Chemically Active Emulsions: Dynamics & Control

• Chemically active emulsions are systems where reaction-driven interfaces sustain non-equilibrium droplet dynamics, enabling self-propulsion and oscillatory behavior.
• They integrate reaction-diffusion processes, Marangoni flows, and phase separation to arrest coarsening, stabilize droplet sizes, and induce synchronized oscillations.
• Their microscale interfacial chemistry and transport dynamics offer tunable control for applications in microfluidics, active matter, and soft-matter engineering.

Searching arXiv for recent and foundational papers on chemically active emulsions. Chemically active emulsions are emulsions in which droplets are maintained away from thermodynamic equilibrium by chemical reactions, interfacial solubilization, or related chemically driven processes that continuously generate fluxes of matter, interfacial stresses, or both. In these systems, droplets are not merely passive compartments: they can act as chemical oscillators, self-propelled particles, phoretic micropumps, or non-equilibrium phase-separated domains whose size, morphology, and interactions are set by the interplay of phase separation, reaction kinetics, diffusion, hydrodynamics, and interfacial constitution. Across the literature, the term encompasses at least three closely connected classes of phenomena: discrete reactive droplets such as Belousov–Zhabotinsky (BZ) water-in-oil oscillators (Thutupalli et al., 2012), solubilization-driven active droplets and micropumps whose interfaces convert chemical gradients into Marangoni flow (Ramesh et al., 2022), and continuum phase-separating mixtures with driven chemical reactions that exhibit arrested coarsening, microphases, and even spatiotemporal chaos (Datt et al., 19 Sep 2025).

1. Concept and definitions

A chemically active emulsion differs from a passive emulsion in that chemical reactions or chemically sustained interfacial processes prevent relaxation to equilibrium and continuously reshape droplet dynamics. In the continuum formulation of binary or ternary mixtures, this nonequilibrium character is introduced when components interconvert through reaction pathways that do not satisfy detailed balance with respect to the phase-separation free energy (Bauermann et al., 27 Jan 2025), or when a fuel is chemo-stated so that reaction fluxes remain driven by a fixed chemical potential difference (Ilker et al., 19 Apr 2026). In discrete-droplet settings, chemical activity may instead arise from oscillatory internal reactions, such as the BZ reaction inside water-in-oil droplets (Thutupalli et al., 2012), or from ongoing interfacial solubilization into micelles, which generates self-sustained compositional gradients and Marangoni stresses (Ramesh et al., 2022).

The term also includes active emulsions in which one component is an active liquid-crystalline material that continuously converts chemical energy into mechanical work. In the continuum framework for binary fluids, this activity appears as additional nonequilibrium terms in the composition, orientational-order, or momentum equations, and the resulting emulsions can display intrinsically driven flows, non-equilibrium steady states, and morphology-dependent active stresses (Cates et al., 2018). A related usage appears in active nematic double emulsions, where droplets containing active nematic shells and passive inclusions self-organize flow and topology patterns under confinement (Negro et al., 2024). A broader experimental review likewise defines active droplets as liquid droplets that spontaneously move in solution without externally applied fields, or droplets that may be immobile individually but acquire motility through interactions with neighboring droplets, with activity arising from self-generated chemical gradients at their interfaces (Birrer et al., 2022).

A useful unifying statement is that chemically active emulsions are systems in which chemical free energy is transduced into interfacial, diffusive, or hydrodynamic driving at the droplet scale. This includes chemically oscillating droplets, self-propelled Marangoni droplets, fuel-driven phase-separated droplets, active nematic droplets, and reactive emulsions stabilized in non-equilibrium microphase-separated states (Thutupalli et al., 2012, Ramesh et al., 2022, Bauermann et al., 27 Jan 2025, Cates et al., 2018).

2. Microscopic constitution and interfacial chemistry

The chemical constitution of droplets and interfaces is central because it determines propulsion mechanism, coupling strength, lifetime, trajectory, and collective interactions (Birrer et al., 2022). In discrete reactive emulsions, one prototypical system is the BZ water-in-oil emulsion: aqueous droplets contain sulfuric acid, sodium bromate, malonic acid, and ferroin, while the continuous oil phase is squalane stabilized by mono-olein (Thutupalli et al., 2012). Mono-olein is important both because it stabilizes droplets via monolayers and bilayers and because its carbon-carbon double bond reacts rapidly with bromine, trapping an inhibitory BZ species and thereby modulating inter-droplet coupling (Thutupalli et al., 2012).

