Spatiotemporal organization of chemical oscillators via phase separation

Artemy Kolchinsky; Giacomo Bartolucci; Jonathan Bauermann

REVIEW 4 major objections 5 minor 58 references

Spatiotemporal organization of chemical oscillators via phase separation

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The sign of an interaction-energy difference decides whether phase separation damps or amplifies chemical oscillations.

desk verdict Useful and well-written extension of the phase-equilibrium dynamics framework to oscillators, but the headline control rule is read off a single numerical scan, not derived. read the letter →

arxiv 2507.16030 v1 pith:UAU34KAZ submitted 2025-07-21 cond-mat.soft nlin.PSphysics.bio-ph

Jonathan Bauermann , Giacomo Bartolucci , Artemy Kolchinsky This is my paper · ORCID

classification cond-mat.softnlin.PSphysics.bio-ph

keywords phaseseparationchemicaloscillationsrock-paper-scissorskineticsFlory-Hugginsfreeenergyreaction-diffusiondynamicsequilibriummetastabilitytravelingwaves

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper studies how phase separation changes the behavior of a simple three-species chemical oscillator. It shows that when diffusion is fast enough to keep the system in global phase equilibrium, the reaction dynamics reduces to a low-dimensional equation in which the two coexisting phases react separately. The authors find that the oscillator's frequency and amplitude can be tuned by the interaction energies between the oscillating species and the phase-forming species: lower interaction energies speed up the oscillation, while the sign of the difference $\chi_{AB} - \chi_{AC}$ decides whether perturbations grow, stay neutral, or decay. The reduced dynamics is validated against a full spatial model, and when reactions become comparably fast to diffusion the model predicts mesoscopic traveling waves in which droplets form and dissolve at wave fronts.

What carries the argument

The key machinery is the phase-equilibrium reaction dynamics, Eq. (7), which describes how the average concentrations evolve when the system is inside the binodal: $d\rho/dt = v\,R(\rho_I) + (1-v)\,R(\rho_{II})$, where the phase compositions $\rho_I, \rho_{II}$ and the volume fraction $v$ are obtained at each instant by minimizing the average Flory-Huggins free energy under the constraint that the total concentration is fixed. This turns the spatial reaction-diffusion problem into a two-dimensional ODE, making it possible to compute the fixed point and its stability in closed form. Stability is determined by the $2\times2$ matrix $M$ of Eq. (9), whose eigenvalues $\lambda_{Re} \pm i\lambda_{Im}$ give the predicted growth/damping rate and oscillation frequency. The paper supplements this ODE with a metastability variable that delays phase formation from the binodal to the spinodal line, and it treats the fast-reaction regime via the full spatial reaction-diffusion equation (Eq. (4)).

What would settle it

A direct test would be to simulate the full spatial reaction-diffusion model (Eq. (4)) for a range of interaction energies and measure the growth or decay rate of small perturbations around the fixed point; if the sign of the relaxation rate does not follow the sign of $\chi_{AC} - \chi_{AB}$, or the predicted neutral stability at $\chi_{AB} = \chi_{AC}$ is absent, the coarse-grained prediction is falsified. An experimental system of phase-separating droplets containing the RPS oscillator could also be compared at $\chi_{AB} > \chi_{AC}$ and $\chi_{AB} < \chi_{AC}$.

Watch

Extended reading notes

Core claim

The central claim is that phase separation provides a general mechanism for controlling the temporal organization of chemical oscillators. For a rock-paper-scissors oscillator ($A+B \to 2A$, $B+C \to 2B$, $C+A \to 2C$) coupled to Flory-Huggins phase separation in which species $A$ forms $A$-rich and $A$-poor phases, the authors find that the real part of the linear relaxation rate at the fixed point has the sign of $\chi_{AC} - \chi_{AB}$: perturbations grow when $\chi_{AB} < \chi_{AC}$, are marginal when $\chi_{AB} = \chi_{AC}$, and are damped when $\chi_{AB} > \chi_{AC}$. The imaginary part of the rate shows that lower $\chi_{AB}$ and $\chi_{AC}$ give faster oscillations. The mechanism is localization: when $B$ is enriched in the $A$-rich phase ($\chi_{AB} < 0$) and $C$ in the $A$-poor phase ($\chi_{AC} > 0$), the production and decay of $A$ are differentially accelerated, either amplifying or damping the oscillation. Including metastability of the homogeneous state changes when phases form and can turn a neutral oscillation into a damped one, and relaxing the timescale separation between reactions and diffusion produces traveling waves of phase equilibria.

