The estimates on the energy functional of an elliptic system with Neumann boundary conditions

Boundary Value Problems20132013:194

DOI: 10.1186/1687-2770-2013-194

Received: 19 June 2013

Accepted: 13 August 2013

Published: 28 August 2013

Abstract

We consider an elliptic system of the form ε 2 Δ u + u = f ( v ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq1_HTML.gif, ε 2 Δ v + v = g ( u ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq2_HTML.gif in Ω with Neumann boundary conditions, where Ω is a C 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq3_HTML.gif domain in R N http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq4_HTML.gif, f and g are nonlinearities having superlinear and subcritical growth at infinity. We prove the existence of nonconstant positive solutions of the system, and estimate the energy functional on a configuration space H ¯ http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq5_HTML.gif by a different technique, which is an important step in the proof of the solution’s concentrative property. We conclude that the least energy solutions of the system concentrate at the point of boundary, which maximizes the mean curvature of Ω.

Keywords

elliptic system estimates energy functional

1 Introduction

We are concerned with the following singularly perturbed system with Neumann conditions:
{ ε 2 Δ u + u = g ( v ) , in  Ω , ε 2 Δ v + v = f ( u ) , u , v > 0 ,  in  Ω , u n = v n = 0 , on  Ω , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ1_HTML.gif
(1.1)

where ε > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq6_HTML.gif is a small parameter, Ω is a C 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq3_HTML.gif bounded domain in R N http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq4_HTML.gif ( N 3 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq7_HTML.gif), 0 Ω http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq8_HTML.gif. f and g are nonlinearities having superlinear and subcritical growth at infinity.

Problem (1.1) arises in many applied models concerning biological pattern formations. For example, the steady states in the Keller-Segel model, the Gierer-Meinhardt model, see [1, 2] for more details. Problem (1.1) has been studied extensively for last twenty years. The motivation for the study of such a problem goes back to the pioneering work of [2, 3] concerning the scalar case (single equation),
{ ε 2 Δ u + u = u p , u > 0 ,  in  Ω , u n = 0 , on  Ω . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ2_HTML.gif
(1.2)

They proved a priori estimates, existence of least energy solutions and the concentrative properties of the solution. Furthermore, in [4, 5], Ni and Takagi proved the existence of a nontrivial solution u ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq9_HTML.gif to problem (1.2) for ε small enough. They showed that u ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq9_HTML.gif attains its maximum value at a point P ε Ω http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq10_HTML.gif, and the subsequences of P ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq11_HTML.gif converge to P, which is the maximum point of mean curvature on Ω.

The subject was studied by many authors for both Neumann and Dirichlet boundary conditions. There are many well-known results about (1.2). Del Pino and Felmer in [6] introduced shorter and more elementary arguments with respect to those in [4, 7]. Wang in [8] obtained multiple solutions of (1.2) by using Ljusternik-Schnirelman method. In [9], Grossi et al., obtained a solution of (1.2) with k maxima points, k is a given positive integer. We refer the reader to [1014] for further references.

As far as we know, Avial and Yang [15] were the first to approach the singularly perturbed system (1.1) with Neumann boundary conditions; they considered (1.1) with special nonlinearities f ( s ) = s p 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq12_HTML.gif, g ( s ) = s q 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq13_HTML.gif ( 2 < p , q < 2 N / N 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq14_HTML.gif). By means of a dual variational formulation, they proved that there exist nontrivial positive solutions u ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq15_HTML.gif and v ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq16_HTML.gif in C 2 ( Ω ¯ ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq17_HTML.gif, which have global maximum point at different points.

A more direct approach was proposed in [1618]. In these papers, the authors extend the idea, which is introduced by Del Pino and Felmer in [6], to system (1.1). In [18], Pistoia and Ramos proved the least energy solutions of system (1.1) concentrate at a point of the boundary, which maximizes the mean curvature of the boundary of Ω. Pistoia and Ramos [19] consider system (1.1) with Dirichlet boundary condition, they proved the existence of the least energy solutions. The solutions are concentrated, as ε goes to zero, at a point of Ω, which is maximized in distance to the boundary of Ω.

Let us recall the idea mentioned in [35], their proof based on the well-known result of Gidas et al. [20], that is, the uniqueness solution of the equation
{ Δ w + w = w p , w > 0 ,  in  R N , w ( 0 ) = max y R N w ( y ) , w 0 ,  as  | y | + , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ3_HTML.gif
(1.3)
here w is radially symmetric and is strictly decreasing, and w ( r ) < 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq18_HTML.gif for r > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq19_HTML.gif. But the uniqueness result for the following system corresponding to (1.1) is not known.
{ Δ u + u = g ( v ) , in  R N , Δ v + v = f ( u ) , in  R N , u , v > 0 , in  R N . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equa_HTML.gif

Besides, it is known that the underlying minimax theorem associated to ground-state level of (1.1) is an infinite-dimensional linking, this is in contrast with (1.2). We refer the reader to [18, 21] for more details on this.

