Active Cahn-Hilliard Theory

• Active Cahn-Hilliard theory is a continuum framework that extends phase separation models by incorporating active currents, non-reciprocal couplings, and reaction terms.
• It derives systematic coarse-grained equations from active Brownian particle models to predict phase boundaries, interfacial tensions, and deviations in domain coarsening.
• The theory underpins practical applications in soft active matter, colloidal assemblies, and biological morphogenesis by capturing motility-induced phase separation and pattern arrest.

Active Cahn-Hilliard Theory provides a continuum framework for describing phase separation, interfacial phenomena, and microstructural dynamics in active matter systems—ensembles of driven particles or fields that consume energy and thereby violate detailed balance at the microscopic or mesoscopic scale. The theory generalizes the passive Cahn-Hilliard equation by introducing active currents, non-reciprocal couplings, reaction terms, and gradient corrections that encode non-equilibrium behaviors such as motility-induced phase separation, pattern arrest, and self-propulsion. Its structure is supported by systematic derivations from microscopic active Brownian-particle models and is further extended to multicomponent, multi-phase, and hydrodynamically-coupled systems. The theory produces quantitative predictions for phase boundaries, coarsening laws, interfacial tensions, and fluctuation-dissipation violations, and it underpins applications from soft active materials and colloidal assemblies to biological morphogenesis and protocell models.

1. Microscopic Foundations and Coarse-Grained Derivation

Active Cahn-Hilliard (CH) theory is often rooted in the dynamics of active Brownian particles (ABPs) with repulsive interactions and self-propulsion. The microscopic equations for particle $k$ are

$$\dot{\mathbf{r}}_k = -\nabla_k U + v_0 \mathbf{e}_k + \boldsymbol{\eta}_k, \qquad \dot{\theta}_k = \sqrt{2r} \xi_k,$$

with $U$ a short-range repulsive potential, $v_0$ the propulsion speed, and $\theta_k$ the particle orientation, subject to translational and rotational diffusion ($D$, $r$) (Speck et al., 2013, Speck et al., 2015).

Through systematic coarse-graining—defining density $\rho(\mathbf{x},t)$ and polarization $\mathbf{p}(\mathbf{x},t)$ fields, and expanding under the assumption of slowly varying spatial fields—the equation hierarchy closes at leading order: $$\partial_t \rho = -\nabla \cdot [v(\rho)\mathbf{p}] + D \nabla^2 \rho, \qquad \partial_t \mathbf{p} = -\tfrac{1}{2}\nabla [v(\rho)\rho] + D \nabla^2 \mathbf{p} - r \mathbf{p},$$ with effective $v(\rho) = v_0 - \zeta \rho$, where $\zeta$ parameterizes "self-trapping" via binary collisions.

Single-field reduction via multiple-scale expansion yields the active CH equation: $$\partial_t \rho = \nabla \cdot [ M(\rho)\nabla \mu ], \quad \mu(\rho, \nabla \rho) = A(\rho - \bar{\rho}) - B(\rho - \bar{\rho})^2 - \kappa \nabla^2 \rho,$$ where $A, B, \kappa$ emerge from the microscopic parameters (Speck et al., 2013, Speck et al., 2015).

2. Mathematical Structure and Gradient Corrections

While the passive Cahn-Hilliard equation arises from a free-energy functional $F$, active CH equations admit additional, non-equilibrium terms: $$\partial_t\rho = \nabla^2 \mu + \nabla \cdot \mathbf{J}_\text{active},$$ where, for example,

$$\mu = f'(\rho) - K(\rho)\nabla^2\rho + \lambda(\rho)|\nabla\rho|^2 + \cdots,$$

and the current includes terms up to $\mathcal{O}(\nabla^4)$ (Bureković et al., 23 Jan 2026). The generalized form includes (editor’s categorization):

Coefficient	Gradient Order	Physical Role
$K(\rho)$	2	Interfacial stiffness
$\lambda(\rho)$	4	Local slope coupling
$\zeta(\rho)$	4	Coupling of curvature and slope
$\nu(\rho)$	4	Cubic active pumping

These density-dependent functions reflect active corrections not captured by simple $\phi^4$ or constant-coefficient models (e.g., Active Model B/B+), requiring a multiple-scale coarse-graining to derive systematically from the microscopic dynamics (Rapp et al., 2019, Bureković et al., 23 Jan 2026).

3. Phase Behavior, Instabilities, and Coarsening Arrest

Active CH theory predicts a spectrum of phase-separation scenarios. The spinodal is shifted due to activity: instability occurs when $f''(0) = \alpha_1 = 0$, yielding a spinodal line $v_0 = v_s(\bar{\rho})$ (Speck et al., 2013, Speck et al., 2015). The cubic nonlinearity in the bulk term, $f(c) = \frac{1}{2}\alpha_1c^2 - \frac{1}{3}g c^3$, determines the nature of the bifurcation. For densities below a critical value $\bar\rho_* \approx 0.29$, phase separation is discontinuous with nucleation and hysteresis; above it, the transition is spinodal-like and continuous.

Active gradient corrections ($\mathcal{O}(\nabla^4)$) can completely arrest coarsening by stabilizing finite-wavelength patterns, as in Turing-type regimes (finite-$k$ stationary instabilities) and/or induce oscillatory (Hopf) instabilities, allowing persistent micellar or traveling-band morphologies (Frohoff-Hülsmann et al., 2020, Saha et al., 2020, Bureković et al., 23 Jan 2026). Coarsening exponents deviate from the passive $1/3$ law when activity-dominated physics dominates domain selection.

