Radial and cylindrical symmetry of solutions to the Cahn–Hilliard equation

The paper is devoted to the classification of entire solutions to the Cahn–Hilliard equation -Δu=u-u3-δ\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$-\Delta u=u-u^3-\delta $$\end{document} in RN\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathbb {R}^N$$\end{document}, with particular interest in those solutions whose nodal set is either bounded or contained in a cylinder. The aim is to prove either radial or cylindrical symmetry, under suitable hypothesis.


Introduction
We consider the entire equation This result agrees with the variational theory, which studies the asymptotic behaviour of the scaled functionals E ε (u, ) = ε 2 |∇u| 2 + W (u) ε dx (1.5) as ε → 0. For instance, Modica proved that, if ε k is a sequence of positive numbers tending to 0 and u ε k is a sequence of minimisers of E ε k (· p, ) under the constraint (1.3) such that u ε k → u 0 in L 1 ( ), then u 0 (x) ∈ {±1} for almost every x ∈ , and the boundary in of the set E := {x ∈ : u 0 (x) = 1} has minimal perimeter among all subsets F ⊂ such that |F| = |E|, where | · p| denotes the volume (see [15], Theorem 1). Further -convergence results relating E ε (· p, ) to the perimeter can be found in [16]. Therefore, given a family {u ε } ε∈(0,ε 0 ) of minimisers under the constraint (1.3), their nodal set is expected to be close to a compact Alexandrov-embedded constant mean curvature surface, at least for ε small. Corollary 2, together with a scaling argument, shows that, for ε small enough, the nodal set of any entire solution to (1.6) in R N such that u > z 2 (ε ) outside a ball is actually a sphere, which is known to be the unique compact Alexandrov-embedded constant mean surface in R N (see [1]). After that, we set and we consider solutions satisfying The aim is to study their symmetry properties and their asymptotic behaviour as δ → 0, with particular interest in solutions which have one periodicity direction.
) be a family of non constant solutions to (1.1) in R N , with N ≥ 2. Assume furthermore that u δ is periodic in x N and, for any δ ∈ (0, 2 (2) u δ is radially symmetric in x .
In view of the aforementioned -convergence results, given a solution u to (1.6) satisfying (1.7), with δ = ε , we expect its nodal set to be close to an Alexandrov-embedded constant mean curvature surface which is contained in a cylinder. This kind of surfaces are fully classified, at least the ones which are embedded in R 3 , in fact it is known that the unique examples are the sphere and Delaunay unduloids, that is a family of non compact revolution surfaces obtained by rotating a periodic curve around a fixed axis in R 3 , which can be taken to be the x 3 -axis, parametrised by a real number τ ∈ (0, 1). We will denote the period of D τ by T τ . For a detailed introduction of Delaunay surfaces, we refer to [12,14]. For any τ ∈ (0, 1), Kowalczyk and Hernandez [11] constructed a family {u τ,ε } ε∈(0,ε 0 ) of solutions to (1.6) in R 3 , with = ε depending on ε, such that (1) ε is positive and bounded uniformly in ε.
We observe that the solutions u ε,τ constructed in [11] are actually negative outside a cylinder, however, in order to obtain the aforementioned family, thanks to the oddness of f , it is enough to replace them with −u ε,τ . An interesting question is uniqueness. In other words, we are interested in the following question.
This would be the counterpart of Corollary 2 for periodic solutions. For now we are not able to give a full answer to this question. However Theorem 3 is a first step in this direction, since it proves that any family {v ε } ε∈(0,ε 0 ) of such solutions has to share many properties with the family {u τ,ε } ε∈(0,ε 0 ) constructed by Hernandez and Kowalczyk. For instance, for ε small, v ε has to satisfy (1), (2), (3) and the scaled functions v ε (εx) tend to −1 uniformly on compact subsets of R N as ε → 0.
The plan of the paper is the following. In Sect. 2 we will state some quite general results, of which the Theorems stated in the introduction are consequences. Section 3 is devoted to the proofs. It is divided into three subsections, dedicated to prove global boundedness, radial symmetry and the asymptotic behaviour for δ small respectively.

Some relevant results
In this section we state some results that are proved in Sect. 3. First we prove boundedness of solutions, which holds irrespectively of the sign of δ.
) and let u δ ∈ L 3 loc (R N ) be a distributional solution to the Cahn-Hilliard equation (1.1). Then a. e. in R N .
Remark 6 • Using Proposition 5, standard elliptic estimates (see [10], Theorem 8.8 and Corollary 6.3) and a bootstrap argument, it is possible to show that any distributional . This parallels the regularity result proved in [6] for the Allen-Cahn equation.
• It follows from the strong maximum principle that either u δ is constant, and in this case it has to be either We observe that Proposition 5 and Remark 6 prove point (1) of Theorem 3, which is actually true for any non constant entire solution. After that, we rule out the case δ ≤ 0, in which only constant solutions are allowed.
We stress that the latter result proves point (1) of Theorem 1 and agrees with the sign of δ obtained by Hernández and Kowalczyk in [11]. Using boundedness and the famous result by Gidas et al. [9], or Theorem 2 of [7], which relies on the moving planes method, we can prove this symmetry result.
and let u δ be a non constant solution to (1.1) such that u δ > z 2 (δ) outside a ball B R , for some R > 0. Then admits a unique solution which is radially symmetric (see [4,17,18]), that is v δ (x) = w δ (|x|).
In view of this fact, we can actually prove the following classification result.
) and let u δ be a non constant solution to (1.1) such that u δ > z 2 (δ) outside a ball B R . Then, up to a translation, u δ = v δ .
In the sequel, we will use the notation W δ (t) := W (t) + δt.