In solubilization-driven active droplets, the constitution is instead organized around oil droplets in a supramicellar aqueous surfactant solution. A representative example uses CB15 droplets in 5 wt% TTAB, well above the critical micelle concentration $0.13\%$, in a Hele–Shaw cell of height $\approx 50\,\mu\mathrm{m}$ (Ramesh et al., 2022). Surfactant monomers and empty micelles coexist in the continuous phase, while the oil continuously solubilizes into micelles, producing filled micelles that act as slowly diffusing chemical products (Ramesh et al., 2022). In related active-droplet experiments, common chemistries include LC oils such as 5CB, 8CB, CB15, and isotropic haloalkanes or aromatic oils in TTAB, CTAB, CTAC, Triton X-100, or mono-olein media, often at $10$–$200\times$ CMC (Birrer et al., 2022, Hokmabad et al., 2020, Khan et al., 2024).

Microscopic particle-based models make the constitution even more explicit. In Brownian-dynamics simulations of active droplets, particles exist in a droplet-forming state $B$, a non-droplet-forming state $A$, and sometimes an inert crowder state $C$ (Berthin et al., 2024). Only $B$–$B$ pairs attract through a truncated-shifted Lennard–Jones potential, while all other pairs are purely repulsive and modeled by Weeks–Chandler–Andersen interactions (Berthin et al., 2024). Chemically active emulsions then emerge because reactions $A \leftrightarrows B$ are split into passive and active pathways, with the active pathway driven by a chemical potential difference $\approx 50\,\mu\mathrm{m}$0 and weighted by local density so that conversion tends to create $\approx 50\,\mu\mathrm{m}$1 in dilute regions and reduce $\approx 50\,\mu\mathrm{m}$2 in dense regions (Berthin et al., 2024). A closely related microscopic model for biomolecular-condensate-like droplets adds client species $\approx 50\,\mu\mathrm{m}$3 and shows how the same chemical drive controls encounter kinetics by modulating residence times and interfacial fluxes (Fries et al., 9 May 2025).

At the continuum level, ternary active-emulsion theories similarly identify a conserved combination of reacting species and a non-conserved reaction extent. In one dynamic theory, a ternary mixture of $\approx 50\,\mu\mathrm{m}$4, $\approx 50\,\mu\mathrm{m}$5, and solvent $\approx 50\,\mu\mathrm{m}$6 is recast in terms of $\approx 50\,\mu\mathrm{m}$7, which is conserved, and $\approx 50\,\mu\mathrm{m}$8, which is non-conserved and directly driven by reactions (Bauermann et al., 27 Jan 2025). In a hydrodynamic theory derived from a ternary solution with fuel-driven reactions, the conserved field is $\approx 50\,\mu\mathrm{m}$9 and the reaction extent is $10$0, with activity entering through a passive pathway and an active pathway whose rates differ by a fuel chemical potential $10$1 (Ilker et al., 19 Apr 2026). These formulations show that chemically active emulsions are not defined by a single chemistry, but by a common constitutional architecture: at least one phase-separating degree of freedom, at least one reactive pathway, and interfaces whose structure or permeability couples chemistry back to phase behavior and flow.

3. Physical mechanisms from chemistry to motion and pattern formation

The dominant transport mechanism in many experimental active droplets is Marangoni driving. Interfacial tension gradients $10$2 produce tangential stresses, which in low-Reynolds-number conditions generate surface flows and bulk recirculation (Birrer et al., 2022, Khan et al., 2024). In pinned active droplets acting as micropumps, the chemical concentration field $10$3 obeys a reaction–advection–diffusion equation,

$10$4

and the coupling between chemical gradients and interfacial flow is described quantitatively by a Brinkman squirmer model under confinement (Ramesh et al., 2022). In this system, increasing droplet radius produces a sequence of flow-mode transitions: posterior vortices migrate to the anterior, a bistability appears between dipolar and quadrupolar modes, and eventually multipolar modes emerge (Ramesh et al., 2022).

