Persistent coherent quantum dynamics in 2D long-range magnets via magnon binding

By combining large-scale neural quantum state simulations with effective theory, this study reveals that persistent coherent quantum dynamics and slow relaxation in 2D long-range quantum magnets arise from the formation of magnon bound states driven by effective attractive interactions, offering a generic mechanism observable in current quantum simulation platforms.

The Gist

The Big Picture: A Quantum Dance That Won't Stop

Imagine you have a giant ballroom filled with thousands of dancers (these are the atoms in a magnet). In a normal situation, if you suddenly change the music (a "quench"), the dancers would eventually get confused, bump into each other, and settle into a chaotic, random shuffle. This is called thermalization—the system loses its memory of how it started and just becomes a hot, messy soup.

However, in this specific type of 2D magnet (where the dancers are far apart but can still "hear" each other), the researchers discovered something amazing: The dancers never stop dancing in a synchronized rhythm. Instead of getting messy, they form pairs and waltz together for a very, very long time.

This paper explains why this happens and proves it using super-computers and clever math.

---

1. The Setting: The "Long-Range" Ballroom

Most magnets are like a crowded dance floor where you can only hold hands with the person standing right next to you. But in these long-range magnets (found in advanced labs using trapped ions or Rydberg atoms), the dancers can "feel" the presence of people across the entire room, even if they are far away. The force gets weaker with distance, but it never truly disappears.

The Experiment:
The scientists started with all the dancers facing the same direction (a "ferromagnetic" state). Then, they suddenly changed the rules (turned on a magnetic field). They expected the dancers to quickly lose their rhythm and become chaotic.

The Surprise:
Instead of chaos, the system kept oscillating (swinging back and forth) for a long time. It was as if the dancers had found a way to stay perfectly coordinated despite the noise.

2. The Secret Weapon: Magnon "Couples"

To understand this, we need to look at the "excitations" or the "mistakes" in the dance.

• Magnons: Imagine one dancer suddenly turns around while everyone else is facing forward. This "spin flip" is called a magnon. In a normal magnet, these magnons are like lone wolves running around, bumping into each other, and causing chaos.
• The Binding: In this long-range system, something magical happens. The long-distance connection between the dancers creates an invisible attraction. When two magnons (two dancers who turned around) get close, they don't just pass each other; they get "stuck" together.

The Analogy:
Think of the magnons as two people in a crowded room.

• In a normal room (short-range): They bump into each other and scatter.
• In this long-range room: There is a magical gravity between them. Even if they are several steps apart, they feel a pull. They form a bound state—a couple that moves together as a single unit.

Because they are bound together, they can't scatter easily. They move as a team, preserving the rhythm of the dance for a long time. This is what the authors call Magnon Binding.

3. The "Extended" Relationship

What makes this paper special is how far these couples stay together.

• In short-range magnets, couples only stick together if they are touching (nearest neighbors).
• In these long-range magnets, the "love" (attraction) is so strong that the couples can stay together even if they are several steps apart.

The researchers found that these "long-distance couples" can exist with gaps of 2, 3, or even 4 steps between them. It's like a couple holding hands across a crowded room without letting go. This creates a whole "zoo" of different types of couples, all dancing in sync.

4. How They Proved It: The "Neural Net" Crystal Ball

Simulating 2D quantum magnets is incredibly hard. The number of possibilities grows so fast that even the world's most powerful supercomputers usually crash when trying to solve it.

• The Old Way: Trying to calculate every single possibility (like counting every grain of sand on a beach).
• The New Way (NQS): The authors used Neural Quantum States. Think of this as training a super-smart AI (a neural network) to "guess" the state of the dancers. Instead of calculating every single grain of sand, the AI learns the pattern of the sand.
• The Result: The AI successfully simulated a 11x11 grid of dancers for a long time, showing that the oscillations (the dancing) didn't die out.

5. The "Effective Theory": The Rulebook

To explain why the AI saw this, the authors wrote a simplified "rulebook" (an effective theory).

• They showed that the long-range forces create a potential well (a valley) that traps the magnon couples.
• They calculated the energy levels of these couples and found they matched perfectly with the frequencies of the oscillations seen in the simulation.
• The Proof: It's like hearing a specific note on a piano and knowing exactly which two keys were pressed to make it. The "notes" (frequencies) of the dancing matched the "keys" (energy gaps) of the bound magnon couples perfectly.

Why Does This Matter?

This isn't just about magnets; it's about keeping quantum information alive.

• Quantum computers are very fragile. Usually, they lose their "coherence" (their ability to do calculations) very quickly because the environment messes them up (thermalization).
• This paper shows a generic mechanism where long-range interactions naturally protect the system. The "bound couples" act as a shield, keeping the quantum rhythm going.
• This suggests that future quantum simulators (using ions or atoms) could be built to exploit this effect, allowing them to store and process information for much longer than we thought possible.

Summary

In a world where quantum systems usually fall apart into chaos, this paper found a way to make them stick together. By using long-range forces, the "particles" (magnons) form long-distance couples that dance in perfect sync for a long time. The researchers used AI to simulate this and math to explain it, revealing a new, robust way to keep quantum systems coherent.

The Takeaway: In the quantum world, sometimes being far apart is actually the best way to stay close.

