Critical scaling and supercritical coarsening in Active Model B+

This study utilizes deterministic simulations to demonstrate that while Active Model B and its extension AMB+ share identical mean-field critical scaling at criticality, their supercritical coarsening behaviors diverge, with AMB exhibiting logarithmic corrections to the classic $t^{1/3}$ growth law and AMB+ showing suppressed corrections that ultimately lead to arrested growth in a microphase-separated state.

The Gist

The Big Picture: The Party That Never Stops

Imagine a crowded dance floor. In a normal, "passive" party (like a standard physics system at equilibrium), people eventually settle down. If the music stops, the crowd stops moving, and if people start grouping together, they form big, solid clusters until the whole floor is one giant group or a few large ones. This is like oil and water separating: eventually, you get a big blob of oil and a big pool of water.

Now, imagine an "Active" party. Here, every dancer has a tiny battery in their shoe. They are constantly burning energy to wiggle, spin, and push themselves around, even when the music stops. This is Active Matter. Because they are constantly using energy, the rules of the game change. They don't just settle; they might form weird, stable patterns that never quite merge into one giant blob.

This paper studies two specific rulebooks for how these "active dancers" (particles) behave when they try to separate into groups. The researchers wanted to know: Does the constant energy use change how fast these groups grow, or does it just add a little bit of chaos?

The Two Rulebooks: Model B vs. Model B+

The scientists looked at two versions of the rules:

1. Active Model B (AMB): This is the basic rulebook. The dancers have a tendency to push away from each other if they are too crowded, but they also have a "self-propulsion" kick that makes them move in weird ways.
2. Active Model B+ (AMB+): This is the upgraded rulebook. It adds a new twist: the dancers can create "currents" or flows that actively fight against the formation of giant clusters. It's like having a bouncer who actively breaks up big groups to keep the party lively and mixed.

The Main Discovery: The "Logarithmic" Speed Bump

In the old, passive world (like oil and water), if you mix them and let them separate, the size of the oil droplets grows over time following a very specific, predictable speed: $L(t) \sim t^{1/3}$. Think of this as a car driving at a steady speed on a highway.

The big question was: Does the "active" energy change this speed?

1. The Critical Moment (The Edge of Chaos)

First, the researchers looked at the exact moment the system is about to separate (the "critical point").

• The Finding: Surprisingly, even with all the extra energy and weird movements, the dancers behave exactly like the passive ones at this specific moment.
• The Analogy: Imagine a tightrope walker. Whether they are wearing a heavy backpack (active) or not (passive), if they are perfectly balanced, they wobble at the exact same speed. The "active" energy didn't change the fundamental rhythm of the wobble.

2. The Growth Phase (The Coarsening)

Next, they watched what happens after the separation starts. The groups (domains) start to grow.

• In Active Model B (AMB): The groups do grow, but not at a perfectly steady speed. The active energy acts like a slow, persistent headwind.

  • Instead of just $t^{1/3}$, the growth looks like $t^{1/3} \times (1 + \text{a tiny bit of } \ln t)$.
  • The Metaphor: Imagine a runner who is supposed to run a mile in 10 minutes. Because of the wind (active energy), they run the first half-mile at 10 minutes, but the second half takes slightly longer, and the third half takes even slightly longer. The wind doesn't stop them, but it makes them slow down logarithmically (very slowly, but noticeably). The paper found that in this model, the "wind" is strong, causing a noticeable delay in the groups merging.

• In Active Model B+ (AMB+): This is where it gets interesting. The new rule (the "bouncer" effect) fights back.

  • The Finding: The "bouncer" (the $\zeta$ term) cancels out the "headwind" (the $\lambda$ term).
  • The Metaphor: In this version, the dancers have a special move that pushes against the wind. The result? The groups grow almost exactly at the normal speed ($t^{1/3}$), with only a tiny, almost invisible delay. The "bouncer" successfully neutralized the chaos.

3. The "Micro-Phase" Arrest (The Frozen Party)

There was one more scenario. If the "bouncer" in Model B+ gets too aggressive (too much positive activity), something wild happens.

