Home Mathematics Concentrating solutions for double critical fractional Schrödinger-Poisson system with p-Laplacian in ℝ3
Article Open Access

Concentrating solutions for double critical fractional Schrödinger-Poisson system with p-Laplacian in ℝ3

  • Shuaishuai Liang , Yueqiang Song and Shaoyun Shi EMAIL logo
Published/Copyright: February 1, 2025

Abstract

In this article, we consider the following double critical fractional Schrödinger-Poisson system involving p-Laplacian in R 3 of the form:

ε s p ( Δ ) p s u + V ( x ) u p 2 u ϕ u p s 2 u = u p s * 2 u + f ( u ) in R 3 , ε s p ( Δ ) s ϕ = u p s in R 3 ,

where ε > 0 is a positive parameter, s ( 3 4 , 1 ) , ( Δ ) p s is the fractional p-Laplacian operator, p ( 3 2 , 3 ) , p s * = 3 p 3 s p is the Sobolev critical exponent, p s = p 2 ( 3 + 2 s ) ( 3 s p ) is the upper exponent in the sense of the Hardy-Littlewood-Sobolev inequality, V ( x ) : R 3 R symbolizes a continuous potential function with a local minimum, and the continuous function f possesses subcritical growth. With the help of well-known penalization methods and Ljusternik-Schnirelmann category theory, we use the topological arguments to attain the multiplicity and concentration of the positive solutions for the above system.

MSC 2010: 35B33; 35J20; 35J60; 47J30; 58E05

1 Introduction

In this article, we are concerned with the multiplicity and concentration of positive solutions for the following fractional p-Laplacian Schrödinger-Poisson system involving double critical nonlinearities in R 3 :

(1.1) ε s p ( Δ ) p s u + V ( x ) u p 2 u ϕ u p s 2 u = u p s * 2 u + f ( u ) in R 3 , ε s p ( Δ ) s ϕ = u p s in R 3 ,

where ε > 0 is a parameter, 3 4 < s < 1 , p s * = 3 p 3 s p , and p s = p 2 ( 3 + 2 s ) ( 3 s p ) are the upper critical exponent in the sense of the Hardy-Littlewood-Sobolev inequality. ( Δ ) p s denotes the fractional p-Laplacian operator and generally defined in the following form:

( Δ ) p s u C ( s ) lim ε 0 + R 3 \ B ε ( x ) u ( x ) u ( y ) p 2 ( u ( x ) u ( y ) ) x y 3 + s p d y in R 3 ,

where C ( s ) is a normalization constant and P.V. is the Cauchy principal value.

Benci and Fortunato [8] proposed the following famous Schrödinger-Poisson system:

(1.2) 2 2 m Δ u e ϕ u = ω u , Δ ϕ = 4 π e u 2 ,

the system 1.2 as the model to describe the solitary waves for the nonlinear stationary Schrödinger equations interacting with the electrostatic field which is not a priori assigned. Here, m stands for the mass, e denotes the charge of the particle, ω means the phase, and is Plank’s constant. And system (1.2) is also known as the nonlinear Schrödinger-Maxwell system. For more results about the background of the Schrödinger-Poisson system, please refer to Benci and Fortunato [8].

In recent years, many scholars focused on the study of the Schrödinger-Poisson system. By using the monotonicity method introduced by Jeanjean [25] and a truncation argument, Azzollini et al. [7] considered a nontrivial positive radial solution for the following Schrödinger-Poisson system:

(1.3) Δ u + q ϕ u = u + u p 1 u in R 3 , Δ ϕ = q u 2 in R 3 ,

which was the first contribution to the existence of solutions for the Schrödinger-Poisson system. In [47], Ruiz considered the classical Schrödinger-Poisson system

Δ u + u + λ ϕ u = u p in R 3 , Δ ϕ = u 2 in R 3 .

Applying the variational method, Ruiz obtained the existence of the solution to the above problem in both cases p ( 1 , 3 ) and p = 5 3 . In some extent, the results of Ruiz extend [35,12,32]. By the Lyapunov-Schmidt reduction method, Jin and Wang [26] considered the existence of infinitely many nonradial positive solutions for the nonlinear Schrödinger-Poisson system. Wang et al. [43] proved the bifurcation and regularity properties of the entire solutions of the planar nonlinear Schrödinger-Poisson system in R 2 , which was the first contribution regarding the bifurcation of the planar nonlinear Schrödinger-Poisson system via the global bifurcation theorem. More precisely, they employed the Trudinger-Moser inequality of Cao [10] to deal with the nonlinearity f ( u ) satisfying exponent growth. Together with the Ljusternik-Schnirelmann theorem of Szulkin [49] and the Hardy-Littlewood-Sobolev inequality to obtain bifurcation and regularity of solutions for the planar nonlinear Schrödinger-Poisson system. In addition, there are some results on the existence and non-existence of solutions, multiplicity of solutions, semiclassical limit, and concentrations of solutions to the Schrödinger-Possion system [11,18,22,33,34,36,37,41,56,57].

Once we turn our attention on the fractional setting, there has some interesting results about the Schrödinger-Possion system. For example, Zhang et al. [55] explored the following fractional nonlinear Schrödinger-Poisson system:

(1.4) ( Δ ) s u + λ ϕ u = g ( u ) in R 3 , ( Δ ) t ϕ = λ u 2 in R 3 .

