Weiss monotonicity and capillary hypersurfaces

Chao Li; Nick Edelen; Otis Chodosh

REVIEW 1 major objections 7 minor 1 cited by

Weiss monotonicity and capillary hypersurfaces

T0 review · 1 major / 7 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper proves that capillary surface density, renormalized by the square of the contact angle, converges to the Weiss energy of the limiting one-phase Bernoulli problem, yielding curvature bounds whose constants depend only on the…

desk verdict The paper proves a genuinely new convergence theorem for capillary densities to the Weiss energy and uses it to derive angle-independent curvature estimates; the main proof is coherent, but one step in Theorem 1.2 leans on a published proposition whose boundary-regularity content needs verification. read the letter →

arxiv 2506.02146 v1 pith:MEEP6C2J submitted 2025-06-02 math.AP math.DG

Otis Chodosh , Nick Edelen , Chao Li This is my paper

classification math.APmath.DG MSC 35R3549Q1053A1035B65

keywords capillarysurfacesAlt–Caffarellifunctionalone-phaseBernoulliproblemWeissmonotonicityformuladensityratiocurvatureestimatesBernstein-typetheoremfreeboundaryregularity

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

Capillary surfaces are mathematical models of fluid interfaces meeting a container wall at a fixed contact angle $\theta$. Earlier work of the same authors showed that as $\theta$ tends to zero, these surfaces can be written as graphs over the container wall and converge, after rescaling by $\theta$, to minimizers of the Alt–Caffarelli energy, the one-phase Bernoulli problem. This paper establishes that in the same limit the capillary area-density ratio, renormalized by $\theta^{-2}$, converges to the Weiss energy density of the limiting problem. That convergence is then used to prove that if a capillary minimizer's density is everywhere close to the flat half-plane value, the interface curvature is bounded by a constant times $\sin\theta$, with the constant depending on the dimension alone. Consequently, global minimizers satisfying the same near-minimal density condition must be capillary half-planes.

What carries the argument

The load-bearing object is the pair of monotone quantities and their regularized versions. For the capillary problem, the varifold $V = [\partial\Omega\cap \mathbb{R}^{n+1}_+] - \cos\theta\,[\partial\Omega\cap \partial\mathbb{R}^{n+1}_+]$ has a density ratio $\Theta_V(x,r) = \|V\|(B_r(x))/(\omega_n r^n)$, monotone in $r$ for boundary-plane points. For the Alt–Caffarelli functional, the Weiss energy $W_v(x,r) = r^{-n}\int_{\{v>0\}\cap B_r(x)}(|Dv|^2+1)\,dy - r^{-n-1}\int_{\partial B_r(x)} v^2\,d\sigma$ is monotone in $r$ for stationary solutions. A smooth cutoff $\zeta$ regularizes both quantities without changing their asymptotics up to the factor $(1-\varepsilon)^n$. The key computation expresses $\omega_n r_i^n\Theta^\zeta_{V_i}(x_i,r_i)$ as $\theta_i^2/2$ times the regularized Weiss integrand in the graphical coordinates $u_i$, plus controlled $O(\theta_i)$ errors; renormalizing by $\theta_i^{-2}$ and taking the limit produces the identity of Theorem 1.1.

What would settle it

Compute the two sides of the limit identity for a family of small-angle capillary minimizers whose limiting Alt–Caffarelli free boundary is a non-flat cone: the theorem predicts that $\theta_i^{-2}\Theta_{V_i}(x_i,r_i)$ converges to the cone's Weiss energy divided by $2\omega_n$, so a discrepancy would refute Theorem 1.1.

