On the absence of time-translation symmetry breaking in some non-reversible interacting particle systems

Using a free energy technique, the authors prove that non-reversible interacting particle systems on one- and two-dimensional lattices with strictly positive rates and a product stationary measure cannot exhibit time-periodic behavior, thereby ruling out time-translation symmetry breaking in these specific non-reversible settings.

The Gist

The Big Picture: Can a Noisy Crowd Dance in Sync?

Imagine a massive crowd of people standing on a grid (like a giant chessboard). Each person can be in one of a few states (e.g., wearing a red shirt or a blue shirt). They constantly change their shirt color based on what their neighbors are doing.

In physics, we usually study systems that eventually settle down and stop changing (like a cup of coffee cooling to room temperature). This is called a stationary state.

However, some systems are "non-reversible," meaning they have a built-in bias or a "push" that keeps them moving. The big question this paper asks is: Can a noisy, chaotic crowd of particles spontaneously organize itself into a perfect, repeating dance routine (a time-periodic orbit) without any external conductor?

If they could, they would be breaking "time-translation symmetry." In plain English: The laws of physics don't care when you start the clock, but if the crowd starts dancing in a loop, the system suddenly cares about the specific time of day. It's like a clock that starts ticking on its own.

The Main Discovery: The "Product Measure" Rule

The author, Jonas Köppl, proves a specific rule for crowds in 1D (a line) and 2D (a flat sheet):

If the crowd has a "random, independent" resting state (a product measure), they cannot spontaneously start dancing in a synchronized loop.

The Analogy: The "Independent Roommates"

Imagine a dormitory where every student (particle) makes decisions based on their neighbors, but there is a special "resting state" where everyone acts completely independently. No one is influenced by anyone else in this specific state; they are just flipping coins to decide their shirt color.

The paper proves that if such a "coin-flipping" state exists for the system, the students can never spontaneously agree to start a synchronized dance (like a wave or a rotating pattern) that repeats every 10 minutes.

Even if the system is "noisy" (random) and "pushy" (non-reversible), the existence of that one independent, random state acts like a brake. It prevents the system from locking into a rhythmic cycle.

Why is this a big deal?

1. The Dimensional Limit:

   • 1D and 2D (Lines and Sheets): The paper says, "Nope, you can't dance here." The noise and the geometry are too chaotic to sustain a perfect rhythm in these dimensions.
   • 3D and up (Volume): The paper admits, "We haven't proven it's impossible here." In fact, physicists suspect that in 3D, these synchronized dances can happen. The paper suggests that 3D is "thick" enough to support a stable rhythm, whereas 1D and 2D are too "thin" and fragile.

2. The "Product Measure" Condition:
   The proof relies on the system having a state where particles are independent. Think of this as a "baseline of chaos." The author shows that if you have this baseline, the system cannot build a "structure of time" (a rhythm) on top of it.

How did they prove it? (The "Energy" Metaphor)

The author uses a mathematical tool called Relative Entropy (or "Free Energy").

• The Metaphor: Imagine the system has a "disorder score."
  • If the system is in a perfect, repeating dance, the disorder score fluctuates in a specific way.
  • The author calculates how this score changes over time.
  • They found that if the system tries to dance in a loop, the "disorder score" would have to behave in a way that is mathematically impossible in 1D and 2D. It's like trying to balance a pencil on its tip while standing on a trampoline; the math says it will inevitably fall back to a random state.

The proof involves a clever trick:

1. They assume the system is dancing in a loop.
2. They measure the "energy cost" of this dance.
3. They show that in 1D and 2D, the "boundary costs" (the edges of the dance floor) are too high compared to the "bulk benefits" (the middle of the dance floor).
4. The math forces the conclusion: The only way to satisfy the energy rules is for the dance to stop immediately and return to the random, independent state.

The "No-Go" Result

The paper concludes with a "No-Go Theorem."

• The Conjecture: Physicists have long suspected that 1D and 2D systems with short-range interactions (where neighbors only talk to immediate neighbors) cannot sustain time-periodic behavior.
• The Proof: This paper proves that suspicion is correct, provided the system has a "product measure" (an independent resting state).

Why should you care?