Reactive self-propulsion can also be mediated by internal oscillatory chemistry. In BZ droplets, the autocatalytic excitatory loop and inhibitory bromine-production loop generate relaxation-type oscillations that are visible as red-to-blue ferroin transitions (Thutupalli et al., 2012). The same reaction modifies interfacial chemistry through bromine transport and bromination of mono-olein, and these chemical changes can synchronize with droplet motion. Large droplets of about $10$5 show internal BZ waves whose propagation direction correlates with droplet movement, while smaller droplets of about $10$6 show speed oscillations synchronized with global color oscillations (Thutupalli et al., 2012).

In chemotactic active emulsions, self-propelling droplets leave chemical trails that act as chemical footprints. For CB15 droplets in TTAB, oil-filled micelles with radius $10$7 diffuse with $10$8, and the trail concentration profile is approximately Gaussian (Hokmabad et al., 2020). Droplet dynamics can then be described by a Chemically Active Polar Particle model,

$10$9

$200\times$0

in which positive couplings $200\times$1 and $200\times$2 encode chemorepulsive drift and reorientation away from trails (Hokmabad et al., 2020). Experimentally extracted median couplings are $200\times$3 and $200\times$4, and the competition between trail age and collision angle controls whether droplets cross or reflect from trails (Hokmabad et al., 2020).

In continuum active-emulsion theories, chemistry instead alters phase separation directly. A minimal scalar model with reactions and hydrodynamics augments Model H by a reaction term and interfacial stresses: $200\times$5

$200\times$6

with $200\times$7 (Datt et al., 19 Sep 2025). Here, phase separation, reactions, and viscous hydrodynamics form a closed feedback loop that can generate steady patterns, arrested coarsening, or spatiotemporal chaos, even in the Stokes regime and without orientational order (Datt et al., 19 Sep 2025). A more systematic hydrodynamic reduction of active ternary mixtures yields an effective conserved-field theory containing Active-Model-B$200\times$8-like gradient terms and additional higher-order terms, showing that chemically active emulsions can generically realize microphases, bubbly phase separation, and dynamic filament phases (Ilker et al., 19 Apr 2026).

4. Theoretical frameworks and mathematical descriptions

Several complementary theoretical frameworks are now used for chemically active emulsions. At the continuum passive baseline lies Model H for binary fluids, in which the composition field $200\times$9 obeys a convective Cahn–Hilliard equation and the fluid velocity obeys incompressible Navier–Stokes with thermodynamic stress $B$0 (Cates et al., 2018). This passive framework is then modified in active systems by non-equilibrium chemical potentials, active deviatoric stresses, or reaction terms that break detailed balance (Cates et al., 2018, Datt et al., 19 Sep 2025, Ilker et al., 19 Apr 2026).

One branch of theory studies chemically active droplets as non-equilibrium phase-separated domains with reactions $B$1. In a ternary mixture, the growth law of a single droplet is obtained in a quasistationary sharp-interface limit by solving reaction–diffusion equations inside and outside the droplet, enforcing local phase equilibrium and local material conservation at the interface, and deriving

$B$2

where $B$3 is the far-field conserved density (Bauermann et al., 27 Jan 2025). For an ensemble of droplets with size distribution $B$4, the kinetics reduce to a continuity equation in radius space,

$B$5

closed by a global conservation law for the conserved density (Bauermann et al., 27 Jan 2025). This theory predicts arrested growth or arrested ripening, stationary monodisperse states, and a late-time scaling collapse of rescaled size distributions (Bauermann et al., 27 Jan 2025).

A closely related minimal ternary model uncovers a critical transition between intensive and extensive active droplets. Using the conserved field $B$6, the non-conserved reaction extent $B$7, and a compositional angle $B$8 plus an activity parameter $B$9, the authors show that the conserved quantity $A$0 controls whether a single droplet reaches a finite stationary size or scales with system size (Bauermann et al., 2024). The critical conserved quantity is

$A$1

and near this threshold the stationary radius diverges as

$A$2

marking a genuine critical transition between intensive and extensive droplets (Bauermann et al., 2024). This distinction is important because intensive droplets in an emulsion interact only weakly and set a monodisperse size, whereas extensive droplets require mutual interactions to arrest coarsening (Bauermann et al., 2024).