Load-bearing premise

All analytical results assume that diffusion is fast enough to keep the system in instantaneous global phase equilibrium, so the dynamics is given by Eq. (7) the moment the average composition enters the binodal; the paper shows this assumption breaks down when reactions become comparably fast, yielding traveling waves instead.

Editorial extensions

If this is right

If the prediction holds, any oscillatory reaction network coupled to a phase-separating species can be sped up, slowed down, amplified, or damped simply by tuning the interaction energies of the oscillating components with the condensate.
The marginal case $\chi_{AB} = \chi_{AC}$ gives a line of neutrally stable oscillations, which could serve as a sensitive switch where a small change in interaction energy flips the system between growth and decay.
Phase separation can buffer oscillation amplitude: it can suppress large-amplitude oscillations entering the binodal, driving the system to a limit cycle that never crosses the binodal.
Metastability introduces hysteresis in the phase-equilibrium dynamics, so the system's history (whether phases already exist) determines whether oscillations are amplified or damped for identical parameters.
When diffusion and reactions act on similar timescales, the model predicts propagating fronts in which droplets are nucleated, grow, and dissolve, a mesoscopic pattern relevant to biological condensates in oscillatory circuits.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The central control parameter is the difference $\chi_{AB} - \chi_{AC}$ of two interaction energies, suggesting a broader design principle: the differential affinity of two competing species for a droplet phase, not their absolute affinities, sets the oscillatory fate. One could test this by engineering synthetic droplets with opposite affinities for the two oscillating species.
The quasi-static phase-equilibrium approximation, validated in the fast-diffusion regime, fails when reactions are fast; the appearance of traveling waves there indicates that spatiotemporal order can still arise, but through a distinct mechanism, so the boundary between the two regimes deserves quantitative mapping.
If extended to oscillatory gene circuits or circadian clocks that involve phase-separated compartments, the model predicts that the relative interaction energies of clock components with condensate scaffolds set both the period and the amplitude of the clock.
The marginal line $\chi_{AB} = \chi_{AC}$ is predicted to be stable only under symmetric noise in interaction energies; any asymmetry would push the system off the line, a signature that could be looked for experimentally.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper does something genuinely useful: it takes the group's earlier phase-equilibrium dynamics framework and applies it to a rock-paper-scissors oscillator, showing that phase separation can damp or amplify oscillations depending on how the interaction asymmetries localize reactants. The clean conceptual punchline is that chi_AB - chi_AC is the control parameter. That, plus the metastability hysteresis effects and the mesoscopic waves in the fast-reaction regime, are genuinely new relative to prior work, which mostly showed that phase separation can trigger oscillations. The coarse-grained model is tractable, the derivation of dE/drho via the implicit-function theorem is legitimate, and the comparison with the spatial model, while qualitative, is honest about the caveats. Citation-wise, Eq. (7) comes from ref. 47, which is a parameter-free derivation from the free energy, so I don't see circularity.

The soft spot is the stability claim. The paper states that the sign of lambda_Re predicts growth when chi_AB < chi_AC, stability at equality, and damping when chi_AB > chi_AC. But the only evidence shown is one numerical scan at psi = 0.6, chi_AS = 4, all rate constants equal to 1. The text promises that the eigenvalues can be computed in closed form, but the appendix only shows how to compute dE/drho; no closed-form expression for lambda_Re appears, and no analytic argument shows its sign is controlled solely by chi_AB - chi_AC across parameter space. So the headline rule is an extrapolation from a single state point. The spatial-model comparison is also only qualitative; it does not quantitatively confirm that the eigenvalue sign matches the damping or growth in the full spatial dynamics. These are real gaps, but they don't sink the paper. The mechanism explanation is plausible and consistent with the examples, and the paper itself acknowledges the quasi-static approximation breaks down in the fast-reaction regime and then explores that regime separately.