Watch

Extended reading notes

Core claim

The central discovery is a limit identity between the monotone quantities of two variational problems. Let $\Omega_i$ be smooth minimizers of the capillary functional $A_{\theta_i}(\Omega) = \mathcal{H}^n(\partial^*\Omega\cap \mathbb{R}^{n+1}_+) - \cos\theta_i\, \mathcal{H}^n(\partial^*\Omega\cap \partial\mathbb{R}^{n+1}_+)$ with $\theta_i\to 0$, and let $V_i = [\partial\Omega_i\cap \mathbb{R}^{n+1}_+] - \cos\theta_i\,[\partial\Omega_i\cap \partial\mathbb{R}^{n+1}_+]$ be the associated capillary varifolds. If $v$ is the limiting Alt–Caffarelli minimizer obtained by the authors' earlier convergence theorem, Theorem 1.1 asserts that $\theta_i^{-2}\Theta_{V_i}(x_i,r_i)\to \frac{1}{2\omega_n} W_v(x,r)$ for $x_i\to x$ in the boundary plane and $r_i\to r$, where $\Theta_{V_i}$ is the density ratio and $W_v$ is the Weiss energy. The proof approximates both quantities by cutoff versions, expands the capillary area in the graphical coordinates $u_i$ using $\operatorname{Lip}(u_i)\le c(n)\theta_i$, and passes to the limit. This identity is the engine for the paper's applications: a priori curvature bounds $|A_M(x)|\le c(n)\sin\theta$ under a density-closeness hypothesis, and the Bernstein-type classification of global minimizers with near-minimal density as capillary half-planes.

Load-bearing premise

The main convergence and curvature estimates depend on the prior regularity result that, for small contact angles, the capillary interface can be written as a function over the container wall whose slope is bounded by a constant times $\theta_i$; if that bound failed or had a different order in $\theta_i$, the renormalized density would not converge to the Weiss energy.

Editorial extensions

If this is right

If a capillary minimizer's density ratio is at every boundary point and radius at most $(1+\varepsilon)(1-\cos\theta)/2$ (up to the negative part of $\cos\theta$), then the curvature of its interface is bounded by $c(n)\sin\theta$ near the boundary, with constants independent of the angle.
A global smooth capillary minimizer satisfying the same near-minimal density condition is necessarily a capillary half-plane: the scale-invariant curvature bound forces the second fundamental form to vanish after rescaling.
When the limiting Alt–Caffarelli minimizer $v$ is regular at a free-boundary point, the rescaled capillary surfaces converge to $v$ in $C^{2,\alpha}$ near that point, interpreted through Hodograph transforms.
In the small-angle blow-up argument, density closeness combined with the Weiss monotonicity formula forces the limiting Alt–Caffarelli minimizer to be linear, which turns the curvature blow-up assumption into a contradiction.

Reading between the lines

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

The same renormalized convergence is likely to hold for complements, with $\theta$ replaced by $\pi-\theta$, following the orientation symmetry noted in Remark 1.4; testing this on explicit capillary half-planes with obtuse angles would be a direct check.
Because the proof uses only the Lipschitz bound and Hausdorff convergence of free boundaries, the identity may extend to sequences of almost-minimizers or to weighted variants of the Alt–Caffarelli functional; the paper does not assert this.
A quantitative refinement seems available: for regular limiting free boundaries, Weiss monotonicity could bound the rate of convergence in the identity, upgrading the curvature estimate to an explicit radius-dependent bound; the paper does not pursue rates.
The density-closeness hypothesis is an $L^\infty$-type condition on a monotone quantity, so a natural testable question is whether the curvature bound survives under an averaged or integral version of the condition, since monotonicity might enforce pointwise closeness at nearby scales.

Share X Bluesky LinkedIn Reddit HN

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 main result here is Theorem 1.1: for capillary minimizers with contact angle θ_i → 0, the renormalized density θ_i^{-2} Θ_{V_i}(x_i, r_i) converges to (1/(2ω_n)) W_v(x, r), the Weiss energy density of the limiting Alt–Caffarelli minimizer. That is a clean, new statement, and the proof is careful. The regularized density expansion is handled with explicit error terms, and the limiting argument via the known Lipschitz bound and the monotonicity of both quantities is sound. This gives a real bridge between capillary monotonicity and one-phase free-boundary monotonicity, and it is a nice step beyond the authors' earlier convergence results.

The applications are exactly what you would hope for: angle-independent curvature estimates (Theorem 1.2), a Bernstein-type theorem (Corollary 1.6), and C^{2,\alpha} regularity in the sense of Hodograph transforms (Corollary 1.7). The curvature estimate, with constants depending only on dimension, would be a major advance if it fully holds. The proof of Theorem 1.2 is a standard contradiction–compactness argument, and most of it reads well. The cone classification lemma is routine but usefully reproduced.

