Rare-Event-Induced Ergodicity Breaking in Logarithmic Aging Systems
Chunyan Li, Qingyang Feng, Tianjie Zhou, Haiwen Liu ⋆ \star, X. C. Xie · 2025
Read Rare-Event-Induced Ergodicity Breaking in Logarithmic Aging Systems by Chunyan Li, Qingyang Feng, Tianjie Zhou, Haiwen Liu ⋆ \star, X. C. Xie (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-12658.txt.
Introduction
Chunyan Li and co-authors report in this 2025 paper a new way for sluggish systems to fall out of equilibrium balance. Glasses and similar materials age and diffuse with extreme slowness, and why their time averages part from group averages was murky. The team studies systems whose spread grows only with the log of time.
In such systems, rare events grow rarer as the system ages. Those events barely move the averages, yet they stretch the system's own clock, so time and group averages of spread split apart and two distinct long-run distributions appear. The pattern differs from the classic Lévy picture. The authors offer the mechanism as groundwork for far-from-equilibrium matter.
Catalog record
| CREATOR | Chunyan Li, Qingyang Feng, Tianjie Zhou, Haiwen Liu ⋆ \star, X. C. Xie |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 46 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-12658
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Rare-Event-Induced Ergodicity Breaking in Logarithmic Aging Systems Chunyan Li, Qingyang Feng, Tianjie Zhou, Haiwen Liu ⋆ \star, X. C. Xie Abstract Ergodicity breaking [ 59 , 60 , 61 , 62 ] and aging effects [ 63 , 64 , 65 , 66 , 67 , 68 ] are fundamental challenges in out-of-equilibrium systems. Various mechanisms [ 63 , 64 , 68 , 69 , 70 ] have been proposed to understand the non-ergodic and aging phenomena, possibly related to observations in systems ranging from structural glass [ 71 ] and Anderson glasses [ 72 ] to biological systems [ 73 , 74 ] and mechanical systems [ 75 , 76 ] . While anomalous diffusion [ 77 , 78 , 79 ] described by Lévy statistics [ 80 ] efficiently captures ergodicity breaking [ 81 ] , the origin of aging and ergodicity breaking in systems with ultraslow dynamics remain unclear. Here, we report a novel mechanism of ergodicity breaking in systems exhibiting log-aging diffusion. This mechanism, characterized by increasingly infrequent rare events with aging, yields statistics deviating significantly from Lévy distribution, breaking ergodicity as shown by unequal time- and ensemble-averaged mean squared displacements and two distinct asymptotic probability distribution function. Notably, although these rare events contribute negligibly to statistical averages, they dramatically change the system’s characteristic time. This work lay the groundwork for microscopic understanding of out-of-equilibrium systems and provide new perspectives on glasses and Griffiths–McCoy singularities. \defbibfilter notinsegment1not segment=1 * Corresponding author. E-mail: [email protected] (H.L.) {refsegment} Introduction Ergodicity, a cornerstone of equilibrium statistical physics, is typically defined as the equivalence between the ensemble average and the time average of observables in the long-time limit [ 82 ] . Nevertheless, ergodicity breaking is widespread in out-of-equilibrium and non-stationary systems, arising from various mechanisms [ 66 , 67 ] . For example, in spin glass systems [ 64 ] ergodicity breaking is primarily driven by the infinitely long time required for the system to explore all possible states. In addition, ergodicity breaking may result from the effect of Anderson localization [ 83 , 84 ] , many-body localization [ 85 ] and integrable Bose gases [ 86 ] . Non-ergodic behavior is also observed in biological systems [ 87 , 88 ] . Aging, a key characteristic in out-of-equilibrium systems, refers to the phenomenon in which the relaxation dynamics slow progressively over time, exhibiting ultraslow behavior [ 66 , 67 , 68 ] . This effect, where the system’s response depends on the elapsed time since its initialization, has been experimentally observed in structural glass [ 69 ] , Anderson glass [ 72 ] , crumpled thin sheets [ 76 ] , and disordered systems [ 75 ] . Moreover, Bernaschi et al. [ 89 ] demonstrates ergodicity breaking in the aging process of spin glasses on a complex energy landscape. A central challenge remains in characterizing the complex dynamics of out-of-equilibrium systems, particularly those exhibiting both aging and ergodicity breaking. Log-aging diffusion provides a unique testing ground to re-examine the mechanisms of ergodicity