{"id":"d9a5ad56-f9b4-4b40-b3a7-88a498510402","arxiv_id":"2608.07450","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For the active Cahn-Hilliard model, the paper derives a curvature-dependent coarsening theory predicting a crossover from t^{1/3} to t^{1/4} growth followed by saturation, and proves well-posedness results for a finite element scheme.","lead":"Active particles that consume energy can separate into dense and dilute phases, and this paper derives how the size of those domains grows and then stops growing. It also proves stability and convergence of a numerical scheme used to simulate the process.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The saturation prediction rests on a polynomial heteroclinic ansatz and a second-order expansion in λ and R^{-1}, yet the only numerical confirmation (λ=2) saturates at R≈3.65, where both parameters are O(1) and the R>>1 assumption fails exactly in the confirming regime.","rationale":"The reader's weakest assumption is the polynomial heteroclinic ansatz, and I agree that this is the most load-bearing point. Section 2.3 and Appendix A derive the growth equation (2.36) by polynomial identity under (2.11)/(A.1), with no existence or approximation theorem establishing that actual static droplet heteroclinics are polynomial. The internal limitation in Section 5.3 concedes that saturation is only simulated for λ=2, while Table 1 gives Rbar≈3.65 for λ=2, violating both R>>1 and small-λ. This makes the numerical confirmation weak in precisely the regime where the perturbative expansion is uncontrolled. The PDE and finite-element analysis in Sections 3-4 appears independent and self-contained, so the paper retains value; however, the headline coarsening prediction remains conditional. I therefore keep the reader's CONDITIONAL verdict unchanged.","tokens_in":54085,"tokens_out":9092,"duration_ms":94958,"concrete_test":"Solve the static droplet two-point boundary-value problem (2.24)-(2.27) by numerical continuation in λ and ξ from the λ=0 tanh profile, for example at λ=0.1 and ξ=0.1, and compare the computed heteroclinic curve χ(f) with the truncated polynomial expression (A.1)-(A.2). If the relative L2 error is not within the nominal O(λ²+ξ²) truncation error, the polynomial identity step underlying (2.36) is invalid. If, instead, the polynomial ansatz survives this check at small parameters, repeat the comparison at the confirming parameters λ=2, R=3.65; a large deviation there would show that the numerical saturation in Figure 5 does not test the theoretical expansion in its controlled regime.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central coarsening claim of Section 2.3 is Equation (2.36), whose coefficients are obtained by inserting the polynomial heteroclinic ansatz (2.11) and its quartic analogue (A.1) into the static droplet ODE and applying polynomial identity. No argument is given that the true heteroclinic trajectory is polynomial, and every subsequent quantity—μ_{s,1}=4/15, the curvature correction ξ=√8/5|λ|, the surface-tension integrals σ and β, and the saturation root Rbar—depends on that ansatz. Thus the predicted crossover to z=4 and the finite-time saturation are only as secure as the ansatz. This is not a purely formal concern: the numerical confirmation in Section 5.3 is performed only at λ=2, where Table 1 gives Rbar≈3.65, so ξ=2/R≈0.55 and λ=2 are both O(1). The R>>1 assumption used to justify the quasi-static reduction (2.24) and the f(r−R) droplet profile is violated at exactly the parameter value used to validate the theory. The paper itself states that saturation for λ<2 is too computationally demanding to simulate, so no direct numerical evidence exists in a regime where the perturbative expansion is controlled. If the true heteroclinic structure differs from the polynomial ansatz, or if O(λ³, λ²ξ, ...) corrections change the root structure of (2.36), the claimed plateau may disappear or occur at a very different length.