In this paper, we prove the existence of nonconstant positive solutions u ε , v ε C 2 ( Ω ) H 1 ( Ω ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq20_HTML.gif of system (1.1), and estimate the energy functional of (1.1) on the configuration space H ¯ http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq5_HTML.gif (defined in Section 2) by a different technique, which is compared with [18]. This estimation is an important step in the proof of H ( P 0 ) = max P Ω H ( P ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq21_HTML.gif, where H ( P ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq22_HTML.gif denotes the mean curvature of Ω at the boundary point P. We conclude the least energy solutions of system (1.1), concentrated at the point of boundary, which maximizes the mean curvature of the boundary of Ω.

2 Statement of main results

The assumption to f , g C 1 ( R N ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq23_HTML.gif is a typical superlinear subcritical one, as in [18], we assume that the following holds.

(S1) f ( s ) = g ( s ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq24_HTML.gif, for s 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq25_HTML.gif. f ( 0 ) = 0 = f ( 0 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq26_HTML.gif, g ( 0 ) = 0 = g ( 0 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq27_HTML.gif. There exist two real numbers l 1 , l 2 > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq28_HTML.gif, 2 < p , q < 2 N / ( N 2 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq29_HTML.gif such that
lim | s | f ( s ) | s | 2 p = l 1 , lim | s | g ( s ) | s | 2 q = l 2 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equb_HTML.gif
(S2) For some δ > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq30_HTML.gif, s R http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq31_HTML.gif, s 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq32_HTML.gif,
f ( s ) s 2 ( 1 + δ ) f ( s ) s > 0 , g ( s ) s 2 ( 1 + δ ) g ( s ) s > 0 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equc_HTML.gif

(S3) f 2 ( s ) 2 f ( s ) F ( s ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq33_HTML.gif, g 2 ( s ) 2 g ( s ) G ( s ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq34_HTML.gif, for s R http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq31_HTML.gif, s 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq32_HTML.gif, where F ( s ) = 0 s f ( t ) d t http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq35_HTML.gif, G ( s ) = 0 s g ( t ) d t http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq36_HTML.gif.

Remark 2.1 Examples of nonlinearities satisfying (S1)-(S3) are
f ( s ) = a 1 | s | b 1 + | s | p 1 , g ( s ) = a 2 | s | b 2 + | s | q 1 , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equd_HTML.gif

where a 1 , a 2 > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq37_HTML.gif, 2 < b 1 p 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq38_HTML.gif, 2 < b 2 q 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq39_HTML.gif, 2 < p , q < 2 N / ( N 2 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq29_HTML.gif.

We should point out that (S1)-(S3) are the natural extension of the assumptions for the scalar case (single equation). Let us recall the assumptions on single equation such as (1.2).

Assume that f : R R http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq40_HTML.gif is continuous and satisfies the following structure assumptions.

(f1) f ( t ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq41_HTML.gif for t 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq42_HTML.gif and f ( t ) = o ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq43_HTML.gif near t = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq44_HTML.gif. f ( t ) = O ( t s ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq45_HTML.gif as t http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq46_HTML.gif, for some 1 < s < ( N + 2 ) / ( N 2 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq47_HTML.gif if N > = 3 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq48_HTML.gif, and s > 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq49_HTML.gif if N = 1 , 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq50_HTML.gif.

(f2) There exists a constant θ > 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq51_HTML.gif such that θ F ( t ) t f ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq52_HTML.gif for t 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq53_HTML.gif, in which F ( t ) = 0 t f ( s ) d s http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq54_HTML.gif.

(f3) The function t f ( t ) / t http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq55_HTML.gif is strictly increasing.

Remark 2.2 Assumption (S1) is the ‘system edition’ of (f1). (f2) is the famous Ambrosetti-Rabinowitz superlinear condition [22], which has appeared in most of studies for superlinear problems. In fact, it implies that the super-quadratic condition on F ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq56_HTML.gif. It has been used in a crucial way not only in establishing the mountain-pass geometry of the functional, but also in obtaining bounds of (PS) sequences. Assumption (S2) implies that f ( s ) s ( 2 + δ ) F ( s ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq57_HTML.gif, g ( s ) s ( 2 + δ ) G ( s ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq58_HTML.gif, which play a important roll in the proof of the existence of system’s solutions. So it is the ‘system edition’ of (f2).

