Gradient flow of phase transitions with fixed contact angle

Kobe Marshall-Stevens; Mayu Takada; Myles Workman; Yoshihiro Tonegawa

REVIEW 2 major objections 5 minor 1 cited by

Gradient flow of phase transitions with fixed contact angle

T0 review · 2 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Under a no-wetting assumption, the singular limit of the Allen–Cahn flow with fixed contact angle is a varifold meeting the boundary at angle $\theta=\arccos(\sigma(1)/c_0)$, and the boundary trace of the limit equals the limit of the…

desk verdict Solid extension of the Neumann-boundary Allen–Cahn convergence theory to fixed contact angles, but a factor-2 error in Proposition 2 limits the stated range to c1<1/√2. read the letter →

arxiv 2411.17979 v2 pith:7FZWKXDL submitted 2024-11-27 math.AP math.DG

Kobe Marshall-Stevens , Mayu Takada , Yoshihiro Tonegawa , Myles Workman This is my paper · ORCID

classification math.APmath.DG MSC 35K5749Q2053C44

keywords Allen-Cahnequationfixedcontactanglegradientflowvarifoldwettingnon-concentrationBVtraceIlmanenmonotonicityformula

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper studies what happens to a diffuse phase interface as its width $\varepsilon$ tends to zero, when the interface meets the container wall at a prescribed contact angle. It claims that, provided the interface does not wet the wall by piling up mass on the boundary, the limiting interface is a varifold that makes the expected fixed contact angle with the wall, and the limiting values of the phase field from the interior and from the boundary agree. A general, unconditional convergence result for the energy measures and the phase field is established first; the sharp contact-angle statement requires an additional non-concentration hypothesis. The paper also derives an Ilmanen-type monotonicity formula valid up to the boundary, which is a tool for future refined convergence results.

What carries the argument

The objects doing the work are the diffuse energy measures $\mu^\varepsilon_{t,1}$ in the interior and $\mu^\varepsilon_{t,2}$ on the boundary, the associated varifolds $V^\varepsilon_{t,1}$ and $V^\varepsilon_{t,2}$, and the discrepancy $\xi^\varepsilon_t=\varepsilon|\nabla u|^2/2-W(u)/\varepsilon$. The first variation formula (Proposition 1) expresses $\delta V^\varepsilon_t$ as the sum of a discrepancy term, a time-derivative term, and a boundary term; passing this identity to the limit under (A5) or (A6) yields the contact-angle identity. Assumption (A6) says the interior measure has no mass in arbitrarily thin boundary tubes, which lets the authors pass the trace of the BV limit and discard the interior part of the first variation on the boundary.

What would settle it

Take a numerical solution of (1) in a disk with $W(s)=(1-s^2)^2/4$, $\sigma(s)=\cos(\theta)\int_{-1}^{s}\sqrt{2W(r)}\,dr$, $\theta=60^\circ$, and a droplet-shaped initial datum with bounded energy. If at some positive time the interface approaches the wall tangentially and the interior energy measure keeps positive mass in every boundary tube while the total energy stays bounded, then (A6) fails and the trace identity of Theorem 3 part 1 would be violated, showing the no-wetting condition is genuinely needed.

Watch

Extended reading notes

Core claim

Under assumptions (A1)–(A4), for almost every time the diffuse solutions converge to a BV limit $u=\pm1$ in the interior and a BV limit $\tilde{u}=\pm1$ on the boundary, with the associated energy measures converging to Radon measures. If the discrepancy does not concentrate on the boundary (A5), the limiting interior varifold $V_{t,1}$ has bounded, time-integrable first variation and satisfies a generalized contact-angle identity. If in addition the interior energy measure itself stays away from the boundary (A6), then the boundary limit is exactly the trace of the interior limit, $Tu=\tilde{u}$, and the boundary part of the first variation equals $-\sigma(1)\int_{\partial^*\{Tu=+1\}} g\cdot n\,dH^{n-2}$ for every test field $g$ tangent to $\partial\Omega$. This is the fixed contact angle condition of [KT17], with angle $\theta=\arccos(\sigma(1)/c_0)$.

