{"id":"47e26b39-ad68-4ae8-83d2-e0fb55014c15","arxiv_id":"2507.23580","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Random identity swaps or momentum kicks between two particle states arrest Ostwald ripening and stabilize small droplets across several Lennard-Jones-type interaction potentials.","lead":"Molecular dynamics simulations of particles with different interaction strengths show that randomly swapping particle identities, or kicking their momenta, stops small droplets from shrinking and large droplets from growing, a process called Ostwald ripening. The result suggests a simple physical reason why protein droplets inside living cells stay small and stable.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The arrest claim rests on a one-microsecond plateau at nonzero swap rates, with no replicate runs or stationarity test; the authors' own text concedes eventual coalescence, so the key assumption that the plateau is a true steady state is untested.","rationale":"The reader's weakest assumption exactly matches the most load-bearing gap I find: the finite-time plateau at nonzero swap rates is not established as a non-equilibrium steady state. This is the keystone of the paper's headline claim, because the abstract and conclusions make a strong universality statement ('in all cases Ostwald ripening is absent only away from equilibrium') that depends on the plateau being the asymptotic state, not a slow coarsening transient. I weight this concern more heavily than secondary issues such as the absence of shipped code or a precise droplet-detection protocol, because those affect reproducibility but not the physical interpretation of the central result. The paper does have independent support in the sense that the qualitative behavior is consistent with prior active-emulsion and chemically active droplet work (e.g., Zwicker et al. 2015; Tjhung et al. 2018), and the use of multiple two-body potentials strengthens the robustness claim within the tested parameter window. However, the strongest evidence for arrest is a single trajectory per condition that runs only to ~1 microsecond, while cellular condensates persist for much longer times; the authors' own admission that eventual coarsening occurs through rare coalescence means 'arrested' is only as strong as the unquantified claim that coalescence is negligible. A 10× longer simulation with multiple seeds and a measured late-time slope would directly distinguish a genuine steady state from a transient, and would also permit error bars on the plateau size. I therefore keep the reader's CONDITIONAL verdict unchanged: the mechanism is plausible and qualitatively credible, but the central claim is not yet quantitatively supported.","tokens_in":9488,"tokens_out":4344,"duration_ms":54224,"concrete_test":"Run the λ = 0.1 swap protocol for 10 independent seeds for 10× longer (1e7 timesteps, ~10 microseconds), recording the largest droplet size L(t) and the full size distribution. Fit L(t) in the late-time window (t > 4e5 steps) to L ~ t^β with uncertainty; if β is significantly positive and L grows by more than ~20% above the 1-microsecond plateau, the plateau is a transient and the arrest claim fails. As a complementary check, double the box length at fixed density and same λ; if the plateau droplet size or droplet number scales with system volume instead of saturating, the apparent arrest is a finite-size effect.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that non-equilibrium driving suppresses Ostwald ripening and stabilizes finite droplets across potentials (abstract; Figs. 3, 4, 8). The decisive evidence is the plateau in largest-droplet size at nonzero swap rates (Figs. 5, 6), but each parameter set is represented by a single one-million-step trajectory (t* = 0.01, mapped via Eq. 9 to roughly 1 microsecond). No error bars, no independent seeds, and no quantitative stationarity check are reported. The plateau could be a slowly coarsening transient: with 27,000 particles and 30% B particles, the system contains about 8,100 B particles, and droplet coalescence is expected to become rare as droplets separate, so a 1-microsecond horizon is short relative to the coalescence-limited ripening timescale. The paper even concedes this in the Conclusions: \"eventual coarsening occurs through rare droplet coalescence.\" If that eventual coarsening occurs on timescales relevant to the motivating biological condensates (seconds to hours), then the title's \"arrested\" and abstract's \"Ostwald ripening is absent\" overstate the result; the observation would be slowed ripening, not a stable non-equilibrium steady state. The load-bearing assumption, therefore, is that the 1-microsecond plateau is the steady state rather than a pre-asymptotic regime. This is distinct from the mechanism claim: it is about whether the measured finite-time behavior supports the stated universality. Because the distinction between a true steady state and a slow transient cannot be settled from the reported data, the CONDITIONAL verdict is appropriate, but the missing long-time and ensemble evidence is the specific gap that must be closed.