Without loss of generally, we may assume that 0 Ω http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq8_HTML.gif. By the following rescaling:
z = ε x , x Ω ε : = { ε x Ω } , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Eque_HTML.gif
equation (1.1) becomes
{ Δ u + u = g ( v ) , in  Ω ε , Δ v + v = f ( u ) , u , v > 0  in  Ω ε , u n = v n = 0 , on  Ω ε . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ4_HTML.gif
(2.1)
To simplify the notations, we define Ω : = Ω ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq59_HTML.gif. Associated with (2.1) is the energy functional
I ( u , v ) = Ω ( u , v + u v ) Ω ( F ( u ) + G ( v ) ) , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ5_HTML.gif
(2.2)

(2.2) is a C 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq3_HTML.gif functional defined over the Hilbert space H : = H 1 ( Ω ) × H 1 ( Ω ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq60_HTML.gif.

We define the norm
( u , v ) 2 : = u 2 + v 2 , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equf_HTML.gif

where u 2 = Ω ( | u | 2 + u 2 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq61_HTML.gif.

It can be observed that the following orthogonal splitting holds: H = H H + http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq62_HTML.gif, here, H : = { ( ϕ , ϕ ) , ϕ H 1 ( Ω ) } http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq63_HTML.gif, H + : = { ( ϕ , ϕ ) , ϕ H 1 ( Ω ) } http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq64_HTML.gif. We set H ˜ = { t ( u , v ) t R + , u , v H 1 ( Ω ) , u v } http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq65_HTML.gif, H ¯ = H H ˜ http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq66_HTML.gif.

Theorem 2.3 Assume (S1)-(S3), then there exists ε 0 > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq67_HTML.gif, such that for any 0 < ε < ε 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq68_HTML.gif, system (1.1) has nonconstant positive solutions u ε , v ε C 2 ( Ω ) H 1 ( Ω ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq69_HTML.gif. If u ε v ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq70_HTML.gif, the estimation of the energy functional on H ¯ http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq5_HTML.gif is
I ( u ε , v ε ) = sup H ¯ I . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ6_HTML.gif
(2.3)

Remark 2.4 We point out that, in contrast with Theorem 2.3, only constant positive solutions are expected to exist for large values of ε [3, 15].

Remark 2.5 The estimation (2.3) is an important step in the proof of
H ( P 0 ) = max P Ω H ( P ) , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equg_HTML.gif

where H ( P ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq22_HTML.gif denotes the mean curvature of Ω at the boundary point P. So we can conclude the least energy solutions of system (1.1) concentrate at a point of the boundary, which maximizes the mean curvature of the boundary of Ω.

3 Proof of Theorem 2.3

To prove Theorem 2.3, we need the following lemma. Let u ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq9_HTML.gif, v ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq71_HTML.gif be the solutions of (1.1), in order to simplify the notations, we define u : = u ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq72_HTML.gif, v : = v ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq73_HTML.gif,
I 1 ( t ) : = I ( ( 1 t ) ϕ + t u , ( 1 t ) ( ϕ ) + t v ) , ϕ H 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ7_HTML.gif
(3.1)

and u ¯ = ϕ + t ( u ϕ ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq74_HTML.gif, v ¯ = ϕ + t ( v + ϕ ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq75_HTML.gif.

Lemma 3.1 Supposing the assumptions in Theorem 2.3 hold, I 1 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq76_HTML.gif admits the following properties:

(i)1 I 1 ( 1 ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq77_HTML.gif, I 1 ( 0 ) 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq78_HTML.gif, I 1 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq79_HTML.gif,

(ii)1 I 1 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq80_HTML.gif, t http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq81_HTML.gif,

(iii)1 I 1 ( 1 ) < 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq82_HTML.gif.

Proof Proof of (i)1. By the definition of I 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq83_HTML.gif, I 1 ( t ) = I ( u ¯ , v ¯ ) ( u ϕ , v + ϕ ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq84_HTML.gif, then I 1 ( 1 ) = I ( u , v ) ( u ϕ , v + ϕ ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq85_HTML.gif, where u and v are the solutions of (1.1), that follows I 1 ( 1 ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq77_HTML.gif.

Again by (3.1),
I 1 ( 0 ) = I ( ϕ , ϕ ) = Ω ( ϕ , ( ϕ ) + ϕ ( ϕ ) ) Ω ( F ( ϕ ) + G ( ϕ ) ) = ϕ 2 Ω ( F ( ϕ ) + G ( ϕ ) ) , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equh_HTML.gif

by (S2), we have F 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq86_HTML.gif, G 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq87_HTML.gif, so I 1 ( 0 ) 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq78_HTML.gif.

