🔬 atomic physics

Fate of Discrete Time Crystals in Finite-Size Cavity QED Systems

This paper demonstrates that finite-size non-mean-field fluctuations in the open Dicke model induce non-monotonic decay of transient Discrete Time Crystals, a phenomenon accurately captured by the stochastic Truncated Wigner Approximation but missed by second-order cumulant expansions, revealing three distinct dynamical regimes with significant implications for experiments on finite atomic ensembles.

Original authors: Haimi Nguyen, David R. Reichman

Published 2026-10-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Haimi Nguyen, David R. Reichman

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

In the quiet corners of physics where matter refuses to settle into a comfortable, static state, a strange new kind of order has recently been discovered. Imagine a clock that does not tick at the same speed as the hand that winds it, but instead beats at half that speed, defying the rhythm of the force driving it. This is the essence of a discrete time crystal, a phase of matter that breaks the symmetry of time itself. Unlike a normal object that eventually settles down when pushed, a time crystal keeps oscillating in a perfect, repeating pattern, even as energy is pumped into it and lost to the environment. For years, scientists have wondered if this behavior could survive in the real world, where systems are never infinitely large and perfect, but are instead made of a finite, countable number of atoms. The question was not just whether these crystals could exist, but how their stability changes as the number of atoms grows from a handful to a vast crowd.

A team of researchers at Columbia University has now mapped out this journey, revealing that the path to a stable time crystal is far more winding than anyone expected. They studied a theoretical model where a group of identical atoms interacts with a cavity, a container for light, which is periodically shaken by an external force. In an idealized world with an infinite number of atoms, this system was known to form a robust time crystal that resists the chaos of energy loss. However, the researchers found that when they simulated systems with a realistic, finite number of atoms, the story changed dramatically. Instead of the stability simply improving as more atoms were added, the system displayed a complex, non-monotonic behavior. There were specific ranges of atom counts where the time crystal's rhythm would suddenly weaken or disappear, only to reappear and strengthen as the group grew even larger.

To uncover these details, the scientists had to overcome a significant hurdle: the standard mathematical tools used to predict the behavior of such large groups of atoms simply failed. These traditional methods, which rely on averaging out the behavior of individual particles, assumed that the system would behave in a smooth, predictable way. The researchers showed that these methods could not capture the subtle, chaotic fluctuations that occur when the number of atoms is finite. Instead, they turned to a more sophisticated approach that treats the system as a collection of many possible paths, or trajectories, evolving simultaneously. By running thousands of these simulated paths and averaging the results, they could see the true dynamics of the system. This method proved remarkably accurate, matching the results of exact calculations for small groups of atoms and allowing the team to simulate much larger groups that would otherwise be impossible to compute.

The simulations revealed three distinct fates for the time crystal, depending on the specific conditions of the system and the number of atoms involved. In some cases, the time crystal was unstable, its rhythm collapsing quickly into chaos. In others, it existed only for a short time, a transient phase that eventually faded away. But in a third regime, the system entered a state that the researchers call "pre-thermal," where the time crystal remained robust and stable for incredibly long periods, seemingly immune to the decay that usually destroys such order. The most surprising discovery was how the transition between these states depended on the size of the atomic group. As the number of atoms increased, the stability of the time crystal did not just improve steadily; it dipped and rose in a non-linear fashion. There were specific sizes, such as around one hundred to five hundred atoms, where the time crystal's signal would weaken significantly before recovering as the group grew even larger.

This behavior arises from a delicate competition between the size of the group and the boundaries of the system's possible states. The researchers found that the initial state of the atoms could straddle the line between two different stable rhythms. When the group is small, the fluctuations are large enough to blur this line, causing the system to lose its rhythm. As the group grows, these fluctuations shrink, but they do not disappear in a simple, straight line. Instead, the interplay between the size of the group and the complex geometry of the system's possible states creates these dips in stability. The researchers also noted that the loss of energy from the cavity, which usually acts as a damping force, actually played a dual role. While it eventually dampens the motion, it also acts as a cooling mechanism that helps preserve the time crystal's order by draining away the excess energy injected by the driving force.

The implications of these findings extend to the real-world experiments currently being conducted with trapped ions and superconducting circuits, where the number of atoms is often in the tens or hundreds. The researchers suggest that experiments in the range of N ∼10–50 are likely to see enhanced stability, as they fall into a regime where the time crystal is most robust. However, for systems with intermediate sizes around N ∼100–500, the stability may dip significantly before recovering. For much larger systems, such as those involving millions of atoms in a Bose-Einstein condensate, the stability should approach the ideal, infinite limit, though the path to get there is not a straight line. The work also highlights the importance of using the right tools to study these systems, showing that the advanced simulation method they used is a powerful way to understand the behavior of complex quantum systems far beyond the reach of simpler approximations. Ultimately, this study provides a clear map of how time crystals behave in the messy, finite reality of the laboratory, showing that their survival depends on a subtle balance of size, energy, and the intricate dance of quantum fluctuations.