Load-bearing premise

The load-bearing premise is that the diffuse interface never piles up on the container wall—for almost every time the interior energy measure has arbitrarily small mass in sufficiently thin boundary strips—and the paper notes that even the weaker discrepancy version of this control has not been derived from the equations.

Editorial extensions

If this is right

For almost every time, under (A6), the boundary trace of the limiting phase field exists as a BV function on $\partial\Omega$ and equals the limit of the boundary traces of the diffuse solutions.
The limiting interior varifold has a boundary first variation supported on the reduced boundary of the trace set, with density $\sigma(1)$; this is exactly the fixed-angle condition of [KT17].
Under the weaker assumption (A5), the limiting varifold already has bounded first variation, integrable in time, and satisfies a generalized contact-angle identity involving the boundary limit $\tilde{u}$.
The Ilmanen-type monotonicity formula holds up to the boundary for all solutions with bounded initial energy and non-negative boundary energy, provided the discrepancy term is controlled.
Under (A6), the limiting interior and boundary varifolds are unique, with the boundary varifold given by $\sigma(1)H^{n-1}$ on the trace set.

Reading between the lines

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

Beyond the paper: if (A6) could be derived from the equations, these results would open the door to a Brakke-flow convergence theorem with fixed contact angle; the paper does not prove this.
Beyond the paper: the monotonicity formula's discrepancy term suggests that under (A5) or (A6) one could obtain local upper-density bounds near the boundary, leading to partial regularity; the authors do not pursue this here.
Beyond the paper: the trace identity says wetting is the only obstruction to a clean free-boundary condition, and when wetting occurs the interface should pop off the wall instantaneously, an event that could be studied as a singular free-boundary phenomenon.

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 paper is a real step forward: it extends the Mizuno–Tonegawa convergence program to the nonzero Robin boundary case, proves BV compactness on the boundary, derives generalized contact-angle conditions for the flow limit, and constructs the boundary Ilmanen monotonicity formula. The structural arguments are sound and the exposition is careful. The authors are also honest that the non-concentration assumptions (A5)/(A6) are not derived from the equations (Remark 6), so the main theorems are conditional — that is a limitation, not a flaw.

The stress-test note is correct. In the proof of Proposition 2, the bound from (A2) should be σ′(u)²/ε ≤ 2c1² W(u)/ε, not c1² W(u)/ε. The absorption then gives (1−2c1²) on the boundary-energy term, so the argument as written only closes for c1 < 1/√2. This affects Theorem 1 part 3, and consequently Theorems 2 and 3 for the full stated range c1 ∈ [0,1). The standard example from Remark 1 has c1 = |cos θ|, so contact angles θ ≤ π/4 are excluded. This is a genuine overclaim, but it is localized: the rest of the proof structure is coherent, and the range restriction may be fixable or at least should be restated.

Everything else I checked holds up: the first-variation computation, the BV convergence arguments, and the monotonicity formula are consistent with the cited literature. The appendices on measure uniqueness are honest about what is not known. The paper is a useful reference for anyone working on phase-field approximations of capillarity or mean curvature flow with contact angle.

Who is this for? Specialists in geometric measure theory and free boundary problems, especially those building on Ilmanen, Mizuno–Tonegawa, and Kagaya–Tonegawa. It deserves a serious referee: the topic is timely, the new results are substantial, and the main weakness is a technical range restriction rather than a faulty idea. I would recommend acceptance after the authors fix Proposition 2 or clearly state the c1 < 1/√2 range.

Referee Report

2 major / 5 minor

Summary. This paper studies the ε→0 limit of the Allen–Cahn equation with nonlinear Robin boundary condition (1), which is the gradient flow of the Modica–Mortola energy with boundary contact energy (2). Under assumptions (A1)–(A4) it proves compactness of the associated energy measures and varifolds, L1 convergence of solutions to a BV function taking values ±1, and boundary L1/BV compactness of the traces. With the discrepancy non-concentration assumption (A5) it obtains a first-variation identity corresponding to a generalized fixed contact angle, and with the stronger measure non-concentration assumption (A6) it proves that the trace of the interior limit equals the limit of traces and that the interior limit varifold meets the boundary at the prescribed angle in the sense of [KT17]. Appendix B derives an Ilmanen-type boundary monotonicity formula.