Who should read this: anyone working on condensate-associated reaction networks, biological clocks, or synthetic pattern-forming systems. It deserves a serious referee, but the revision should either provide the promised closed-form derivation or a systematic scan over psi, chi_AS, and rate constants, plus a more quantitative spatial-model check. No code is shipped, which makes the numerical claims harder to verify, though the model is simple enough to reimplement.

My recommendation: send it to peer review, but push for the stability claim to be substantiated rather than just stated from a single scan.

Referee Report

4 major / 5 minor

Summary. This paper studies the rock-paper-scissors chemical oscillator (A+B→2A, B+C→2B, C+A→2C) coupled to Flory-Huggins phase separation. The authors formulate a coarse-grained dynamics at instantaneous global phase equilibrium (Eq. 7), in which average concentrations evolve according to volume-weighted reactions in coexisting phases while phase fractions and compositions are re-equilibrated. They analyze the linear stability of the fixed point inside the binodal, claiming that the sign of the real part of the eigenvalues is controlled by χAB−χAC (amplification for χAB<χAC, neutral for equality, damping for χAB>χAC), and that oscillation frequency increases when χAB and χAC are more negative. They illustrate damping and enhancement, show a buffering mechanism for fixed points outside the binodal, introduce a metastability hysteresis rule, and finally study the fast-reaction regime where traveling waves of phase equilibria appear. A spatial reaction-diffusion model is used for comparison.

Significance. The central idea—reducing a phase-separating reactive system to a low-dimensional quasi-static dynamics at phase equilibrium—is useful and goes beyond fitting: Eq. (7) is a parameter-free consequence of the free energy and mass-action kinetics, and the stability framework of Eqs. (8)–(11) is a sensible tool for studying oscillator control by phase separation. The paper also makes a falsifiable prediction (the χAB−χAC sign rule) and provides qualitative spatial-model evidence (Fig. 7). If the sign rule is confirmed beyond a single state point, the work would be a valuable contribution to the understanding of chemical oscillations in condensates. The paper is clearly written and the four-quadrant comparison is a good start. However, the central quantitative claim is not yet supported by a derivation or systematic validation.

major comments (4)

[Phase separation controls the oscillations (Fig. 3 and Appendix: Response to perturbation of phase equilibria)] The central stability claim—'The sign of λ_Re predicts that perturbations grow when χAB < χAC, are stable for χAB = χAC and are damped when χAB > χAC'—is presented as a general result, but it is supported only by the numerical scan in Fig. 3 at a single state point (ψ = 0.6, χAS = 4, kAB = kBC = kCA = 1). The text states that 'the eigenvalues λ1/2 of M can be computed in closed form,' yet no closed-form expression for λ_Re is given, and the Appendix derives only the implicit-function response dE/dρ via Eq. (11). Please provide the analytic expression (or at least an analytic sign condition) for λ_Re, or alternatively restrict the claim and add systematic scans over ψ, χAS, and rate-constant ratios to demonstrate the generality asserted in the text.
[Phase separation controls the oscillations (mechanism paragraph)] The mechanism paragraph contains an inconsistency with the stated sign rule and with Fig. 2. It says 'For χAB > χAC, the production of A is accelerated more than its decay. Thus, the amplitude increases over time,' whereas the sign rule and the examples in Fig. 2 (e.g., χAB = 1, χAC = −1) show damping for χAB > χAC. In that case χAB > 0 decelerates the production of A and χAC < 0 accelerates its decay, so the amplitude should decrease. Please correct the inequality and the accompanying physical explanation.
[Abstract and Appendix: Comparison between the two dynamics] The abstract and main text state that the coarse-grained analysis is 'validated with a spatial model,' but the Appendix comparison (Fig. 7) is explicitly qualitative: it reports that both dynamics describe the same four-quadrant behavior, while the spatial model includes interface effects, Laplace pressure, finite nucleation times, and ripening that are neglected in the ODE. No quantitative comparison of the predicted amplitude or frequency as a function of χAB and χAC is provided, so the predictive content of the sign rule is not tested against the spatial model. Please add quantitative measures (e.g., oscillation amplitude and frequency from the spatial model versus the ODE) or revise the validation claim to 'qualitative agreement.'
[Dynamics with metastability] The metastability hysteresis rule changes the stability behavior qualitatively: for χAB = χAC = 1, the non-metastable dynamics has constant amplitude, whereas the metastable dynamics damps to a limit cycle (Fig. 5). Because real spatial systems have nucleation barriers, the idealized instantaneous-binodal switching rule behind Eq. (7) and Fig. 3 may not describe the spatial model in the same parameter regime. The paper should clarify which switching rule underlies the sign rule, and how the sign rule is expected to manifest when nucleation is finite.

