Minimizers of the Allen-Cahn energy with sub-quadratic growth
Ovidiu Savin, Chilin Zhang · 2025
Read Minimizers of the Allen-Cahn energy with sub-quadratic growth by Ovidiu Savin, Chilin Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-02029.txt.
Introduction
Ovidiu Savin and Chilin Zhang prove results in a 2025 paper on a classic energy model from math and physics. The model balances smooth spread against a pull toward two set values. Its best shapes, called minimizers, can grow without bound. The authors ask what shapes are allowed when growth stays below a square rate.
The paper proves limit theorems for such global best shapes. The finds extend past work tied to a known claim about flat level sets. Part of the proof rests on smooth behavior for a linked edge and area problem. That linked work came from an earlier team named in the text. Readers who study phase change and free edges gain new tools here. The results narrow which wild shapes can occur at large scale.
Catalog record
| CREATOR | Ovidiu Savin, Chilin Zhang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-02029
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Minimizers of the Allen-Cahn energy with sub-quadratic growth Ovidiu Savin, Chilin Zhang Abstract . We establish Liouville theorems for global minimizers u u of the Allen-Cahn energy ∫ | ∇ u | 2 + W ( u ) 𝑑 x , \int|\nabla u|^{2}+W(u)\,dx, which have subquadratic growth at infinity. In particular we extend the results of [ S1 , S3 ] concerning the De Giorgi’s conjecture to the setting of unbounded solutions. Part of the analysis relies on the regularity of minimizers for a Dirichlet/perimeter functional which was studied by Athanasopoulous-Caffarelli-Kenig-Salsa in [ ACKS ] . 1. Introduction The typical energy functional associated with phase field models combines the Dirichlet integral of a density u together with a potential term (or free energy) W(u) , (1.1) J(u):=\int_{\Omega}\frac{|\nabla u|^{2}}{2}+W(u)\,dx. In the classical example of the Allen-Cahn energy [ AlC ] , the term W is a double-well potential of the form (1.2) W(t)=(1-t^{2})^{2}, which is relevant in the theory of phase-transitions and minimal surfaces. In their celebrated result, Modica and Mortola [ MM ] showed that 0 -homogenous rescalings of bounded minimizers |u|\leq 1 , converge up to subsequences to a \pm 1 configuration separated by a minimal surface, i.e. u_{\epsilon}(x)=u\left(\frac{x}{\epsilon}\right)\quad\to\quad\chi_{E}-\chi_{E^{c}}\quad\mbox{in $L^{1}_{loc}$, \quad as $\epsilon\to 0$,} with E a set of minimal perimeter. At the level of the energy, this result is expressed in terms of the Gamma-convergence of the rescaled energies to a multiple of the perimeter functional c_{0}Per_{\Omega}(E) . The connection between the level sets of u at large scales, \{u=t\} for a fixed t\in(-1,1) , and minimal surfaces is a subject of great interest that was intensively studied in the literature, see for example [ M , St , HT , AC , GG , CM ] . A key difference between the two objects is that minimal surfaces remain invariant under dilations, while the level sets of u do not. On the other hand, the level sets of u are smooth for a.e. t while minimal surfaces could have singularities or higher multiplicities. Some natural problems that arise in this context, known as De Giorgi type conjectures, ask whether or not global solutions of the two problems have the same rigidity properties. The original formulation of De Giorgi made in 1978 refers precisely to this question for bounded monotone solutions of the Allen Cahn equation (1.3) \triangle u=W^{\prime}(u), and the corresponding Bernstein problem for minimal graphs. In [ S1 ] , the first author proved the following version of the De Giorgi’s conjecture concerning minimizers of the Allen-Cahn energy. Theorem 1.1 ( [ S1 ] ) . Let u be a global minimizer of ( 1.1 )-( 1.2 ) with |u|\leq 1 . Then u is one-dimensional if n\leq 7 . The term global minimizer refers to a function u defined in the whole space \mathbb{R}^{n} , that minimizes the energy in any ball subject to its own boundary conditions. The notion that u is one-dimensional means that u(x)=g(x\cdot\nu) with \nu a unit direction and g a function of one-variable which solves the ODE ( 1.3 ). The dimension n=7 turns out to be sharp just as in the case of area minimizing minimal surfaces, by a construction of Liu-Wang-Wei [ LWW ] . A consequence of Theorem 1.1 is that it implies the graphical version of the De Giorgi conjecture up to dimension n\leq 8 under the natural assumptions |u|\leq 1,\quad u_{x_{n}}>0,\quad\quad\mbox{$\{u=0\}$ is a graph over $\mathbb{R}^{n-1}$ in the $x_{n}$ direction.