Under the Berestycki-Lions conditions (see Zhang et al. [55] (H2)–(H4)) combined with the perturbation approach, they obtained the existence of positive solutions and the asymptotics of solutions for system (1.4) as the nonlinearity g ( u ) satisfying subcritical and critical growth. Teng [51] was interested in the existence of ground-state solutions for the fractional Schrödinger-Poisson system with critical nonlinearity

( Δ ) s u + V ( x ) u + ϕ u = μ u q 1 u + u 2 s in R 3 , ( Δ ) t ϕ = u 2 in R 3 ,

where s , t ( 0 , 1 ) . Together with the method of the Pohozaev-Nehari manifold, the monotonic trick, and global compactness lemma, Teng obtained the existence of a nontrivial ground-state solution for the above system. For more results on the Schrödinger-Possion with critical nonlinearity, please refer [6,7,28,45].

On the other hand, to our best knowledge, most of the scholars consider the existence of solutions for Schrödinger-Possion with the nonlinearity ϕ u p 2 u satisfying the subcritical growth instead of critical growth. Therefore, there are few results on the double-critical Schrödinger-Poisson system. Qu and He [44] explored the following fractional Schrödinger-Poisson system with doubly critical growth:

ε 2 s ( Δ ) s u + V ( x ) u = ϕ u 2 s * 3 u + u 2 s * 2 u + f ( u ) , x R 3 , ε 2 s ( Δ ) s ϕ = u 2 s * 1 , x R 3 .

By the well-known Ljusternik-Schnirelmann category theory and the method of generalized Nehari mainfold from Szulkin [49], they obtain the concentration of solutions for the fractional Schrödinger-Poisson system. Meng and He [40] were concerned with the following Schrödinger-Poisson system:

(1.5) Δ u ϕ u 3 u = λ u + μ u q 2 u + u 4 u in R 3 , Δ ϕ = u 5 in R 3 .

With help of the truncation technique and the genus theory, they obtained the existence and multiplicity of normalized solutions for system (1.5). Feng [20] considered the critical Schrödinger-Poisson-type system of the form:

(1.6) Δ u + V ( x ) u ϕ u 3 u = u 4 u + f ( u ) in R 3 , Δ ϕ = u 5 in R 3 ,

where the partial potential V ( x ) = x 1 2 + x 2 2 + 1 , x = ( x 1 , x 2 , x 3 ) R 3 , and 2 = 6 is the Sobolev critical exponent. With the aid of the concentration-compactness principle and the mountain pass theorem, Feng obtained the existence of positive ground-state solution for system (1.6). There are also some results of the p-Laplacian Schrödinger system involving double critical nonlinearity, and we merely give some related on this topic. Bhakta et al. [9] investigated the following Schrödinger system with double critical nonlinearities:

(1.7) ( Δ ) p s u = u p s * 2 u + γ α p s * u α 2 u v β in Ω , ( Δ ) p s v = v p s * 2 v + γ β p s * u β 2 v u γ in Ω ,

where s ( 0 , 1 ) , p ( 1 , ) with N > s p , α , β > 1 such that α + β = p s * and Ω = R N or Ω R N is a smooth bounded domain. Employing the Nehari manifold, they obtained the existence of positive solutions of system (1.7) in both the cases γ = 1 and γ > 1 . Liu and Chen [38] considered the Choquard equation with double-critical exponents of the form:

(1.8) Δ p u + A x θ u p 2 u = ( I α * F ( u ) ) f ( u ) , x R N ,

where f satisfies the double critical growth due to the Hardy-Littlewood-Sobolev inequality. Making the suitable assumptions on f , together with the refined Sobolev inequality with Morrey norm and the generalized version of the Lions-type theorem, they proved the existence of ground-state solutions for the above problem. For more results, please see Liang and Rădulescu [27] and Su [48]. However, to our best knowledge, there are few results on the concentration of solutions for the Schrödinger-Poisson system with the fractional p-Laplacian and double critical nonlinearities.

Inspired by the aforementioned works, we consider concentrating solutions for the double critical fractional Schrödinger-Poisson system with p-laplacian in R 3 . To more precisely state our results, we assume that the potential V ( x ) satisfies the following conditions due to Del Pino and Felmer [14] and Rabinowitz [46]:

  1. V ( x ) : R 3 R is a continuous function and there exists V 0 , such that

    0 < V 0 inf x R 3 V ( x ) V ( x ) for all x R 3 .

  2. There is an open bounded domain Λ R 3 such that V 0 < min Λ V , Π = { x Π : V ( x ) = V 0 } 0 , and

    Π δ = { x R 3 : dist ( x , Π ) δ } Λ .

And the nonlinearity f C ( R , R ) fulfills the following conditions:

  1. lim t 0 + f ( t ) t 2 p 1 = 0 .

  2. There exist q 2 ( 2 p , p s * ) and μ > 0 such that

    f ( t ) μ t q 2 1 for any t > 0 .

  3. There exist c 0 > 0 and θ ( 2 p , p s * ) such that F ( t ) c 0 t p and f ( t ) t θ F ( t ) > 0 for all t > 0 , where F ( t ) = 0 τ f ( s ) d s .