Technical Summary

1. Problem Statement

The dynamics of two-dimensional (2D) long-range quantum magnets represent a significant frontier in condensed matter physics, relevant to platforms like Rydberg atom arrays, trapped ions, and dipolar quantum gases. While 1D long-range systems are well-understood (exhibiting phenomena like prethermal plateaus and confined quasiparticles), extending these insights to 2D remains a central open problem due to two main challenges:

• Numerical Intractability: The exponential growth of the Hilbert space and rapid entanglement buildup make standard methods like exact diagonalization (ED) and tensor networks (tDMRG/MPS) ineffective for large 2D systems.
• Theoretical Gap: It is unclear how 2D geometry combined with power-law interactions ($1/r^\alpha$) alters elementary excitations, whether they form bound states over multiple lattice sites, and how this impacts thermalization.

The authors aim to determine the mechanism behind slow relaxation and persistent coherent dynamics in 2D long-range magnets, specifically investigating the 2D transverse-field Ising model.

2. Methodology

The authors employ a hybrid approach combining large-scale numerical simulations with an analytical effective field theory:

• Neural Quantum States (NQS) Simulations:

  • To overcome the limitations of ED, the authors use a variational representation of the wavefunction using artificial neural networks (specifically a Convolutional Neural Network based on the ResNet framework).
  • Time evolution is governed by the Time-Dependent Variational Principle (TDVP).
  • This allows them to simulate system sizes (up to $11 \times 11$ and $101 \times 101$ for spectral analysis) and evolution times ($Jt \approx 60$) far beyond the reach of exact diagonalization.
  • They monitor longitudinal magnetization $\langle \hat{S}^z_i(t) \rangle$ and connected spin-spin correlators $\tilde{C}(d, t)$ after a global quench from a fully polarized ferromagnetic state.

• Effective Few-Body Theory (Schrieffer-Wolff Transformation):

  • The authors construct an effective low-energy Hamiltonian by treating the transverse field ($g$) as a perturbation to the long-range Ising interaction ($J$).
  • Using a Schrieffer-Wolff (SW) transformation up to second order in $g$, they integrate out couplings between different magnon sectors (sectors with different numbers of spin flips).
  • This yields an effective Hamiltonian acting within specific magnon sectors (e.g., the two-magnon sector), revealing the nature of interactions between quasiparticles.

3. Key Contributions

• Discovery of Magnon Binding in 2D: The paper identifies that long-range interactions in 2D generate an effective attractive potential between magnons (spin flips). Unlike the 1D case where domain walls are confined, or the short-range 2D case where dynamics are governed by Hilbert space fragmentation, the 2D long-range regime features extended magnon bound states.
• Mechanism for Slow Relaxation: The authors demonstrate that these bound states are the quasiparticles responsible for the observed slow relaxation and persistent oscillations. The system fails to thermalize rapidly because the excitations are trapped in a manifold of bound states rather than scattering into a continuum.
• Validation of Theory: The work provides a rigorous quantitative link between the spectral peaks observed in NQS time-evolution data and the energy gaps calculated from the effective few-magnon Hamiltonian.

4. Key Results

A. Observation of Persistent Coherent Dynamics

• Oscillations: After quenching from a ferromagnetic state, the system exhibits long-lived, non-decaying oscillations in magnetization and correlation functions. This contrasts with the expectation of rapid thermalization in generic quantum systems.
• Spectral Analysis: Fourier transforms of the time-series data reveal sharp, well-defined frequency peaks. These peaks correspond to discrete energy gaps between few-magnon sectors (e.g., $0 \to 1$, $0 \to 2$, $1 \to 2$ magnon transitions), confirming that the dynamics are dominated by a sparse set of few-body excitations rather than a many-body continuum.

B. The Effective Hamiltonian and Magnon Binding

• Attractive Potential: The effective theory reveals that two magnons separated by distance $d$ experience an attractive potential:
  $$U(d) \approx -\left(2J - \frac{g^2}{2J}\right) \frac{1}{N_\alpha |d|^\alpha}$$
  This attraction is mediated by the long-range Ising coupling and persists over several lattice sites.
• Bound State Spectrum: The competition between this long-range attraction and the correlated pair hopping (kinetic term) creates a rich spectrum of bound states.
  • Inverse Participation Ratio (IPR): Analysis of eigenstates shows high IPR values for states with large mean separation $\bar{d}$, indicating localization (binding) even at distances $d > 1$.
  • Dependence on $\alpha$:
    • For strongly long-range interactions ($\alpha = 2, 3$), bound states exist up to separations of $d \approx 4$ or more, with many quasi-localized states.
    • For short-range interactions ($\alpha = 6, \infty$), binding is restricted to nearest-neighbor sites ($d \approx \sqrt{2}$), and the dynamics become monochromatic.

C. Quantitative Agreement

• The energy gaps calculated from the diagonalization of the effective few-magnon Hamiltonian match the peak positions in the Fourier spectra of the NQS simulations with high precision.
• This agreement validates the effective theory up to interaction strengths of $g/J = 0.5$ and confirms that the slow dynamics are governed by the formation and evolution of these bound pairs.

5. Significance and Outlook

• Generic Mechanism: The paper establishes that magnon binding is a generic mechanism for persistent quantum coherence in 2D long-range magnets, applicable to various platforms (Rydberg atoms, trapped ions, dipolar gases) where power-law interactions are present.
• Experimental Relevance: The findings suggest that current quantum simulation platforms can directly observe these spatially extended bound magnons using site-resolved correlation spectroscopy.
• Theoretical Impact: The work bridges the gap between 1D confinement physics and 2D many-body dynamics, showing that long-range interactions fundamentally alter the quasiparticle landscape in higher dimensions, preventing thermalization through the formation of stable, extended bound states.
• Future Directions: The framework is expected to apply to 3D systems and other kinetic processes, offering a new paradigm for understanding non-equilibrium dynamics in long-range interacting quantum matter.