minor comments (5)

[Appendix, Eq. (10)] The pressure-balance condition appears to contain a typo: the last term reads µi(ρI/II)(ρI_i − ρI_i) and should presumably be µi(ρI/II)(ρI_i − ρII_i).
[Phase separation controls the oscillations, text after Eq. (9)] The phrase 'Once found, however, the eigenvalues λ1/2 of M can be computed in closed form' is never followed by the expression; if the closed form is not provided, this sentence should be removed or replaced by a reference to the appendix.
[Fig. 3 caption] Fig. 3 does not specify the color scale; since the sign of λ_Re is central, a colorbar and a highlighted zero contour would make the sign rule directly readable.
[Introduction and Chemical Oscillator] The paper sets kAB = kBC = kCA = 1 at the outset; the frequency and stability predictions are therefore only for equal rate constants, and the claim that the techniques can be straightforwardly generalized to other networks should be accompanied by a statement of how rate asymmetries would enter the analysis.
[Dynamics with metastability] The metastability switching rule is described only verbally; a precise algorithmic statement (hysteresis variable, spinodal and binodal crossing conditions) would improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central sign rule is computed from the model, and the coarse-grained rate equation is physically stated and checked against a spatial model.

full rationale

The derivation chain is self-contained. Eq. (7) is introduced via the stated assumption that fast diffusion maintains global phase equilibrium; the expression dρ/dt = vR(ρI)+(1−v)R(ρII) is a direct consequence of reactions occurring independently in each phase and is cited to prior work [47] for its earlier use, not fitted to data. The stability analysis computes M by differentiating the phase-equilibrium constraints through the Jacobian J (Appendix), so the eigenvalue sign rule in Fig. 3 is a computed prediction of the model, not an input. The fixed point and eigenvalues are obtained numerically at stated parameters (ψ=0.6, χAS=4, kij=1), and the sign rule is checked against finite-amplitude trajectories in Fig. 2. The spatial model provides an independent qualitative validation (Appendix), with the paper explicitly listing neglected interface and nucleation effects. The only same-group citation is [47] for the phase-equilibrium reaction dynamics, but the present conclusion—that the sign of χAB−χAC controls damping versus amplification—is not assumed by that reference and is not definitionally equivalent to any fitted parameter. No uniqueness theorem is imported, and no known result is renamed. Thus there is no circular step; the open issue is numerical generality at a single state point, which is a correctness/completeness concern, not circularity.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities. It builds on Flory-Huggins mean-field theory and mass-action kinetics, with several parameters chosen by hand. The only ad hoc ingredient is the metastability hysteresis switching rule.

free parameters (6)

chi_AS = 4
Interaction strength between A and solvent; chosen above the critical value 2 to ensure A-rich/A-poor phase separation in all examples.
psi = 0.6
Total average concentration, set by initial conditions in all main-text examples; conserved by the dynamics.
chi_AB, chi_AC = varied in [-1,1]
Interaction energies controlling localization of B and C; the key control parameter is chi_AB - chi_AC. Values are chosen to illustrate damping and amplification regimes.
kappa_i = 1.5
Gradient coefficient in the spatial model free energy; set for interface width in simulations.
gamma = 1
Mobility coefficient in the spatial diffusion term.
k_AB = k_BC = k_CA = 1 (main text), 0.0005 or 0.1 (spatial)
Reaction rate constants; set to 1 in the coarse-grained analysis to fix time units, and varied in spatial simulations to control timescale separation.