  4. The function t f ( t ) t p 1 is strictly increasing in (0, ).

  5. There exists σ ( 2 p , p s * ) such that

    lim t f ( t ) t σ 1 = 0 .

In order to obtain the multiplicity of positive solutions of system (1.1), we first recall the result of the Ljusternik-Schnirelmann category. If Y is a closed set of a topological space X , we use cat X Y to denote the Ljusternik-Schnirelmann category of Y in X , namely, the least number of closed and contractible sets in X which cover Y . Let us define the sets

Γ { x R 3 : V ( x ) = V 0 } and Π δ { x R 3 : d ( x , Π ) δ } .

Now, we are ready to state our results in this article.

Theorem 1.1

Assume that ( f 1 )–( f 5 ) and ( V 1 )–( V 2 ) hold. Then, for any δ > 0 , there exists ε δ > 0 such that for any ε ( 0 , ε δ ) , system (1.1) has at least cat Π δ ( Π ) positive solutions. Furthermore, if ( u ε , ϕ ε ) denotes one of these solutions and η ε R 3 is a global maximum point of u ε , then

lim ε 0 V ( η ε ) = V 0 .

Remark 1.1

There is no doubt that we shall encounter the following difficulties when we deal with system (1.1):

  1. The nonlinearity satisfying critical growth in system (1.1) means that there does not exist the convergence of bounded (PS) sequences. It is more complicated to obtain the critical value of the mountain pass, due to the appearance of double critical terms in system (1.1), which is the major novelty in this article. We shall use the concentration-compactness principle to overcome this obstacle. In addition, there are two nonlocal critical convolution terms in the Schrödinger-Poisson system, and it is natural to consider how the interaction between the nonlocal term and the nonlinear term will affect the existence of system (1.1).

  2. The nonlinearity f is merely a continuous function in this article. Therefore, the corresponding Nehari manifold is not differentiable; therefore, we cannot directly use the Ljusternik-Schnirelmann category theory to obtain the existence of multiple solutions to system (1.1). In order to overcome this obstacle, we shall take full advantage of the differentiability of the unit ball instead of the Nehari manifold and the unit sphere to look for the existence of solutions. As far as we know, there is no literature on the existence and concentration of solutions for the fractional p-Laplacian Schrödinger-Poisson system with double critical nonlinearities.

  3. We assume that V ( x ) satisfies the local condition proposed by Del Pino and Felmer [14], and there is no information about the potential V ( x ) on the infinity. Therefore, we shall use the penalized arguments of Del Pino and Felmer [14] to overcome these difficulties.

Remark 1.2

In this article, our work considers the new Poisson term involving the upper exponent p s in the sense of the Hardy-Littlewood-Sobolev inequality. It is also different from doubly critical growth in Qu and He [44], as described in Remark 1.1, the critical exponent in the Poisson term is an important difficulty that we need to deal with in this article.

This article is organized as follows: In Section 2, we introduce the variational setting and give preliminary lemmas. In Section 3, we will consider the modified problem by using truncated function. In Section 4, we study the autonomous problem. In Section 5, we are devoted to the proof of Theorem 1.1.

2 Preliminaries

In this section, first, we introduce some notations

  • C , C 1 , C 2 , denote positive constants possibly different from line to line.

  • p denotes the usual norm of the Lebesgue measurable space in R 3 , for all p [ 1 , + ) .

  • o n ( 1 ) denotes the real sequence with o n ( 1 ) 0 as n + .

  • and stand for the strong and weak convergence in the related function spaces, respectively.

  • C c ( R 3 ) denotes the set of infinitely differentiable functions with compact support in R 3 .

Next, we define the fractional Sobolev space W s , p ( R 3 )

W s , p ( R 3 ) { u L p ( R 3 ) : [ u ] s , p < } ,

and the D s , p ( R 3 ) as the closure of C c ( R 3 ) in regard to

[ u ] s , p p = R 3 R 3 u ( x ) u ( y ) p x y 3 + s p d x d y .

We also know that the fractional Sobolev space W s , p ( R N ) is a uniformly convex Banach space (similar to Puzzi et al. [42]) equipped with norm

u W s , p ( R 3 ) = ( u p p + [ u ] s , p p ) 1 p ,

which is equivalent to the following norm:

u V 0 = ( u p , V 0 p + [ u ] s , p p ) 1 p , u p , V 0 p = R 3 V 0 u ( x ) p d x .

Especially p = 2, we have the following:

u D s , 2 ( R 3 ) 2 = R 3 R 3 u ( x ) u ( y ) 2 x y 3 + 2 s d x d y .

Now we introduce the workspace framework as the follow,

H ε = u W s , p ( R 3 ) : R 3 V ε ( x ) u p d x <

and H ε is equipped with the norm u p = [ u ] s p + R 3 V ε ( x ) u p d x , where V ε ( x ) = V ( ε x ) . Let us recall the following embedding property for the space D s , p ( R 3 ) and W s , p ( R 3 ) .