Next, we want to show I 1 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq79_HTML.gif.
I 1 ( 1 ) = Ω ( u , v + u v ) Ω ( F ( u ) + G ( v ) ) = Ω ( 1 2 u , v + 1 2 u v ) + Ω ( 1 2 u , v + 1 2 u v ) Ω ( F ( u ) + G ( v ) ) = 1 2 Ω Δ u + u , v + 1 2 Ω Δ v + v , u Ω ( F ( u ) + G ( v ) ) = 1 2 Ω g ( v ) v + 1 2 Ω f ( u ) u Ω ( F ( u ) + G ( v ) ) = Ω ( 1 2 f ( u ) u F ( u ) ) + Ω ( 1 2 g ( v ) v G ( v ) ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ8_HTML.gif
(3.2)
From (S2), we can deduce there exist δ > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq30_HTML.gif, such that f ( t ) t ( 1 + δ ) f ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq88_HTML.gif, then
0 s f ( t ) t ( 1 + δ ) 0 s f ( t ) = ( 1 + δ ) F ( s ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equi_HTML.gif
The left side in last inequality is equal to f ( s ) s F ( s ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq89_HTML.gif, so f ( s ) s ( 2 + δ ) F ( s ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq90_HTML.gif. By the same way, g ( s ) s ( 2 + δ ) G ( s ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq91_HTML.gif. Then (3.2) can change to
I 1 ( 1 ) ( 1 2 1 2 + δ ) Ω ( f ( u ) u + g ( v ) v ) > 0 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equj_HTML.gif

So, we get the properties (i)1.

To prove (ii)1 is equal to show
I ( u ¯ , v ¯ ) , t . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equk_HTML.gif
We only need to prove for 2 < p < 2 N / ( N 2 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq92_HTML.gif,
lim inf t 1 | 1 t | p Ω ( F ( u ¯ ) + G ( v ¯ ) ) > 0 , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ9_HTML.gif
(3.3)
By (S1), there exist C such that
1 | 1 t | p Ω F ( ( 1 t ) ϕ + t u ) = C | 1 t | p Ω | ( 1 t ) ϕ + t u | p + o ( 1 ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equl_HTML.gif
As the same,
1 | 1 t | p Ω G ( ( t 1 ) ϕ + t v ) = C | 1 t | p Ω | ( t 1 ) ϕ + t v | p + o ( 1 ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equm_HTML.gif
We prove (3.3) by contradiction. Assume that there exists a subsequence t n http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq93_HTML.gif such that
1 | 1 t n | p Ω ( F ( ϕ + t n ( u ϕ ) ) + G ( ϕ + t n ( v + ϕ ) ) ) 0 , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equn_HTML.gif
that is,
Ω ( | ϕ + t n 1 t n u | p + | ϕ + t n 1 t n v | p ) + o ( 1 ) = 0 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equo_HTML.gif
So, for any ϕ H 1 ( Ω ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq94_HTML.gif, we have
Ω ( | ϕ + t n 1 t n u | p + | ϕ + t n 1 t n v | p ) = o ( 1 ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equp_HTML.gif
Take ϕ = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq95_HTML.gif in the last equality, then
| t n 1 t n | p Ω ( u p + v p ) = o ( 1 ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equq_HTML.gif

For p is a number between 2 and 2 N / ( N 2 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq96_HTML.gif, u 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq97_HTML.gif, we deduce u = v http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq98_HTML.gif. It contradicts with the original assumption of u v http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq99_HTML.gif.

Now, we turn to the prove of (iii)1. By (3.1),
I 1 ( t ) = Ω ( ( ϕ + t ( u ϕ ) ) , ( ϕ + t ( v + ϕ ) ) + ( ϕ + t ( u ϕ ) ) ( ϕ + t ( v + ϕ ) ) ) Ω ( F ( u ¯ ) + G ( v ¯ ) ) = Ω ( ϕ , ϕ + ϕ , t ( v + ϕ ) + t ( u ϕ ) , ϕ ) + Ω t ( u ϕ ) , t ( v + ϕ ) + Ω ( ϕ ( ϕ ) + t ϕ ( v + ϕ ) t ϕ ( u ϕ ) ) + Ω t 2 ( u ϕ ) ( v + ϕ ) Ω ( F ( u ¯ ) + G ( v ¯ ) ) , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equr_HTML.gif
then
I 1 ( t ) = 2 Ω ( ( u ϕ ) , ( v + ϕ ) + ( u ϕ ) ( v + ϕ ) ) Ω f ( u ¯ ) ( u ϕ ) 2 Ω g ( v ¯ ) ( v + ϕ ) 2 = 2 Ω ( u , v + u , ϕ + ϕ , v + ϕ , ϕ ) + 2 Ω ( u v + u ϕ ϕ v ϕ 2 ) Ω f ( u ¯ ) ( u ϕ ) 2 Ω g ( v ¯ ) ( v + ϕ ) 2 = Ω ( ( Δ v + v ) u + ( Δ u + u ) v + 2 ( Δ u + u ) ϕ + 2 ( Δ v v ) ϕ ) + 2 Ω ( | ϕ | 2 ϕ 2 ) Ω f ( u ¯ ) ( u ϕ ) 2 Ω g ( v ¯ ) ( v + ϕ ) 2 = 2 ϕ 2 + Ω ( f ( u ) u + g ( v ) v + 2 g ( v ) ϕ 2 f ( u ) ϕ ) Ω f ( u ¯ ) ( u ϕ ) 2 Ω g ( v ¯ ) ( v + ϕ ) 2 , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equs_HTML.gif
take t = 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq100_HTML.gif, then
I 1 ( 1 ) = 2 ϕ 2 + Ω ( f ( u ) ( u 2 ϕ ) f ( u ) ( u ϕ ) 2 ) + Ω g ( v ) ( v + 2 ϕ ) Ω g ( v ) ( v ( ϕ ) ) 2 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ10_HTML.gif
(3.4)
By (S2), for some δ > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq30_HTML.gif,
Ω f ( u ) ( u 2 ϕ ) f ( u ) ( u ϕ ) 2 Ω δ f ( u ) u , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equt_HTML.gif
in the same way,
Ω g ( v ) ( v + 2 ϕ ) g ( v ) ( v ( ϕ ) ) 2 Ω δ g ( v ) v . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equu_HTML.gif