Remark 10
It is possible to see that, for any δ ∈ (0, 2 is strictly decreasing, since Thus, using that, by Proposition 8, v δ is decreasing, which yields that w δ (0) < 0.
In particular, in view of Remark 10, which yields that the nodal set of v δ is neither empty nor a singleton, Corollary 2 is true.
Considering solutions that are approaching a positive limit just with respect to N − 1 variables, we can prove the following.
We note that this proves point (2) of Theorem 3. Even in this case, our result agrees with the construction of [11], where the authors prove the existence of a family of solutions fulfilling the symmetries of the Delaunay surface D τ , hence, in particular they are periodic in x N , radially symmetric and radially increasing in x . Here we show that any periodic solution has to be radially symmetric and radially increasing in x . Finally, in order to prove point (3) of Theorem 3, we need the following result, which shows that the phase transition has to be complete.

Proposition 12 For any
) such that, for any δ ∈ (0, δ 0 ) and This result somehow parallels Lemma 2.5 of [8]. The proof relies on both the moving planes and the sliding method. For a detailed proof of point (3) of Theorem 3, we refer to Sect. 3.

Boundedness
In order to prove boundedness for distributional solutions to (1.1), we will rely on a result proved by Brezis [2].

Remark 14
A similar argument is used in [5] to prove boundedness for solutions to a class of vectorial equations of the form with 0 < k 1 < · · · < k n . The scalar Allen-Cahn equation is included in this class. Here we prove that a similar result is true for a slightly different non linearity, due to the presence of δ.
Now we can prove Proposition 7, using boundedness and a result of [6] where nonexistence f ground states for some special non linearies is proved.

Radial symmetry
The aim of this subsection is to prove Proposition 11. In order to do so, we need some decay at infinity of the solution. From now on, we denote the variables by x := (x 1 , x ) ∈ R × R N −1 . For λ ∈ R, we set λ := {x ∈ R 3 : x 1 < λ}. (3.2) This changing of notation is justified by the fact that several times this section x N is the periodicity variable, hence we are not allowed to start the moving planes in that direction.

Lemma 15
Let u δ be a solution to (1.1). Assume furthermore that u δ > z 2 (δ) in the half-space R N \ λ , for some λ ∈ R. Then Proof The statement is trivial if u δ is constant (see Remark 6), hence we can assume that it is non constant. We apply Lemma 2.3 of [6] to w := u δ −z 2 (δ) in the half space R N \ λ , where, by Lemma 13, 0 < w < β. This is possible since the non linearity g(t) : is positive in (0, β) and g (0) > 0. We recall that the constants α and β are defined in the Proof of Proposition 5. The conclusion is that and the limit is uniform in the other variables.
Using the fact that f (z 3 (δ)) < 0, we can actually prove a better result about the decay rate of z 3 (δ) − u δ .

3.4)
Proof We compare the bounded function v := z 3 (δ) − u δ with the barrier μe −γ x 1 , for γ ∈ (0, − f (z 3 (δ))), in the half-space R N \ M , with M > 0 large enough. In fact, on Note that here we use the fact that v ∈ L ∞ , which is true by Lemma 13. Moreover, setting h δ (v) : uniformly with respect to x . Thus, by the maximum principle for possibly unbounded domains (see Lemma 2.1 of [3]), we conclude that (3.4) is true in R N \ M . Changing, if necessary, the constant C(γ ), the required inequality is fulfilled in the whole space.
In order to prove Proposition 11, we need to apply Theorem 2 of [7], which we recall, for the reader's convenience.

Theorem 17 ([7]) Let v > 0 be a bounded entire solution to
Then v is radially symmetric in y, that is, up to a translation, v(y, z) = w(|y|, z), and radially decreasing in y, that is ∂ y j v(y, z) < 0 for any x = (y, z) ∈ R M × R N −M with y = 0.
Proof By Proposition 5, z 1 (δ) < u δ < z 3 (δ) and, by Remark 6, u δ is smooth. By Lemma 15, it converges to z 3 (δ) as |x | → ∞, uniformly in x N . Since u δ is periodic, in order to conclude that it is radially symmetric in x and radially decreasing, it is enough to apply Theorem 17