Significance. The results are potentially valuable: they extend varifold convergence theory for Allen–Cahn to boundary contact-angle flows in a time-dependent setting, and the boundary monotonicity formula is a useful new tool. The paper is careful and transparent about the fact that (A5)/(A6) are additional geometric hypotheses not derived from the equations (see Remark 6). However, the central range of the main theorems is broader than the proofs currently support: the factor-of-2 error in Proposition 2 limits the boundary compactness argument to c1<1/√2 rather than c1<1, and the trace identity has a missing δ→0 step. With those repaired, the paper would make a solid contribution.

major comments (2)

[§2, Proposition 2 (proof, displayed estimate after (A2))] The displayed estimate after applying (A2) reads ∫∂Ω σ′(u)²/ε dH^{n−1} ≤ c1²∫∂Ω W(u)/ε dH^{n−1}, but (A2) gives |σ′(u)| ≤ c1√(2W(u)), so the correct bound is ∫ σ′(u)²/ε ≤ 2c1²∫ W(u)/ε. Consequently the absorption step gives (1−2c1²)∫∂Ω(ε|∇u|²/2 + W(u)/ε)dH^{n−1} ≤ 1/2∫Ω ε(∂tu)² dx + C, which is usable only when c1 < 1/√2. Since Proposition 2 is the only estimate behind the boundary BV compactness in Theorem 1 part 3 and the boundary bounds in Theorems 2 and 3, those results are not established for the full stated range c1 ∈ [0,1). For the example in Remark 1 this excludes contact angles θ ≤ π/4. The authors should either restrict (A2) accordingly or supply a different argument for the claimed range.
[§3.2, proof of Theorem 3 part 1] After fixing δ ∈ (0,κ), the terms ∫_{Nδ} X·d[Dw] and ∫_{Nδ} ∇w_i·X are only bounded by ||X||L∞ times a liminf/limsup of ||V_i_t,1||(Nδ); for fixed δ these bounds need not vanish as j→∞. To conclude ∫∂Ω f T w dH^{n−1} = ∫∂Ω f w̃ dH^{n−1}, one must let δ→0 (or choose δ depending on the threshold in (A6)) after passing to the limit in j, using (A6) to make limsup_i ||V_i_t,1||(Nδ) arbitrarily small. As written the proof omits this δ→0 step, so the trace identity is not yet justified.

minor comments (5)

[Title page] The fourth author's name is typeset as 'TONEGA W A' in the arXiv header; this should be corrected to 'TONEGAWA'.
[§2, Proposition 2] Once the factor of 2 is fixed, use a single symbol for the constant in the display (e.g. write 1−2c1² consistently) instead of interchanging c1 and c1².
[§3.1, proof of Theorem 2 part 1] The convergence of the boundary varifolds V_j_t,2 to σ(1)H^{n−1}⌊{ũ(·,t)=+1} is used implicitly; stating it explicitly would help the reader.
[Appendix B, Proposition 3] The statement says the constants depend only on n, Ω, and E0, while the proof introduces dependence on the mean curvature of ∂Ω; this is fine since that is determined by Ω, but the sentence should say so.
[Figure 1 caption] The caption would benefit from a note that the 'popping' behavior depicted is heuristic and not derived from the assumptions.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's conditional convergence statements follow from assumptions (A1)-(A6) via independent estimates and published theorems, not by construing the conclusion into the input.