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the active Cahn–Hilliard equation (active Model B), Eq. (1.1), with regular and singular double-well potentials. In Section 2 the authors introduce a phase-plane method based on assumed polynomial heteroclinic trajectories to derive static kink and droplet properties, obtaining μ_{s,1}=4/15, the droplet curvature parameter ξ=√(8/5)|λ|, and an ODE (2.36) for the characteristic droplet radius. From this ODE they predict a crossover in domain growth from L(t)∼t^{1/3} to L(t)∼t^{1/4}, followed by finite-time saturation of the domain length, with numerical evidence reported in Section 5.3. Sections 3–4 contain formal a-priori estimates, a finite element scheme, stability and convergence analysis, and, for the singular potential, local-in-time existence and uniqueness in d=2,3 and global well-posedness in d=1 under a smallness condition on λ. The central coarsening claim, however, rests on a formal polynomial heteroclinic ansatz and on a quasi-static R≫1 reduction, and the only direct numerical confirmation is performed at λ=2, where the asymptotic assumptions are not satisfied.","tokens_in":54465,"tokens_out":4728,"duration_ms":52201,"significance":"If the coarsening prediction is correct, it would resolve an open debate in the active matter literature: active Model B would exhibit a qualitatively different late-time behavior from passive Model B, with the characteristic domain size saturating at a finite length instead of following Lifshitz–Slyozov growth indefinitely. The rigorous parts of the paper are substantial: the finite element analysis in Sections 3–4, including the discrete Gagliardo–Nirenberg inequality of Lemma 4.4, the discrete Bihari-type inequality of Lemma 4.5, the existence and uniqueness result of Theorem 4.10, and the global one-dimensional well-posedness of Theorem 3.2 are careful and valuable contributions. The significance of the paper as a whole, however, is conditional on the coarsening theory, which is currently formal and validated only in a parameter regime where its own assumptions fail.","major_comments":[{"comment":"The central derivation assumes, without proof, that the heteroclinic trajectory connecting the two saddle points exists and, at every perturbative order, has the polynomial form (2.11), (2.20), or (A.1). All subsequent quantities—μ_{s,1}=4/15, ξ=√(8/5)|λ|, the surface-tension integrals σ and β, and the droplet-radius ODE (2.36)—are obtained by inserting this ansatz and applying polynomial identity. No argument establishes that the true heteroclinic structure is polynomial, and no error estimate controls the truncation at order O(λ², ξ², λξ). If the true heteroclinic differs from the ansatz, the predicted crossover and saturation could disappear or occur at a different length scale. I ask the authors either to provide a proof of the polynomial form, or to verify the derived coefficients by direct numerical solution of the ODEs (2.7) and (2.24) in the phase plane, or to present an explicit error bound showing the neglected terms cannot change the root structure of Eq. (2.36).","section":"Section 2.1–2.3, Eqs. (2.11), (2.20), (A.1)"},{"comment":"The quasi-static droplet picture relies on R≫1: the factor 1/r is approximated by ξ=2/R, the profile is written as f(r−R), and the interface boundary conditions are imposed at ±∞. However, the numerical confirmation in Section 5.3 and Table 1 is performed only at λ=2, where the predicted saturation radius is Rbar≈3.65. At this value ξ=2/R≈0.55 and λ=2 are both O(1), so the expansion parameters are not small in exactly the regime used for validation. The paper itself states that saturation for λ<2 is too computationally demanding to simulate. Consequently, the numerical evidence does not test the asymptotic expansion in any controlled regime. Please provide evidence for small λ, or a direct numerical test of the full droplet ODE without the R≫1 reduction, or a quantitative estimate of the neglected O(λ³, λ²ξ, λξ²) terms.","section":"Section 2.3, Eqs. (2.24), (2.36); Table 1"},{"comment":"The saturation of L(t) is inferred from an ensemble of 20 runs in a 256² domain at λ=2, but no confidence intervals or ensemble spread are shown, and the saturation radius Rbar≈3.65 is of the same order as the interface width. In this regime the measured L(t) may be affected by finite-size effects, by the small number of droplets of size comparable to the interface thickness, or by the resolution of the structure-factor computation. The claim that the plateau is the saturation predicted by Eq. (2.36) would be much stronger if the time window were extended, if smaller λ values were accessed, or if the same plateau were observed for different domain sizes with convergence in the numerical