{"id":"536a107e-578c-4df3-9826-c54445280d70","arxiv_id":"2412.02854","paper_version":3,"verdict":"UNVERDICTED","confidence":"MODERATE","novelty_score":2.0,"correctness_risk":"low","formal_verification":"none","parameter_count":4,"one_line_summary":"A synthesis of active phase separation, centered on Active Model B+ and H, showing that multiple interfacial tensions, some negative, generate reverse Ostwald ripening, microphase separation, bubbly phases, and active foams.","lead":"This is a review of how fluid-fluid phase separation changes when the constituents are active, self-propelled particles instead of equilibrium molecules. It argues that activity can create several different interfacial tensions at once, some negative, which produce new patterns such as microphase separation.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The negative-tension phenomenology is secure inside AMB+, but its transfer to real active suspensions hinges on the ζ term surviving coarse-graining with the required sign; the review's own QSM lacks ζ and the ABP coarse-graining has uncontrolled approximations.","rationale":"The review is internally consistent: the AMB+ construction, the ψ-transformation for binodals, the sharp-interface derivation of σB,D, and the capillary-wave result σcw = (σB + σD)/2 are coherent, and the paper honestly reports open questions and numerical limitations. The reader's weakest assumption—that realistic suspensions reduce to a local, Markovian, constant-mobility, lowest-order scalar field theory—correctly identifies a real limitation, but I do not think it is the single load-bearing point. The more precise vulnerability is the ζ term: it is the sole source of negative Ostwald tensions in AMB+, yet its microscopic generality and sign are not established. The QSM counterexample and the 2D/3D asymmetry in repulsive ABP simulations already appear in the review, so this is not a fatal objection to the 'can' claim, but it does weaken the broader 'generic' narrative. Since the reader's verdict is UNVERDICTED and my concern is about transferability rather than an internal inconsistency, the appropriate disposition is unchanged: the review should be read as a well-posed theoretical framework whose microscopic reach remains open. Credit is due for the explicit comparisons with independent particle simulations (e.g., Caporusso et al. bubble statistics) and for flagging the very gaps I identify.","tokens_in":49733,"tokens_out":6500,"duration_ms":79105,"concrete_test":"Use the coarse-grained theory of Section IIIH for repulsive ABPs, with v(ρ) and p_C(ρ) measured from particle simulations, to compute the density-dependent coefficients K(ρ), λ(ρ), ζ(ρ), M(ρ). Then evaluate the generalized σB and σD (extending Eq. (21) to density-dependent coefficients) across the 2D and 3D MIPS parameter ranges. If σB remains positive in 3D while negative in 2D at comparable reduced parameters, the reverse-Ostwald claim is dimension-specific; if σB is positive everywhere, the negative-tension phenomenology does not transfer to repulsive ABPs even though the continuum model itself is valid.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The review's central synthesis is that active phase separation generically produces multiple interfacial tensions, some negative, leading to reverse Ostwald ripening and microphase separation. Within AMB+ this is internally consistent: Section IIE shows that negative σB,D arise only when the ζ term is present, and Figure 2 confirms the associated regimes numerically. The load-bearing step is the bridge from AMB+ to real systems. Section IIIA's bottom-up quorum-sensing model has no ζ term and gives σne > 0, so it cannot exhibit the new phenomenology. Section IIIH's coarse-graining of repulsive ABPs does produce a ζ term, but the authors state that the density-dependent coefficients 'do not allow for easily assessing their signs' and that the approach involves 'several uncontrolled approximations'. The review also notes in Section IIID that 3D repulsive ABP simulations report no vapour bubbles, suggesting the negative-σB regime may be dimension- or model-dependent. Thus the generic claim is not yet tied to a robust microscopic mechanism: if after a more careful coarse-graining ζ is absent, small, or of the wrong sign in a given system, the reverse-Ostwald and microphase predictions disappear even though the system is genuinely active. The constant-mobility and Markovian assumptions flagged by the reader are part of this issue, but the crux is the status of the ζ term, not the mobility prefactor alone.