full rationale

The derivation chain is not circular. Theorem 1 is proved by energy-monotonicity compactness (Lemma 1), a Phi-change of variables, and the boundary BV estimate of Proposition 2; Theorem 2 derives the boundary first-variation identity from Proposition 1, assumption (A5), and the Radon-Nikodym/L^2 representation; Theorem 3 derives the trace identity and contact-angle identity from (A6) plus the divergence theorem. Citations to [MT15], [KT17], [KT18], [Ilm93], [Ton03] are used as computational templates, definitions, or rectifiability/integrality theorems: they are independent published results, not the target theorems. The paper explicitly disclaims (Remark 6) that (A5) has not been deduced, so the main results are honestly conditional rather than being forced by an input. The skeptic's factor-2 issue in Proposition 2 (Section 2) is a real-looking internal correctness gap: from |sigma'(s)| <= c1 sqrt(2W(s)) one gets sigma'(u)^2/epsilon <= 2 c1^2 W(u)/epsilon, so the displayed absorption step would need c1 < 1/sqrt(2), not merely c1 < 1. This is a proof gap, not a circular definition or fitted-input prediction; no step of the derivation reduces by construction to its own conclusion.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

No free parameters or invented entities. The paper introduces no new postulates beyond standard PDE assumptions; its central results are conditional on the stated non-concentration assumptions (A5)/(A6), listed as domain axioms.

assumptions (7)

domain assumption Existence of smooth solutions u_i to (5) for each epsilon_i on Omega x [0, inf)
The paper assumes (A3) a sequence of smooth solutions; no existence or long-time regularity proof is provided, only a priori estimates.
domain assumption Uniform energy and sup-norm bounds at initial time (A4)
Theorems 1-3 all assume (A4); it is standard and not derived.
domain assumption (A2) |sigma'(s)| <= c1 sqrt(2W(s)) with c1 < 1
This controls boundary energy and ensures the contact energy density is below the interior energy density; used in Proposition 2 and Theorem 2.
standard math Standard results of BV theory, Radon measure compactness, and varifold first variation (e.g., [EG15], [Lah17])
Used throughout; accepted background.
domain assumption (A5) discrepancy non-concentration
Needed for Theorem 2; not derived (Remark 6).
domain assumption (A6) measure non-concentration near boundary
Needed for Theorem 3; stronger than (A5) and precludes wetting.
domain assumption Smoothness and compactness of the domain boundary, and sigma >= 0 for the monotonicity formula
The paper assumes Omega is bounded with smooth boundary; Appendix B additionally assumes sigma >= 0.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Gradient flow of phase transitions with fixed contact angle." pith.science (2026). https://pith.science/paper/7FZWKXDL

@misc{pith2026241117979,
  author       = {Pith},
  title        = {Pith review of: Gradient flow of phase transitions with fixed contact angle},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7FZWKXDL}},
  note         = {Machine review of arXiv:2411.17979}
}

read the original abstract

We study the gradient flow of the Allen-Cahn equation with fixed boundary contact angle in Euclidean domains for initial data with bounded energy. Under general assumptions, we establish both interior and boundary convergence properties for the solutions and associated energy measures. Under various boundary non-concentration assumptions, we show that, for almost every time, the associated limiting varifolds satisfy generalised contact angle conditions and have bounded first variation, as well as deducing that the trace of the limit of the solutions coincides with the limit of their traces. Moreover, we derive an Ilmanen type monotonicity formula, for initial data with bounded energy, valid for the associated energy measures up to the boundary.

Figures

Figures reproduced from arXiv: 2411.17979 by the authors.

**Figure 1.** In both graphics the interior solid curves depict a portion of ∂Ω, and the dashed curves depict smooth boundaries of interface regions in the interior of Ω. The left-hand graphic depicts a positive time along the flow at which the interface comes into tangential contact with ∂Ω, i.e. the interface ‘wetting’. The right-hand graphic depicts the interface at any later positive time after which the ‘popping’ of the inte… view at source ↗

**Figure 2.** The thicker dashed line in the graphic depicts a smooth boundary of the region {u = +1} for the limiting function u (at some fixed time) which makes contact angle θ with respect to the set {T u = +1}, depicted with the thin dashed line, in ∂Ω. The angle θ corresponds to the angle formed between the unit inward pointing co-normal, −⃗n∂{u=+1}, to the boundary of {u = +1} in ∂Ω, and the outward pointing unit normal, ⃗n… view at source ↗

Discussion (0). Continue with ORCID 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

The vector-valued Allen-Cahn equation with potentials of high-dimensional double-wells under Robin boundary conditions
math.AP 2025-05 conditional novelty 6.0 of 10

The vector-valued Allen-Cahn equation with Robin boundary conditions converges locally in time to mean curvature flow with contact angle, and the limits solve harmonic heat flow and a minimal pair condition.