Microscopic particle-based descriptions complement these continuum theories by explicitly representing droplets as clusters of particles. In Brownian-dynamics simulations, particle positions obey overdamped Langevin dynamics,

$A$3

while reactions $A$4 occur with passive rates obeying detailed balance and active rates driven by $A$5 (Berthin et al., 2024). This framework reveals how non-equilibrium driving controls droplet size distributions, shape anisotropy $A$6, lifetime distributions, and phase portraits $A$7 with both an unstable critical volume $A$8 and a stable selected volume $A$9 (Berthin et al., 2024). An extension of the same microscopic philosophy to reactive encounter kinetics introduces a compartmentalized Smoluchowski theory in which $C$0, $C$1, and the residence time $C$2 determine whether active droplets accelerate or decelerate molecular encounters (Fries et al., 9 May 2025).

Finally, in active-nematic emulsions and active liquid-crystalline droplets, the coupled fields are phase fields $C$3, the velocity $C$4, and the tensor order parameter $C$5. Activity enters through the stress

$C$6

or, in polar emulsions, through analogous polar active stresses and self-advection (Negro et al., 2024, Carenza et al., 2019, Cates et al., 2018). Although these formulations do not always model explicit chemical reactions, they are used to describe chemically powered active gels and therefore belong to the broader theoretical landscape of chemically active emulsions (Cates et al., 2018).

5. Characteristic behaviors and morphologies

Chemically active emulsions exhibit a richer phenomenology than passive emulsions because reactions and interfacial activity can both oppose and reshape phase separation. One recurrent outcome is arrested coarsening and monodispersity. In microscopic simulations, the average droplet volume grows approximately as $C$7 at equilibrium $C$8, consistent with Ostwald ripening, whereas for $C$9 it rapidly saturates to a finite value and the stationary distribution $B$0 narrows as $B$1 increases (Berthin et al., 2024). Continuum droplet-size theory likewise predicts stationary monodisperse states $B$2, with exponential relaxation to the selected radius and a universal scaling collapse of late-time distributions when plotted in terms of $B$3 (Bauermann et al., 27 Jan 2025).

A second distinctive behavior is reverse or suppressed Ostwald ripening. In the encounter-kinetics model, activity-driven conversion inside dense droplets and passive regeneration outside create a steady exit flux $B$4 of non-condensing species, and increasing $B$5 decreases the residence time $B$6 of particles in droplets (Fries et al., 9 May 2025). In the hydrodynamic active-emulsion theory, negative effective interfacial energy $B$7 destabilizes macroscopic coarsening and instead stabilizes microphases or bubbly phase separation (Ilker et al., 19 Apr 2026). In the critical intensive–extensive framework, extensive droplets still show modified ripening, but it arrests because droplet interactions and conservation laws eventually balance growth (Bauermann et al., 2024).

A third family of behaviors concerns oscillations and synchronization. BZ droplets coupled by surfactant bilayers can exhibit target waves, traveling waves, spiral waves, and strict anti-phase oscillations depending on topology and malonic-acid concentration (Thutupalli et al., 2012). In large hexagonal lattices, waves propagate across the network; in imperfect lattices, defects refract waves and generate spiral patterns; and in one-dimensional chains with $B$8 malonic acid, inhibitory-dominated coupling produces strict anti-phase behavior (Thutupalli et al., 2012). These behaviors establish chemically active droplets as discrete reaction–diffusion oscillators organized by network geometry.

Further nonequilibrium morphologies include liquid shells, bubbling states, and filament phases. In a chemically active binary mixture with driven reactions, a central spinodal instability inside a droplet can create a dilute core surrounded by a dense shell, with stable shells coexisting in active emulsions and a large-system shell width

$B$9

in the asymptotic limit (Bauermann et al., 2023). In another continuum model, noisy active phase separation yields bubbly states in which a macro-phase coexists with many small bubbles, and stronger driving produces a dynamic active filament phase whose onset coincides with a kink-like singularity in the entropy-production rate (Ilker et al., 19 Apr 2026).