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports molecular dynamics simulations of a binary mixture of particles in states A and B, interacting through Lennard-Jones-type pair potentials. The B-B interaction is either a Lennard-Jones potential plus a shifted Gaussian (the 'double-well' model) or five times the A-A Lennard-Jones potential. The system is driven out of equilibrium either by randomly exchanging A-B identities at fixed composition or by random momentum kicks to a fraction of particles. In the equilibrium (no-swap) limit the largest droplet grows and the droplet-size distribution coarsens; at nonzero swap rates or with momentum kicks, the largest droplet size appears to plateau after a few hundred thousand steps and a stable distribution of small droplets is reported. The authors conclude that Ostwald ripening is suppressed or absent away from equilibrium and propose this as a mechanism for stabilizing biomolecular condensates. The paper contains no analytical theory; the evidence is entirely from LAMMPS simulations, with one million timesteps per run mapped to roughly one microsecond of physical time.","tokens_in":9768,"tokens_out":6063,"duration_ms":66388,"significance":"If the reported plateau is a genuine non-equilibrium steady state, the result would be significant: it would support a generic, potential-independent mechanism by which local energy input arrests Ostwald ripening, relevant to active emulsions and biological condensates. The study's strengths are the breadth of interaction potentials considered (Gaussian-modified Lennard-Jones with three Gaussian centers, a deeper Lennard-Jones potential) and the use of two distinct driving protocols (identity swaps and momentum kicks). The paper does not provide code or raw data, but the model is simple enough to reproduce. The significance is conditional on establishing that the observed plateau is a true steady state and not a slowly coarsening transient, which the current single-run, one-microsecond simulations do not yet establish.","major_comments":[{"comment":"The central claim that Ostwald ripening is 'absent' or 'arrested' rests on a single one-million-step trajectory for each parameter set, corresponding to about one microsecond of physical time as stated. No replicate runs, error bars, or quantitative stationarity test are reported, and the Conclusions explicitly concede that 'eventual coarsening occurs through rare droplet coalescence.' With roughly 8,100 B particles, coalescence-limited ripening can naturally be slower than the simulated time horizon, so the plateau in the largest droplet size could be a pre-asymptotic transient rather than a non-equilibrium steady state. The manuscript needs multiple independent seeds, longer simulations or a stationarity criterion (e.g., no drift in the droplet-size distribution over several droplet turnover times), and ideally a measurement of the monomer evaporation/condensation flux to distinguish slowed ripening from true arrest.","section":"Figs. 5-6; Eq. (9); Conclusions"},{"comment":"The paper does not operationally distinguish 'no growth of the largest droplet' from 'Ostwald ripening is absent.' The authors themselves describe the growth in the no-swap control as 'likely due to coalescence,' and at finite swap rates the largest droplet could remain nearly constant even if Ostwald ripening continues to transfer material from smaller to larger droplets. To support the title's claim, the analysis should separate coalescence from Ostwald ripening, for example by tracking individual droplet volumes, monitoring the evolution of the full droplet-size distribution against Lifshitz-Slyozov scaling, or measuring monomer exchange between droplets. Without such an analysis, the observations are consistent with 'slowed ripening,' not uniquely with 'arrested Ostwald ripening.'","section":"Figs. 3-6; Analysis and Conclusions"},{"comment":"The robustness claim 'across all interaction potentials considered' is based on two functional forms: the Lennard-Jones plus shifted Gaussian with H = -0.7 and three values of a, and the five-times Lennard-Jones potential, plus the momentum-kick protocol. The Gaussian parameters are explicitly chosen so that 'the special results ... do not occur' otherwise, which suggests the arrest may be parameter-sensitive. A systematic scan over H, delta, rho*, T*, or the B-particle fraction would be needed to substantiate the claim that arrested ripening is generic. At minimum, the abstract and conclusions should temper 'in all cases' to 'in the parameter