🔬 condensed matter

Discrete distributions and statistical mechanics of small systems

This paper establishes a theoretical framework linking discrete probability distributions to the statistical mechanics of small systems, demonstrating how concepts like infinite divisibility and scalability underpin quasiparticle descriptions, virial expansions, and phase transitions, while extending these principles to non-extensive thermodynamics and deformed distributions.

Original authors: Lev B. Klebanov, Michal Šumbera

Published 2026-07-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Lev B. Klebanov, Michal Šumbera

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

Imagine you are standing in a kitchen, watching a pot of water boil. To a physicist, that bubbling pot is a chaotic dance of trillions of invisible particles, all bumping into each other. For over a century, scientists have used a set of rules called "statistical mechanics" to predict how these massive crowds behave. They treat the particles like a giant, smooth fluid, assuming that if you have enough of them, the individual quirks don't matter. But what happens when you shrink that pot down to the size of a single grain of sand? Or when you look at a tiny cluster of atoms in a particle collider? The old rules start to crack. This is the world of "small systems," where the number of particles is so small that you can't just average them out; you have to count them one by one.

In this tiny world, the number of particles isn't a fixed, smooth number. It's a roll of the dice. Sometimes you have 10 particles, sometimes 12, sometimes 8. Scientists call this a "probability distribution." Think of it like a weather forecast: instead of saying "it will rain," the forecast gives you a percentage chance. In the microscopic world, the "weather" is the number of particles, and the "forecast" is a mathematical map showing how likely you are to find a specific number of them. The big question is: what shape does this map take? Does it look like a perfect bell curve, or something stranger? Understanding this shape is crucial because it tells us how energy flows, how matter changes state (like turning from liquid to gas), and how the universe behaves when things get really, really small.

This paper, written by Lev Klebanov and Michal Sumbera, is like a detective story that connects two seemingly different worlds: the math of counting dice rolls and the physics of boiling water. The authors are trying to figure out the exact "shape" of the particle count in these tiny systems. They discovered that the number of particles in these small systems often follows very specific, elegant mathematical patterns. One of their biggest findings is that these patterns are "infinitely divisible." Imagine you have a bag of marbles. If you can split that bag into two smaller bags, and then split those smaller bags into even tinier bags, and keep doing this forever without the marbles ever turning into something that isn't a marble, that's infinite divisibility. The authors show that the particle counts in these systems behave exactly like that.

They also introduce a concept called "scalability." This is a bit more technical than just resizing a toy. In the paper, scalability is a property of the mathematical formula (the probability generating function) that describes the particle count, not the physical count itself. It means that if you take this mathematical formula and stretch or shrink it by multiplying it by any non-negative real number—like 0.5 or 2.5—the result is still a valid mathematical map for a discrete distribution of whole numbers. It's as if the underlying rules of the game are so flexible that even if you mathematically "stretch" the probability landscape by a fraction, the system still resolves into a clean, countable set of outcomes without breaking the rules of the game. The paper suggests that the rules governing the probability of finding particles in these small systems allow for this unique mathematical scaling, preserving their discrete character even when the distribution is "stretched" or "shrunk" by real numbers.

The authors apply these math tricks to real-world physics problems, like the moment a liquid turns into a gas (the critical point). They found that right at this chaotic tipping point, the distribution of particles doesn't look like a normal bell curve. Instead, it settles into a specific, strange shape called the "Discrete Stable distribution." They argue that this shape is a "fixed point," meaning it's the ultimate destination for the system's behavior no matter how you zoom in or out. It's like a whirlpool in a river; no matter where you drop a leaf, it eventually gets sucked into that same spinning pattern.

Furthermore, the paper tackles a modern twist on physics called "Tsallis thermodynamics," which deals with systems where particles interact over long distances, like a crowd where everyone can shout to everyone else, not just their neighbors. In this weird, long-range world, the usual rules of "infinite divisibility" break down, but the authors show that "scalability" still holds true. Even in this chaotic, long-range environment, the mathematical description of the particle counts can still be scaled by real numbers while keeping their fundamental discrete character. They also show how familiar distributions, like the "Negative Binomial," are actually just "deformed" versions of the simple "Poisson" distribution, much like how a rubber band is just a stretched version of a spring.

In short, this paper builds a bridge between abstract number theory and the messy reality of tiny physical systems. It suggests that even in the chaotic, fluctuating world of the very small, there is a hidden order. By using these mathematical tools, scientists can better predict how small systems behave, from the tiny droplets in a cloud to the subatomic particles created in high-energy collisions. The authors don't just describe the chaos; they provide a new map to navigate it, showing that the rules of the game are more consistent and scalable than we previously thought, even when the system is too small to be described by the old, giant-system rules.