assumptions (6)

domain assumption Mass-action kinetics with second-order rate constants
The RPS oscillator dynamics in Eq. (2) is assumed to follow deterministic mass-action kinetics.
domain assumption Flory-Huggins free energy with only A interacting
Free energy Eq. (3) models interactions only between A and S,B,C; B and C do not phase separate by themselves. This is a restricted modeling choice.
domain assumption Fast diffusion limit: global phase equilibrium is maintained instantaneously
Section 'Reaction dynamics at phase equilibrium' derives Eq. (7) under this assumption. This is the weakest load-bearing assumption.
domain assumption Phase equilibrium is the global minimum of the average free energy
Used to determine (v, rho_I, rho_II) from Eqs. (5)-(6). The global minimum may be multi-valued; the model assumes a unique choice.
ad hoc to paper Metastability hysteresis switching rule
Section 'Dynamics with metastability' introduces a switching variable that enters the phase-separated dynamics only when the spinodal is crossed and leaves at the binodal. This is an ad hoc model for nucleation barriers.
domain assumption Mobility matrix Gamma_ij = gamma rho_i (delta_ij - rho_j)
Chosen so that diffusion reduces to Fick's law in the ideal limit; this is a standard, but non-unique, choice.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Spatiotemporal organization of chemical oscillators via phase separation." pith.science (2026). https://pith.science/paper/UAU34KAZ

@misc{pith2026250716030,
  author       = {Pith},
  title        = {Pith review of: Spatiotemporal organization of chemical oscillators via phase separation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UAU34KAZ}},
  note         = {Machine review of arXiv:2507.16030}
}

read the original abstract

We study chemical oscillators in the presence of phase separation. By imposing timescale separation between slow reactions and fast diffusion, we define a dynamics at phase equilibrium for the relevant degrees of freedom. We demonstrate that phase separation affects reaction kinetics by localizing reactants within phases, allowing for control of oscillator frequency and amplitude. The analysis is validated with a spatial model. Finally, relaxing the timescale separation between reactions and diffusion leads to waves of phase equilibria at mesoscopic scales.

Figures

Figures reproduced from arXiv: 2507.16030 by the authors.

**Figure 1.** Chemical oscillator in the presence of coexisting phases. (a) Spatial profiles (left) of ρ and corresponding spatial averages sketched in concentration space (right) at five time points in a spatial model. (b) Trajectory of average (volume) concentrations in ρ in concentration space (orange). Within the binodal domain (grey shaded), phases coexist at equilibrium; outside the binodal domain (white), equilibrium is w… view at source ↗

**Figure 3.** Phase separation affects stability and oscillation frequency at the fixed point. (a) Real and (b) Imaginary part of the relaxation rates of perturbations to the fixed point of Eq. (7) as a function of χAB and χAC . ψ = 0.6, χAS = 4; over time, and there are longer and longer intervals in which the mixture is composed mainly of a single solute, i.e. ψ ∼ ρi . When this solute is A, two phases coexist; otherwise, the … view at source ↗

**Figure 2.** Phase separation controls oscillations. Left: oscillations of the average concentrations ρ and volume v. Right: trajectories of the average concentrations, with region within binodal domain in grey. (a,b) For χAB = 1, χAC = −1, oscillations are damped. (c,d) For χAB = −1, χAC = 1, oscillation are enhanced. ψ = 0.6, χAS = 4. in the concentration of each individual reactant, the dynamics in Eq. (7) differ from Eq. (2)… view at source ↗

Figures from the paper (3 more)

**Figure 5.** Figure 5: Metastability affects oscillations. Without metastability, phases form as soon as the binodal line is crossed (light colors); with metastability, phases form only once the spinodal (darker colors) are crossed. (a) Oscillations of the average concentration ρA and volume…

**Figure 6.** Figure 6: Mesoscopic waves of phase equilibria emerge in the fast reaction regime. (a) Spatial concentrations in a large system. (b) Right: local total concentration ψ in a zoomed-in region of space (grey dashed box in top plots). Average concentrations in five mesoscopic envir…

**Figure 7.** Figure 7: Qualitative agreement between the dynamics at phase equilibrium and the spatial dynamics in the fast diffusion regime: Dynamics of the average concentrations ρ for four cases in a twodimensional spatial model (a) and the dynamics at phase equilibrium (b). For the dyna…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 50 canonical work pages

[1]

C.P.Brangwynne,C.R.Eckmann,D.S.Courson,A.Rybarska, C.Hoege,J.Gharakhani,F.Jülicher,andA.A.Hyman,Science 324, 1729 (2009)

work page 2009
[2]