Thus, we obtain I 1 ( 1 ) < 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq82_HTML.gif. □

Proposition 3.2 (Theorem 1.1 in [18])

Under assumptions (H), there exists ε 0 > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq67_HTML.gif such that for any 0 < ε < ε 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq68_HTML.gif, problem (1.1) has nonconstant positive solutions u ε , v ε C 2 ( Ω ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq101_HTML.gif. Moreover, both functions u ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq9_HTML.gif and v ε http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq71_HTML.gif attain their maximum value at some unique and common point P ε Ω http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq102_HTML.gif. (The assumption (H) is composed of (S1), (S3) and the following (3.5).)

Remark 3.3 We will compare our assumptions (S1)-(S3) with the conditions (H) of Proposition 3.2 in the following proof of Theorem 2.3.

Proof of Theorem 2.3 The existence of solutions of (1.1) can follow the steps of Theorem 1.1 in [18]. They use some ideas introduced by Del Pino and Felmer [6], and differ from the method of Ni and Takagi. It needs to be pointed out that (S2) implies the following conditions:

For some δ > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq30_HTML.gif, s R http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq31_HTML.gif, s 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq32_HTML.gif,
f ( s ) s ( 2 + δ ) F ( s ) > 0 , g ( s ) s ( 2 + δ ) G ( s ) > 0 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ11_HTML.gif
(3.5)

The assumption (H) in Proposition 3.2 is composed of (S1), (S3) and (3.5). By Proposition 3.2, the existence of solutions can be proved under (H). So, we can get the existence of solutions of (1.1) under (S1)-(S3).

The rest of the paper is devoted to the proof of (2.3). By the definition of space H ¯ http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq5_HTML.gif, we only need to prove for any ϕ H 1 ( Ω ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq94_HTML.gif, t R http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq103_HTML.gif, t > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq104_HTML.gif the following holds,
I ( ϕ + t u , ϕ + t v ) I ( u , v ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ12_HTML.gif
(3.6)
Obviously t 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq105_HTML.gif. By (3.1), (3.6) is equal to
I 1 ( t ) I 1 ( 1 ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ13_HTML.gif
(3.7)
We prove (3.7) by contradiction, suppose that the maximum point of I 1 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq76_HTML.gif is t 0 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq106_HTML.gif, and t 0 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq107_HTML.gif.
I 1 ( t ) = Ω ( ϕ , ( v + ϕ ) + ( u ϕ ) , ϕ + 2 t ( u ϕ ) , ( v + ϕ ) ) + Ω ( ϕ ( v + ϕ ) ϕ ( u ϕ ) + 2 t ( u ϕ ) ( v + ϕ ) ) Ω f ( u ¯ ) ( u ϕ ) Ω g ( v ¯ ) ( v + ϕ ) = Ω ( ( ϕ + t ( u ϕ ) ) , ( v + ϕ ) + ( u ϕ ) , ( ϕ + t ( v + ϕ ) ) ) + Ω ( ( u ϕ ) ( ϕ + t ( v + ϕ ) ) + ( v + ϕ ) ( ϕ + t ( u ϕ ) ) ) Ω f ( u ¯ ) ( u ϕ ) Ω g ( v ¯ ) ( v + ϕ ) = Ω ( u ¯ , ( v + ϕ ) + ( u ϕ ) , v ¯ + ( u ϕ ) v ¯ + ( v + ϕ ) u ¯ ) Ω ( f ( u ¯ ) ( u ϕ ) + g ( v ¯ ) ( v + ϕ ) ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equv_HTML.gif
Let I 1 ( t ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq108_HTML.gif, we get
Ω ( u ¯ , ( v + ϕ ) + ( u ϕ ) , v ¯ + ( u ϕ ) v ¯ + ( v + ϕ ) u ¯ ) = Ω ( f ( u ¯ ) ( u ϕ ) + g ( v ¯ ) ( v + ϕ ) ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ14_HTML.gif
(3.8)
Thus, by (3.8), we have
2 I ( u ¯ , v ¯ ) = Ω ( u ¯ , ( ϕ + t ( v + ϕ ) ) + u ¯ ( ϕ + t ( v + ϕ ) ) + ( ϕ + t ( u ϕ ) ) , v ¯ ) + Ω ( ϕ + t ( u ϕ ) ) v ¯ 2 Ω ( F ( u ¯ ) + G ( v ¯ ) ) = Ω ( u ¯ , ( ϕ ) + u ¯ ( ϕ ) + ϕ , v ¯ + ϕ v ¯ ) + Ω u ¯ , t ( v + ϕ ) + Ω ( u ¯ t ( v + ϕ ) + t ( u ϕ ) , v ¯ + t ( u ϕ ) v ¯ ) 2 Ω ( F ( u ¯ ) + G ( v ¯ ) ) = Ω ( u ¯ , ( ϕ ) + u ¯ ( ϕ ) + ϕ , v ¯ + ϕ v ¯ ) + Ω f ( u ¯ ) t ( u ϕ ) + Ω g ( v ¯ ) t ( v + ϕ ) 2 Ω ( F ( u ¯ ) + G ( v ¯ ) ) = : 2 I 2 ( t ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ15_HTML.gif
(3.9)

