🔬 condensed matter

Identifying Non-Ideal Reaction-Diffusion Systems Unable to Maintain Diffusion Out-of-Equilibrium

This paper introduces a general method using a kinetic potential as a Lyapunov function to demonstrate that diffusion inevitably equilibrates in non-ideal reaction-diffusion systems driven by autonomous chemostats, provided their reaction networks are either pseudo-detailed balanced or complex balanced.

Original authors: Francesco Avanzini, Timur Aslyamov, Massimiliano Esposito

Published 2026-07-08

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Francesco Avanzini, Timur Aslyamov, Massimiliano Esposito

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 a bustling city where molecules are the citizens. Some citizens are "internal" (they live and work within the city limits), while others are "external" (they constantly arrive from and leave to the outside world, like commuters). These citizens interact, react with one another, and move around (diffuse) to find new neighbors.

This paper is about understanding when this molecular city settles down into a calm, uniform state versus when it stays in a chaotic, constantly moving state.

The Big Question: Why Don't Things Ever Settle?

In nature, molecules often form complex structures (like droplets inside a cell). Usually, if you leave a system alone, diffusion (random movement) smooths everything out until the concentration is the same everywhere—like a drop of ink spreading evenly in water.

However, in living cells, chemical reactions are constantly driven by energy, keeping the system "out of equilibrium." Sometimes, this driving force is so strong that diffusion never gets a chance to smooth things out; the molecules stay clumped in specific patterns. The authors wanted to figure out: Under what specific rules does diffusion always win and smooth everything out, regardless of how hard the reactions are pushing?

The Tool: The "Kinetic Potential" (The Energy Slope)

To answer this, the authors invented a mathematical tool called a Kinetic Potential. Think of this as a giant, invisible landscape or a hill.

The Goal: They wanted to prove that this "hill" always slopes downward.
The Analogy: Imagine a ball rolling down a hill. As long as the hill slopes down, the ball (the system) will keep rolling until it reaches the very bottom (a steady, calm state).
The Proof: If they can show this "hill" always goes down, they prove that the system must eventually stop moving chaotically and settle into a uniform state where diffusion has done its job.

The Two Types of Molecular Cities

The authors tested their "hill" method on two specific types of molecular networks (blueprints for how molecules react):

1. The "Pseudo-Detailed Balanced" City

The Rule: In this city, the rules for how molecules react depend on both the people starting the reaction (reactants) and the people finishing it (products).
The Result: The authors proved that if the reaction rules follow this specific "both sides" logic, the "hill" always slopes down. No matter how hard the external forces push, the molecules will eventually spread out evenly. Diffusion wins.
The Catch: If the reaction rules depend only on the starting people (ignoring the finishers), or if the rules change based on how crowded the neighborhood is in a weird way, the "hill" might have a flat spot or a bump. In that case, the system can get stuck in a chaotic, non-uniform state.

2. The "Complex Balanced" City

The Rule: This is a stricter type of city. Here, the reaction rules depend only on the starting people (reactants). It's like a factory assembly line where the speed depends only on the raw materials you put in, not what comes out.
The Result: The authors found that even with this stricter rule, the "hill" still slopes down. Diffusion wins, and the system becomes uniform.
The Surprise: This was a generalization of previous work. Before, scientists thought this only happened with very specific types of reactions (like simple "Arrhenius" rules). The authors showed it works for a broader range of "reactant-only" rules.

The "Why" Behind the Rules

Why do these specific rules matter? The authors explain it using the concept of Lumped Reactions.

Imagine several different roads leading from "Point A" to "Point B." In a "Pseudo-Detailed Balanced" system, all these roads must behave in a way that is consistent with the entire trip (start to finish).
If the roads behave differently depending on which specific path you take (even if they go from A to B), the "hill" gets bumpy. The system can get stuck in a traffic jam (a non-equilibrium state) where diffusion can't clear the congestion.

The Bottom Line

The paper provides a general method to predict when a chemical system will calm down and become uniform.

If the reaction rules follow the specific "Pseudo-Detailed" or "Complex Balanced" patterns: The system will eventually smooth out. Diffusion cannot be kept out of equilibrium.
If the reaction rules break these patterns (e.g., reacting differently based on local crowding in a way that doesn't fit the rules): The system can stay chaotic and non-uniform forever, even with constant external driving.

The authors also discovered a new mathematical "floor" (a lower bound) for how much energy is wasted (entropy production) in these systems when they are forced to stay out of equilibrium.

In short: They built a mathematical "gravity" test. If a molecular network passes the test (by following specific stoichiometric rules), gravity (diffusion) will always pull it into a calm, uniform state. If it fails the test, it can stay in a chaotic, patterned state forever.

Technical Summary: Identifying Non-Ideal Reaction-Diffusion Systems Unable to Maintain Diffusion Out-of-Equilibrium

Problem Statement
The paper addresses the physical principles governing the self-organization of biomolecular condensates, which arise in complex, non-equilibrium environments where constituents undergo continuous chemical reactions. While microscopic and mesoscopic approaches have elucidated various phenomena (e.g., spatial concentration profiles, instabilities, and coarsening), many existing models rely on phenomenological descriptions that lack systematic thermodynamic consistency. A specific gap exists in understanding when diffusion in non-ideal reaction-diffusion (RD) systems necessarily equilibrates (i.e., reaches a homogeneous steady state) versus when it can be sustained out of equilibrium by autonomous homogeneous chemostats. Previous work by the authors [30] identified classes of networks where diffusion equilibrates but relied on specific assumptions regarding reaction fluxes that appeared to contradict particle-based simulations [9, 12] where diffusion remained out of equilibrium. The central problem is to resolve this apparent contradiction and establish general conditions under which diffusion equilibrates in non-ideal RD systems.