parameters.","section":"Section 5.3, Figure 5"},{"comment":"Remark 2.1 admits that including the O(λ³) term changes the coefficient ξ0 and hence the values in Table 1, but asserts without proof that the qualitative behavior is unchanged. Since Eq. (2.36) is only a second-order expansion in (λ, ξ), there is no remainder estimate showing that higher-order terms cannot remove the positive root Rbar or move it to a completely different scale. This is load-bearing because the finite-time crossover and the saturation plateau are both read off from the root structure of this truncated ODE. Please provide either a rigorous remainder bound or a numerical check that the root Rbar is stable when additional terms are included.","section":"Section 2.3, Remark 2.1"}],"minor_comments":[{"comment":"The caption of Figure 3 lists parameter values λ=0,1,2 while the text in Section 5.1 says λ ranges over {0,0.1,1,2}; please clarify which values are actually shown and label the rows consistently.","section":"Figure 3 vs. Section 5.1"},{"comment":"There are several typographical errors: 'Lifshit-Slyozov' in the Introduction should be 'Lifshitz–Slyozov', 'spyral' in Section 2.2 should be 'spiral', 'tat' in Section 3.2 should be 'that', and 'Bihary' in Section 4.1 should be 'Bihari'.","section":"Throughout"},{"comment":"The singular potential is regularized with α=0.001, but no sensitivity study with respect to α is reported; a sentence justifying this choice or showing robustness in α would be useful.","section":"Section 5.2"},{"comment":"The abstract states that the method 'recovers the exact values of key quantities' and that the power-law shift is 'explained'; given the formal nature of the heteroclinic ansatz, the wording should be softened to indicate that the values are obtained within the assumed polynomial ansatz.","section":"Abstract and Introduction"}],"recommendation":"major_revision","confidential_remarks":"The rigorous numerical analysis portions (Sections 3–4) are solid and publishable on their own. My main reservation is that the paper's headline claim—the coarsening crossover and saturation—is supported only by a formal ansatz and by numerical evidence in a regime where the expansion parameters are O(1). This is a load-bearing issue that cannot be resolved by editorial changes alone. If the authors can either justify the polynomial heteroclinic ansatz, supply a controlled numerical test at small λ, or provide rigorous error estimates for the truncated ODE, the paper would be acceptable; otherwise the central claim should be substantially downgraded in presentation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know about arXiv:2608.07450. First, the rigorous numerical analysis is real: they build a fully discrete finite element scheme with lumped mass for the active Cahn-Hilliard equation, prove well-posedness and conditional stability for regular and singular potentials, and obtain convergence to a local-in-time weak solution in d=2,3 with a singular potential, plus global well-posedness in d=1 under a smallness condition. The discrete Gagliardo-Nirenberg inequality and the discrete Bihari lemma are new tools. That part is careful and deserves a serious referee. Second, the coarsening theory in Section 2.3 is a formal asymptotic construction, not a proof. It gives a concrete ODE for the droplet radius, Eq. (2.36), and predicts a crossover from z=3 to z=4 and then saturation. The derivation has no fitted parameters, which is nice, but it rests on two assumptions: the heteroclinic trajectory is a polynomial at each order, and the droplet radius satisfies R>>1. Neither is justified. The paper calls the polynomial ansatz a 'guess' in Section 2.1, and everything downstream—the surface tension correction, the saturation radius—depends on it.\n\nYour stress-test note is on target. The only numerical confirmation of saturation is at lambda=2, where Table 1 gives Rbar approx 3.65. So xi = 2/R approx 0.55 and lambda = 2 are both O(1), and the R>>1 assumption used in the quasi-static reduction fails exactly where the theory is tested. The paper itself says in Section 5.3 that saturation for lambda<2 is computationally too demanding to simulate. That is an honest statement, but it means the central claim is not yet backed by numerics in a controlled regime. If the true heteroclinic is not polynomial, the saturation may disappear or move. This does not sink the paper, but it does mean the coarsening part should be read as a conjecture with a plausible mechanism, not a confirmed result.