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This review article argues that phase separation in active matter is governed not by a single positive interfacial tension, as in equilibrium fluids, but by several distinct nonequilibrium tensions that control different phenomena and can individually become negative. The authors develop this claim within Active Model B+ (AMB+) for a single conserved scalar field, and its momentum-conserving counterpart Active Model H (AMH), showing how negative values of the Ostwald tensions σD, σB and of the capillary-wave tension σcw generate reverse Ostwald ripening, microphase separation, bubbly phase separation, and active foam states. They then survey the connection to particle-based models, emphasizing that only some microscopic coarse-graining procedures produce the crucial ζ term, and close with extensions to multiple species, nonconserved dynamics, flocking, and active nematics.","tokens_in":49965,"tokens_out":11184,"duration_ms":110222,"significance":"The review's central synthesis—that activity splits the equilibrium interfacial tension into independent nonequilibrium tensions, some of which can be negative—provides a coherent and falsifiable framework for a wide range of active phase separation phenomena. Its strengths include the explicit derivation of coexistence conditions and Ostwald dynamics from AMB+, the candid discussion of numerical and conceptual limitations (e.g., no d=3 AMB+ simulations, unresolved coarsening exponents, undetermined signs in the bottom-up ζ term, absence of vapour bubbles in 3D repulsive ABPs), and the careful comparison of field-theoretic predictions with particle-model and experimental observations. If the microscopic-transfer questions are resolved in the directions suggested by the paper, the review will have identified a generic, low-order route to finite-length-scale patterning in active matter that is distinct from Turing mechanisms. The paper is a review rather than a source of new derivations, but its pedagogical value and its honest demarcation of open problems justify publication after correction of the technical issues below.","major_comments":[{"comment":"The two expressions for the transformation ψ(φ) in Eq. (15) are inconsistent by an overall sign. With A=(ζ−2λ)φ/K, the first expression reads K/(ζ−2λ)(1−e^A), while the second reads K/(ζ−2λ)(e^A−1). Since Eqs. (17) and (21) rely on ψ, a reader using the second form would obtain binodals and interfacial tensions with the wrong sign. Please correct the second equality or, if that form is intended, adjust the first and verify the λ̄→0 limit.","section":"Sec. II.C, Eq. (15)"},{"comment":"Eq. (19) gives Ṙ=(d−1)Mσne/(R(Δφ)²)(1/Rs−1/R), but the derivation using (28), (31), and (32) yields a prefactor (d−1)(d−2), not (d−1). The sentence 'The factor (d−2) in eq. (19) and (32) was omitted in [177]' does not resolve the issue because (19) as printed still lacks the (d−2) factor. This alters the d-dependence of the Ostwald rate and is especially important because many numerical tests of the new states are performed in d=2, where the formula requires a logarithmic correction.","section":"Sec. II.E, Eq. (19); Sec. II.I, Eqs. (28)–(32)"},{"comment":"The abstract states as a generic property that 'the fluid-fluid interfaces created by active phase separation can have several distinct interfacial tensions... some of which can be negative.' The bottom-up evidence reviewed in Secs. IIIA, IIID, and IIIH is more guarded: the quorum-sensing model has no ζ term and gives σne>0; the coarse-graining of repulsive ABPs that does produce a ζ term has undetermined signs and 'several uncontrolled approximations'; and 3D repulsive ABP simulations do not show vapour bubbles. Please qualify the abstract and the introductory claim to distinguish the AMB+-internal prediction from the still-incomplete microscopic-transfer argument.","section":"Abstract and Sec. III"}],"minor_comments":[{"comment":"The phrase 'we discuss the impact of activity on nucleation in Sec. III and on interfacial roughening in Sec. IIJ' contains a wrong cross-reference: nucleation kinetics is treated in Sec. II.I, not Sec. III.","section":"Sec. II.E, paragraph before Sec. II.F"},{"comment":"There is a typo: 'is called Acitve Model B (AMB)' should read 'Active Model B'.","section":"Sec. II.A"},{"comment":"In the sentence 'andsσcw also becomes negative', the word 'ands' appears to be a typographical artifact; it should read 'and σcw'.","section":"Sec. II.G"},{"comment":"The text contains 'based on a a quasi-equilibrium approximation'; the doubled article 'a a' should be corrected.","section":"Sec. III.E"},{"comment":"The phrase 'borne out by particle-baseds models' contains a typo: 'particle-baseds' should be 'particle-based'.","section":"Sec. I.B"}],"recommendation":"minor_revision","confidential_remarks":"This is a strong, well-organized review that should be published after the local corrections. The sign inconsistency in Eq. (15) and the factor error in Eq. (19) are important enough to require author attention before typesetting; the abstract should also be aligned with the explicit caveats in Sec. III. I would not require new simulations or a new derivation, as the open issues are already disclosed by the authors."