🔬 condensed matter

Tuning the strength of emergent correlations in a Brownian gas via batch resetting

This paper demonstrates that a non-interacting gas of Brownian particles subject to batch resetting develops long-range correlations in its nonequilibrium stationary state, where the correlation strength can be tuned by the batch size and undergoes a transition at a critical particle number of six, a phenomenon predicted to be observable in optical-trap experiments.

Original authors: Gabriele de Mauro, Satya N. Majumdar, Gregory Schehr

Published 2026-09-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Gabriele de Mauro, Satya N. Majumdar, Gregory Schehr

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 statistical physics, scientists study how vast collections of tiny particles move and interact. A central question in this field is how order can emerge from chaos. Usually, if particles do not push or pull on one another, they move independently, like strangers walking through a crowded room without ever noticing each other. However, a relatively new idea called stochastic resetting has shown that even without direct contact, particles can become deeply linked. This happens when an external force periodically interrupts their random wandering and sends them back to a starting point. When this happens to a group of particles all at once, they develop a strange, long-range connection: the position of one particle becomes statistically tied to the position of every other particle, creating a unified state that defies the usual rules of independent motion.

This phenomenon has sparked intense interest because it reveals how a shared, fluctuating environment can bind a system together. But a crucial question remained unanswered: what happens if the reset does not happen to everyone at the same time? If only a few particles are sent back while the rest continue wandering, does the group still stay connected, or does the link break? This is the puzzle that Gabriele de Mauro, Satya N. Majumdar, and Grégory Schehr set out to solve. They studied a gas of particles moving along a line, where, at random intervals, a specific number of them are chosen at random and returned to the origin, while the others keep moving. They wanted to know how the strength of the connection between particles changes as they vary the size of the group being reset.

The researchers discovered that the answer is far more complex than a simple "more is better" rule. They found that even when only a small fraction of the particles are reset at any given moment, the entire group still develops strong, long-range correlations. The particles, though never touching, act as if they are part of a single, coordinated unit. This happens because the random selection process creates a shared history; every particle is eventually affected by the reset events, and the timing of these events weaves their paths together. The team developed a new mathematical framework to track these connections exactly, allowing them to calculate the strength of the bond between any two particles for any size of the group and any number of total particles.

One of the most striking findings concerns how the strength of this connection changes over time. In the case where all particles are reset simultaneously, the correlation between them grows steadily until it reaches a stable maximum. However, when only a subset is reset, the story changes. The correlation builds up quickly at first, but then it begins to fade before eventually settling into a steady state. The researchers identified the cause of this dip: a specific type of "mismatch" event. Sometimes, one particle is reset while its neighbor is not. This asynchronous event breaks the link they had just built, acting like a temporary eraser of their shared history. This creates a unique rhythm where the connection strengthens, weakens, and then stabilizes, a behavior that never occurs when the whole group is reset together.

The study also revealed a surprising threshold in the number of particles required to see a specific kind of behavior. The team found that the relationship between the number of particles being reset and the final strength of the connection depends on the total size of the gas. If the gas contains fewer than six particles, the connection gets stronger the more particles are reset at once. But if the gas contains six or more particles, this rule breaks down. In larger groups, the strongest connection does not occur when the maximum number of particles are reset, nor when the minimum are reset. Instead, the peak connection strength appears when a very small, specific number of particles are reset at a time. This transition happens precisely when the total number of particles reaches six, a critical point where the balance between the individual wandering of particles and their collective resetting shifts.

These results are not just theoretical curiosities; they describe a physical reality that can be tested in the laboratory. The authors point out that modern experiments using optical traps—laser beams that hold tiny particles in place—can already implement this kind of partial resetting. By modulating the strength of the traps for only a few particles at a time, scientists can create the exact conditions described in the paper. The findings suggest that researchers can tune the behavior of these particle gases simply by adjusting how many particles are reset in each event, offering a new way to control the collective properties of matter without changing the particles themselves. The work provides a complete, exact description of how these emergent connections form, fade, and stabilize, turning a complex mathematical problem into a clear picture of how randomness can create order.

Problem Statement
This work investigates a gas of $N$ non-interacting Brownian particles on a line subject to a novel "batch resetting" protocol. In this setup, at a constant rate $r$ , a uniformly random subset of $m$ particles is instantaneously reset to the origin, while the remaining $N-m$ particles continue diffusing. The study addresses a fundamental gap in the theory of stochastic resetting: while the case of fully synchronous resetting ( $m=N$ ) is well-understood and solvable via standard renewal theory, the intermediate regime ( $1 < m < N$ ) presents a significant mathematical challenge. In this regime, the system never returns to a fully renewed configuration after a reset event, causing standard renewal methods to fail. The central question is how the strength of emergent long-range correlations, which arise purely from the dynamics of the shared fluctuating environment (the resetting events), depends on the batch size $m$ and the total number of particles $N$ .