Now the soft spot. In Case 1a of Theorem 1.2, after rescaling, the point x'_i has bounded distance to the free boundary, so the projected point can land on ∂{v>0} in the limit. The proof asserts convergence of θ_i^{-1}|A_{M'_i}(x'_i)| to |D^2v(0)|, using the improved convergence of [1, Prop 4.11]. But if v is the half-plane solution (y·n)_+, it is not C^2 across the free boundary, so ordinary C^{2,α} convergence of the graph functions cannot hold at boundary-approaching points. The paper does not state whether [1, Prop 4.11] provides convergence of the second fundamental form in the Hodograph sense up to the free boundary, and Corollary 1.7's caveat suggests the ordinary statement is only away from it. If the proposition does not cover boundary-approaching points, the contradiction in Case 1a does not go through as written. This is a verification concern about a published dependency, not an observed falsehood; someone should read the statement of [1, Prop 4.11] carefully before trusting Theorem 1.2.

The paper is for specialists in geometric measure theory and free-boundary regularity. The convergence theorem is solid and deserves to be known. The curvature estimate is important but currently rests on a subtle external dependence. I would send this to a serious referee, and specifically ask them to check that dependency.

Recommendation: engage with it. The core ideas are good, the main new theorem is real, and the soft spot is a checkable point.

Referee Report

1 major / 7 minor

Summary. This paper continues the authors' study of the small-angle limit of minimizing capillary hypersurfaces in a half-space. The main result, Theorem 1.1, states that for a sequence of capillary minimizers with contact angle θ_i → 0, the renormalized capillary area density θ_i^{-2}Θ_{V_i}(x_i,r_i) converges to (1/(2ω_n)) W_v(x,r), the Weiss energy density of the limiting Alt–Caffarelli minimizer. The authors use this convergence to prove an a priori curvature estimate (Theorem 1.2) under the assumption that the capillary density is uniformly close to the flat half-plane value, with constants depending only on the dimension. They then derive a Bernstein-type classification of global minimizers with near-minimal density (Corollary 1.6) and an improved regularity statement for the graphical convergence, formulated via Hodograph transforms near the free boundary (Corollary 1.7). The proofs combine a regularized density expansion with compactness and improved-regularity results from the authors' prior work [1] and related capillary regularity theory [2,3].

Significance. If the main theorems are correct, the paper provides a sharp quantitative bridge between capillary variational problems and the one-phase Bernoulli problem: the monotone quantity of the capillary problem converges to the Weiss energy, and one obtains dimension-dependent curvature bounds that remain uniform as the contact angle tends to zero. The explicit dimension dependence in Theorem 1.2 is a notable strength, and the monotone-quantity convergence in Theorem 1.1 is a natural and potentially widely useful result. The arguments are largely coherent, and the paper leans on external benchmarks rather than ad hoc assumptions. The main caveat, discussed below, is the precise sense in which the improved convergence from [1, Proposition 4.11] is used in the proof of Theorem 1.2; this is a localized but load-bearing point that needs to be clarified or repaired.

major comments (1)

[Section 2, Case 1a of Theorem 1.2 (around Eq. (14))] The proof invokes [1, Proposition 4.11] to obtain θ_i^{-1}u'_i → v in C^{2,α}_{loc}(ℝ^n) and then uses the normalization (14) to assert 1 = θ_i^{-1}|A_{M'_i}(x'_i)| → |D^2v(0)|. As written, this step is not justified. The limiting function v is a one-phase minimizer with a free boundary, and the rescaling point 0 may lie on ∂{v>0}; in the model case v(y)=(y·n)_+, the function is not C^1 at the free boundary, so ordinary C^{2,α} convergence of the graph functions cannot hold in a neighborhood of such a point. If [1, Proposition 4.11] only provides C^{2,α} convergence away from the free boundary, then the points x'_i, which are allowed to approach the interface, are not covered and the curvature contradiction in Case 1a does not follow. The authors should either quote a version of [1, Proposition 4.11] that gives convergence of the surfaces at boundary-approaching points (for instance, in the Hodograph sense mentioned in Corollary 1.7), or rewrite the normalization step so that the limit of the second fundamental form is obtained from ambient convergence of the hypersurfaces rather than from the Hessian of the graph function at a potentially nonsmooth point. This is a load-bearing gap in the small-angle curvature estimate.

minor comments (7)