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is the review of active phase separation that Cates and Nardini were clearly meant to write. It is a thorough, well-organized synthesis of the AMB+/AMH framework, built around the key idea that activity breaks the degeneracy of interfacial tension: separate tensions govern Ostwald ripening, capillary waves, and mechanical forcing, and each can go negative. That central claim is internally consistent within AMB+, and the paper explains it about as clearly as it can be explained. It also does the right thing in flagging its own soft spots: unverified d=3 AMB+ simulations, unresolved coarsening exponents, and the barely analyzed active foam states. Credit is due for the pedagogical value of the field-transformation sections and the careful distinction between quasi-passive activity effects and genuinely TRS-breaking terms.\n\nThe soft spot is the bridge from AMB+ to real particles, and here the paper is necessarily more hand-wavy. The quorum-sensing model, which is the cleanest controlled coarse-graining, has no ζ term and gives positive σ_ne. The coarse-graining of repulsive ABPs does generate ζ, but the authors admit the density-dependent coefficients cannot be sign-assessed and the approximations are uncontrolled. And the 3D repulsive ABP simulations show no vapour bubbles, suggesting reverse Ostwald may be specific to two dimensions in those models. That does not kill the review's thesis—the phenomenology is real within minimal field theories, and the 2D ABP simulations do show matching bubble statistics—but it does mean the 'generic' claim is more a research program than an established fact. The authors themselves say the bottom-up connection is 'very much in its infancy,' which is the honest statement.\n\nThe paper leans heavily on the authors' own prior work. That is normal for a review by the field's architects, and they cite independent simulations and experiments where they exist, but a reader should be aware that the synthesis is also a promotion of the AMB+ paradigm.\n\nNet: this deserves a serious referee. It will be the standard entry point for graduate students and researchers moving into active phase separation, and it is honest enough about what is and is not established. I would send it out.","headline":"Authoritative synthesis of the multiple-interfacial-tension framework for active phase separation; the generic transfer to real systems rests on the ζ term, whose bottom-up status remains shaky.","tokens_in":50561,"tokens_out":1845,"would_cite":true,"duration_ms":20031,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This review argues that active phase separation is governed by several distinct interfacial tensions, some negative, which can reverse Ostwald ripening, destabilize capillary waves, and create microphase-separated and foam states.","keywords":["active matter","phase separation","detailed balance","interfacial tension","Ostwald ripening","microphase separation","Active Model B+","Active Model H"],"falsifier":"Measure, in the same active phase-separating suspension, the growth/evaporation dynamics of droplets and bubbles together with the relaxation of imposed interface height perturbations: if a single positive interfacial tension governs both processes and no regime shows reverse Ostwald ripening or capillary-wave destabilization as activity increases, the multiple-negative-tension picture is refuted.","tokens_in":49474,"feed_emoji":"🫧","tokens_out":4056,"duration_ms":45850,"temperature":0.7,"pith_summary":"The paper argues that, when microscopic detailed balance is broken, a fluid-fluid interface no longer has a single positive interfacial tension. Instead, different properties—Ostwald ripening, capillary-wave relaxation, mechanical forcing—are controlled by different tensions, and some of those can become negative while the interface remains stable. From this starting point, the review builds a unified account of phenomena that are impossible in equilibrium: reverse Ostwald ripening, microphase separation, bubbly phase separation, and