} We remark however that the original formulation of the De Giorgi conjecture was stated only for bounded monotone solutions of ( 1.3 ) without the assumption that \{u=0\} is a graph over the whole \mathbb{R}^{n-1} . Under these weaker hypotheses the conjecture is known to be true only in dimensions 2 and 3 by the works of Ghoussoub-Gui [ GG ] , Ambrosio-Cabrè [ AmC ] , and to be false in dimension n\geq 9 by a counterexample due to Del Pino, Kowalczyk and Wei [ DKW ] . Without the graphicality condition the rigidity question is closely related to the classification of global stable solutions in one dimension lower. In this paper we extend the result of Theorem 1.1 to minimizers that are not necessarily bounded but that have subquadratic growth at infinity. Our main result is the following. Theorem 1.2 . Let u be a global minimizer of ( 1.1 ) with (1.4) W(t)=\left\{\begin{array}[]{l}(1-t^{2})^{2},\quad\quad\mbox{if}\quad|t|\leq 1,\\ \par\\ 0,\quad\quad\quad\quad\quad\mbox{if}\quad|t|>1,\\ \end{array}\right. and assume that u=o(|x|^{2}) as |x|\to\infty . Then u is one-dimensional if n\leq 7 . Notice that the potential W in Theorem 1.2 is extended trivially outside the interval [-1,1] and therefore the second term in the energy is relevant only in the range |u|<1 . Similarly, Theorem 1.2 implies the graphical version of the De Giorgi conjecture with subquadratic growth in one dimension higher. Theorem 1.3 . Let u be a global solution to ( 1.3 ) with W as in ( 1.4 ). If u_{x_{n}}>0 , u=o(|x|^{2}) as |x|\to\infty , and \{u=0\} is a graph over \mathbb{R}^{n-1} in the x_{n} direction, then u is one-dimensional if n\leq 8 . A few comments are in order regarding our main results. A first remark is that the Modica-Mortola Gamma-convergence result no longer applies under the hypotheses of Theorems 1.2 and 1.3 since u is not necessarily bounded. Now the blow-down analysis and the corresponding singular limit problems depend strongly on the behavior of u at infinity. A heuristic computation of a possible singular limit problem in this setting was carried out by Athanasopoulous, Caffarelli, Kenig, Salsa in [ ACKS ] . They considered minimizers of ( 1.1 ),( 1.4 ) which are not necessarily bounded, in order to motivate and introduce the Dirichlet/perimeter energy functional \mathcal{F} (see ( 1.5 ) below) which they studied in [ ACKS ] . The formal computation carried out in their paper suggests that if \max_{B_{R}}u\sim\sqrt{R} then the limiting problem for the blow-down rescalings u(Rx)/\sqrt{R} as we let R\to\infty , consists in minimizing an energy of the form (1.5) \mathcal{F}(v):=\int_{\Omega}|\nabla v|^{2}dx+Per_{\Omega}(\{v>0\}). The regularity of minimizers of \mathcal{F} was investigated in [ ACKS ] . In particular the authors established the Lipschitz continuity of such minimizers and the smoothness of their free boundaries \partial\{v>0\} outside a set of Hausdorff dimension n-8 . Not surprisingly, these results will play an important role in our analysis. On the other hand it is not at all clear apriori how u behaves at infinity. For example there are one-dimensional minimizers of J which grow linearly at infinity, and possibly two dimensional minimizers which grow barely quadratically. The strategy to prove Theorem 1.2 is to show that either |u|\leq 1 , in which case the prior result Theorem 1.1 in [ S1 ] applies, or that eventually u is asymptotically linear for a sequence of radii R_{k}\to\infty . Heuristically, if u grows sufficiently fast at infinity then the solution at large scales behaves close to a harmonic function. The desired conclusion then follows from an improvement of flatness result which states that u is better and better approximated by one-dimensional solutions as we restrict to smaller scales. Next we discuss some possible extensions of our results that we intend to explore in the future