S.F.Banani,H.O.Lee,A.A.Hyman,andM.K.Rosen,Nature Reviews Molecular Cell Biology18, 285–298 (2017)

work page 2017
[3]

Boeynaems, S

S. Boeynaems, S. Alberti, N. L. Fawzi, T. Mittag, M. Poly- menidou, F. Rousseau, J. Schymkowitz, J. Shorter, B. Wolozin, L. Van Den Bosch, P. Tompa, and M. Fuxreiter, Trends in Cell Biology28, 420–435 (2018)

work page 2018
[4]

A. A. Hyman, C. A. Weber, and F. Jülicher, Annual Review of Cell and Developmental Biology30, 39–58 (2014)

work page 2014
[5]

D. M. Mitrea and R. W. Kriwacki, Cell Communication and Signaling14, 10.1186/s12964-015-0125-7 (2016)

work page doi:10.1186/s12964-015-0125-7 2016
[6]

C.A.Strulson, R.C.Molden, C.D.Keating,andP.C.Bevilac- qua, Nature Chemistry4, 941–946 (2012)

work page 2012
[7]

K. K. Nakashima, M. A. Vibhute, and E. Spruĳt, Frontiers in Molecular Biosciences6, 10.3389/fmolb.2019.00021 (2019)

arXiv 2019
[8]

C. A. Weber, D. Zwicker, F. Jülicher, and C. F. Lee, Reports on Progress in Physics82, 064601 (2019)

work page 2019

Show all 58 references

[9]

Oparin,The Origin of Life(Dover Publications, New York, 1953)

A. Oparin,The Origin of Life(Dover Publications, New York, 1953)

1953
[10]

J. B. S. Haldane, Rationalist Annual , 3 (1929)

1929
[11]

Q.-X. Liu, A. Doelman, V. Rottschäfer, M. de Jager, P. M. Herman, M.Rietkerk,andJ.vandeKoppel,Proceedingsofthe National Academy of Sciences110, 11905 (2013)

2013
[12]

K.Siteur,Q.-X.Liu,V.Rottschäfer,T.vanderHeide,M.Rietk- erk,A.Doelman,C.Boström,andJ.vandeKoppel,Proceedings oftheNationalAcademyofSciences 120,e2202683120(2023)

2023
[13]

Aslyamov, F

T. Aslyamov, F. Avanzini, É. Fodor, and M. Esposito, Physical Review Letters131, 138301 (2023)

2023
[14]

Avanzini, T

F. Avanzini, T. Aslyamov, É. Fodor, and M. Esposito, The Jour- nal of Chemical Physics161(2024)

2024
[15]

Kirschbaum and D

J. Kirschbaum and D. Zwicker, Journal of The Royal Society Interface18, 20210255 (2021)

2021
[16]

Y.ChoandW.M.Jacobs,TheJournalofChemicalPhysics 159, 10.1063/5.0166824 (2023)

2023 doi
[17]

S. Laha, J. Bauermann, F. Jülicher, T. C. T. Michaels, and C. A. Weber, Physical Review Research6, 10.1103/physrevre- search.6.043092 (2024)

2024 doi
[18]

S. C. Glotzer, E. A. Di Marzio, and M. Muthukumar, Physical 6 Review Letters74, 2034–2037 (1995)

1995
[19]

Zwicker, A

D. Zwicker, A. A. Hyman, and F. Jülicher, Physical Review E 92, 10.1103/physreve.92.012317 (2015)

2015 doi
[20]

J. D. Wurtz and C. F. Lee, Physical Review Letters 120, 10.1103/physrevlett.120.078102 (2018)

2018 doi
[21]

Kumar and S

A. Kumar and S. A. Safran, Physical Review Letters 131, 10.1103/physrevlett.131.258401 (2023)

2023 doi
[22]

Bauermann, G

J. Bauermann, G. Bartolucci, C. A. Weber, and F. Jülicher, arXiv:2501.16299 10.48550/ARXIV.2501.16299 (2025)

2025 doi
[23]

Ziethen, J

N. Ziethen, J. Kirschbaum, and D. Zwicker, Physical Review Letters130, 10.1103/physrevlett.130.248201 (2023)

2023 doi
[24]