regimes studied.'","section":"Models, Eq. (2), Figs. 5-6"},{"comment":"The protocol is asserted to produce a non-equilibrium steady state because A-B transitions occur at equal rates, but the composition is held fixed by construction and the energy exchange during swaps is stated to average to zero. The manuscript also says the simulations run 'under constant number, volume, and energy (NVE) conditions, with a Langevin thermostat applied,' which is internally inconsistent because a Langevin thermostat exchanges energy. Since the paper's central comparison is equilibrium versus non-equilibrium, the authors should provide a quantitative check of non-equilibrium behavior, such as broken detailed balance in particle trajectories, non-zero entropy production, or a measured distribution that differs from the equilibrium ensemble at the same T* and rho*, and they should clarify whether the thermostat is active during production runs.","section":"Models; 'randomly exchanging identities' paragraph"}],"minor_comments":[{"comment":"The time mapping appears inconsistent: one million steps with a reduced timestep of 0.01 corresponds to 10^4 reduced time units, which with tau = 4.1e-10 s gives approximately 4 microseconds, not the stated 'approximately 1 microsecond.' Please check the conversion and state the physical time consistently.","section":"Eq. (9) and simulation parameters"},{"comment":"There are several language issues, including 'This phenomena' (should be 'This phenomenon'), the inconsistent hyphenation of 'non-equilibrium'/'nonequilibrium', and the phrase 'interacting with each other via several central potential' (should be 'several central potentials').","section":"Abstract and Introduction"},{"comment":"The word 'non-equlibrium' in the caption is a typo for 'non-equilibrium.'","section":"Figure 3 caption"},{"comment":"The equilibrium phase diagram is shown for the plain Lennard-Jones potential and the double-well potential with a=2.2 only; the five-times-Lennard-Jones case is not shown, so the statement that the chosen operating point lies inside the coexistence region for all studied B-B potentials is not fully supported by the figure.","section":"Fig. 2 and phase-diagram text"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for physics.bio-ph and addresses a topical problem. My main concern is the gap between the finite-time simulation evidence and the strong wording 'arrested' and 'absent'; this should be resolved with additional simulations and statistical analysis. I would also encourage the editor to ask for a data/code availability statement, as the manuscript currently provides neither trajectories nor input scripts, which limits reproducibility."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is worth a serious look, but its headline is stronger than its data. What is actually new is modest but real: the authors show that a very simple driving protocol—random A/B identity swaps or random momentum kicks—produces a plateau in largest-droplet size for several two-body central potentials. Prior work had already reported activity-suppressed or reversed Ostwald ripening, so the phenomenon is not new. The contribution is the simplicity of the model and the demonstration that the effect survives across potential shapes. That is a legitimate extension, and the qualitative result is credible.\n\nWhat the paper does well: the operating point is checked against Monte Carlo phase diagrams, the swap-rate dependence is shown, the B-fraction switching experiment is a nice control, and the authors honestly concede in the conclusions that eventual coarsening occurs through rare droplet coalescence. There is no circularity: the arrest is read directly from simulation output, not produced by fitting. The mechanism discussion is qualitative but sensible.\n\nThe soft spots are real and, for the central claim, load-bearing. Each parameter set is a single one-million-step trajectory mapped to about one microsecond. There are no replicate runs, no error bars, and no quantitative stationarity test. The plateau could be a slow transient: with roughly 8,100 B particles, coalescence-limited coarsening can easily be slower than one microsecond. The authors' expectation that longer runs would not change the outcome is not a substitute for a test. So the abstract's “in all cases Ostwald ripening is absent” and the title's “arrested” overstate what is shown. What is shown is suppression of ripening on a microsecond timescale. That is still interesting for a minimal model, and the biological relevance is plausibly stated, but the distinction matters for a physics audience.\n\nMinor issues: no code or data shipped, which slows verification; Eq. 11 is not actually used; the Gaussian parameter selection is acknowledged and is not itself a flaw.