in Two Dimensions

Problem Statement
The paper investigates the hydrodynamic behavior of conserved charges (defects) in two-dimensional lattice models subject to local hard-core constraints, specifically focusing on square ice and dimer models. While standard hydrodynamic theory posits that conserved quantities generically diffuse, deviations such as subdiffusion are often attributed to disorder, integrability, or fine-tuning to critical points. This work addresses a fundamental question: how do constraints encoded by emergent $U(1)$ gauge theories manifest in the transport of defects in the absence of disorder, integrability, or fine-tuning? The authors specifically examine the motion of a single monopole defect in a background of Ising spins on a square lattice, where the spins obey the "ice rule" (two arrows in, two arrows out per vertex).

Methodology
The study employs a multi-faceted approach combining numerical simulations with analytical and phenomenological frameworks:

Microscopic Simulations: The authors perform Monte Carlo simulations of a stochastic hopping model for a dynamical charge (monopole) on a square lattice. The hopping is constrained by the local ice rule: a hop is permitted only if the link arrow points in the direction of the hop and the target site is empty.
Continuous-Time Random Walk (CTRW) Mapping: The microscopic dynamics are analyzed by mapping the defect's trajectory to a Loop-Erased Random Walk (LERW). The authors treat the "erased loops" (retractions) as waiting times in a CTRW framework.
Height Field Theory: The system is mapped to a coarse-grained height model, where the defect corresponds to a vortex-antivortex pair. This allows for a continuum description involving a logarithmic potential arising from the residual entropy of the ground state manifold.
Non-Equilibrium Growth Analysis: The growth of the cluster traced by the defect is analyzed through the lens of Schramm-Loewner Evolution (SLE) to determine the fractal dimension of the frontier.
Finite Density Analysis: The dynamics are extended to a finite density of mobile defects to observe crossover behaviors.

Key Contributions and Results

Robust, Tunable Subdiffusion: The primary finding is that a single defect exhibits robust, non-transient subdiffusion characterized by a mean-squared displacement scaling as $\langle r^2 \rangle \sim t^\beta$ . Numerical results yield $\beta \approx 0.86$ . Crucially, this anomalous transport arises purely from the geometric constraints and the residual entropy of the ground state, without any external disorder or fine-tuning.
Dynamical Caging and Fractional Diffusion: The subdiffusion is attributed to "dynamical caging," where the defect undergoes anomalously long retractions. The distribution of these waiting times follows a power law $\psi(\tau) \sim \tau^{-(1+\beta)}$ with $\beta \approx 0.84$ , leading to a divergent mean waiting time. This is captured effectively by a fractional diffusion equation involving a Riemann-Liouville fractional operator.
Logarithmic Potential and Tunability: Through the height field mapping, the defect's motion is described as a random walk in an emergent, entropic logarithmic potential $V(r) \sim \ln(r)$ . The strength of this potential is determined by the stiffness of the height field, which is linked to the algebraic decay exponent of ground-state correlations. The authors demonstrate that by introducing static charges (violations of the ice rule), one can continuously tune the ground-state correlations and, consequently, the subdiffusion exponent $\beta$ .
Fractal Dimension and SLE Universality: The frontier of the cluster traced by the defect is found to have a fractal dimension $d_f \approx 1.25$ (or $5/4$ ). This value matches that of a Loop-Erased Random Walk (LERW) and corresponds to an SLE( $\kappa$ ) description with $\kappa = 2$ . This contrasts with simple random walks or self-avoiding walks, which typically fall into the Kardar-Parisi-Zhang (KPZ) universality class with $d_f = 4/3$ (SLE( $8/3$ )). The authors conjecture that the scaling limit of the frontier is described by SLE(2).
Crossover at Finite Density: In the presence of a finite density of mobile defects, the subdiffusive behavior persists only up to a length scale corresponding to the average inter-particle spacing (analogous to a Debye screening length). Beyond this scale, the system crosses over to ordinary diffusion.

Significance and Claims
The paper claims that this work identifies a unique mechanism for anomalous transport in simple, paradigmatic models of geometric frustration. The significance lies in demonstrating that:

Constraint-Induced Anomalies: Subdiffusion can emerge purely from local constraints and the resulting emergent gauge structure, without the need for disorder or critical fine-tuning.
Tunability: The dynamical exponent is tunable via the manipulation of ground-state correlations (e.g., by adding static charges), offering a potential platform for controlling transport in synthetic systems.
Universality Class Shift: The defect dynamics define a new universality class for non-equilibrium growth, distinct from standard KPZ, characterized by SLE(2) and LERW statistics.
Relevance to Quantum Simulation: These findings are directly relevant to platforms such as artificial spin ice and quantum simulators aiming to realize discrete link models and emergent gauge theories, providing a theoretical baseline for interpreting defect dynamics in these systems.

The authors emphasize that this "frustration/constraint-induced subdiffusion" provides a distinct alternative to conventional diffusive hydrodynamics, rooted in the non-local nature of fractionalized defects within the constrained manifold.