Diffusion in interacting two-dimensional systems under a uniform magnetic field

This paper demonstrates that the fermionic truncated Wigner approximation accurately captures the diffusive relaxation dynamics of interacting two-dimensional fermions in a uniform magnetic field at infinite temperature, revealing that while strong interactions suppress magnetic-field effects, comparable interaction and kinetic energies significantly reduce diffusion in sufficiently large systems.

The Gist

Imagine you are trying to understand how a crowd of people moves through a giant, crowded dance floor. Now, add two tricky rules to this scenario:

1. The Magnetic Spin: Everyone is holding a tiny, invisible compass that makes them want to spin in circles instead of walking straight.
2. The Personal Space: Everyone is also trying to avoid bumping into their neighbors, but they are also slightly attracted to them, creating a complex social dance.

This is essentially what physicists are trying to figure out in this paper, but instead of people, they are studying electrons (tiny particles) in a 2D grid (like a chessboard) under a magnetic field.

Here is the breakdown of their discovery, using simple analogies:

1. The Problem: The "Math Nightmare"

Usually, when scientists want to predict how these particles move, they use math. But when you add a magnetic field and make the particles interact with each other, the math becomes impossibly hard. It's like trying to predict the exact path of every single drop of water in a hurricane while they are all bumping into each other.

For a long time, scientists could only do this accurately for very small groups or in a single line (1D). But real life is 2D (like a flat sheet), and the math breaks down there.

2. The Solution: The "Crowd Simulator" (fTWA)

The authors developed a clever shortcut called the fermionic Truncated Wigner Approximation (fTWA).

• The Analogy: Imagine you want to know how a crowd will disperse after a concert. Instead of tracking every single person (which is too hard), you create a "cloud" of thousands of possible scenarios. You run a simulation where you randomly assign people to different starting spots and let them move according to the rules. Then, you look at the average result of all those simulations.
• The Surprise: The authors found that this "average cloud" method works surprisingly well for 2D systems, even when the particles are interacting. It was expected to fail, but it actually captured the physics perfectly for a wide range of interaction strengths. It's like guessing the weather by looking at a cloud, and realizing the cloud is actually a perfect map of the storm.

3. The Discovery: The "Magnetic Traffic Jam"

They used this simulator to watch how "density waves" (ripples of crowdedness) spread out over time. This is called diffusion.

• No Magnetic Field: Without the magnetic field, the particles spread out smoothly, like ink dropping in water.
• With Magnetic Field: When they turned on the magnetic field, the particles started spinning in circles (orbiting). This made it much harder for them to get from point A to point B.
  • The Result: The diffusion slowed down significantly. The magnetic field acted like a traffic jam, forcing the particles to take a winding, inefficient path instead of a straight line.

4. The Twist: When "Friends" Overpower the "Spin"

They also tested what happens when the particles interact very strongly (when they really care about their neighbors).

• Weak Interactions: If the particles are just casually interacting, the magnetic field is the boss. It stops them from moving efficiently.
• Strong Interactions: If the particles are very strongly connected (like a group of friends holding hands tightly), they start moving together as a unit. In this case, the magnetic field's attempt to make them spin becomes less important. The "social bond" (interaction) overrides the "magnetic spin."
  • The Analogy: If you are walking alone in a crowd, a strong wind (magnetic field) might blow you off course. But if you are in a tight huddle with a group of friends holding hands, the wind can't push you as easily; your group momentum keeps you moving straight.

5. The Catch: You Need a Big Enough Dance Floor

One of the most important findings is about size.
To see these magnetic effects clearly, you can't just look at a tiny 2x2 grid. You need a massive dance floor (at least 400+ spots).

• Why? If the floor is too small, the particles hit the walls and bounce back, which messes up the data. It's like trying to study ocean waves in a bathtub; the waves just hit the sides and don't behave like they do in the ocean. The authors showed that you need a large enough system to see the true "magnetic length" (how far a particle spins before it moves forward).

Why Does This Matter?

This isn't just abstract theory. The authors suggest that ultracold atom experiments (using lasers to trap atoms in grids) can actually test this right now.

• The Takeaway: They have provided a reliable "cheat code" (the fTWA method) that allows scientists to predict how quantum materials behave in magnetic fields without needing a supercomputer the size of a city. This helps us understand how to build better electronics, sensors, or even quantum computers that rely on controlling how electrons move in 2D layers.

In a nutshell: They found a way to simulate a complex quantum dance floor, discovered that magnetic fields act like a traffic jam for particles, and learned that if the particles hold hands tightly enough, they can ignore the traffic jam.

Technical Summary

1. Problem Statement

The dynamics of interacting particles in orbital magnetic fields within two-dimensional (2D) systems are notoriously difficult to study theoretically. This difficulty arises from two main factors:

• Electronic Correlations: Interacting 2D systems lack straightforward theoretical methods for handling strong correlations.
• Finite-Size Effects: To correctly resolve the physics of magnetic fields (specifically the magnetic length and flux per plaquette), simulations require system sizes significantly larger than those accessible to exact numerical methods (like exact diagonalization). For instance, a flux of $\Phi = 1/6$ per plaquette requires a lattice larger than $6 \times 6$, pushing the limits of exact solvability.
• Dimensionality Gap: While 1D systems have been extensively studied, 2D systems with magnetic fields remain poorly understood, particularly regarding the interplay between interaction strength and magnetic flux on diffusive transport.

The authors aim to investigate the diffusive relaxation dynamics of interacting spinless fermions in 2D lattice systems under uniform magnetic fields, specifically in the infinite temperature regime.

