🔬 condensed matter

Gap-controlled thermalization in a SSH model version of the Fermi-Pasta-Ulam-Tsingou chain

This study demonstrates that in a dimerized Fermi-Pasta-Ulam-Tsingou chain, the acoustic-optical band gap acts as a tunable, momentum-selective filter for nonlinear energy transport, where a sharp isolation threshold emerges at half-maximum dimerization due to the onset of specific umklapp processes that either activate or suppress long-time thermalization depending on whether the gap permits three-wave scattering.

Original authors: José A. Aké, Gerardo G. Naumis

Published 2026-09-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: José A. Aké, Gerardo G. Naumis

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 world of solid materials, atoms are often imagined as tiny balls connected by stiff springs, vibrating in perfect rhythm. For decades, physicists have studied what happens when these springs are not perfectly linear—when they get slightly stiffer or looser as they stretch, a condition known as nonlinearity. In such systems, energy usually flows from one vibration pattern to another until it spreads out evenly among all possible movements, a state scientists call thermal equilibrium. However, a famous puzzle from the 1950s showed that under certain conditions, this energy does not spread out at all. Instead, it gets stuck in a few patterns, cycling back and forth for a very long time. This behavior challenged the idea that all complex systems naturally settle into a calm, mixed state. Today, researchers are asking whether we can control this behavior by changing the structure of the material itself, specifically by introducing gaps in the allowed frequencies of vibration, much like how a radio tuner blocks out certain stations to isolate a single signal.

A team of researchers has now explored this question using a computer model of a chain of atoms connected by alternating springs. Imagine a line of identical weights, but instead of every spring being the same, the stiffness alternates: one spring is strong, the next is weak, the next is strong, and so on. This arrangement creates a natural separation between low-frequency vibrations, which travel easily through the chain, and high-frequency vibrations, which are harder to excite. The researchers wanted to see if this separation, or gap, would stop the energy from mixing between these two types of vibrations. They simulated a chain of sixty-four particles, starting with a single low-frequency vibration and watching how the energy moved over time. By adjusting the difference in stiffness between the strong and weak springs, they could widen or narrow the gap between the low and high frequencies.

The study revealed a sharp and surprising threshold. When the difference in spring stiffness was small, the energy flowed freely from the low-frequency vibrations to the high-frequency ones, and the system eventually reached a balanced state where energy was shared among all modes. However, as the researchers increased the difference in stiffness, they found a critical point where this flow was strongly suppressed. Once the stiffness difference reached a specific value, the gap became too wide for the primary three-wave energy transfer mechanism to occur. The low-frequency vibrations remained largely trapped in their own group, and the high-frequency vibrations were only marginally excited, even after millions of oscillation cycles. The researchers discovered that this barrier is a strict rule for these specific three-wave collisions: the geometry of the system dictates that two low-frequency waves can only merge to create a single high-frequency wave if the gap is narrow enough. Once the gap widens beyond a certain limit, these specific resonant conditions vanish, effectively isolating the two groups of vibrations from one another, though other slower processes may still allow for minimal transfer.

The researchers also found that the way the ends of the chain are held affects this process. When the ends of the chain were fixed in place, the system behaved as expected, with the gap acting as a filter that allowed for partial relaxation even above the isolation threshold. But when the ends were connected to form a loop, the behavior became more complex. In this looped setup, the system got stuck in long-lived states for almost all levels of stiffness difference, except for a very narrow range where the gap was just right. In these specific cases, the energy managed to cross the gap quickly. In other cases, the energy remained trapped in a few patterns for a very long time, a phenomenon the researchers describe as a "sticky" state. This suggests that the boundaries of a material can fundamentally change how heat and energy move through it, creating situations where energy gets stuck in specific patterns rather than spreading out.

The study confirms that the gap in the vibration frequencies acts as a tunable filter for energy transport. By simply changing how different the springs are from one another, one can switch the system from a state where energy mixes freely to a state where it is strictly separated. This finding offers a new way to think about controlling heat and sound in materials. If scientists can engineer materials with specific gaps, they might be able to design systems that keep energy localized or allow it to flow only under very specific conditions. The work highlights that the path to thermal equilibrium is not always a straight line; it can be blocked by the very structure of the material itself. The researchers noted that while their results are based on computer simulations, they provide a clear picture of how these gaps control the flow of energy, suggesting a general method for managing thermalization in complex, structured materials.

Technical Summary: Gap-controlled thermalization in a SSH model version of the Fermi-Pasta-Ulam-Tsingou chain

Problem Statement
The Fermi-Pasta-Ulam-Tsingou (FPUT) problem historically highlighted the unexpected persistence of long-lived recurrences and the lack of thermalization in classical nonlinear lattices under specific initial conditions. While recent research has explored FPUT generalizations with spatial heterogeneity or topological features, the specific influence of a robust spectral band gap on long-time energy redistribution under nonlinear driving remains largely unexplored. This work addresses a critical open question in nonlinear dynamics and topological physics: whether dimerized spring constants in a classical Su-Schrieffer-Heeger (SSH) analogue can preserve an acoustic-optical gap and delay thermalization when finite-amplitude nonlinearities are introduced. Unlike previous studies focusing on mass-dimer models (alternating masses), this study investigates spring-dimer models (alternating spring constants), which exhibit distinct resonance manifolds.