\n\nWho is this for? Active-matter and biomolecular-condensate researchers who want a quick, clean illustration that simple non-equilibrium driving can stabilize small droplets. It deserves a serious referee and likely revision before acceptance: add multiple seeds and error bars, extend or at least probe the plateau timescale, and rewrite the claims to say what is actually observed. I would not cite it in its current form, but I would read a revised version.","headline":"A clean minimal demonstration that identity swaps or momentum kicks suppress droplet coarsening in LJ fluids, but the 'Ostwald ripening is absent' claim outruns the evidence: one-microsecond single runs, no error bars, and a conceded eventual-coalescence loophole.","tokens_in":10412,"tokens_out":1777,"would_cite":false,"duration_ms":22768,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Random identity swaps or momentum kicks arrest Ostwald ripening and stabilize finite-size droplets across several two-body potentials.","keywords":["Ostwald ripening","non-equilibrium steady state","liquid-liquid phase separation","molecular dynamics","biomolecular condensates","arrested coarsening","two-state particle model","Lennard-Jones potential"],"falsifier":"Run the same two-state simulations to $10^{7}$ or $10^{8}$ timesteps at fixed nonzero swap rates and track the largest and mean droplet sizes; if growth resumes with a power-law or logarithmic time dependence after the plateau, the arrest is a finite-time effect. Also check the plateau droplet size in boxes of increasing volume: if the plateau grows with system size, the apparent arrest may be a finite-size artifact.","tokens_in":12,"feed_emoji":"🫧","tokens_out":6877,"duration_ms":132041,"temperature":0.7,"pith_summary":"In a mixture of two particle states that attract each other differently, equilibrium droplets coarsen through Ostwald ripening: small droplets dissolve and feed larger ones. The paper tries to establish that this ripening disappears when the system is driven away from equilibrium by random A↔B identity swaps or by random momentum kicks, and that the effect is generic rather than tied to one interaction potential. Using molecular dynamics with Lennard-Jones, Lennard-Jones-plus-Gaussian double-well, and deeper Lennard-Jones interactions, the authors observe a steady distribution of small droplets under non-equilibrium driving, while the same systems coarsen at equilibrium. If true, the result would identify a simple physical mechanism—local energy input that breaks time-reversal symmetry—by which biomolecular condensates in living cells might avoid ripening and keep a finite size.","feed_headline":"Driving droplets out of equilibrium halts Ostwald ripening","feed_subtitle":"In simulations, random A-B identity swaps and momentum kicks keep small droplets stable across several interaction potentials.","key_machinery":"The central object is a two-state particle model with a fixed A:B ratio, where A-A and A-B pairs interact via a Lennard-Jones potential and B-B pairs interact via a more attractive potential, either a double-well Lennard-Jones-plus-Gaussian or a fivefold-deeper Lennard-Jones. An identity swap changes a particle's interaction potential with its neighbors in place. The system is driven out of equilibrium by random A↔B identity swaps at a rate λ, or by random momentum kicks, which inject or remove local energy and break time-reversal symmetry while preserving total energy on average. This local energy transfer is the mechanism that suppresses the Gibbs-Thomson-driven diffusive transport from small to large droplets, leaving coalescence as the only, infrequent, growth route.","core_discovery":"At equilibrium, droplets of the more-attractive B particles grow over time by coarsening and coalescence. When the system is held out of equilibrium—by exchanging the identities of 10% of A and B particles every 1,000 timesteps, or by giving random momentum kicks to 30% of particles—the droplet-size distribution stops changing and small droplets persist for the full one-million-step run. The paper reports this arrest for the modified Lennard-Jones potential, the Lennard-Jones-plus-Gaussian double-well potential, and the fivefold-deep Lennard-Jones potential, and concludes that across all interaction potentials considered, Ostwald ripening is absent away from equilibrium. The physical picture offered is that local energy deposition or extraction destabilizes large droplets, which contain many bulk particles, more than intermediate-sized droplets, so the thermodynamic driving force for ripening is neutralized.","pith_inferences":["If the arrest is generic, then any sustained source of random energy input, not just ATP-driven enzymatic cycles, should stabilize finite-size condensates; this could be tested in vitro by adding an enzymatic fuel that maintains a non-equilibrium steady state.","The one-million-step