Cho and W

Y. Cho and W. M. Jacobs, Physical Review Letters 130, 10.1103/physrevlett.130.128203 (2023)

2023 doi
[25]

Zwicker, R

D. Zwicker, R. Seyboldt, C. A. Weber, A. A. Hyman, and F. Jülicher, Nature Physics13, 408 (2017)

2017
[26]

J.Bauermann,C.A.Weber,andF.Jülicher,AnnalenderPhysik 534, 10.1002/andp.202200132 (2022)

2022 doi
[27]

Demarchi, A

L. Demarchi, A. Goychuk, I. Maryshev, and E. Frey, Physical Review Letters130, 10.1103/physrevlett.130.128401 (2023)

2023 doi
[28]

Häfner and M

G. Häfner and M. Müller, ACS Nano18, 16530–16544 (2024)

2024
[29]

Goychuk, arXiv:2506.07753 10.48550/ARXIV.2506.07753 (2025)

A. Goychuk, arXiv:2506.07753 10.48550/ARXIV.2506.07753 (2025)

2025 doi
[30]

M. S. Heltberg, A. Lucchetti, F.-S. Hsieh, D. P. M. Nguyen, S.-h. Chen, and M. H. Jensen, Cell185, 4394 (2022)

2022
[31]

Zhuang, Z

Y. Zhuang, Z. Li, S. Xiong, C. Sun, B. Li, S. A. Wu, J. Lyu, X. Shi, L. Yang, Y. Chen,et al., Cell186, 3245 (2023)

2023
[32]

Tariq, N

D. Tariq, N. Maurici, B. M. Bartholomai, S. Chandrasekaran, J.C.Dunlap,A.Bah,andB.R.Crane,Elife 12,RP90259(2024)

2024
[33]

I. S. Haugerud, H. D. Vuĳk, J. Boekhoven, and C. A. Weber, arXiv:2503.11604 10.48550/ARXIV.2503.11604 (2025)

2025 doi arXiv
[34]

Sastre, A

J. Sastre, A. Thatte, A. M. Bergmann, M. Stasi, M. Tena- Solsona, C. A. Weber, and J. Boekhoven, Nature Communi- cations16, 10.1038/s41467-025-57240-8 (2025)

2025 doi
[35]

Spruĳt, ChemSystemsChem7, 10.1002/syst.202400056 (2024)

I.B.A.Smokers,B.S.Visser,W.P.Lipiński,K.K.Nakashima, and E. Spruĳt, ChemSystemsChem7, 10.1002/syst.202400056 (2024)

2024 doi
[36]

J. M. Smith,Evolution and the theory of games(Cambridge University Press, Cambridge, England, 2012)

2012
[37]

Sinervo and C

B. Sinervo and C. M. Lively, Nature380, 240–243 (1996)

1996
[38]

Onsager, Physical Review37, 405–426 (1931)

L. Onsager, Physical Review37, 405–426 (1931)

1931
[39]

S.R.DeGrootandP.Mazur, Non-equilibriumThermodynamics, Dover Books on Physics (Dover Publications, Mineola, NY, 2003)

2003
[40]

E. J. Kramer, P. Green, and C. J. Palmstrøm, Polymer25, 473–480 (1984)

1984
[41]

S.Bo,L.Hubatsch,J.Bauermann,C.A.Weber,andF.Jülicher, PhysicalReviewResearch 3,10.1103/physrevresearch.3.043150 (2021)

2021 doi
[42]

Reichenbach, M

T. Reichenbach, M. Mobilia, and E. Frey, Nature 448, 1046–1049 (2007)

2007
[43]

T.Reichenbach,M.Mobilia,andE.Frey,JournalofTheoretical Biology254, 368–383 (2008)

2008
[44]

Q. He, M. Mobilia, and U. C. Täuber, Physical Review E82, 10.1103/physreve.82.051909 (2010)

2010 doi
[45]

Safran,Statistical thermodynamics of surfaces, interfaces, and membranes(CRC Press, London, England, 2019)

S. Safran,Statistical thermodynamics of surfaces, interfaces, and membranes(CRC Press, London, England, 2019)

2019
[46]