Methodology
The authors develop a general method based on the construction of a kinetic potential, denoted as $\hat{F}[\mathbf{c}]$ , which acts as a Lyapunov function for the system. The method operates within a macroscopic, mean-field framework for open, non-ideal RD systems at constant temperature and volume.

The core of the method involves two steps:

Construction of the Potential: A kinetic potential is defined as $\hat{F}[\mathbf{c}] = F[\mathbf{c}] + \sum_x \Delta_x \int_V d\mathbf{r} c_x(\mathbf{r})$ , where $F[\mathbf{c}]$ is the nonequilibrium Helmholtz free energy and $\{\Delta_x\}$ are concentration-independent coefficients encoding kinetic properties.
Verification of Non-Negativity: The authors derive the time derivative of this potential, $\dot{\hat{F}}[\mathbf{c}] = -T(\dot{\hat{\Pi}}[\mathbf{c}] + \dot{\Sigma}_{dff}[\mathbf{c}])$ . They aim to prove that the reaction contribution $\dot{\hat{\Pi}}[\mathbf{c}]$ is non-negative. If $\dot{\hat{\Pi}}[\mathbf{c}] \geq 0$ and $\dot{\Sigma}_{dff}[\mathbf{c}] \geq 0$ (the latter being the entropy production due to diffusion), then $\dot{\hat{F}}[\mathbf{c}] \leq 0$ . Since $\hat{F}$ is lower-bounded, this implies the system relaxes to a steady state where both $\dot{\hat{\Pi}}$ and $\dot{\Sigma}_{dff}$ vanish, meaning diffusion equilibrates.

The method is applied to two topological classes of Chemical Reaction Networks (CRNs): pseudo-detailed balanced and complex balanced networks. The analysis focuses on identifying the specific constraints on reaction fluxes required for the construction of the Lyapunov function to hold.

Key Contributions and Results

Generalization of Previous Results: The paper generalizes the findings of Ref. [30] by relaxing assumptions on reaction fluxes. It demonstrates that diffusion equilibrates in non-ideal RD systems driven by autonomous homogeneous chemostats if the CRN belongs to one of two classes and the reaction fluxes satisfy specific stoichiometric constraints.
Conditions for Pseudo-Detailed Balanced CRNs:
- For a CRN to be pseudo-detailed balanced, cycles must consist of sequences of reactions interconverting the same pair of complexes (lumped reactions).
- The authors prove that diffusion equilibrates if reaction fluxes are controlled by the stoichiometry of both reactants and products within a lumped reaction. Specifically, the fluxes must satisfy Eq. (22), where the dependence on dynamical concentrations is determined solely by the lumped reaction $\varepsilon(\rho)$ , not the specific reaction $\rho$ .
- This condition allows for fluxes derived from transition state theory (Eq. 24) and specific forms used in Ref. [41] (Eq. 25), provided the symmetry factors are consistent across the lumped reaction.
- If fluxes depend on the specific reaction within a lumped set (e.g., different transition rates for reactions involving chemostatted species vs. those that do not, as seen in Refs. [9, 12, 41]), the condition fails, and diffusion can remain out of equilibrium.
Conditions for Complex Balanced CRNs:
- For a CRN to be complex balanced, every cycle must correspond to a loop in the graph of complexes (deficiency $\delta=0$ ).
- The authors show that diffusion equilibrates if reaction fluxes are controlled only by the stoichiometry of the reactants (the complex acting as a reactant). This is formalized in Eq. (37), where the flux depends on the complex $\kappa(\rho)$ but not the specific reaction $\rho$ .
- This constraint is more restrictive than that for pseudo-detailed balanced networks. It excludes fluxes where the dependence on concentrations is determined by both reactants and products (e.g., Eq. 24).
- Under these conditions, the kinetic potential can be constructed using coefficients derived from a homogeneous steady state ( $\Delta_x = -\mu_x(\mathbf{c}^h_{ss})$ ), ensuring $\dot{\hat{\Pi}} \geq 0$ .
Novel Lower Bound on Entropy Production:
- The method reveals a tighter lower bound for the entropy production rate due to reactions ( $\dot{\Sigma}_{rct}$ ) than the standard non-negativity condition.
- For complex balanced networks, this bound (Eq. 47) relates the entropy production to the affinity of reactions at the homogeneous steady state multiplied by the instantaneous integrated currents.

Significance and Claims
The paper claims to resolve the apparent contradiction between mean-field theories (which predicted diffusion equilibration) and particle-based simulations (which observed non-equilibrium diffusion) in pseudo-detailed balanced systems. The authors attribute this discrepancy to the specific form of reaction fluxes used in simulations, which violated the necessary stoichiometric constraints (Eq. 22) derived in this work.

The significance lies in providing a rigorous, thermodynamically consistent framework to determine when diffusion must equilibrate in non-ideal, driven RD systems. The authors modestly state that while they have identified the conditions for two specific classes of networks, a general procedure to construct such a kinetic potential for arbitrary CRNs remains an open question. They emphasize that the different constraints on fluxes for the two network classes originate from their distinct stoichiometric properties (specifically, the relationship between the stoichiometric matrix and the incidence matrix). The work does not propose new experimental setups but rather clarifies the theoretical boundaries within which non-equilibrium self-organization via diffusion can be sustained.