Active droplets with orientational order introduce topological morphologies and dynamic defect networks. A single active nematic double emulsion with one passive core undergoes a transition from translational to rotational and then meandering motion as the active Ericksen number $B$0 increases, while remaining defect-free and topologically trivial (Negro et al., 2024). By contrast, a two-core emulsion is topologically non-trivial and necessarily hosts disclination loops. With increasing activity, a single charged loop powers a rotor-like state; at higher activity the loop stretches, writhes, and recombines to form an active living polymer (Negro et al., 2024). This suggests that chemically active emulsions can serve as topological matter under controlled confinement.

Finally, strong hydrodynamic coupling can drive spatiotemporal chaos in chemically active mixtures without inertia or orientational order. In the Stokes regime, a reactive binary fluid described by active Cahn–Hilliard–Stokes equations shows positive maximal Lyapunov exponents, persistent deformation and re-formation of target and spiral patterns, and amplitude equations identical to those of Rayleigh–Bénard convection with mean flow (Datt et al., 19 Sep 2025). This establishes interfacial-stress-driven chaos as a generic possibility in scalar chemically active emulsions (Datt et al., 19 Sep 2025).

6. Collective behavior, control, and applications

Collective behavior in chemically active emulsions arises from several distinct couplings: diffusive exchange through interfaces, reaction-mediated chemical fields, hydrodynamic flow fields, and geometric or topological constraints. In BZ emulsion networks, bilayers are the conduits for direct diffusive exchange, and network topology determines whether the collective state is a pacemaker-driven target pattern, a traveling wave, a spiral, or an anti-phase chain (Thutupalli et al., 2012). In chemotactic droplets, each swimmer continuously modifies its environment by leaving trails of filled micelles; at sufficient density, overlapping repulsive trails transiently cage droplets, and the mean-squared displacement develops a plateau despite extremely low droplet volume fraction (Hokmabad et al., 2020). In confined 2D cells, caging appears for droplet number densities $B$1, while in 3D it occurs already near $B$2 and $B$3 because chemical fields, not steric collisions, form the barriers (Hokmabad et al., 2020).

Control of collective states can be achieved chemically, geometrically, or mechanically. Chemical-drive control is explicit in microscopic models, where increasing $B$4 shifts the droplet size distribution and narrows polydispersity (Berthin et al., 2024), and in the encounter-kinetics theory, where an optimal reaction regime appears when the residence time in droplets matches the encounter time inside droplets,

$B$5

so that droplets confine particles long enough to accelerate encounters but not so long that they merely trap them in different compartments (Fries et al., 9 May 2025). This provides a concrete design principle for biomolecular condensates as tunable regulators of intracellular reaction kinetics (Fries et al., 9 May 2025).

Confinement is another major control axis. Pinned active droplets in Hele–Shaw cells simplify hydrodynamic analysis and act as micropumps, while varying droplet radius tunes the Péclet number and therefore selects dipolar, quadrupolar, or multipolar modes (Ramesh et al., 2022). Patterned activity in active emulsions of polar gels localizes the scale at which chemical energy is converted into mechanical energy, and the resulting system shows multiscale dynamics without an inertial cascade, drag reduction, and a spatially modulated effective slip length under pressure-driven flow (Carenza et al., 2019). This suggests that chemical or external “doping” of activity can be used to program flow topology and rheology (Carenza et al., 2019).

A particularly direct external control strategy is to embed anisotropic solid boundaries inside active droplets. Ferromagnetic FePt clusters inserted into self-propelling droplets couple bidirectionally to the internal Marangoni flow: the flow orients the cluster, and a magnetic field realigns the cluster, which in turn reorients the flow (Khan et al., 2024). Measured cluster reorientation occurs in about $B$6, while droplet steering takes about $B$7, allowing trajectories to be programmed by weak fields (Khan et al., 2024). Rotating magnetic fields continuously perturb the flow and produce chiral curling motion even in achiral emulsions; increasing the rotation frequency suppresses net translation and can split or disperse collective states (Khan et al., 2024). Because this control works across multiple chemistries, it suggests a route to chemically active micromachines whose chemotactic interactions are preserved but whose flow fields are externally steerable (Khan et al., 2024).