horizon leaves open whether the plateau is a true steady state; a direct test is to run much longer and see whether the largest-droplet growth curve remains flat, and to check whether the plateau size depends on box size.","The simulations suggest a mean-field prediction that the plateau droplet size should scale with the swap rate and the curvature contribution to chemical potential; an analytic reaction-diffusion model with random state switching could be compared with the measured size distributions.","The mechanism may extend to other phase-separating mixtures, such as colloid-polymer systems, alloys, or emulsions, wherever random compositional or kinetic changes are present, though the paper only simulates the specific Lennard-Jones-type models."],"forward_implications":["Droplet size can be tuned by the swap rate: increasing λ gives progressively smaller largest-droplet sizes, with no ripening at nonzero rates.","Because the arrest appears for all tested B-B potentials, sustained local energy input, not the specific chemistry of the interactions, is the essential ingredient for suppressing Ostwald ripening.","Coalescence remains the only growth mechanism in the non-equilibrium steady state, so any eventual coarsening is slow and occurs on timescales long compared with physiological ones.","Momentum kicks reproduce the same arrest in a single-state system, so the mechanism does not require internal states or potential switching; any form of random local energy transfer suffices."],"supporting_citations":[{"why":"Active-emulsion theory showing that chemical reactions can suppress Ostwald ripening, the prior result this paper extends to two-body potentials.","marker":"[10]"},{"why":"Experimental observation that active coacervate droplets resist Ostwald ripening, the biological phenomenon the paper aims to explain.","marker":"[22]"},{"why":"Review of active emulsions establishing that out-of-equilibrium processes can arrest phase separation and stabilize droplet sizes.","marker":"[24]"},{"why":"Active-fluid simulation showing reversal of the Ostwald process, a related precedent for activity arresting coarsening.","marker":"[31]"},{"why":"Earlier molecular dynamics model using a three-body potential to produce small stable droplets, which this paper generalizes to two-body potentials.","marker":"[40]"},{"why":"Theory of chemical-reaction-controlled phase-separated drops, supporting the idea that reaction rates set droplet size and coarsening behavior.","marker":"[41]"}],"fun_headline_variants":["Non-equilibrium freezes droplet ripening","Droplets stop coarsening when driven out of equilibrium","Out-of-equilibrium halts Ostwald ripening in simulations","Off-balance droplets defy Ostwald ripening"],"cache_read_input_tokens":12416,"weakest_assumption_plain":"The load-bearing premise is that the flat droplet-size plateau seen over one million timesteps is a genuine non-equilibrium steady state rather than a slow transient; the simulations last about one microsecond, so if ripening resumes on longer timescales the central claim collapses.","fun_headline_variants_meta":{"raw":{"variants":["Non-equilibrium freezes droplet ripening","Droplets stop coarsening when driven out of equilibrium","Out-of-equilibrium halts Ostwald ripening in simulations","Off-balance droplets defy Ostwald ripening"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000805,"raw_usage":{"total_tokens":3458,"prompt_tokens":789,"completion_tokens":2669,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":405,"completion_tokens_details":{"reasoning_tokens":2617}},"tokens_in":405,"tokens_out":2669,"duration_ms":20056,"temperature":1.0,"reasoning_tokens":2617,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T10:35:08.446814+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same two-state simulations to $10^{7}$ or $10^{8}$ timesteps at fixed nonzero swap rates and track the largest and mean droplet sizes; if growth resumes with a power-law or logarithmic time dependence after the plateau, the arrest is a finite-time effect. Also check the plateau droplet size in boxes of increasing volume: if the plateau grows with system size, the apparent arrest may be a finite-size artifact.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Experimental observation that active coacervate droplets resist Ostwald ripening, the biological phenomenon the paper aims to explain."},{"cited_title":"Osmanovi´ c and Y","cited_arxiv_id":null,"evidence_quote":"Earlier molecular dynamics model using a three-body potential to produce small stable droplets, which this paper generalizes to two-body potentials."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Theory of chemical-reaction-controlled phase-separated drops, supporting the idea that reaction rates set droplet size and coarsening behavior."}],"review_version":1}