We claim that the function I 2 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq109_HTML.gif, defined in (3.9), has the following properties:

(i)2 I 2 ( 0 ) 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq110_HTML.gif, I 2 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq111_HTML.gif, I 2 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq112_HTML.gif;

(ii)2 For any t 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq53_HTML.gif, if I 2 ( t ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq113_HTML.gif, then I 2 ( t ) 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq114_HTML.gif;

(iii)2 I 2 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq115_HTML.gif, as t http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq81_HTML.gif.

First, we prove (i)2. In fact, by (3.1) and (3.9),
I 2 ( 0 ) = 1 2 { Ω 2 ( ϕ , ( ϕ ) + ϕ ( ϕ ) ) 2 Ω ( F ( ϕ ) + G ( ϕ ) ) } = ϕ 2 Ω ( F ( ϕ ) + G ( ϕ ) ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equw_HTML.gif

Following from assumption (S2), we obtain I 2 ( 0 ) 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq110_HTML.gif.

Again by (3.9) and Lemma 3.1, I 2 ( 1 ) = I 1 ( 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq116_HTML.gif and I 1 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq79_HTML.gif, so I 2 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq111_HTML.gif. Next, we want to compute I 2 ( 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq117_HTML.gif. From (3.9),
2 I 2 ( t ) = Ω ( ϕ , ( ϕ ) + ( t ( u ϕ ) ) , ( ϕ ) + ϕ ( ϕ ) + t ( u ϕ ) ( ϕ ) ) + Ω ( ϕ , ( ϕ ) + ϕ , ( t ( v + ϕ ) ) + ϕ ( ϕ ) + ϕ t ( v + ϕ ) ) + Ω ( f ( u ¯ ) t ( u ϕ ) + g ( v ¯ ) t ( v + ϕ ) ) 2 Ω ( F ( u ¯ ) + G ( v ¯ ) ) , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ16_HTML.gif
(3.10)
2 I 2 ( t ) = Ω ( ( u ϕ ) , ( ϕ ) + ( u ϕ ) ( ϕ ) ) + Ω ( ϕ , ( v + ϕ ) + ϕ ( v + ϕ ) ) + Ω ( f ( u ¯ ) t ( u ϕ ) 2 f ( u ¯ ) ( u ϕ ) + g ( v ¯ ) t ( v + ϕ ) 2 g ( v ¯ ) ( v + ϕ ) ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ17_HTML.gif
(3.11)
Then
2 I 2 ( 1 ) = Ω ( ( u ϕ ) , ( ϕ ) + ( u ϕ ) ( ϕ ) ) + Ω ( ϕ , ( v + ϕ ) + ϕ ( v + ϕ ) ) + Ω ( f ( u ) ( u ϕ ) 2 f ( u ) ( u ϕ ) + g ( v ) ( v + ϕ ) 2 g ( v ) ( v + ϕ ) ) = 2 ϕ 2 + Ω ( Δ u , ϕ + u ( ϕ ) + Δ v , ϕ + v ϕ ) + Ω ( f ( u ) ϕ + g ( v ) ( ϕ ) ) + Ω ( f ( u ) ( u ϕ ) 2 + g ( v ) ( v + ϕ ) 2 f ( u ) u g ( v ) v ) = 2 ϕ 2 + 2 Ω ( f ( u ) ϕ + g ( v ) ( ϕ ) ) + Ω ( f ( u ) ( u ϕ ) 2 + g ( v ) ( v + ϕ ) 2 ) Ω ( f ( u ) u + g ( v ) v ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equx_HTML.gif

Combined with (3.4), 2 I 2 ( 1 ) = I 1 ( 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq118_HTML.gif. By Lemma 3.1, I 1 ( 1 ) < 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq82_HTML.gif, so I 2 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq112_HTML.gif.