2. Methodology

The study employs a combination of exact numerical benchmarks and a semiclassical approximation method to overcome the exponential growth of the Hilbert space.

• Model: The system is modeled as spinless interacting fermions on a 2D lattice with nearest-neighbor hopping ($J$) and nearest-neighbor interaction ($V$). A uniform magnetic field is introduced via the Peierls phase in the hopping term along the $x$-direction using the Landau gauge. The flux per plaquette is defined as $\Phi = \gamma/2\pi$.
• Primary Method: Fermionic Truncated Wigner Approximation (fTWA):
  • The authors utilize fTWA, a semiclassical method that maps quantum operators to phase-space variables (Weyl symbols).
  • Mechanism: Quantum evolution is governed by Hamilton's equations of motion in phase space. Initial conditions are sampled from a Wigner function (approximated as a product of Gaussians for product states), and observables are calculated by averaging over an ensemble of trajectories.
  • Advantage: fTWA scales polynomially with system size, allowing simulations of hundreds of lattice sites, whereas exact methods scale exponentially.
  • Approximation: Interactions are treated approximately (truncated at a certain order), while magnetic fields are treated exactly within the formalism.
• Benchmarking: The reliability of fTWA is validated against the Lanczos method (exact diagonalization) on small "ladder" systems ($10 \times 2$).
• Initial State & Analysis:
  • The system is initialized with a sinusoidal density wave along the $x$-direction: $\langle \hat{n}_r(0) \rangle \approx 0.5 + A_0 \sin(2\pi m/\lambda)$.
  • The relaxation is analyzed by tracking the decay of the amplitude $A(t)$.
  • If the decay is exponential ($A(t) \propto e^{-\alpha t}$) and the decay rate scales as $\alpha \propto 1/L_x^2$, the system is identified as diffusive, allowing the extraction of the diffusion constant $D$ via $\alpha = 4\pi^2 D / \lambda^2$.

3. Key Contributions

• Validation of fTWA in 2D: The paper demonstrates that fTWA, which often fails in 1D integrable systems, provides unexpectedly high accuracy for 2D interacting systems with intermediate interaction strengths ($V/J \lesssim 2.5$). This accuracy holds even for small ladder systems where exact benchmarks are possible.
• Systematic Study of Finite-Size Effects: The authors establish that resolving magnetic field effects requires significantly larger system sizes than previously assumed. They show that for fluxes like $\Phi = 1/6$, system sizes must exceed $\sim 400$ sites (e.g., $L_x \times L_y \approx 32 \times 12$) to observe saturation of the diffusion constant and eliminate finite-size artifacts.
• Interaction-Dependent Transport Regimes: The study maps out how the interplay between interaction strength ($V$) and magnetic flux ($\gamma$) dictates transport behavior.

4. Key Results

A. Benchmarking and Reliability

• fTWA results for amplitude decay match Lanczos results closely for $V/J \leq 2.5$ in 2D ladder systems.
• The error between fTWA and exact results decreases as the system size increases (a common feature in semiclassical methods due to the suppression of higher-order quantum corrections).
• In contrast to 1D, where fTWA deviates significantly for interacting systems, the 2D geometry allows fTWA to capture the correct diffusive dynamics.

B. Magnetic Field Effects on Diffusion

• Strong Suppression: The presence of a magnetic field significantly suppresses the diffusion constant $D$.
  • For $\gamma = \pi$ (maximum flux), $D/J \approx 0.5$, compared to $D/J \approx 2.0$ in the zero-field case.
• Finite-Size Saturation: For intermediate flux values ($\gamma \in (0, \pi)$), the diffusion constant does not stabilize until the system size is large enough to accommodate the magnetic length in all directions. Specifically, for $\Phi = 1/6$ ($\gamma = \pi/3$), the $y$-direction size must be at least 12 sites to resolve the full flux per plaquette.

C. Role of Interaction Strength ($V$)

• Weak to Intermediate Interactions ($V/J \lesssim 1$): The magnetic field has a strong effect on diffusion. The orbital effects of the flux significantly hinder particle transport.
• Strong Interactions ($V/J > 1$): As interactions exceed the hopping energy, the influence of the magnetic field on diffusion is suppressed.
  • The diffusion constant follows a scaling law $D \propto V^{-2}$, consistent with a scattering rate growing as $V^2$.
  • In this regime, the magnetic length scale becomes less relevant, and the dynamics are dominated by interaction-induced scattering rather than orbital magnetic effects.

5. Significance and Outlook

• Theoretical Breakthrough: The work provides a validated theoretical tool (fTWA) to study non-equilibrium dynamics in 2D interacting systems under magnetic fields, a regime previously inaccessible to exact numerical methods.
• Experimental Relevance: The authors suggest that their findings are directly testable on current optical lattice platforms using ultracold atoms. These platforms can already generate synthetic magnetic fields and create density waves, making the predicted suppression of diffusion by magnetic fields and the interaction-dependent crossover experimentally verifiable.
• Physical Insight: The results clarify that while magnetic fields strongly inhibit diffusion in weakly interacting 2D systems, strong interactions can effectively "wash out" these orbital magnetic effects, leading to a transport regime dominated by interaction scattering.

In summary, this paper bridges the gap between theoretical limitations and experimental capabilities in 2D quantum transport, demonstrating that semiclassical methods can accurately capture complex many-body dynamics in magnetic fields provided the system size is sufficiently large and interactions are not excessively strong.