Lemma 2.1

(see Adams [1]) Let 0 < s < 1 and 1 < p < 3 , then there exists a constant S * > 0 such that, for any u D s , p ( R 3 ) ,

(2.1) S * = inf u D s , p ( R 3 ) u 0 R 3 R 3 u ( x ) u ( y ) p x y 3 + s p d x d y R 3 u p s * d x p p s * ,

so that

u p s * p S * 1 [ u ] s , p p .

Moreover, W s , p ( R 3 ) is continuously embedded in L t ( R 3 ) for any t [ p , p s * ] and compactly in L loc t ( R 3 ) , for any t [ 1 , p s * ) .

Lemma 2.2

(see Lions [31]) Let 0 < s < 1 , 1 < p < 3 , and r ( p , p s * ) . If { u n } is a bounded sequence in W 1 , p ( R 3 ) and if

lim n sup y R 3 B R ( y ) u n r d x = 0 ,

where R > 0 , then u n 0 in L t ( R 3 ) for all t ( p , p s * ) .

From the Lax-Milgram theorem, for each u W 1 , p ( R 3 ) and the given ε > 0 , there exists a unique ϕ u D s , p ( R 3 ) such that

( Δ ) s ϕ u = u p s .

For the weak solution ϕ u , the form is as follows:

(2.2) ϕ u = C * R 3 u ( y ) p s x y 3 2 s d y ,

where C * = Γ ( 3 2 s 2 ) 2 2 s ε 2 π 3 2 Γ ( s ) , Γ is the Gamma function (see He [21] for instance). In the sequel, we often omit the constant C * for convenience.

To deal with the difficulties caused by the critical Poisson term, we need to use the following well-known Hardy-Littlewood-Sobolev inequality.

Proposition 2.1

(see Lieb and Loss [29]) Let t , r > 1 and 0 < μ < N with 1 t + μ N + 1 r = 2 . Let f L t ( R N ) and g L r ( R N ) , then there exists a sharp positive constant C ( t , N , μ , r ) , such that

(2.3) R N R N f ( x ) g ( y ) x y μ d x d y C ( t , N , μ , r ) f t g r .

Now, we consider that if F ( u ) = u β 0 for some β 0 > 0 , according to the Hardy-Littlewood-Sobolev inequality we know the integral

R N R N F ( u ( y ) ) F ( u ( x ) ) x y μ d y d x

has well defined and we need that t β 0 [ p , p s * ] , such that the following inequality holds:

(2.4) p 2 2 μ N = p t β 0 p s * t = p s * 2 2 μ N = p s ,

where t = 2 N ( 2 N μ ) , p 2 2 μ N is called the lower critical exponent and p s p s * 2 2 μ N is the upper critical exponent in the sense of the Hardy-Littlewood-Sobolev inequality. In particular, we will use μ = 3 2 s and N = 3 in the whole article, and thus, the right side of the above inequality can be obtained p s p ( 3 + 2 s ) 2 ( 3 s p ) . Therefore, for u D s , p ( R 3 ) , we can obtain

R 3 R 3 u ( x ) p s u ( y ) p s x y 3 2 s d x d y p 2 p s C ( r , t ) p 2 p s u p s * p .

Now, we use S H , L to denote the best constant as follows:

(2.5) S H , L inf u D s , p ( R 3 ) u 0 R 3 R 3 u ( x ) u ( y ) p x y 3 + s p d x d y R 3 R 3 u ( x ) p s u ( y ) p s x y 3 2 s d x d y p 2 p s

and we have the following relationship for the constant S H , L (see He et al. [24]):

S H , L = S * C r , t p 2 p s .

Applying the change of variable x ε x , system (1.1) can be written as follows:

( Δ ) p s u + V ( ε x ) u p 2 u ϕ u u p s 2 u = f ( u ) + u p s * 2 u in R 3 , ( Δ ) s ϕ = u p s in R 3 ,

where ϕ u = C * R 3 u ( y ) p s x y 3 2 s d y .

Next, we introduce the following results which are similar to Du et al. [15]. For reader’s benefit, here we give a detailed proof.

Lemma 2.3

For any u W s , p ( R 3 ) , then

  1. for each t > 0 , there holds ϕ t u = t p s ϕ u .

  2. ϕ u D s , 2 ( R 3 ) S * 1 2 u p s * p s

    and

    R 3 ϕ u u p s d x S * 1 u p s * 2 p s ,

    where S * inf u D s , 2 ( R 3 ) u 0 R 3 R 3 u ( x ) u ( y ) 2 x y 3 + 2 s d x d y R 3 u 2 s * d x 2 2 s * .

  3. If u n u in W s , p ( R 3 ) , then ϕ u n ϕ u in D s , 2 ( R 3 ) and

    R 3 ϕ u n u n p s 2 u n η ˜ d x R 3 ϕ u u p s 2 u η ˜ d x for e a c h η ˜ W s , p ( R 3 ) .

  4. If u n u in D s , 2 ( R 3 ) and u n u a.e. in R 3 , then

    R 3 ϕ u n u n p s d x R 3 ϕ u n u u n u p s d x R 3 ϕ u u p s d x 0

    in ( D s , 2 ( R 3 ) ) * , where ( D s , 2 ( R 3 ) ) * is the dual space of D s , 2 ( R 3 ) .

Proof

( i ) By (2.2), through simple calculations, we can obtain the conclusion ( i ) .