breaking, aging and the subtle role of rare events beyond conventional statistical descriptions. Anomalous diffusion, observed in various complex systems [ 90 , 91 ] , is characterized by the mean squared displacement (MSD) growing asymptotically as a power law, \langle\Delta^{2}x\rangle\propto t^{\alpha} (with \alpha=1 corresponding to normal diffusion). Previous studies [ 92 , 93 ] suggest that ergodicity breaking is related to the occurrence of Lévy statistics [ 77 ] , and dominated by a few rare events [ 94 , 95 ] . In this context, theories have been proposed to explain out-of-equilibrium dynamics [ 70 , 96 , 97 ] . However, the origin of aging and its influence on ergodicity remain unclear, especially in systems with log-aging diffusion characterized by \langle\Delta^{2}x\rangle\propto\ln{(t/t_{0})} with t_{0} representing the initial time [ 98 , 99 , 100 , 101 ] , related to various physical systems [ 72 , 74 , 73 , 75 ] . Through analysis of log-aging diffusion, we demonstrate ergodicity breaking governed by an unconventional mechanism. We find that increasingly infrequent rare events paradoxically control the long-time dynamics, resulting in long-tail distributions and infinite characteristic time. These findings pertain to ultraslow dynamics modeled via rugged landscapes [ 63 ] and demonstrate dynamics analogous to Griffiths-McCoy singularities [ 102 ] . Scale invariance breaking due to aging Figure 1: Scheme of Log-Aging Diffusion and Subdiffusion . (a) The forward waiting time distribution of log-aging diffusion governs the dynamics of the process, thereby introducing aging effects. Simulations of the first 5 steps’ random waiting times with \alpha=0.5 for t_{0}=10^{2} and 10^{3} are displayed in the top panel. In the bottom panel, the forward waiting time distributions at different t_{a} values collapse into a single curve after scaling by the factor t_{a} . (b) The waiting time distribution of subdiffusion exhibits a power-law tail, ensuring scaling invariance. In the top panel, the waiting times are distributed according to \frac{\alpha}{(t+1)^{1+\alpha}} , \alpha=0.5 . As shown in the bottom panel, data in the interval 10^{2}-10^{3} , when scaled by a factor of 10^{2} (solid triangles), reproduce the features observed in the interval 10^{4}-10^{5} (blue dots). The log-aging diffusion [ 98 , 99 ] , based on continuous-time random walk (CTRW) [ 103 , 78 , 104 ] , offers a statistical framework to analyze complex systems. In log-aging diffusion, both the jump length and the waiting time between successive jumps are continuous and random. While the jump length distribution is trivial, the waiting time distribution is flexible and depends on the present time t_{a} , also referred to as the forward waiting time distribution [ 98 ] , and is expressed as \psi_{1}(\tau\mid t_{a})=\frac{\sin{\pi\alpha}}{\pi}\frac{t_{a}^{\alpha}}{\tau^{\alpha}(\tau+t_{a})} . One consequence of this is that the process is sensitive to the initial time, t_{0} , of the experiment, such as the time when the external field is withdrawn. The logarithmic time evolution of log-aging diffusion for initial times t_{0}=10^{2} and 10^{3} is presented, and the statistics of waiting times at different present times t_{a} are shown in Fig 1 a. The system undergoes aging: over time, its state continues to depend on the current time, making it more likely to experience longer waiting periods and thus slowing down progressively. Moreover, the forward waiting time distribution decays with a power-law tail, leading to an infinite average waiting time. The collapse of the curves for different t_{a} when plotted with t/t_{a} on the x-axis and t_{a}\psi_{1} on the y-axis indicates that the system’s scale varies over time, thereby breaking its scale invariance. These features make log-aging diffusion a powerful tool for analyzing the dynamics of out-of-equilibrium disordered systems, which are characterized by a nontrivial energy landscape and an infinite phase space of lifetime quantities [ 63 ] . For subdiffusion, often modeled as a renewal process, the waiting time distribution \psi(\tau) exhibits an asymptotic power-law tail, \psi(\tau)\sim A_{\alpha}\tau^{-1-\alpha} . This heavy tail leads to time-independent scale invariance in the system’s properties (Supplementary Fig. S4 ). As demonstrated by data collapsing across different time scales (Fig. 1 b), this scale invariance governs key characteristics, such as constant relative fluctuations [ 81 ] and specific scaling forms for probability distributions [ 105 ] . Thus,