\n\nWho should read this? People working on active Model B or on finite element methods for fourth-order phase-field equations. The numerical analysis half is solid and citable. The coarsening half will be useful if it later gets a proof or better numerical support. The paper deserves peer review—the rigorous results alone justify it. The referees should push the authors to either prove the ansatz in a simpler case or run smaller-lambda simulations, and to clearly mark the saturation prediction as conditional.\n\nRecommendation: send it to review, but expect the coarsening section to need revision or a sharpened caveat.","headline":"Rigorous FE analysis plus a formal, unproven coarsening theory—worth refereeing, but the saturation prediction is not confirmed by the numerics.","tokens_in":54928,"tokens_out":3289,"would_cite":true,"duration_ms":34557,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35K55","35K35","65M60","35B40"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that in the active Cahn–Hilliard model, activity and interface curvature together renormalize surface tension, so coarsening crosses over from $L(t)\\sim t^{1/3}$ to $L(t)\\sim t^{1/4}$ and then saturates at a finite…","keywords":["active Cahn–Hilliard equation","active Model B","phase separation","coarsening dynamics","Lifshitz–Slyozov growth","heteroclinic trajectories","surface tension","finite element method"],"falsifier":"For a concrete check, compute the static spherical droplet solution of the active Cahn–Hilliard equation numerically to high precision at $\\lambda=1$, where the paper's Table 1 predicts a saturation radius $\\bar R\\approx 95.24$, and compare the interface curvature $\\xi=2/R$ with $\\sqrt{8}|\\lambda|/5$; a significant deviation would indicate that the polynomial heteroclinic ansatz is incomplete. Alternatively, simulate two-dimensional coarsening for $\\lambda=1$ well beyond the time at which the solution of the droplet-radius ODE reaches its plateau and test whether $L(t)$ continues growing as $t^{1/4}$ instead of levelling off near the predicted saturation value.","tokens_in":53868,"feed_emoji":"🧫","tokens_out":6108,"duration_ms":60614,"temperature":0.7,"pith_summary":"The paper studies the active Cahn–Hilliard equation (active Model B), a diffuse-interface model for motility-induced phase separation, and argues that activity changes coarsening qualitatively. Using phase-plane heteroclinic trajectories instead of the earlier Newton-mapping construction, it recovers exact static-kink and spherical-droplet quantities and derives how activity plus local interface curvature renormalize surface tension. The central prediction is a crossover in domain growth $L(t)\\sim t^{1/z}$ from the passive Lifshitz–Slyozov value $z=3$ to $z=4$, followed by saturation at a finite length controlled by $1/\\lambda$. The paper also constructs a lowest-order finite element scheme, proves local-in-time existence and uniqueness of weak solutions in dimensions $d=2,3$ and global well-posedness in $d=1$ under a smallness condition on activity, and reports two-dimensional simulations showing the $z=3\\to 4$ crossover before saturation.","feed_headline":"Coarsening in active fluids shifts to t^{1/4}, then stalls","feed_subtitle":"A new phase-plane derivation predicts activity caps domain size at a finite length, altering the classic Ostwald growth law.","key_machinery":"The central object is the phase-plane heteroclinic trajectory: static planar kinks and large spherical droplets are viewed as orbits connecting saddle equilibria in the $(\\phi,\\chi)$ plane, and the orbit is sought as a polynomial in $(\\phi-\\phi_a)(\\phi-\\phi_c)$ at each order in the small parameters $\\lambda$ and $\\xi=(d-1)/R$. Imposing polynomial identity in the first-order ODE system yields exact coefficients, including $\\mu_{s,1}=4/15$ and $\\xi=\\sqrt{8}|\\lambda|/5$, and, at second order, explicit corrections to the surface tension $\\sigma$ and the interfacial moment $\\beta$ that enter the interface balance. This machinery converts activity and curvature into the modified coefficients of the droplet-radius ODE, from which the growth-law crossover and the saturation length follow.","core_discovery":"The