This is about understanding the limits of order in chaos.

• In Nature: It helps explain why we don't see certain types of rhythmic patterns in thin films or 1D chains of atoms, even if the physics allows for "active" movement.
• In Technology: It sets boundaries for designing materials or networks that need to oscillate or pulse. If you are building a 2D sensor network, this paper tells you that you can't rely on the system to spontaneously synchronize its rhythm just by letting the particles interact; you need an external clock.

Summary in One Sentence

In a noisy, 1D or 2D world where particles interact with their neighbors, if there is a state where they act completely independently, they are mathematically forbidden from spontaneously organizing into a perfect, repeating time-loop. They can only dance if the room is 3D or larger.

Technical Summary

1. Problem Statement and Context

The paper addresses the phenomenon of Spontaneous Time-Translation Symmetry Breaking (TTSB) in stochastic interacting particle systems.

• The Phenomenon: TTSB occurs when a system, governed by time-independent (autonomous) dynamical equations, settles into a non-trivial time-periodic orbit rather than a time-stationary state. Mathematically, this means there exists a probability measure $\mu_0$ and a period $\tau > 0$ such that $\mu_0 P_\tau = \mu_0$, but $\mu_0 P_s \neq \mu_0$ for $0 < s < \tau$.
• The Gap: While TTSB is well-understood in mean-field approximations and certain long-range interaction models, its existence in systems with short-range interactions in low dimensions ($d=1, 2$) has been a contentious issue in physics.
  • Physics Conjecture: It is widely believed that TTSB cannot occur in $d=1, 2$ for short-range systems, even if the dynamics are non-reversible (non-equilibrium).
  • Mathematical Status: Previous rigorous results were limited to reversible systems or systems with finite-range interactions in $d=1$. The case of non-reversible systems in $d=2$ remained open.
• Goal: The author aims to prove that for a specific class of non-reversible interacting particle systems on $\mathbb{Z}^d$ ($d=1, 2$) with strictly positive rates, the existence of a product measure as a stationary state precludes the existence of any non-trivial time-periodic orbits.

2. Methodology and Framework

The proof relies on a free energy technique utilizing relative entropy and relative entropy loss, adapted to handle non-reversible dynamics.

A. Model Setup

• State Space: $\Omega = \{1, \dots, q\}^{\mathbb{Z}^d}$ (finite local state space).
• Dynamics: Continuous-time Markov processes defined by a generator $L$ with transition rates $c_x(\eta, \xi_x)$.
• Key Assumptions:
  • Strict Positivity (R3): Rates are bounded away from zero ($c_x > c > 0$).
  • Short-Range Decay (R4): The influence of distant spins decays sufficiently fast (faster than $|x-y|^{-2d}$), ensuring short-range interactions.
  • Product Measure Stationarity: The system admits a product measure $\mu$ (where spins are independent) as a time-stationary measure.
  • Non-Reversibility: The system is not assumed to be reversible; the proof works for general non-equilibrium dynamics.

B. Core Technical Strategy

The proof proceeds by contradiction. It assumes the existence of a time-periodic orbit and shows that the relative entropy loss must vanish, forcing the orbit to be trivial (constant).

1. Finite-Volume Relative Entropy Loss:
   The author defines the relative entropy loss in a finite volume $\Lambda$:
   $$g^L_\Lambda(\nu|\mu) = \frac{d}{dt}\bigg|_{t=0} h_\Lambda(\nu_t | \mu)$$
   where $h_\Lambda$ is the relative entropy.

2. Time-Averaged Entropy Loss Principle (Proposition 4.1):
   The central technical lemma states that if a measure $\nu$ evolves along a trajectory where the time-averaged relative entropy loss is zero, and if the measure satisfies a "positive-mass" property (strict positivity on cylinder sets), then $\nu$ must be the stationary product measure $\mu$.

   • Crucial Innovation: Unlike previous works (e.g., [JK25a]) that relied on reversibility to relate rates to the stationary measure, this paper uses the assumption that $\mu$ is a product measure to derive a specific decomposition of the entropy loss.