[Equation (1)] There is a typo: 'satsfying' should be 'satisfying'.
[Remark 1.3] 'The a salient aspect' should be 'A salient aspect'.
[Proof of Theorem 1.1] In the display after (8), 'similary' should be 'similarly'.
[Theorem 1.1 statement] The phrase 'any xi ∈ ∂ℝ^{n+1}_+ → x ∈ B^n_1' is not a well-formed convergence statement; it should say 'any sequence xi → x with xi ∈ ∂ℝ^{n+1}_+'.
[Introduction, after (1)] The statement that the convergence θ_i^{-1}u_i → v is 'in fact C^{2,α}_{loc}(B_1)' should be qualified as in Corollary 1.7, since for a one-phase minimizer with a free boundary this cannot hold in the usual sense at boundary points.
[Lemma 2.1, n = 1 case] 'If Ω≠ ∅ or ℝ^{2}_+' should presumably be 'If Ω≠ ∅ and Ω≠ ℝ^{2}_+'; otherwise the sentence is ambiguous.
[Case 2a of Theorem 1.2] 'Defined the rescaled domains' should be 'Define the rescaled domains'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorem 1.1 is an explicit expansion result against the known Weiss energy, and the prior results it invokes are independent support rather than restatements of the target claim.

full rationale

The paper's central claim, Theorem 1.1, is not circular. The capillary density ratio is expanded directly from the graphical representation obtained in the authors' prior work [1], using |u_i| + |Du_i| = O(theta_i). The resulting theta_i^{-2} terms match, by an explicit computation, the Alt-Caffarelli integrand |Dv|^2 + 1 and the Weiss boundary term involving v^2. The normalizations are fixed by the known regular-point values (1-cos theta)/2 for the capillary density and omega_n/2 for the Weiss energy, so no parameter is fitted to make the limits agree. The convergence theta_i^{-1} u_i -> v, inherited from [1], is an independent compactness/regularity result whose assumptions do not include the density-to-Weiss convergence proved here. Theorem 1.2's use of [1, Proposition 4.11] and [2, Lemma 4.2] is likewise reliance on previously established regularity and cone-classification results, not on the theorem being proved. Even the cone classification in Lemma 2.1 is reproduced with a proof rather than merely imported. The skeptical concern about whether [1, Proposition 4.11] provides convergence at points approaching the free boundary is a verification question about a published dependency, not a circularity: the paper's own derivation does not define its conclusion in terms of that proposition, nor does it rename a fitted input as a prediction. Overall, the derivation chain is self-contained modulo legitimate prior results and external tools such as Weiss's monotonicity formula and De Philippis-Maggi compactness.

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

The central claim rests on previously established monotonicity, compactness, and regularity results, several from the authors themselves. These are not circular but they are load-bearing; the paper's own contribution is the convergence theorem and its applications on top of these foundations.

assumptions (6)

domain assumption Weiss energy W_v(x,r) is monotone in r for minimizers of the Alt-Caffarelli functional.
Used in the proof of Theorem 1.1 to control W^ζ_v and to pass ε → 0 via continuity (cited to [5, Lemma 9.1]).
domain assumption Capillary varifold density ratio Θ_V(x,r) is monotone in r for free-boundary stationary varifolds.
Gives the monotonicity used throughout; cited to [3,2].
domain assumption Compactness of capillary minimizers and varifolds under uniform density bounds.
Used in Theorem 1.2 Cases 2a and 2b; cited to [3, Theorem 2.9] and [2, Lemma 3.12].
domain assumption Uniform Lipschitz bound and convergence of the graphical representation of small-angle minimizers.
Provides Lip(u_i) ≤ c θ_i and W^{1,2}∩C^α convergence used in Theorem 1.1 and Theorem 1.2; from [1].
domain assumption Improved C^{2,α} convergence of rescaled graphs under a uniform curvature bound.
Used in Case 1a and Corollary 1.7; from [1, Proposition 4.11].
domain assumption Dilation-invariant cone classification: minimizers with Θ_V(0) ≤ (1-cos θ)/2 are trivial or capillary half-planes.
Used in Case 2a of Theorem 1.2; Lemma 2.1, essentially [2, Lemma 4.2].

how reviews work

0 comments

Cite this review

Pith. "Pith review of Weiss monotonicity and capillary hypersurfaces." pith.science (2026). https://pith.science/paper/MEEP6C2J

@misc{pith2026250602146,
  author       = {Pith},
  title        = {Pith review of: Weiss monotonicity and capillary hypersurfaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MEEP6C2J}},
  note         = {Machine review of arXiv:2506.02146}
}

read the original abstract

Previous work of the authors established the rigorous limiting behavior of minimizing capillary surfaces to minimizers of the Alt--Caffarelli functional as the capillary angle tends to zero. We prove here that in this limit, the capillary area-density converges to the Weiss energy density. We apply this to obtain angle-independent curvature estimates and regularity results for capillary minimizers.