Technical Summary: Discrete Probability Distributions and Statistical Mechanics of Small Systems

Problem Statement
Classical statistical mechanics traditionally addresses macroscopic systems with an enormous number of degrees of freedom. However, extending these principles to small, non-macroscopic systems (e.g., colloidal particles, polymers, or systems formed in ultra-relativistic nuclear collisions) requires a rigorous treatment of ensembles with a finite, countable number of particles. A significant gap exists in the literature regarding the particle-number distribution ( $p_k = P(N=k)$ ) in the grand canonical ensemble. While often assumed to be Poissonian for simplicity, this assumption fails to capture observed multi-particle correlations and may be inconsistent with the behavior of thermodynamic quantities near phase transitions. The paper aims to bridge discrete probability theory and the statistical mechanics of small systems by constructing an equation of state based on general particle-number distributions rather than relying on uncertain microscopic interaction dynamics.

Methodology
The authors employ the apparatus of probability generating functions (p.g.f.) to develop a theoretical framework for non-negative integer-valued random variables. The methodology proceeds through several interconnected stages:

Probabilistic Foundations: The paper establishes the properties of specific families of discrete distributions, including Power Series Distributions (PSD), infinitely divisible distributions, and compound distributions. It utilizes p.g.f.s to define moments, cumulants, and combinants.
Kinetic Modeling: Stationary solutions of birth-death (B-D) equations are analyzed to identify "Markovian distributions," which satisfy one-step recurrence relations. This links the probabilistic structure to kinetic models of particle creation and annihilation.
Statistical Mechanical Application: The theoretical results are applied to the Grand Canonical Ensemble (GCE). The partition function is treated as a power series in fugacity, defining the particle-number distribution as a PSD. The authors investigate conditions for infinite divisibility and scalability within this ensemble.
Phase Transition Analysis: The behavior of particle-number distributions near the critical point of the liquid-vapor phase transition is studied using renormalization group concepts, specifically looking for fixed points of scaling transformations.
Non-Extensive Thermodynamics: The framework is extended to Tsallis thermodynamics to examine systems with long-range interactions, testing the preservation of scalability versus infinite divisibility.

Key Contributions and Results

Power Series and Infinitely Divisible Distributions: The authors demonstrate that particle-number distributions in the GCE naturally form Power Series Distributions. They derive conditions under which these distributions are infinitely divisible, linking this property to a "quasiparticle picture" of an interacting gas.
Combinants and Virial Expansion: A novel connection is established between the "combinants" of a distribution (coefficients in the expansion of $\log H(w)$ ) and the virial expansion in statistical mechanics. The paper shows that the virial equation of state can be derived directly from the properties of these combinants.
Markovian Distributions: The study identifies that stationary solutions of birth-death processes correspond to Markovian distributions. This class includes well-known distributions (Poisson, Binomial, Negative Binomial, Logarithmic, Sibuya) and provides a kinetic justification for their appearance in statistical mechanics.
Critical Point and Discrete Stable Distribution: In the vicinity of the critical point of a liquid-vapor phase transition, the particle-number distribution converges to the Discrete Stable distribution. This distribution is identified as the fixed point of a renormalization semi-group transformation, arising from the requirement of self-decomposability (splitting a system into independent fractions).
Scalability vs. Infinite Divisibility: A crucial distinction is drawn between infinite divisibility and scalability. While infinite divisibility (the ability to decompose a variable into a sum of i.i.d. variables) is central to standard statistical mechanics, it may fail in systems with long-range interactions. However, the paper proves that scalability (the ability to scale the distribution by a real factor while preserving its discrete character) is preserved even in Tsallis non-extensive thermodynamics.
Deformed Distributions: The framework naturally yields deformed discrete distributions. For instance, the Negative Binomial distribution is shown to arise as a $q$ -deformation of the Poisson distribution within the context of non-extensive thermodynamics.

Significance and Claims
The paper claims to fill a theoretical gap by providing a rigorous probabilistic foundation for the statistical mechanics of small systems. By moving beyond the assumption of Poissonian distributions, the authors offer a mechanism to describe multi-particle correlations and phase transitions more accurately.

Specifically, the work asserts that:

The infinite divisibility of particle-number distributions provides a natural quasiparticle description for interacting gases.
The Discrete Stable distribution is the unique fixed point for particle-number evolution near critical points, offering a new perspective on critical phenomena in small systems.
Scalability is a more robust property than infinite divisibility, remaining valid even in non-extensive thermodynamic regimes (Tsallis statistics) where standard infinite divisibility breaks down.
The use of combinants offers a direct link between the microscopic probability structure and macroscopic equations of state (virial expansion).

The authors conclude that these probabilistic tools are essential for a consistent description of small systems, particularly where traditional thermodynamic limits ( $V \to \infty$ ) are replaced by limits involving the maximal number of particles $N(V) \to \infty$ in a fixed volume, or where interactions are long-ranged.

More like this