Reference graph

Works this paper leans on

18 extracted references · 17 canonical work pages · cited by 1 Pith paper

[1]

Convergence of the A llen-- C ahn equation with a nonlinear R obin boundary condition to mean curvature flow with contact angle close to 90^

Helmut Abels and Maximilian Moser. Convergence of the A llen-- C ahn equation with a nonlinear R obin boundary condition to mean curvature flow with contact angle close to 90^ . SIAM Journal of Mathematical Analysis , 54(1):114--172, 2022

work page 2022
[2]

John W. Cahn. Critical point wetting. The Journal of Chemical Physics , 66(8):3667--3672, 1977

work page 1977
[3]

Regularity of free boundaries in anisotropic capillarity problems and the validity of Y oung’s law

Guido De Philippis and Francesco Maggi. Regularity of free boundaries in anisotropic capillarity problems and the validity of Y oung’s law. Archive for Rational Mechanics and Analysis , 216:473--568, 2015

work page 2015
[4]

Evans and Ronald F

Lawrence C. Evans and Ronald F. Gariepy. Measure theory and fine properties of functions, revised edition. Chapman and Hall/ CRC , 2015

work page 2015
[5]

Morton E. Gurtin. Some results and conjectures in the gradient theory of phase transitions. Metastability and incompletely posed problems ( M inneapolis, M innesota, 1985), IMA Volumes in Mathematics and its Applications, Springer, New York , 3:135--146, 1985

work page 1985
[6]

BV solutions for mean curvature flow with constant contact angle: A llen-- C ahn approximation and weak-strong uniqueness

Sebastian Hensel and Tim Laux. BV solutions for mean curvature flow with constant contact angle: A llen-- C ahn approximation and weak-strong uniqueness. Indiana University Mathematics Journal , to appear, 2021

work page 2021
[7]

Convergence rates for the A llen– C ahn equation with boundary contact energy: the non-perturbative regime

Sebastian Hensel and Maximilian Moser. Convergence rates for the A llen– C ahn equation with boundary contact energy: the non-perturbative regime. Calculus of Variations and Partial Differential Equations , 61(201), 2022

work page 2022
[8]

Hutchinson and Yoshihiro Tonegawa

John E. Hutchinson and Yoshihiro Tonegawa. Convergence of phase interfaces in the van der W aals- C ahn- H illiard theory. Calculus of Variations and Partial Differential Equations , 10(1):49--84, 2000

work page 2000

Show all 18 references

[9]

Convergence of the A llen-- C ahn equation to B rakke’s motion by mean curvature

Tom Ilmanen. Convergence of the A llen-- C ahn equation to B rakke’s motion by mean curvature. Journal of Differential Geometry , 38(2):417--461, 1993

1993
[10]

Convergence of the A llen-- C ahn equation with a zero N eumann boundary condition on non-convex domains

Takashi Kagaya. Convergence of the A llen-- C ahn equation with a zero N eumann boundary condition on non-convex domains. Mathematische Annalen , 373:1485--1528, 2019

2019
[11]

A fixed contact angle condition for varifolds

Takashi Kagaya and Yoshihiro Tonegawa. A fixed contact angle condition for varifolds. Hiroshima Mathematical Journal , 47:139--153, 2017

2017
[12]

A singular perturbation limit of diffused interface energy with a fixed contact angle condition