Kardar,Statistical physics of fields(Cambridge University Press, Cambridge, England, 2007)

M. Kardar,Statistical physics of fields(Cambridge University Press, Cambridge, England, 2007)

2007
[47]

Bauermann, S

J. Bauermann, S. Laha, P. M. McCall, F. Jülicher, and C. A. Weber, Journal of the American Chemical Society144, 19294 (2022)

2022
[48]

S. H. Strogatz,Nonlinear Dynamics and Chaos(CRC Press, 2018)

2018
[49]

Deviri and S

D. Deviri and S. A. Safran, Proceedings of the National Academy of Sciences of the United States of America118, e2100099118 (2021)

2021
[50]

Zechner and F

C. Zechner and F. Jülicher, Cell Systems16, 101168 (2025)

2025
[51]

Klosin, F

A. Klosin, F. Oltsch, T. Harmon, A. Honigmann, F. Jülicher, A. A. Hyman, and C. Zechner, Science367, 464–468 (2020)

2020
[52]

Bray, Advances in Physics43, 357–459 (1994)

A. Bray, Advances in Physics43, 357–459 (1994)

1994
[53]

SeeSupplementalMaterialat[url]formoviesofspatialsystems with phase separation
[54]

Grisel, M

F.Pedregosa,G.Varoquaux,A.Gramfort,V.Michel,B.Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J.Vanderplas,A.Passos,D.Cournapeau,M.Brucher,M.Perrot, and E. Duchesnay, Journal of Machine Learning Research12, 2825 (2011)

2011
[55]

Halatek and E

J. Halatek and E. Frey, Nature Physics14, 507–514 (2018)

2018
[56]

J.Halatek,F.Brauns,andE.Frey,PhilosophicalTransactionsof theRoyalSocietyB:BiologicalSciences 373,20170107(2018)

2018
[57]

A. W. Fritsch, A. F. Diaz-Delgadillo, O. Adame-Arana, C. Hoege, M. Mittasch, M. Kreysing, M. Leaver, A. A. Hy- man, F. Jülicher, and C. A. Weber, Proceedings of the National Academy of Sciences118, 10.1073/pnas.2102772118 (2021)

2021 doi
[58]

J. F. Robinson, T. Machon, and T. Speck, Physical Review E 111, 10.1103/1bdc-1bjb (2025). 7 APPENDIX Interaction parameters In the Flory-Huggins free energy density, as introduced in Eq. (3), molecular interactions between componentAand the other components j = B, C, Sare para...

2025 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

C.P.Brangwynne,C.R.Eckmann,D.S.Courson,A.Rybarska, C.Hoege,J.Gharakhani,F.Jülicher,andA.A.Hyman,Science 324, 1729 (2009)

work page 2009

[2] [2]

S.F.Banani,H.O.Lee,A.A.Hyman,andM.K.Rosen,Nature Reviews Molecular Cell Biology18, 285–298 (2017)

work page 2017

[3] [3]

Boeynaems, S

S. Boeynaems, S. Alberti, N. L. Fawzi, T. Mittag, M. Poly- menidou, F. Rousseau, J. Schymkowitz, J. Shorter, B. Wolozin, L. Van Den Bosch, P. Tompa, and M. Fuxreiter, Trends in Cell Biology28, 420–435 (2018)

work page 2018

[4] [4]

A. A. Hyman, C. A. Weber, and F. Jülicher, Annual Review of Cell and Developmental Biology30, 39–58 (2014)

work page 2014

[5] [5]

D. M. Mitrea and R. W. Kriwacki, Cell Communication and Signaling14, 10.1186/s12964-015-0125-7 (2016)

work page doi:10.1186/s12964-015-0125-7 2016

[6] [6]

C.A.Strulson, R.C.Molden, C.D.Keating,andP.C.Bevilac- qua, Nature Chemistry4, 941–946 (2012)

work page 2012

[7] [7]

K. K. Nakashima, M. A. Vibhute, and E. Spruĳt, Frontiers in Molecular Biosciences6, 10.3389/fmolb.2019.00021 (2019)

arXiv 2019

[8] [8]

C. A. Weber, D. Zwicker, F. Jülicher, and C. F. Lee, Reports on Progress in Physics82, 064601 (2019)

work page 2019