( i i ) Since ϕ u is a unique weak solution of ( Δ ) s ϕ = u p s , thus in the sense of weak convergence, we have

ϕ u D s , 2 ( R 3 ) 2 = R 3 ϕ u u p s d x ϕ u 2 s * u p s * p s S * 1 2 ϕ u D s , 2 ( R 3 ) u p s * p s

and

R 3 ϕ u u p s d x ϕ u 2 s * u p s * p s S * 1 2 ϕ u D s , 2 ( R 3 ) u p s * p s S * 1 u p s * 2 p s .

( i i i ) Since u n u in W s , p ( R 3 ) as n , then u n p s u p s in L p s * p s ( R 3 ) as n . Thus, for any η ˜ C 0 ( R 3 ) , the uniqueness of weak solution of the Poisson equation implies that

R 3 ϕ u n η ˜ d x = R 3 u n p s η ˜ d x R 3 u p s η ˜ d x = R 3 ϕ u η ˜ d x .

That is, ϕ u n ϕ u in D s , 2 ( R 3 ) . It is now simple to conclude

(2.6) R 3 ( ϕ u n ϕ u ) u p s 2 u η ˜ d x 0 as n .

Let r ˆ 1 = p s * p s , using the Hölder inequality, we have

R 3 ( ϕ u n [ u n p s 2 u n u p s 2 u ] ) r ˆ 1 d x C ( ϕ u n p s * r ˆ 1 u n p s * r ˆ 1 ( p s 1 ) + ϕ u n p s * r ˆ 1 u p s * r ˆ 1 ( p s 1 ) ) < + .

Note that u n ( x ) u ( x ) a.e. in R 3 can be inferred from weak convergence of u n u in W s , p ( R 3 ) , which implies that

(2.7) R 3 ϕ u n [ u n p s 2 u n u p s 2 u ] η ˜ d x 0 as n .

Thus, from (2.6) and (2.7), we know that conclusion ( i i i ) holds.

( i v ) Set v n = u n u , then v n 0 in D s , 2 ( R 3 ) L p s * ( R 3 ) and v n 0 a.e. in R 3 . By ( i i i ) , we have ϕ v n 0 in D s , 2 ( R 3 ) . Since u n u a.e. in R 3 , and u n u L p s * ( R 3 ) , we infer that u n p s u p s in L p s * p s ( R 3 ) , and so u n p s u n u p s u p s 0 in L p s * p s holds. Consequently, by the weak convergence of { u n } , the Hölder inequality as n , we obtain

R 3 ϕ u n u n p s d x R 3 ϕ u u p s d x R 3 ϕ v n v n p s d x = R 3 [ ϕ u n ϕ u ϕ v n ] u n p s d x + R 3 ϕ v n u p s d x + R 3 ϕ u [ u n p s u p s ] d x + R 3 ϕ v n [ u n p s u p s u n u p s ] d x 0 .

So, ( i v ) can be proved.□

3 The modified problem

In this section, we mainly investigate the existence of positive solutions to system (1.1). We use the change of variable x ε x and look for solutions to the following equation (3.1):

(3.1) ( Δ ) p s u + V ε ( x ) u p 2 u ϕ u u p s 2 u = f ( u ) + u p s * 2 u in R 3 .

Now, we consider the modified equation, there exist l 0 > p and a > 0 such that f ( a ) + a p s * 1 = V 0 l 0 a p 1 . Then, we define the following function:

f ˜ ( t ) = f ( t ) + ( t + ) p s * 1 if t < a , V 0 l 0 t p 1 if t a ,

and

g ( x , t ) = χ Λ ( x ) f ( t ) + ( 1 χ Λ ( x ) ) f ˜ ( t ) if t > 0 , 0 if t 0 ,

where χ Λ denotes the characteristic function. According to the properties of f , we see that Carathéodory function g satisfies the following conditions:

  1. (i) G ( x , t ) c 0 t p and 0 < θ G ( x , t ) g ( x , t ) t for any x Λ and for any t > 0 ;

  2. (ii) 0 < p G ( x , t ) g ( x , t ) t V 0 l 0 t p for any x R 3 \ Λ and for any t > 0 ;

  3. for each x R 3 the function t g ( x , t ) t p 1 is increasing in ( 0 , ) and for each x R 3 \ Λ the function t g ( x , t ) t p 1 is increasing in ( 0 , a ) ;

  4. lim t 0 + g ( x , t ) t p 1 = 0 uniformly in x R 3 .

According to defintion of g , we have the following properties:

(3.2) g ( x , t ) f ( t ) for all t > 0 , x R 3 ,

(3.3) g ( x , t ) = 0 for all t < 0 , x R 3 .