• The Finding: The groups stop growing entirely. They don't merge into one giant blob, nor do they stay as one big mess. They get stuck in a state of micro-phase separation.
• The Metaphor: Imagine the bouncer is so good at breaking up groups that he keeps the dancers in small, perfect circles of 5 people. No matter how long the party goes on, you never get a crowd of 50. The system gets "arrested" or frozen in a pattern of small, stable bubbles. This is a state that doesn't exist in normal physics; it's a unique feature of active matter.

Why Does This Matter?

For a long time, scientists thought active matter might follow completely new, mysterious laws of physics. They wondered if the "active" energy would break the old rules entirely.

This paper says: "No, the old rules mostly still apply, but with a twist."

• The Twist: The active energy doesn't change the type of growth (it's still the $t^{1/3}$ law).
• The Correction: It just adds a "logarithmic correction"—a slow, subtle drag on the process.
• The Control: In the more advanced model (AMB+), you can tune the system to either make this drag stronger, cancel it out completely, or even stop the growth entirely to create stable, tiny patterns.

Summary in One Sentence

The researchers discovered that while active, energy-hungry particles behave differently than passive ones, they still follow the same fundamental growth laws as normal matter, just with a "logarithmic speed bump" that can be tuned, canceled out, or used to freeze the system into tiny, stable patterns.

Technical Summary

1. Problem Statement

Active matter systems, such as motility-induced phase separation (MIPS), violate detailed balance and time-reversal symmetry (TRS), leading to non-equilibrium phenomena distinct from equilibrium phase transitions. While equilibrium phase separation (Model B) follows the Lifshitz-Slyozov (LS) growth law ($L(t) \sim t^{1/3}$) for conserved order parameters, active systems exhibit complex behaviors.

• Active Model B (AMB) and its minimal extension, Active Model B+ (AMB+), are continuum field theories describing these systems.
• The Core Question: Recent Functional Renormalization Group (FRG) analyses suggest that in two dimensions ($d=2$), the active coupling parameters ($\lambda, \nu, \zeta$) are marginal operators. This implies they do not change the universality class but may introduce logarithmic corrections to the scaling laws.
• The Conflict: Previous numerical studies reported effective dynamic exponents ($z_c$) that varied continuously with activity, suggesting a breakdown of the standard $t^{1/3}$ law. The authors aim to resolve whether this is a new scaling law or a manifestation of logarithmic corrections predicted by FRG.

2. Methodology

The authors employ deterministic numerical simulations in two dimensions ($d=2$) to study AMB and AMB+.

• Models:
  • AMB: Includes a gradient-nonlinear term $\lambda(\nabla\phi)^2$ in the chemical potential.
  • AMB+: Adds a non-integrable active current term $\zeta(\nabla^2\phi)\nabla\phi$ and a $\nu \nabla^2(\phi^2)$ term.
  • The dynamics are governed by Eq. (1): $\partial_t\phi = -\nabla \cdot J$, where $J$ contains both equilibrium and active contributions.
• Simulation Details:
  • Noise: Simulations are performed in the noiseless limit ($D=0$) to isolate deterministic critical dynamics and coarsening.
  • Numerical Scheme: Pseudo-spectral method with periodic boundary conditions and Euler-time updates.
  • Parameters: System sizes up to $L=1024$; parameters varied include $\lambda \in [-2, 2]$, $\nu=0.5$, and $\zeta \in \{0, 1\}$.
• Analysis Techniques:
  • Critical Scaling: Finite-size scaling (FSS) analysis of the order parameter $m(t)$ to extract critical exponents ($\alpha, \beta, z, \nu$).
  • Phase Diagram: Analytical derivation of binodal densities using a generalized equal-area construction (pseudodensity/pseudopotential method) and numerical validation.
  • Coarsening: Calculation of the characteristic domain size $L(t)$ via the structure factor $S(k,t)$ to test growth laws.