In the presence of non-reciprocal (antisymmetric) couplings—for example, in active mixtures or non-reciprocal Cahn-Hilliard (NRCH) models—the system exhibits traveling density waves, global polar order, and the possibility of microphase-separated active smectics (Saha et al., 2020, Saha, 2024). Detailed balance violation is reflected in the absence of a global Lyapunov functional and in entropy production at the field-theoretic level (Johnsrud et al., 4 Feb 2025).

4. Hydrodynamic and Multiphase Extensions

The CH-Navier–Stokes (CHNS) or Cahn-Hilliard–Darcy frameworks introduce hydrodynamic coupling via passively and actively generated stresses. The hydrodynamically-extended active CH equation couples to incompressible flow as

$$\partial_t \phi + \mathbf{v} \cdot \nabla \phi = M \nabla^2 \mu + \nabla \cdot (\chi(\phi) \nabla \phi),$$

$$\rho (\partial_t \mathbf{v} + \mathbf{v} \cdot \nabla \mathbf{v}) = -\nabla p + \eta \nabla^2 \mathbf{v} + \phi \nabla \mu + \nabla \cdot \Sigma^A,$$

where $\Sigma^A$ is an active stress tensor (e.g., contractile or extensile, scaling as $$\propto -\zeta[\partial_i \phi \partial_j \phi - \frac{1}{2}\delta_{ij}|\nabla\phi|^2]$$) (Padhan et al., 12 Mar 2025). This yields active turbulence, arrested coarsening, and self-propelled droplet regimes, manifesting distinct regime diagrams in terms of Péclet, active, and capillary numbers.

In multiphase and multicomponent extensions (including chemotaxis, active transport, and reversible reactions), active fluxes—e.g., $\nabla \cdot (\chi_\varphi \sigma \nabla \varphi)$—and nonlocal or Oono-type terms mediate additional patterning, interface motion, and phase equilibrium (He et al., 2022, Garcke et al., 2015, Garcke et al., 12 Mar 2025).

5. Binodals, Interfacial Tensions, and Nonequilibrium Thermodynamics

In general, binodal (coexistence) densities in active CH theory are not determined solely by a common-tangent construction on a bulk $f(\rho)$ due to gradient and active corrections. The appropriate coexistence criteria demand the matching of two "pseudopressures" derived from the steady-state condition ($J=0$) and the full gradient structure of the current. Three distinct interfacial tensions can be defined:

1. $\sigma_V$ (vapor-side Laplace tension)
2. $\sigma_L$ (liquid-side Laplace tension)
3. $\sigma_\text{cw}$ (capillary-wave tension) with explicit integrals over the interface profile and coefficient functions (Bureković et al., 23 Jan 2026). These yield different curvature-dependences for Ostwald ripening and capillary wave relaxation, beyond the predictions of active Model B/B+.

In NRCH models, an "active pressure" is constructed—a spatially constant combination of intensive variables and nonreciprocal corrections—which generalizes mechanical balance to non-equilibrium mixtures, including cases where interfacial tension and Laplace pressure admit active (i.e., non-reciprocal or curvature-independent) components (Saha, 2024).

At the stochastic field-theoretic level, fluctuation-dissipation violations are tied directly to entropy production, and the structure of two-point correlation and response functions in the non-reciprocal or active sector encodes genuinely nonequilibrium information (Johnsrud et al., 4 Feb 2025).

6. Applications and Physical Realizations

Active Cahn-Hilliard frameworks quantitatively reproduce experimental and simulation results for motility-induced phase separation, active colloidal motion (including artificial Janus particles and isotropic chemical colloids), tunable self-organization in active mixtures, protocell growth/division models, and tumor growth with active membrane transport. The theory is validated by direct comparison to particle-resolved simulations and, in contexts such as chemotactic transport and growth-division cycles, matches phase-field and sharp-interface asymptotics (Speck et al., 2013, Dobson et al., 24 Feb 2025, Garcke et al., 12 Mar 2025, He et al., 2022, Garcke et al., 2015).

Key predictions, such as hysteresis lines in $(\phi, v_0)$-space, critical exponents for cluster coarsening, and the emergence of stable microphases or traveling bands in active mixtures, have received both computational and experimental support. Adjusting control parameters (e.g., fuel or light intensity for active colloids, parameters in the effective free energy, or coupling strengths in chemotactic or non-reciprocal fluxes) enables the realization of predicted transitions and patterning regimes.

7. Limitations, Generalizations, and Perspectives

Current active Cahn-Hilliard theories are limited by closure relations arising from systematic coarse-graining (e.g., neglect of higher orientational moments), assumptions of local interactions, and the difficulty of capturing hydrodynamic long-range effects or inertia. Mapping to an effective passive free energy is generally only possible under restrictions (e.g., absence of cross-gradient couplings, specific symmetry conditions in multicomponent systems), beyond which active Laplace pressures and nontrivial mechanical imbalances arise (Saha, 2024).

Recent work demonstrates the necessity of including all gradient-order terms and density dependence in the coefficient functions to accurately describe phase behavior away from weak-separation limits (Bureković et al., 23 Jan 2026). Analytical understanding of fluctuation-dissipation relations and entropy production in these models is advancing, with rigorous field-theoretic and diagrammatic frameworks now available (Johnsrud et al., 4 Feb 2025).

The active Cahn-Hilliard framework continues to be the minimal and quantitative deterministic description for scalar active matter, with ongoing developments in its extension to more complex mixtures, hydrodynamic coupling, and rigorous statistical mechanical underpinnings.