It is clear that equation (3.1) and the following equation (3.4) are equivalent,

(3.4) ( Δ ) p s u + V ε ( x ) u p 2 u ϕ u u p s 2 u = g ε ( x , u ) in R 3 ,

where g ε ( x , u ) = g ( ε x , u ) . Clearly, if u ε is a solution of equation (3.4), then u ε is also a solution to equation (3.1). We denote that Λ ε = { x R 3 : ε x Λ } . So, the solution of equation (3.1) was the critical point of the following energy functional:

J ε ( u ) = 1 p R 3 R 3 u ( x ) u ( y ) p x y 3 + s p d x d y + 1 p R 3 V ε ( x ) u p d x 1 2 p s R 3 ϕ u u p s d x R 3 G ε ( x , u ) d x ,

where G ( ε x , t ) = 0 t g ( ε x , ρ ) d ρ . From ( g 3 ) and (3.2), we can obtain that J ε ( u ) C 1 ( H ε , R ) , and thus we give the Fréchet derivative of J ε ( u )

J ε ( u ) , φ = R 6 u ( x ) u ( y ) p 2 ( u ( x ) u ( y ) ) ( φ ( x ) φ ( y ) ) x y 3 + s p d x d y + R 3 V ε ( x ) u p 2 u φ d x R 3 ϕ u u p s 2 u φ d x R N g ε ( x , u ) φ d x

for any u , φ H ε . Now, we define the Nehari manifold (see Willem [54]) associated with J ε by

N ε = { u H ε \ { 0 } : [ u ] s , p p + R 3 V ε ( x ) u p d x R 3 ϕ u u p s d x = R 3 g ε ( x , u ) u d x } .

Next, we suppose that ( g 1 ) ( g 3 ) and ( V 1 ) ( V 2 ) are satisfied. Then, for all ε > 0 , we prove that the energy function satisfies the mountain pass geometry.

Lemma 3.1

The energy functional J ε has the following properties:

  1. there exist α , β > 0 satisfying J ε ( u ) α with u ε = β ;

  2. there exists e H ε such that e ε > β and J ε ( e ) < 0 .

Proof

According to ( g 3 ) , ( f 5 ) , and (3.2), we know that for each ζ > 0 small enough, there exists C ζ > 0 such that

(3.5) g ( x , t ) ζ t p 1 + C ζ t σ 1

and

(3.6) G ( x , t ) ζ t p + C ζ t σ for ( x , t ) R 3 × R .

By (3.5) and Lemma 2.3-(ii), we have

(3.7) J ε ( u ) = 1 p u ε p 1 2 p R 3 ϕ u u p s d x R 3 G ε ( x , u ) d x 1 p u ε p S p s 2 p s u 2 p s ζ u ε p C ζ u ε σ .

Since p < 2 p s , we choose u ε = β ( 0 , 1 ) , there exist α , β > 0 , such that J ε ( u ) α with u ε = β .

(ii) We fix the function u 0 H ε \ { 0 } with u 0 = 1 , by ( g 1 ) - ( i ) , we have

(3.8) J ε ( t u 0 ) = t p p u 0 ε p t 2 p s 2 p s R 3 ϕ u u 0 p s d x R 3 G ε ( x , t u 0 ) d x t p p u 0 ε p t 2 p s 2 p s R 3 ϕ u u 0 p s d x c 0 t p u 0 ε p .

Taking e = t u 0 and t large enough, we know that ( i i ) holds.□

We assume that { u n } H ε is a Palais-Smale sequence at the level c for J ε , J ε ( u n ) c and J ε ( u n ) 0 . Let

H ε + = { u H ε : supp ( u + ) Λ ε > 0 }

and

S ε + = S ε H ε + ,

where S ε is the unit sphere in H ε . Then, we obtain the following lemma:

Lemma 3.2

Suppose that ( V 1 ) ( V 2 ) , ( g 1 ) ( g 3 ) , (3.2), and (3.3) hold. Then

  1. For each u H ε + , let h u : R + R be defined by h u ( t ) = J ε ( t u ) . Then, there is a unique t u > 0 such that

    h u ( t ) > 0 i n ( 0 , t u ) , h u ( t ) < 0 i n ( t u , ) ;

  2. There exists τ > 0 independent of u such that t u τ for any u S ε + . Moreover, for each compact set K S ε + , there is a positive constant C K such that t u C K for any u K ;

  3. The map m ˇ ε : H ε + N ε given by m ˇ ε ( u ) = t u u is continuous and m ε m ˇ ε S ε + is a homeomorphism between S ε + and N ε . Moreover m ε 1 ( u ) = u u ε ;

  4. If there is a sequence { u n } n N S ε + such that dist ( u n , S ε + ) 0 then m ε ( u n ) ε and J ε ( m ε ( u n ) ) .

Proof

( i ) In view of Lemma 3.1, we consider the following cases: h u ( t ) J ε ( t u ) for t 0 , h u ( 0 ) = 0 , h u ( t ) > 0 for t sufficiently small, and h u ( t ) < 0 for t sufficiently large. That means when t = t u > 0 , we deduce that max t > 0 h u ( t ) is achieved at t u , so h ( t u ) = 0 and t u u N ε . There exist t 1 and t 2 and set 0 < t 2 < t 1 , such that t 1 u , t 2 u N ε , then we have the associated equations of t 1 u and t 2 u

(3.9) t 1 p 1 u ε p t 1 2 p s 1 R 3 ϕ u u p s d x = R 3 g ε ( x , t 1 u ) u d x

and

(3.10) t 2 p 1 u ε p t 2 2 p s 1 R 3 ϕ u u p s d x = R 3 g ε ( x , t 2 u ) u d x .