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Smoothness and stability in the Alt-Phillips problem
math.AP 2025-07 conditional novelty 8.0 of 10

For the Alt-Phillips free boundary problem, the paper proves smoothness of regular free boundaries for all exponents, derives a stability inequality for negative exponents, and rules out nontrivial axially symmetric s...

Reference graph

Works this paper leans on

7 extracted references · 6 canonical work pages · cited by 1 Pith paper

[1]

Otis Chodosh, Nick Edelen, and Chao Li, Improved regularity for minimizing capillary hypersurfaces, Ars Inven. Anal. (2025), Paper No. 2, 27. MR 4888466

work page 2025
[2]

Luigi De Masi, Nick Edelen, Carlo Gasparetto, and Chao Li, Regularity of minimal surfaces with capillary boundary conditions , (2024), arXiv:2405.20796

work page Pith review arXiv 2024
[3]

De Philippis and F

G. De Philippis and F. Maggi, Regularity of free boundaries in anisotropic capillarity prob- lems and the validity of Young’s law , Arch. Ration. Mech. Anal. 216 (2015), no. 2, 473–568. MR 3317808

work page 2015
[4]

Alberto Pacati, Giorgio Tortone, and Bozhidar Velichkov, Some remarks on singular capil- lary cones with free boundary , (2025), arXiv:2502.07697

arXiv 2025
[5]

28, Springer, Cham, [2023] ©2023

Bozhidar Velichkov, Regularity of the one-phase free boundaries, Lecture Notes of the Unione Matematica Italiana, vol. 28, Springer, Cham, [2023] ©2023. MR 4807210

work page 2023
[6]

Georg Sebastian Weiss, Partial regularity for a minimum problem with free boundary , J. Geom. Anal. 9 (1999), no. 2, 317–326. MR 1759450

work page 1999
[7]

Brian White, A local regularity theorem for mean curvature flow , Ann. of Math. (2) 161 (2005), no. 3, 1487–1519. MR 2180405 Department of Mathematics, Stanford University, Building 380, Stanford, CA 94305, USA Email address : ochodosh@stanford.edu Department of Mathematics, University of Notre Dame, Notre Dame, IN 46556 USA Email address : nedelen@nd.edu...

work page 2005

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Otis Chodosh, Nick Edelen, and Chao Li, Improved regularity for minimizing capillary hypersurfaces, Ars Inven. Anal. (2025), Paper No. 2, 27. MR 4888466

work page 2025

[2] [2]

Luigi De Masi, Nick Edelen, Carlo Gasparetto, and Chao Li, Regularity of minimal surfaces with capillary boundary conditions , (2024), arXiv:2405.20796

work page Pith review arXiv 2024

[3] [3]

De Philippis and F

G. De Philippis and F. Maggi, Regularity of free boundaries in anisotropic capillarity prob- lems and the validity of Young’s law , Arch. Ration. Mech. Anal. 216 (2015), no. 2, 473–568. MR 3317808

work page 2015

[4] [4]

Alberto Pacati, Giorgio Tortone, and Bozhidar Velichkov, Some remarks on singular capil- lary cones with free boundary , (2025), arXiv:2502.07697

arXiv 2025

[5] [5]

28, Springer, Cham, [2023] ©2023

Bozhidar Velichkov, Regularity of the one-phase free boundaries, Lecture Notes of the Unione Matematica Italiana, vol. 28, Springer, Cham, [2023] ©2023. MR 4807210

work page 2023

[6] [6]

Georg Sebastian Weiss, Partial regularity for a minimum problem with free boundary , J. Geom. Anal. 9 (1999), no. 2, 317–326. MR 1759450

work page 1999

[7] [7]

Brian White, A local regularity theorem for mean curvature flow , Ann. of Math. (2) 161 (2005), no. 3, 1487–1519. MR 2180405 Department of Mathematics, Stanford University, Building 380, Stanford, CA 94305, USA Email address : ochodosh@stanford.edu Department of Mathematics, University of Notre Dame, Notre Dame, IN 46556 USA Email address : nedelen@nd.edu...

work page 2005