3. Key Contributions

1. Verification of Mean-Field Criticality: The authors confirm that at the critical point ($r_c=0$), both AMB and AMB+ exhibit mean-field critical exponents identical to equilibrium Model B ($\alpha=1/4, z=4, \beta=1/2$), despite the presence of non-equilibrium currents.
2. Analytical Phase Diagram for AMB+: They derive exact analytical expressions for the binodal densities ($\phi_\pm$) in AMB+ using a generalized equal-area construction, validating these against numerical simulations.
3. Resolution of Growth Law Discrepancy: The paper demonstrates that the apparent continuous variation of dynamic exponents in previous studies is an artifact of logarithmic corrections to the standard $t^{1/3}$ law, rather than a fundamental change in the scaling exponent.
4. Role of the $\zeta$ Term: They identify that the $\zeta$ term in AMB+ acts to suppress the logarithmic corrections found in AMB and can lead to the arrest of coarsening (microphase separation) under specific conditions.

4. Key Results

A. Critical Behavior ($r_c = 0$)

• Exponents: Both AMB and AMB+ follow the scaling relation $m(t) \sim t^{-\alpha}$ with $\alpha = 1/4$ and dynamical exponent $z=4$.
• Universality: The transition belongs to the same universality class as equilibrium Model B (Ising-like static exponents, mean-field dynamic exponents).
• Finite-Size Effects: The critical point shifts in finite systems as $r_c(L) \sim -L^{-2}$, consistent with the linear dispersion relation.

B. Phase Diagram (Binodals)

• A generalized equal-area construction yields the coexistence densities $\phi_\pm$.
• The phase diagram in the $(r, \phi, \lambda)$ space shows that while $\lambda$ affects the width of the coexistence region, the critical point remains near $r=0$.
• Time-reversal symmetry breaking leads to an asymmetry in the binodal curve ($\phi_+ \neq -\phi_-$).

C. Supercritical Coarsening ($r < 0$)

The study of domain growth $L(t)$ reveals the following:

• Active Model B ($\zeta=0$):
  • The growth law deviates from pure $t^{1/3}$.
  • The data fits perfectly to the form:
    $$L(t) = L_0 t^{1/3} \left(1 + \frac{c}{\ln t}\right)$$
  • The coefficient $c$ is large and depends on $\lambda$. This explains why previous studies observed "effective" exponents varying with time/activity; the logarithmic term mimics a power law over intermediate time scales.
• Active Model B+ ($\zeta=1$):
  • Suppression of Corrections: For negative $\lambda$, the $\zeta$ term significantly reduces the magnitude of the logarithmic correction ($c$ is much smaller than in AMB), restoring behavior closer to the classic $t^{1/3}$ law.
  • Microphase Separation: For positive $\lambda$ (specifically when the effective coupling $b = 2\lambda + \nu - \zeta > 0$), the active current opposes Ostwald ripening. The domain growth saturates at a finite length scale, leading to a long-lived microphase-separated state (bubbly phase) rather than macroscopic phase separation.

5. Significance

• Theoretical Validation: The results provide strong numerical evidence supporting recent Functional Renormalization Group (FRG) predictions that active couplings are marginal in $d=2$. This confirms that activity introduces marginal corrections (logarithmic) rather than altering the fundamental universality class or the asymptotic growth exponent.
• Mechanism of Arrest: The paper elucidates how the interplay between different active terms ($\lambda, \nu, \zeta$) can either enhance logarithmic slowing (AMB) or cancel them out (AMB+ with specific parameters), or even reverse Ostwald ripening to stabilize finite-sized clusters.
• Resolution of Controversy: It resolves the ambiguity in the literature regarding "continuously varying exponents" in active coarsening, attributing them to pre-asymptotic logarithmic regimes rather than a new scaling law.
• Experimental Relevance: The findings suggest that experimental observations of MIPS in 2D systems (e.g., active colloids) may exhibit slow coarsening or arrested states depending on the specific nature of the active currents, which can be tuned by the effective coupling parameters.

In summary, the paper establishes that while active matter breaks time-reversal symmetry, its critical dynamics in 2D remain in the mean-field universality class of Model B, with the primary non-equilibrium effect being logarithmic corrections to coarsening that can be tuned or suppressed by specific active terms.