paper claims that the active term $\\lambda|\\nabla\\phi|^2$ modifies the effective surface tension both through a constant activity-induced shift and through a curvature-dependent correction, so the usual Ostwald-ripening balance is altered. When the shrinking-droplet ODE for the typical droplet radius $R_0(t)$ is derived to second order in $\\lambda$ and $1/R$, the metastable radius $\\bar R(t)$, instead of diverging as supersaturation vanishes, approaches a finite positive value: coarsening crosses from $R_0\\sim t^{1/3}$ to $R_0\\sim t^{1/4}$ and then plateaus. This is offered as an explanation of the power-law crossover conjectured in the literature and as evidence that active Model B saturates at a finite domain size, unlike passive Model B, whose domains grow without bound. The same phase-plane method also fixes the static kink chemical potential $\\mu_{s,1}=4/15$ and the spherical-droplet curvature $\\xi=\\sqrt{8}\\,|\\lambda|/5$, previously obtained only approximately by Newton mapping.","pith_inferences":["A testable extension: if saturation is real, the final domain size for fixed $\\lambda$ should be independent of the initial supersaturation and scale roughly as $\\lambda^{-1}$; measuring $L(\\infty)$ for several $\\lambda$ values would separate genuine saturation from extremely slow growth.","The polynomial heteroclinic ansatz is stated perturbatively, so the exact static droplet solution at finite $\\lambda$ may be transcendental; a rigorous existence proof for that droplet would be needed to make the predicted saturation length quantitative beyond leading order.","The same second-order surface-tension machinery could be applied to active Model B+ or to droplets in other geometries, yielding analogous saturation lengths and providing a template for curvature-aware coarsening theories in other non-equilibrium phase-separating systems.","The numerical observation that the crossover time decreases with $\\lambda$ suggests a possible scaling collapse of $L(t)/L_{\\mathrm{sat}}$ against $t\\lambda^\\alpha$, though the paper does not propose such a collapse."],"forward_implications":["If the central claim is correct, active Model B does not follow passive Lifshitz–Slyozov growth indefinitely: the characteristic domain size stops growing at a finite length set by the activity parameter.","The observed $z=4$ growth is a finite-time transient, not the true asymptotic law; simulations that see $L(t)\\sim t^{1/4}$ are consistent with the system approaching a plateau rather than growing forever.","The static-kink chemical potential $\\mu_{s,1}=4/15$ and the droplet curvature $\\xi=\\sqrt{8}|\\lambda|/5$ provide exact reference values that earlier Newton-mapping treatments had only approximated numerically.","In the singular-potential case, the finite element analysis proves that discrete solutions remain in $(-1,1)$, giving local-in-time weak solutions in $d=2,3$ and global-in-time solutions in $d=1$ under small activity, which supports reliable late-time simulations.","Regular and singular potentials are shown numerically to produce qualitatively identical phase-separation dynamics, so polynomial-potential simulations are indicative of the behavior with the physically motivated Flory–Huggins potential."],"supporting_citations":[{"why":"Introduces the active Cahn–Hilliard equation, proves static kink and droplet existence by Newton mapping, and supplies the numerical results and uncommon-tangent construction that the heteroclinic method here recovers exactly.","marker":"[36]"},{"why":"Conjectures the $z=3$ to $z=4$ crossover in domain growth and explicitly asks for a curvature-aware surface-tension theory, providing the numerical evidence the present theory explains.","marker":"[27]"},{"why":"Provides the interface-dynamics and Ostwald-ripening framework, including the shrinking-droplet balance and the definition of $R_0(t)$, on which the droplet-radius ODE is built.","marker":"[6]"},{"why":"Gives the Lifshitz–Slyozov growth law $R_0\\sim t^{1/3}$ that the active model is shown to depart from.","marker":"[24]"},{"why":"Demonstrates reverse Ostwald ripening and saturation in active Model B+, the analogous saturation phenomenon that the paper's results mirror for active Model B.","marker":"[34]"},{"why":"Provides the active Model B+ saturation result that the paper cites as a precedent for a finite