Thus, according to the definition of ( g 2 ) and subtract equation (3.9) from equation (3.10), we can obtain

(3.11) 0 = [ t 1 ( 2 p s p ) s t 2 ( 2 p s p ) ] R 3 ϕ u u p s = R 3 g ε ( x , t 1 u ) ( t 1 u ) p 1 g ε ( x , t 2 u ) ( t 2 u ) p 1 u p d x > 0 .

The above inequality is impossible, so we prove ( i ) is true.

( i i ) Set u S ε + . By ( i ) , there exists t u > 0 such that h u ( t u ) = 0 . Fix ζ > 0 . According to the definition of g and the Sobolev embedding theorems of Di Nezza et al. [17], we have that

(3.12) t u p 1 u ε p t u 2 p s 1 R 3 ϕ u u p s d x = R 3 g ε ( x , t u u ) u d x ζ C 1 t p 1 + C 2 C ζ t σ 1 .

Thus, there exists τ > 0 , independent of u , such that t u τ . Now, we assume that K S ε + is a compact set and there exists { u n } n N K such that t n t u n . Therefore, there exists u K such that u n u in H ε . According to Lemma 3.1 ( i i ) , we know that

(3.13) J ε ( t n u n ) .

Fix v n N ε and v n t n u n . According to J ε ( v n ) , v n = 0 and combining with ( g 1 ) , we can obtain

(3.14) J ε ( v n ) = J ε ( v n ) 1 θ J ε ( v n ) , v n = 1 p 1 θ v n ε p + 1 θ 1 2 p s R 3 ϕ v n v n p s d x + 1 θ R 3 ( g ε ( x , v n ) v n θ G ε ( x , v n ) ) d x 1 p 1 θ v n ε p + 1 θ R 3 \ Λ ( g ε ( x , v n ) v n θ G ε ( x , v n ) ) d x 1 p 1 θ v n ε p θ p θ p R 3 \ Λ V ε ( x ) v n p l 0 1 p 1 θ v n ε p θ p θ p R 3 \ Λ v n ε p l 0 = 1 1 l 0 θ p θ p v n ε p 0 .

Combining with (3.13) and v n = t n u n , we obtain a contradiction. Thus, we finish the proof of ( i i ) .

( i i i ) We define the maps m ˇ ε , m ε , and m ε 1 by

m ˇ ε ( t u u ) = t u u and m ε = m ˇ ε S ε + .

Combining ( i ) and ( i i ) with Szulkin and Weth [50], we can obtain m ε is a homeomorphism between m ˇ ε S ε + and N ε , and the inverse of m ε is given by m ε 1 ( u ) = u u ε .

( i v ) Let { u n } S ε + be a sequence such that dist ( u n , S ε + ) 0 , then for each v S ε + , we have u n u n v a.e. in Λ ε , by ( V 1 ) , ( V 2 ) , and Sobolev embedding, there is a constant C t > 0 such that

u n + L t ( Λ ε ) inf v S ε + u n v L t ( Λ ε ) C t inf v S ε + Λ ε ( [ u n v ] s , p p + V ε ( x ) ( u n v ) p ) 1 p C t dist ( u n , S ε + )

for all t ( p , p s * ) . Combining ( g 1 ) , ( g 3 ) , (3.2), and Sobolev embeddings, we can obtain that for each t > 0 ,

R 3 G ε ( x , t u n ) Λ ε F ( t u n ) + t p l 0 R 3 \ Λ ε V ε ( x ) u n p C 1 t 2 p Λ ε ( u n + ) 2 p + C 2 t p s * Λ ε ( u n + ) p s * + t p u ε p l 0 C 3 t 2 p dist ( u n , S ε + ) 2 p + C 4 dist ( u n , S ε + ) p s * + t p l 0 .

Thus, we can obtain that

limsup n R 3 G ε ( x , t u n ) t p l 0 , t > 0 .

According to the definition of J ε ( m ε ( u n ) ) , we obtain

liminf n J ε ( m ε ( u n ) ) liminf n J ε ( t u n ) liminf n t p p u n ε p + t p s p s R 3 ϕ u n u n p s t p l 0 l 0 p p l 0 t p .

Thus, J ε ( m ε ( u n ) ) , m ε ( u n ) ε as n . Thus, we complete the proof.□

Let us consider the maps

ψ ˇ ε : H ε + R and ψ ε : S ε + R ,

by ψ ˇ ε ( u ) J ε ( m ˇ ε ( u ) ) and ψ ε ψ ˇ ε S ε + . According to Lemma 3.2, we can obtain the next results.