Methodology
To overcome the breakdown of renewal theory for $m < N$ , the authors develop a general Fokker-Planck framework in Fourier space. The core technical innovation is the identification of an exact recursive closure for the marginal probability density functions (PDFs).

Fokker-Planck Equation: The authors derive the exact Fokker-Planck equation for the joint PDF in Fourier space. Crucially, they demonstrate that the evolution equation for an $n$ -point marginal involves only marginals of order $k \le n$ . This contrasts with generic many-body systems where a BBGKY-type hierarchy couples lower-order marginals to higher-order ones, rendering the problem unsolvable without approximation.
Recursive Solution: Leveraging this structure, the authors obtain exact solutions for the stationary state and time-dependent marginals. Specifically, they derive closed-form expressions for the one-point marginal $\tilde{P}^1_m(k)$ and the two-point marginal $\tilde{P}^2_m(k_1, k_2)$ for arbitrary $N$ and $m$ .
Correlation Analysis: Using these exact marginals, the authors compute the time-dependent two-point correlation function $C_{ij}(t) = \langle x_i^2(t)x_j^2(t) \rangle - \langle x_i^2(t) \rangle \langle x_j^2(t) \rangle$ . They define a normalized correlator $a(t; m, N)$ to quantify the strength of correlations, which interpolates between 0 (uncorrelated) and 1 (fully correlated).

Key Results

Exact Solvability: The paper provides exact analytical expressions for the stationary joint PDF and its marginals for all $1 \le m \le N$ . For $m=N-1$ and $m=1$ , the full stationary JPDF is explicitly constructed.
Tunable Correlations: The system generates robust, long-range attractive correlations for any $1 < m < N$ . The strength of these correlations is tunable via the parameter $m$ .
Non-Monotonic Time Dependence: A novel phenomenon is discovered for $1 < m < N$ $1 < m < N$ : the normalized correlation function $a(t)$ $a (t)$ exhibits non-monotonic behavior. It initially grows (correlation-building), reaches a maximum at a finite time $t^*$ $t^{*}$ , and then decays (decorrelation phase) before saturating to a stationary value. This behavior arises from a competition between:
- Correlation injection via simultaneous resets of pairs (probability $\propto m-1$ ).
- Correlation loss via diffusion and asynchronous resets (where one particle is reset and the other is not, probability $\propto N-m$ ).
- The timescale for this decorrelation diverges as $m \to N$ (the synchronous limit), explaining why the non-monotonicity vanishes in that limit.
Stationary Transition at $N_c = 6$ : The stationary amplitude of the unnormalized correlation strength, $V(m, N)$ $V (m, N)$ , displays a qualitative transition as a function of $N$ $N$ .
- For $N < 6$ , the correlation strength increases monotonically with $m$ , peaking at the fully synchronous limit ( $m=N$ ).
- For $N > 6$ , the behavior becomes non-monotonic. The global maximum shifts from $m=N$ to $m=2$ . Additionally, a shallow local minimum appears at $m=N-1$ .
- At the critical value $N_c = 6$ , the maximum is degenerate, occurring at $m=2, 3,$ and $N$ .
Dimensional Robustness: The analysis is extended to arbitrary spatial dimensions $d$ . The normalized correlation function remains non-monotonic in time, and the critical value $N_c = 6$ for the stationary transition remains unchanged, indicating the robustness of the results.

Significance and Claims
The paper claims to provide the first exact solution for a system of diffusing particles subject to partial batch resetting, a regime previously inaccessible to standard renewal techniques. The primary significance lies in the discovery of a new "intrinsic decorrelation mechanism" driven by asynchronous resetting events, which leads to non-monotonic temporal correlations. Furthermore, the identification of a critical particle number $N_c = 6$ governing the stationary correlation strength reveals a complex interplay between single-particle fluctuations and multi-particle correlations.

The authors assert that their predictions are experimentally testable using existing optical-trap setups (specifically referencing setups capable of modulating trap stiffnesses), where selective resetting of subsets of particles can be implemented. The work also establishes that the exact recursive closure of marginals is likely a generic property of batch-resetting systems, extending beyond simple diffusion. The paper concludes by noting that these results offer a unified framework for understanding dynamically emergent correlations in systems where degrees of freedom are coupled only through a shared, fluctuating environment acting on subsets of the system.

More like this