characteristic domain size caused by activity.","marker":"[10]"},{"why":"Supplies the phase-plane heteroclinic-trajectory technique for bistable traveling waves that is adapted here to static kinks and droplet profiles.","marker":"[21]"},{"why":"Gives the structure-factor definition used to extract the numerical domain length $L(t)$ in the coarsening simulations.","marker":"[1]"}],"fun_headline_variants":["Active coarsening: t^{1/3} to t^{1/4}, then finite size stalls","Phase-plane method predicts active coarsening stall","Activity shifts Cahn-Hilliard coarsening law to t^{1/4}","Active surface tension caps domain growth at finite size","Cahn-Hilliard activity: from Ostwald growth to saturation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The coarsening derivation assumes that the static spherical droplet profile can be represented as a polynomial heteroclinic orbit in $(\\phi-\\phi_a)(\\phi-\\phi_c)$ at each order in $\\lambda$ and $1/R$, with the droplet radius large compared with the interface width; if the true heteroclinic orbit is not of that polynomial form, the derived surface-tension corrections and the predicted saturation length do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Active coarsening: t^{1/3} to t^{1/4}, then finite size stalls","Phase-plane method predicts active coarsening stall","Activity shifts Cahn-Hilliard coarsening law to t^{1/4}","Active surface tension caps domain growth at finite size","Cahn-Hilliard activity: from Ostwald growth to saturation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000288,"raw_usage":{"total_tokens":1775,"prompt_tokens":1120,"completion_tokens":655,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":736,"completion_tokens_details":{"reasoning_tokens":572}},"tokens_in":736,"tokens_out":655,"duration_ms":6328,"temperature":1.0,"reasoning_tokens":572,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T04:26:45.852987+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a concrete check, compute the static spherical droplet solution of the active Cahn–Hilliard equation numerically to high precision at $\\lambda=1$, where the paper's Table 1 predicts a saturation radius $\\bar R\\approx 95.24$, and compare the interface curvature $\\xi=2/R$ with $\\sqrt{8}|\\lambda|/5$; a significant deviation would indicate that the polynomial heteroclinic ansatz is incomplete. Alternatively, simulate two-dimensional coarsening for $\\lambda=1$ well beyond the time at which the solution of the droplet-radius ODE reaches its plateau and test whether $L(t)$ continues growing as $t^{1/4}$ instead of levelling off near the predicted saturation value.","supporting_citations":[{"cited_title":"Wittkowski, A","cited_arxiv_id":null,"evidence_quote":"Introduces the active Cahn–Hilliard equation, proves static kink and droplet existence by Newton mapping, and supplies the numerical results and uncommon-tangent construction that the heteroclinic method here recovers exactly."},{"cited_title":"Pattanayak, S","cited_arxiv_id":null,"evidence_quote":"Conjectures the $z=3$ to $z=4$ crossover in domain growth and explicitly asks for a curvature-aware surface-tension theory, providing the numerical evidence the present theory explains."},{"cited_title":"Bray, Theory of phase ordering kinetics, Adv","cited_arxiv_id":null,"evidence_quote":"Provides the interface-dynamics and Ostwald-ripening framework, including the shrinking-droplet balance and the definition of $R_0(t)$, on which the droplet-radius ODE is built."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the Lifshitz–Slyozov growth law $R_0\\sim t^{1/3}$ that the active model is shown to depart from."},{"cited_title":"Tjhung, C","cited_arxiv_id":null,"evidence_quote":"Demonstrates reverse Ostwald ripening and saturation in active Model B+, the analogous saturation phenomenon that the paper's results mirror for active Model B."},{"cited_title":"I: Cellular Physiology, Springer New York, NY, ISSN 0939-6047, 2010","cited_arxiv_id":null,"evidence_quote":"Supplies the phase-plane heteroclinic-trajectory technique for bistable traveling waves that is adapted here to static kinks and droplet profiles."},{"cited_title":"Agosti, P .F","cited_arxiv_id":null,"evidence_quote":"Gives the structure-factor definition used to extract the numerical domain length $L(t)$ in the coarsening simulations."}],"review_version":1}