Now, we turn to the proof of (ii)2. Let I 2 ( t ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq113_HTML.gif, by (3.11),
t Ω ( ( u ϕ ) , ( ϕ ) + ( u ϕ ) ( ϕ ) ) + t Ω ( ϕ , ( v + ϕ ) + ϕ ( v + ϕ ) ) = Ω ( f ( u ¯ ) t 2 ( u ϕ ) 2 t f ( u ¯ ) ( u ϕ ) + g ( v ¯ ) t 2 ( v + ϕ ) 2 t g ( v ¯ ) ( v + ϕ ) ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equy_HTML.gif
Then by (3.10),
2 I 2 ( t ) = ϕ 2 Ω ( f ( u ¯ ) t 2 ( u ϕ ) 2 2 t f ( u ¯ ) ( u ϕ ) + 2 F ( u ¯ ) ) Ω ( g ( v ¯ ) t 2 ( v + ϕ ) 2 2 t g ( v ¯ ) ( v + ϕ ) + 2 G ( v ¯ ) ) . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ18_HTML.gif
(3.12)
By (S3), we get
f ( u ¯ ) t 2 ( u ϕ ) 2 + 2 F ( u ¯ ) 2 t ( u ϕ ) 2 f ( u ¯ ) F ( u ¯ ) 2 t ( u ϕ ) f ( u ¯ ) , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equz_HTML.gif
similarly,
g ( v ¯ ) t 2 ( v + ϕ ) 2 + 2 G ( v ¯ ) 2 t ( v + ϕ ) g ( v ¯ ) , http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equaa_HTML.gif

then by (3.12), I 2 ( t ) 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq114_HTML.gif. We proved the property (ii)2.

The proof of (iii)2 is similar to (ii)1. Then we complete the proof of the claim.

Suppose that the maximum point of I 1 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq76_HTML.gif is t 0 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq106_HTML.gif, t 0 1 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq107_HTML.gif, then either t 0 ( 0 , 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq119_HTML.gif or t 0 ( 1 , + ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq120_HTML.gif. If t 0 ( 0 , 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq121_HTML.gif, I 1 ( t 0 ) > I 1 ( 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq122_HTML.gif, by (3.1), (3.9) and property (i)2 of I 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq123_HTML.gif, I 2 ( t 0 ) > I 2 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq124_HTML.gif, I 2 ( 0 ) 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq110_HTML.gif. By the property (iii)2 of I 2 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq123_HTML.gif, there exist t ˆ > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq125_HTML.gif, such that
I 2 ( t ˆ ) = 0 and I 2 ( t ˆ ) > 0 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ19_HTML.gif
(3.13)
If t 0 ( 1 , + ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq120_HTML.gif, then I 1 ( t 0 ) > I 1 ( 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq122_HTML.gif. By Lemma 3.1, I 1 ( 1 ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq126_HTML.gif, I 1 ( 1 ) < 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq127_HTML.gif, so there exist t 1 ( 1 , t 0 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq128_HTML.gif, such that I 1 ( t 1 ) < I 1 ( 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq129_HTML.gif, and I 1 ( t 1 ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq130_HTML.gif. Thus, I 2 ( t 1 ) < I 2 ( 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq131_HTML.gif. I 2 ( 1 ) > 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq112_HTML.gif, so there exist t ˜ ( 1 , t 1 ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq132_HTML.gif, such that
I 2 ( t ˜ ) = 0 and I 2 ( t ˜ ) > 0 . http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_Equ20_HTML.gif
(3.14)

In fact, (3.13) and (3.14) is a contradiction to the nature (ii)2 of I 2 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq109_HTML.gif, that is, for any t 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq53_HTML.gif, if I 2 ( t ) = 0 http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq113_HTML.gif, the value of I 2 ( t ) http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq109_HTML.gif must be smaller than 0.

Having reached a contradiction, this completes the proof of Theorem 2.3. □

Declarations

Acknowledgements

The author is supported by the project of ‘Youth Innovation,’ funded by the Department of Science and Technology of Fujian province (2011J05003), and supported by the Projects A of the Educational Department of Fujian Province (JA11053).