Potential applications are correspondingly broad. Chemically active emulsions have been proposed as model systems for biological synchronization, reaction–diffusion pattern formation, and active matter (Thutupalli et al., 2012), as tunable micropumps and mixers in confined microfluidics (Ramesh et al., 2022), as synthetic analogues of biomolecular condensates with controllable size, lifetime, and function (Berthin et al., 2024, Fries et al., 9 May 2025), and as platforms for programmable topological flows and active living polymers (Negro et al., 2024). A broader review also emphasizes Janus droplets, Pickering emulsions, multiple emulsions, and boundary-shaped trajectories as a general design space for active-droplet research (Birrer et al., 2022).

7. Relation to passive emulsions, controversies, and open problems

A persistent misconception is that active emulsions are merely passive emulsions with faster dynamics. The literature instead shows several qualitative departures from passive behavior. Passive Ostwald ripening follows classical growth laws and proceeds toward a single equilibrium phase-separated state, whereas chemically active systems can arrest coarsening, reverse ripening, or stabilize stationary droplet sizes through reaction–diffusion balances (Bauermann et al., 27 Jan 2025, Berthin et al., 2024, Bauermann et al., 2024). Passive binary fluids are governed by free-energy minimization, but active ternary mixtures require non-variational terms in the effective flux, and their entropy production remains strictly positive because fuel-driven reactions maintain them out of equilibrium (Ilker et al., 19 Apr 2026). Likewise, active scalar mixtures can display spatiotemporal chaos in the Stokes regime without inertia or orientational order, distinguishing them both from classical turbulence and from active nematic turbulence (Datt et al., 19 Sep 2025).

A second misconception is that self-propelled droplets are controlled only by hydrodynamics. Several studies show instead that chemical constitutions and interfacial kinetics are indispensable. In solubilizing droplets, the competition between fast monomer diffusion and slow filled-micelle diffusion sets multistability, saturation, and eventual arrest (Ramesh et al., 2022). In chemotactic emulsions, collective caging depends on trail diffusion and chemotactic torque, not merely on hydrodynamic interactions (Hokmabad et al., 2020). In microscopic condensate models, the same geometry can either accelerate or decelerate molecular encounters depending on whether residence times match encounter times (Fries et al., 9 May 2025). This suggests that purely hydrodynamic descriptions are insufficient whenever interfacial chemistry is state-dependent or composition-sensitive.

Several open problems recur across the literature. The permeability of bilayer membranes between BZ droplets is crucial but was not measured directly in the oscillation-network experiments (Thutupalli et al., 2012). Solubilization-driven droplets display rich higher-order flows whose full nonlinear selection and switching remain only partially understood, particularly at high Péclet number or long times (Ramesh et al., 2022). Microscopic stochastic models establish size, shape, and lifetime fluctuations, but they neglect hydrodynamics and are limited in accessible time scales and system sizes (Berthin et al., 2024). Continuum active-emulsion theories identify effective Active-Model-B$B$8-like structures, but multicomponent reactions, viscoelasticity, and realistic biochemistry remain largely unresolved (Ilker et al., 19 Apr 2026, Datt et al., 19 Sep 2025). The critical intensive–extensive transition suggests a route to repeated droplet division, but experimental observation of successive division events in chemically active emulsions remains a future goal (Bauermann et al., 2024). Active shells and living disclination polymers demonstrate the role of topology, yet extensions to more cores, chirality, and mixed contractile/extensile activity remain open (Negro et al., 2024).

A plausible implication is that chemically active emulsions are best viewed not as a single material class but as a hierarchy of non-equilibrium soft-matter architectures. At one end are discrete droplets whose interfaces transduce chemistry into motion; at the other are reactive phase-separating mixtures whose continuum fields realize active phase separation, pattern selection, and hydrodynamic chaos. What connects them is not a unique composition, but the coexistence of interfaces, conserved material transport, and chemically sustained departures from equilibrium (Thutupalli et al., 2012, Ramesh et al., 2022, Datt et al., 19 Sep 2025, Ilker et al., 19 Apr 2026).