Takashi Kagaya and Yoshihiro Tonegawa. A singular perturbation limit of diffused interface energy with a fixed contact angle condition. Indiana University Mathematics Journal , 67(4):1425--1437, 2018

2018
[13]

Equality of the usual definitions of B rakke flow

Ananda Lahiri. Equality of the usual definitions of B rakke flow. https://arxiv.org/abs/1705.08789, 2017

2017 arXiv
[14]

Gradient theory of phase transitions with boundary contact energy

Luciano Modica. Gradient theory of phase transitions with boundary contact energy. Annales de l'I.H.P. Analyse non linéaire , 4(5):487--512, 1987

1987
[15]

Convergence of the A llen-- C ahn equation with N eumann boundary conditions

Masashi Mizuno and Yoshihiro Tonegawa. Convergence of the A llen-- C ahn equation with N eumann boundary conditions. SIAM Journal on Mathematical Analysis , 47(3):1906--1932, 2015

1906
[16]

Owen and Peter Sternberg

Nicholas C. Owen and Peter Sternberg. Gradient flow and front propagation with boundary contact energy. Proceedings of the Royal Society of London. Series A: Mathematical and Physical Sciences , 437(1901):715--728, 1992

1901
[17]

Time-global existence of generalized BV flow via the A llen-- C ahn equation

Kiichi Tashiro. Time-global existence of generalized BV flow via the A llen-- C ahn equation. Interfaces and Free Boundaries , published online first, 2024

2024
[18]

Integrality of varifolds in the singular limit of reaction-diffusion equations

Yoshihiro Tonegawa. Integrality of varifolds in the singular limit of reaction-diffusion equations . Hiroshima Mathematical Journal , 33(3):323--341, 2003

2003

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Convergence of the A llen-- C ahn equation with a nonlinear R obin boundary condition to mean curvature flow with contact angle close to 90^

Helmut Abels and Maximilian Moser. Convergence of the A llen-- C ahn equation with a nonlinear R obin boundary condition to mean curvature flow with contact angle close to 90^ . SIAM Journal of Mathematical Analysis , 54(1):114--172, 2022

work page 2022

[2] [2]

John W. Cahn. Critical point wetting. The Journal of Chemical Physics , 66(8):3667--3672, 1977

work page 1977

[3] [3]

Regularity of free boundaries in anisotropic capillarity problems and the validity of Y oung’s law

Guido De Philippis and Francesco Maggi. Regularity of free boundaries in anisotropic capillarity problems and the validity of Y oung’s law. Archive for Rational Mechanics and Analysis , 216:473--568, 2015

work page 2015

[4] [4]

Evans and Ronald F

Lawrence C. Evans and Ronald F. Gariepy. Measure theory and fine properties of functions, revised edition. Chapman and Hall/ CRC , 2015

work page 2015

[5] [5]

Morton E. Gurtin. Some results and conjectures in the gradient theory of phase transitions. Metastability and incompletely posed problems ( M inneapolis, M innesota, 1985), IMA Volumes in Mathematics and its Applications, Springer, New York , 3:135--146, 1985

work page 1985

[6] [6]

BV solutions for mean curvature flow with constant contact angle: A llen-- C ahn approximation and weak-strong uniqueness

Sebastian Hensel and Tim Laux. BV solutions for mean curvature flow with constant contact angle: A llen-- C ahn approximation and weak-strong uniqueness. Indiana University Mathematics Journal , to appear, 2021

work page 2021

[7] [7]

Convergence rates for the A llen– C ahn equation with boundary contact energy: the non-perturbative regime

Sebastian Hensel and Maximilian Moser. Convergence rates for the A llen– C ahn equation with boundary contact energy: the non-perturbative regime. Calculus of Variations and Partial Differential Equations , 61(201), 2022

work page 2022

[8] [8]

Hutchinson and Yoshihiro Tonegawa

John E. Hutchinson and Yoshihiro Tonegawa. Convergence of phase interfaces in the van der W aals- C ahn- H illiard theory. Calculus of Variations and Partial Differential Equations , 10(1):49--84, 2000

work page 2000