active foams. These phenomena are shown to follow from minimal continuum field theories for a conserved scalar order parameter, extended by the lowest-order terms that break time-reversal symmetry. The review matters because it proposes that a wide range of active systems, from self-propelled colloids to biomolecular condensates and ecological patterns, share a generic nonequilibrium mechanism rather than requiring separate explanations for each.","feed_headline":"Negative tensions reverse Ostwald ripening in active fluids","feed_subtitle":"Without detailed balance, an interface carries several tensions, some negative, yielding microphases and foams.","key_machinery":"The workhorse is Active Model B+ (AMB+), the minimal extension of conserved Model B obtained by adding the two lowest-order time-reversal-symmetry-breaking terms, λ|∇ϕ|² and ζ(∇²ϕ)∇ϕ, to the diffusive current while keeping mobility and noise constant. The key identity is the decomposition of interfacial physics into separate tensions: σ_B and σ_D governing the Ostwald process for bubbles and droplets, σ_cw governing capillary waves, and σ_M governing mechanical forcing, with σ_cw given by the symmetric average of σ_B and σ_D for constant mobility. A nonlinear change of variables (ϕ to ψ) and a pseudo-pressure construction fix the binodals without solving the full interfacial profile, and a quasi-static sharp-interface ansatz yields the droplet growth law and capillary-wave dispersion. In the wet case, AMH couples the same diffusive current to a Navier-Stokes equation with an active deviatoric stress ∝ (K̃−K)S, whose sign controls the mechanical tension σ_M.","core_discovery":"The central claim is that activity, by breaking detailed balance at the microscale, splits the equilibrium concept of a single positive interfacial tension into several distinct nonequilibrium tensions. For a dry active system described by Active Model B+ (AMB+), the droplet and bubble Ostwald tensions σ_D and σ_B can each become negative, reversing the Ostwald process so that small droplets or bubbles grow rather than shrink; the capillary-wave tension σ_cw = (σ_B + σ_D)/2 can also become negative, destabilizing height fluctuations while the interface remains stable to normal perturbations. These negative tensions give rise, respectively, to microphase separation, bubbly phase separation, and the active foam state. In wet systems the momentum-conserving extension, Active Model H (AMH), introduces a further independent mechanical tension σ_M, which when negative causes droplets to split by self-shearing, producing additional microphase-separated steady states. The review argues that these effects are generic, and that minimal single-scalar field theories capture the essential phenomenology, with further new behaviour appearing in multi-species, non-conserved, and orientational systems.","pith_inferences":["Editorial inference: if the framework is correct, reporting a single 'surface tension' for an active interface is incomplete; tensions should be defined operationally per phenomenon, and mismatches between them are direct signatures of broken detailed balance.","Editorial inference: the same negative-tension logic should generalize to nonreciprocal multi-species mixtures and reaction-coupled condensates; the review's own NRCHM and Model AB+ extensions already show reverse-Ostwald-like mechanisms, suggesting a common principle beyond the single-scalar case.","Editorial inference: the theory implies that finite-size clusters in active colloids, currently often attributed to long-ranged interactions or arrests, could instead be steady states maintained by negative Ostwald tensions; this is testable by checking whether cluster size stays finite when system size is increased and whether bubbles nucleate inside dense clusters.","Editorial inference: a concrete experimental probe would be to measure the relaxation of an imposed sinusoidal height perturbation on an active interface and independently measure the growth rate of droplets; equality of the extracted tensions would support the equilibrium-like picture, whereas a discrepancy would support the multiple-tension scenario."],"forward_implications":["Where the nonequilibrium tension σ_D or σ_B is negative, Ostwald ripening is reversed and the system reaches a microphase-separated state of finite-size droplets or bubbles instead of full phase separation.","When σ_cw becomes negative, the liquid-vapour interface is linearly unstable to height fluctuations yet remains stable to normal ones, producing an active foam state with system-spanning liquid filaments.","Bubbly phase separation arises