Technical Summary: Fate of Discrete Time Crystals in Finite-Size Cavity QED Systems

Problem Statement
Discrete Time Crystals (DTCs) represent a non-equilibrium phase of matter characterized by robust subharmonic oscillations that break discrete time-translation symmetry. While DTC behavior has been demonstrated in the periodically driven, open Dicke model within the thermodynamic limit ( $N \to \infty$ ), experimental realizations in platforms such as trapped ions, superconducting qubits, and atom-cavity systems are inherently finite-sized. A critical open question remains regarding how finite-size effects influence the stability and lifetime of the DTC phase. Previous work by Gong et al. predicted that the DTC decay rate would decrease monotonically with the number of atoms ( $N$ ) and suggested that the second-order cumulant expansion (CE) would be a viable method for studying these finite-size fluctuations. This paper investigates the validity of these predictions and the reliability of standard approximation methods in the finite- $N$ regime.

Methodology
The authors study an ensemble of $N$ indistinguishable two-level atoms coupled to a lossy cavity mode with a periodically time-varying coupling $\lambda(t)$ . The system evolves under a Lindblad master equation accounting for cavity loss at rate $\kappa$ . To explore the dynamics across different system sizes, the authors employ three complementary numerical approaches:

Exact Numerical Simulations: For very small systems ( $N \le 10$ ), the Lindblad master equation is solved exactly using the QuTiP package, representing the atom ensemble in a collective spin basis. For intermediate sizes ( $10 < N \le 20$ ), the Monte Carlo Wave Function (MCWF) quantum jump method is used, averaging over 5,000 stochastic trajectories to ensure convergence.
Truncated Wigner Approximation (TWA): For large $N$ , a stochastic version of the TWA is employed. This semi-classical method evolves an ensemble of classical trajectories sampled from the Wigner distribution of the initial quantum state. A key technical advantage highlighted is that the computational cost of this specific TWA formulation remains independent of $N$ by exploiting the symmetry of the initial state and the nature of the noise, allowing for the simulation of large ensembles.
Second-Order Cumulant Expansion (CE): The authors compare their results against both the Partial Second-Order Cumulant Expansion (PSC) and the Full Second-Order Cumulant Expansion (FSC). They note that the PSC, which removes $Z_2$ symmetry-breaking terms, is fundamentally unsuitable for DTC dynamics. The FSC, while widely used, constrains the system to a Gaussian form, which may fail to capture non-Gaussian statistics arising from transient symmetry breaking.

Key Contributions and Results

Failure of Second-Order Cumulant Expansion: The authors demonstrate that the FSC is unreliable for describing finite-size DTC dynamics. Because the FSC forces the density matrix into a Gaussian form, it fails to capture fluctuations spanning two symmetric manifolds, which are expected during transient symmetry breaking. Consequently, the FSC often fails to converge or produces inaccurate steady-state results for symmetry-breaking order parameters like $\langle \sigma_x \rangle$ .
Accuracy of Stochastic TWA: Despite its approximate semi-classical nature, the stochastic TWA is shown to be remarkably accurate, particularly for $N > 20$ . The authors attribute this success to the open nature of the system, where dissipation rapidly destroys quantum coherence between symmetry-broken states, suppressing superposition states in favor of statistical mixtures. The TWA consistently reproduces exact results within a 5% error margin for $N > 20$ .
Non-Monotonic Finite-Size Scaling: Contradicting the prediction that DTC stability improves monotonically with $N$ $N$ , the authors identify a complex, non-monotonic behavior in the amplitude and decay rates of transient DTCs as $N$ $N$ increases.
- Dynamical Regimes: Three distinct regimes are identified: (1) Unstable DTC, where order is absent or decays rapidly; (2) Transient DTC, where order persists for hundreds of cycles but exhibits clear decay; and (3) Pre-thermal DTC, where order remains robust over long simulation times with vanishing decay rates.
- The "Dip" in Stability: In both transient and pre-thermal regimes, there exists an intermediate region of $N$ (e.g., $N \sim 100-500$ for pre-thermal) where DTC stability dips significantly before recovering and asymptotically approaching the mean-field limit. This manifests as a drop in DTC amplitude or an increase in decay rate before eventual stabilization.
Mechanism of Non-Monotonicity: The non-monotonic behavior is explained by the interplay between the size of the initial phase-space cloud (which decreases algebraically with $N$ ) and the intricate, foliated boundary separating two degenerate Floquet attractors (in-phase and out-of-phase). For finite $N$ , the initial cloud may straddle these boundaries, leading to trajectory scrambling and amplitude cancellation upon averaging. This effect is most pronounced when the mismatch between spin and cavity field angular velocities causes trajectories to dephase.

Significance
The paper establishes that finite-size effects in the open Dicke model lead to a non-trivial, non-monotonic approach to the thermodynamic limit, challenging the intuition that larger systems uniformly exhibit more stable DTC behavior. The work validates the stochastic TWA as a robust and efficient tool for describing long-time dynamics in finite-size cavity QED systems, a capability that extends beyond exact methods. These findings have direct implications for interpreting experiments on finite-sized atomic assemblies, suggesting that the observation of DTC order may be highly sensitive to the specific number of atoms in the ensemble, with potential "dips" in stability at intermediate scales. The authors note that while the specific model requires the absence of the diamagnetic $A^2$ term (which typically destroys the superradiant transition), the predicted effects may be observable in non-photonic platforms or specific Bose-Einstein condensate setups where this term is negligible or replaced.

More like this