Authors’ Affiliations

(1)
School of Mathematics and Computer Sciences, Fujian Normal University

References

  1. Gierer A, Meinhardt H: A theory of biological pattern formation. Kybernetik 1972, 12: 30-39. 10.1007/BF00289234View Article
  2. Lin CS, Ni WM, Takagi I: Large amplitude stationary solutions to a chemotaxis system. J. Differential Equations 1988, 72: 1-27. 10.1016/0022-0396(88)90147-7MathSciNetView Article
  3. Ni WM, Takagi I: On the Neumann problem for some semilinear elliptic equations and systems of activator-inhibitor type. Trans. Amer. Math. Soc. 1986, 297: 351-368. 10.1090/S0002-9947-1986-0849484-2MathSciNetView Article
  4. Ni WM, Takagi I: On the shape of least-energy solutions to a semilinear Neumann problem. Comm. Pure Appl. Math. 1991, 44: 819-851. 10.1002/cpa.3160440705MathSciNetView Article
  5. Ni WM, Takagi I: Locating peaks of least-energy solutions to a semilinear Neumann problem. Duke Math. J. 1993, 70: 247-281. 10.1215/S0012-7094-93-07004-4MathSciNetView Article
  6. Del Pino M, Felmer P: Spike-layered solutions of singularly perturbed elliptic problems in a degenerate setting. Indiana Univ. Math. J. 1999, 48: 883-898.MathSciNetView Article
  7. Ni WM, Wei J: On the location and profile of spike-layer solutions to singularly perturbed semilinear Dirichlet problems. Comm. Pure Appl. Math. 1995, 48: 731-768. 10.1002/cpa.3160480704MathSciNetView Article
  8. Wang ZQ: On the existence of multiple single-peaked solution for a semilinear Neumann problem. Arch. Ration. Mech. Anal. 1992, 120: 375-399. 10.1007/BF00380322View Article
  9. Grossi M, Pistoia A, Wei J: Existence of multipeak solutions for a semilinear Neumann problem via nonsmooth critical point theory. Calc. Var. Partial Differential Equations 2000, 11: 143-175. 10.1007/PL00009907MathSciNetView Article
  10. Grossi M, Pistoia A: On the effect of critical points of distance function in superlinear elliptic problems. Adv. Differential Equ 2000, 5: 1397-1420.MathSciNet
  11. Li YY, Nirenberg L: The Dirichlet problem for singularly perturbed elliptic equations. Comm. Pure Appl. Math. 1998, 51: 1445-1490. 10.1002/(SICI)1097-0312(199811/12)51:11/12<1445::AID-CPA9>3.0.CO;2-ZMathSciNetView Article
  12. Wei J: On the boundary spike layer solutions to a singularly perturbed Neumann problem. J. Differential Equations 1997, 134: 104-133. 10.1006/jdeq.1996.3218MathSciNetView Article
  13. Wei J: On the interior spike layer solutions of a singularly perturbed semilinear Neumann problem. Tohoku Math. J. 1998, 50: 159-178. 10.2748/tmj/1178224971MathSciNetView Article
  14. Wei J: On the interior spike solutions for some singular perturbation problems. Proc. Roy. Soc. Edinb. Sect A 1998, 128: 849-874. 10.1017/S030821050002182XView Article
  15. Avila AI, Yang J: On the existence and shape of least energy solutions for some elliptic systems. J. Differential Equations 2003, 191: 348-376. 10.1016/S0022-0396(03)00017-2MathSciNetView Article
  16. Busca J, Sirakov B: Symmetry results for semilinear elliptic systems in the whole space. J. Differential Equations 2000, 163: 41-56. 10.1006/jdeq.1999.3701MathSciNetView Article
  17. Clement P, De Figueiredo DG, Mitidieri E: Positive solutions of semilinear elliptic systems. Comm. Partial Differential Equations 1992, 17: 923-940. 10.1080/03605309208820869MathSciNetView Article
  18. Pistoia A, Ramos M: Locating the peaks of the least energy solutions to an elliptic system with Neumann boundary conditions. J. Differential Equations 2004, 201: 160-176. 10.1016/j.jde.2004.02.003MathSciNetView Article
  19. Pistoia A, Ramos M: Locating the peaks of the least energy solutions to an elliptic system with Dirichlet boundary conditions. Nonlinear Differ. Equ. Appl. 2008, 15: 1-23. 10.1007/s00030-007-4066-8MathSciNetView Article
  20. Gidas B, Ni WM, Nirenberg L:Symmetry of positive solutions of non-linear elliptic equations in R n http://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2013-194/MediaObjects/13661_2013_Article_440_IEq133_HTML.gif. Adv. in Math. Suppl. Stud. 7. Mathematical Analysis and Applications. Part A 1981, 369-402.
  21. Abbondandolo A, Felmer P, Molina J: An estimate on the relative Morse index for strongly indefinite functionals. Electron. J. Differ. Equ. Conf. 2001, 6: 1-11.MathSciNet
  22. Ambrosetti A, Rabinowitz PH: Dual variational methods in critical point theory. J. Funct. Anal. 1973, 14: 349-381. 10.1016/0022-1236(73)90051-7MathSciNetView Article

Copyright

© Zeng; licensee Springer 2013

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://​creativecommons.​org/​licenses/​by/​2.​0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.