Methodology
The authors study a one-dimensional FPUT chain consisting of $N=64$ equal masses connected by alternating spring constants $\kappa_A = 1 + \Delta\kappa$ and $\kappa_B = 1 - \Delta\kappa$ . The system includes cubic ( $\alpha$ -type) nonlinearity, where the nonlinear coefficients are locally scaled by the spring constants to reflect a common interatomic potential origin.

The study employs the following methodological framework:

Analytical Approach: The linear dispersion relation is derived using Bloch's theorem, identifying acoustic and optical branches separated by a frequency gap $\Delta\omega$ . The nonlinear Hamiltonian is transformed into normal modes to analyze three-wave (triad) resonances. The authors derive coupling coefficients ( $\Gamma$ ) for the acoustic-acoustic-optical ( $--+$ ) channel, which is the primary driver for interbranch energy transfer.
Numerical Simulation: Long-time symplectic integrations (8th-order Kahan-Li algorithm) are performed to preserve Hamiltonian structure and phase-space volume. Simulations cover a wide time window ( $T_{max} = 10^8 \Delta t$ ) to capture slow thermalization dynamics.
Boundary Conditions: Both Fixed Boundary Conditions (FBC) and Periodic Boundary Conditions (PBC) are examined. Initial excitations involve single low-frequency acoustic modes (typically the second mode for PBC to avoid center-of-mass translation, and the first for FBC).
Observables: Thermalization is quantified via modal energy evolution, spectral entropy $S(t)$ (using time-averaging to smooth recurrences), and the finite-time maximum Lyapunov exponent (ftMLE) to assess chaoticity.

Key Contributions and Results
The study establishes that the phononic gap acts as a tunable, momentum-selective filter for nonlinear energy transport. The primary findings are:

Isolation Threshold: A sharp isolation threshold exists at a dimerization strength of $\Delta\kappa = 0.5$ (corresponding to a spring ratio $\eta = \kappa_B/\kappa_A = 1/3$ ).
- Below Threshold ( $\Delta\kappa < 0.5$ ): Three-wave acoustic-acoustic-optical scattering is active. The system thermalizes, with the optical branch becoming activated on timescales of order $10^4$ oscillation periods.
- Above Threshold ( $\Delta\kappa > 0.5$ ): The spectral gap becomes too wide to satisfy the resonance condition for three-wave processes. Specifically, the condition $\omega_+(0) \leq 2\omega_-(\pi)$ cannot be met. Consequently, the acoustic and optical bands remain dynamically isolated, and thermalization is suppressed.
Role of Umklapp Processes: The onset of thermalization in the gapped regime is governed by the availability of Umklapp processes. The threshold $\Delta\kappa = 0.5$ coincides with the first allowed Umklapp process, where two zone-boundary acoustic phonons ( $k=\pm\pi$ ) fuse into a zone-center optical phonon ( $k=0$ ). For $\Delta\kappa > 0.5$ , this process is kinematically forbidden.
Interaction Vertex and Resonance Manifold: The efficiency of energy transfer depends on two factors: the density of the resonant manifold and the magnitude of the interaction vertex $|\Gamma|$ .
- As $\Delta\kappa$ increases, the resonant manifold density decreases.
- Simultaneously, the interaction vertex norm for the $--+$ channel decreases as the system approaches the "dimer liquid" limit ( $\Delta\kappa \to 1$ ).
- The effective scattering rate $R(\eta)$ , which integrates the overlap of the vertex and the resonant manifold, vanishes for $\eta < 1/3$ ( $\Delta\kappa > 0.5$ ).
Boundary Condition Dependence:
- FBC: Shows a monotonic decrease in thermalization efficiency as $\Delta\kappa$ increases, with partial relaxation possible even above the isolation threshold due to higher-order processes.
- PBC: Exhibits more complex behavior. For $\Delta\kappa \notin \{0.1, 0.2\}$ , the system becomes trapped in long-lived "sticky states" with low spectral entropy. This trapping is attributed to specific initial conditions (e.g., second-mode excitation) that align with phase-space trapping regions, a phenomenon reminiscent of the $N=3$ case.
Lyapunov Exponents: The ftMLE analysis reveals that while the gap controls global energy transfer (thermalization), it does not necessarily alter the local phase-space geometry (chaotic divergence) in the same way. In the PBC case, all $\Delta\kappa$ values collapse onto the same power-law decay for the Lyapunov exponent, suggesting the gap isolates energy sectors without necessarily changing the local chaotic nature of the trajectories.

Significance
The paper claims to establish the phononic gap as a tunable control parameter for nonlinear energy transport in dimerized lattices. By demonstrating that a specific threshold ( $\Delta\kappa = 0.5$ ) can effectively switch off three-wave resonant energy transfer, the work provides a general route to controlling thermalization in systems with topological or structural gaps. The results complement existing bulk-boundary pictures of SSH-like lattices by probing acoustic-mode spectral transfer within the bulk. The authors note that while the results elucidate the bulk mechanism for the $\alpha$ -FPUT chain, the behavior of $\beta$ -type nonlinearities and the origin of sticky states for larger $N$ remain open questions for future investigation. The study does not propose new experimental setups but offers a theoretical framework for understanding how structural dimerization influences nonlinear dynamics in classical lattices.

More like this