3. Decomposition of Entropy Loss (Lemma 5.1):
   The author derives a new representation of the entropy loss that separates:

   • Bulk Contribution: A negative term involving a convex function $F(x) = x \log x - x + 1$, which measures the deviation of the local density from the product measure.
   • Boundary/Error Terms: Terms arising from the finite volume approximation, controlled by the interaction range.

4. Inequality Chain (Lemmas 5.2 – 5.6):

   • Lemma 5.2: Establishes an inequality linking the bulk deviation (measured by coefficients $\alpha_\Lambda$) to the boundary terms (measured by $\gamma_\Lambda$).
   • Lemma 5.4: Uses the short-range assumption (R4) to bound the boundary coefficients $\gamma_\Lambda$.
   • Lemma 5.5: Relates the $L^1$ deviation ($\beta$) to the $L^2$ deviation ($\alpha$) using Cauchy-Schwarz.
   • Lemma 5.6 (Monotonicity): Proves that the bulk deviation $\alpha_\Lambda$ is non-decreasing with respect to the volume $\Lambda$. This step critically relies on $\mu$ being a product measure.

5. Dimensional Analysis (Lemma 5.7):
   The proof culminates in a recursive inequality involving the sum of deviations over expanding volumes $\Lambda_n$.

   • The inequality takes the form: $(\sum \delta_k)^2 \leq C n^{d-1} \sum \delta_k$.
   • For dimensions $d=1$ and $d=2$, the growth of the boundary term ($n^{d-1}$) is too slow to support a non-zero sequence of deviations $\delta_k$. This forces $\delta_k = 0$ for all $k$, implying the measure is constant.
   • For $d \geq 3$, the boundary term grows fast enough ($n^{d-1} \geq n^2$) to potentially allow non-trivial solutions, consistent with the existence of TTSB in higher dimensions.

3. Key Results

Main Theorem (Theorem 2.1)

Let $d \in \{1, 2\}$. If an interacting particle system satisfies the regularity conditions (R1–R4) and admits a product measure $\mu$ as a time-stationary measure, then the set of time-periodic orbits $\mathcal{O}$ coincides with the set of time-stationary measures $\mathcal{S}$, and both are equal to $\{\mu\}$.
$$\mathcal{O} = \mathcal{S} = \{\mu\}$$
Corollary 2.2: Consequently, such systems cannot exhibit non-trivial time-periodic behavior (TTSB).

Specific Contributions

1. First Result for Non-Reversible Systems in $d=2$: This is the first rigorous proof ruling out TTSB in two dimensions for non-reversible, short-range interacting particle systems.
2. Generalization of State Space: Extends previous results (which were often binary) to general finite local state spaces $\{1, \dots, q\}$.
3. Relaxation of Reversibility: Successfully removes the reversibility assumption, which was a major barrier in previous literature.
4. Product Measure Condition: Demonstrates that the existence of a product stationary measure is a strong constraint that enforces time-stationarity in low dimensions.

4. Significance and Implications

• Validation of Physics Conjectures: The paper provides rigorous mathematical support for the long-standing physics conjecture that TTSB is impossible in low-dimensional ($d=1, 2$) short-range systems, even far from equilibrium.
• Role of Dimensionality: The proof explicitly highlights the critical role of dimension. The "no-go" result holds for $d \leq 2$ but fails for $d \geq 3$ due to the scaling of boundary terms versus bulk terms in the entropy loss inequality. This aligns with numerical simulations suggesting TTSB is possible in $d \geq 3$.
• Methodological Advancement: The development of a free energy technique that does not require reversibility opens the door for analyzing other non-equilibrium phenomena in interacting particle systems.
• Limitations and Future Work:
  • The result relies on the stationary measure being a product measure. The author notes in the "Outlook" section that extending this to general Gibbs measures (quasilocal measures) remains an open problem, though it is conjectured that such measures are also time-stationary in $d=1, 2$.
  • The proof does not rule out TTSB in $d \geq 3$, consistent with known counter-examples in higher dimensions.

In summary, Köppl's work establishes a fundamental limitation on the long-time behavior of non-reversible particle systems in low dimensions, proving that the presence of a product stationary state acts as a "trap" preventing the emergence of coherent time-periodic oscillations.