The asymptotic behaviour for ı small
First we show that if a solution lies between 1/ √ 3 and z 3 (δ), then it is constant. This is proved by the moving planes method.
In order to prove this fact, we assume by contradiction that there exists λ ∈ R such that the open set λ := {x ∈ λ : v − v λ < 0} is nonempty, and we observe that, in any connected component ω of λ we have due to the strict monotonicity of f δ in [1/ √ 3, 1) (for the definition of h δ , see the Proof of Lemma 16). As a consequence, by the maximum principle for possibly unbounded domains, Composing v with any rotation of R N , we conclude that v is a constant solution to (1.1), thus v ≡ 0.
Moreover, we take a smooth cutoff function χ : R → [0, 1] such that χ = 1 in (−∞, −1) and χ = 0 in (0, ∞) and we set We will denoteW :=W 0 . It is possible to see thatW δ enjoys the following properties: In the sequel, we will be interested in a solution to for δ ≥ 0 small enough and R large. This will be used as a barrier in the Proof of Proposition 12, which relies on a sliding method. This can be obtained in a variational technique, by minimising the functional The case δ = 0 is treated in Lemma 2.4 of [8].

Lemma 19
Let δ 0 > 0 be so small that W δ (z 3 (δ)) < α/2, for any δ ∈ [0, δ 0 ). Then, For any R > 0 and δ ∈ [0, δ 0 ), there exists a minimiser β R,δ ∈ C 2 (B R ) of (3.11) among all functions with trace z 1 (δ) on ∂ B R . Moreover, there exists R 0 > 0 such that, for any R ≥ R 0 and for any δ ∈ [0, δ 0 ), • there exists a solution β R of (3.10) with δ = 0 such that Proof Existence follows from coercivity and weak lower semi continuity. By the fact that W δ ≡ α in (−∞, μ(δ)) and (3.8), we can see the minimiser actually has to satisfy z 1 (δ) ≤ β R,δ ≤ z 3 (δ), thus, due to the strong maximum principle, either (3.12) holds or β R,δ ≡ z 1 (δ). Now we prove (3.13), which, in particular, shows that β R,δ > z 1 (δ) in B R , at least for R ≥ R 0 . In order to do so, we assume that there exists a sequence R k → ∞ and a sequence δ k ∈ [0, δ 0 ) such that It follows that, on the one hand (3.15) where ω N denotes the surface of S N −1 . On the other hand, if, for R > 1 and δ ∈ [0, δ 0 ), we take w R,δ to be equal to z 1 (δ) on ∂ B R and to z 3 (δ) in B R−1 with |∇w R,δ | bounded uniformly in δ, we can see that there exists a constant C > 0 such that, for k large enough, This contradicts the minimality of β R k ,δ k . Finally we prove (3.14). In the forthcoming argument, R > 0 will always be arbitrary but fixed. We observe that, since β R,δ is bounded uniformly in R > 0 and δ > 0, then any sequence δ k → 0 admits a subsequence, that we still denote by δ k , such that β R,δ k converges in Since the convergence is uniform and (3.12) holds, then as δ → 0. Moreover, by (3.13) and the strong maximum principle,

Now we can prove Proposition 12.
Proof It is enough to prove that, if there exists a sequence δ k → 0, a sequence u δ k of solutions to (1.1) and ν > −1 such that then there exists a subsequence δ k such that u δ k ≡ z 3 (δ k ).
Claim For any ε > 0 and ρ > 0, there exists a subsequence, which we still denote by u δ k , and a sequence x k ∈ R N such that Therefore the sequence u k (x) := u δ k (x +x k ) admits a subsequence converging, in C 2 loc (R N ), to a solution u ∞ to the Allen-Cahn equation (3.20) By (3.19), we can see that u ∞ (0) = 1, thus u ∞ ≡ 1. As a consequence, for any ε > 0 (small) and ρ > 0, there exists a subsequence (still denoted by u k ) such that hence the claim is true. In order to prove our result, we first observe that, by (3.13), for δ 0 small as in Lemma 19 and δ ∈ (0, δ 0 ), there exists R > 0 and a solution β R,δ to (3.10) such that Moreover, by (3.14), there exists a solution β R to and δ 1 = δ 1 (R) > 0 such that, for any δ ∈ (0, δ 1 ), we have As a consequence, for any δ ∈ (0,δ), whereδ =δ(R) := min{δ 0 , δ 1 (R)}, we get Now, applying the claim with ρ = R and we can prove the existence of a subsequence, still denoted by u δ k , and a sequence x k in R N such that Sliding β R,δ k , with k ≥ k 0 fixed, we get the lower bound In conclusion, by Lemma 18, u δ k ≡ z 3 (δ k ).
• u δ is periodic in x N . Then u δ → −1 as δ → 0, uniformly on compact subsets of R N . Proof By Lemma 13, the family u δ is uniformly bounded, hence any sequence δ k → 0 admits a subsequence, that we still denote by δ k , such that u δ k converges in C 2 loc (R N ) to a solution u ∞ to the Allen-Cahn equation (3.20). Since u δ are all non constant solutions, then, by Proposition 12, we have inf R N u δ → −1, as δ → 0.