when a microphase-separated region at density φ_BL coexists with excess bulk vapour, with φ_BL lying above the mean binodal density; the bubble size distribution is set by competition between reverse Ostwald, nucleation, and coalescence.","In momentum-conserving Active Model H, a negative mechanical tension σ_M makes droplets split by self-shearing, and balancing this splitting against forward or reverse Ostwald sets a finite steady-state droplet size.","For the critical point of active phase separation between two uniform phases, the active nonlinearities are irrelevant to one-loop order, so the static critical exponents remain those of the equilibrium Model B/Ising class despite broken detailed balance.","The multiplicity of tensions means that measuring only one interfacial quantity in an active emulsion, for example the amplitude of capillary fluctuations, does not determine the coarsening or nucleation behaviour, which are governed by separate tensions.","Tests in particle simulations and experiments should look for the signature competition between reversed Ostwald, nucleation, and coalescence in bubble-size distributions, with near-monodisperse sizes when reverse Ostwald dominates.","For systems with multiple species or with chemical reactions, the same logic suggests that nonreciprocal interactions or birth-death dynamics can independently reverse Ostwald or stabilize finite domains, potentially producing hierarchical microphase-separated structures."],"supporting_citations":[{"why":"Provides the Hohenberg-Halperin classification of equilibrium dynamical models, the template that AMB+ and AMH extend.","marker":"[151]"},{"why":"Introduces AMB+, the λ and ζ terms, the pseudo-pressure transformation, and the reverse Ostwald mechanism.","marker":"[161]"},{"why":"Derives the capillary-wave tension σ_cw = (σ_B+σ_D)/2, the instability condition σ_cw<0, and the resulting microphase and foam states.","marker":"[166]"},{"why":"Gives the quasi-static sharp-interface derivation of the droplet growth law and nucleation kinetics for AMB+.","marker":"[177]"},{"why":"Introduces Active Model H and the self-shearing droplet-splitting instability driven by negative σ_M.","marker":"[228]"},{"why":"Supplies the equilibrium Ostwald and coarsening baseline, the zero-temperature fixed point, and the t^{1/3} law that active theories must reproduce or modify.","marker":"[152]"}],"fun_headline_variants":["Activity splits interfacial tension, enabling reverse Ostwald ripening","Negative tensions in active fluids reverse Ostwald ripening","Microphase separation from negative interfacial tensions","When interfacial tension goes negative, new phases emerge","Multiple interfacial tensions, some negative, reshape active phases"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole edifice rests on the premise that realistic active suspensions can be described, at the scales of interest, by a local Markovian field theory of a single conserved density with constant mobility, white noise, and only the lowest-order gradient terms; if density-dependent mobility, memory, or additional slow fields are generically important, the specific predictions about negative tensions and reverse Ostwald need not transfer to real systems.","fun_headline_variants_meta":{"raw":{"variants":["Activity splits interfacial tension, enabling reverse Ostwald ripening","Negative tensions in active fluids reverse Ostwald ripening","Microphase separation from negative interfacial tensions","When interfacial tension goes negative, new phases emerge","Multiple interfacial tensions, some negative, reshape active phases"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000461,"raw_usage":{"total_tokens":2360,"prompt_tokens":1053,"completion_tokens":1307,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":669,"completion_tokens_details":{"reasoning_tokens":1234}},"tokens_in":669,"tokens_out":1307,"duration_ms":9922,"temperature":1.0,"reasoning_tokens":1234,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:01:28.197044+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure, in the same active phase-separating suspension, the growth/evaporation dynamics of droplets and bubbles together with the relaxation of imposed interface height perturbations: if a single positive interfacial tension governs both processes and no regime shows reverse Ostwald ripening or capillary-wave destabilization as activity increases, the multiple-negative-tension picture is refuted.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces Active Model H and the self-shearing droplet-splitting instability driven by negative σ_M."}],"review_version":1}