Proposition 3.1

Suppose that hypotheses ( V 1 ) ( V 2 ) , ( g 1 ) ( g 3 ) , and (3.2) hold true. Then,

  1. ψ ˇ ε C 1 ( H ε , R ) and

    ψ ˇ ε ( u ) , v = m ˇ ε ( u ) ε u ε J ε ( m ˇ ε ( u ) ) , v f o r a l l v H ε a n d u H ε + ;

  2. ψ ε C 1 ( S ε + , R ) and

    ψ ε ( u ) , v = m ε ( u ) ε J ε ( m ε ( u ) ) , v , f o r e v e r y v T u S ε + { v H ε : u , v ε = 0 } ,

    where

    u , v ε = R 3 R 3 u ( x ) u ( y ) p 2 ( u ( x ) u ( y ) ) ( v ( x ) v ( y ) ) x y N + p s d x d y + R 3 V ( ε x ) u p 2 u v d x ;

  3. Assume that { u n } n N is a ( PS ) c sequence of ψ ε , which means { m ε ( u n ) } n N is also a ( PS ) c sequence for J ε . Assume that { u n } n N N ε is a bounded ( PS ) c sequence of J ε , then { m ε 1 ( u n ) } n N is also a ( PS ) c sequence of the functional ψ ε ;

  4. u is a critical point of ψ ε , if and only if, m ε ( u ) is a nontrivial critical point for J ε . Moreover, the corresponding critical values coincide and

    inf u S ε + ψ ε ( u ) = inf u N ε J ε ( t u ) .

Lemma 3.3

Let { u n } be a ( PS ) c ε sequence for J ε . Then, { u n } u N is bounded in H ε .

Proof

Suppose that { u n } is a Palais-Smale sequence of J ε , which means

J ε ( u n ) c , J ε ( u n ) 0 i n H ε 1 .

Then, according to ( g 1 ) , we have that

c + o n ( 1 ) ( 1 + u n ε ) J ε ( u n ) 1 θ J ε ( u n ) , u n = 1 p 1 θ u n ε p + 1 θ 1 2 p s R 3 ϕ u n u n p s d x + R 3 1 θ g ε ( x , u n ) u n G ε ( x , u n ) d x 1 p 1 θ u n ε p + 1 θ Λ ε g ε ( x , u n ) u n θ G ε ( x , u n ) d x + 1 θ Λ ε c g ε ( x , u n ) u n θ G ε ( x , u n ) d x 1 p 1 θ u n ε p + 1 θ 1 p 1 l 0 u n ε p 1 p 1 θ 1 1 l 0 u n ε p .

Thus, we deduce that { u n } n N is bounded in H ε . We also have

c 1 p 1 θ 1 1 1 l 0 1 u n ε p

and we denote

(3.15) K = c 1 p 1 θ 1 1 1 l 0 1 .

This completes the proof.□

In order to prove that J ε satisfies the ( PS ) c condition at the level c R if any ( PS ) c sequence has a convergent subsequence. As the appearance of a critical term in equation (3.1) inspired the result of the concentration-compactness principle of Lions [30] and Chabrowski [13], we establish the following concentration compactness principle to overcome this difficulty.

Lemma 3.4

Let { u n } be a bounded sequence in D s , p ( R 3 ) converging weakly and a.e. to some u in E and κ , κ , ν , ν be the bounded nonnegative measures. Assume that

(3.16) R 3 u n ( y ) p s x y 3 2 s d y u n ( x ) p s φ

weakly in the sense of a measure where φ is a bounded positive measure on R 3 and define

(3.17) φ = lim R limsup n x R R 3 u n ( y ) p s x y 3 2 s d y u n ( x ) p s d x .

Then, there is a countable sequence of points { z i } i I R 3 and families of positive numbers { ν i : i I } , { φ i : i I } , and { κ i : i I } such that

(3.18) φ = R 3 u ( y ) p s x y 3 2 s d y u ( x ) p s + i I φ i δ z i , i I φ i 1 p s < ,

(3.19) κ R 3 u ( x ) u ( y ) p x y 3 + s p d y + i I κ i δ z i ,

(3.20) ν = u p s * + i I ν i δ z i ,

and

(3.21) S H , L φ i p 2 p s κ i , φ i p s * 2 p s C ( r , t ) p s * 2 p s ν i ,

where δ x is the Dirac-mass of mass 1 concentrated at x R 3 . This completes the proof. For the energy at infinity, we have

(3.22) limsup n R 3 R 3 u n ( y ) p s u n ( x ) p s x y 3 2 s d y d x = R 3 d φ + φ

and

(3.23) C ( r , t ) p s * p s φ p s * p s ν ν + R 3 d ν , S * p C ( r , t ) p p s φ p p s κ ( R 3 d κ + κ ) .

Moreover, if u = 0 and R 3 d κ = S H , L R 3 d φ p 2 p , then φ is concentrated at a single point.

Proof

Since { u n } is a bounded sequence in D s , p ( R 3 ) converging weakly to u , we denote that q n u n u . Then, we have q n ( x ) 0 a.e. in R 3 and { q n } converges weakly to 0 in D s , p ( R 3 ) . Applying Chabrowski [13] and Lions [30], in the sense of measure, we have

R 3 q n ( x ) q n ( y ) p x y 3 + s p d y τ ˆ 1 κ R 3 u ( x ) u ( y ) p x y 3 + s p d y , q n p s * τ ˆ 2 ν u p s * , R 3 q n ( y ) p s x y 3 2 s d y q n ( x ) p s τ ˆ 3 φ R 3 u ( y ) p s x y 3 2 s d y u ( x ) p s .

In order to prove the possible concentration at finite points, we first show that

(3.24) R 3 R 3 ϒ ˆ q n ( y ) p s x y 3 2 s d y ϒ ˆ q n ( x ) p s d x R 3 R 3 q n ( y ) p s x y 3 2 s d y ϒ ˆ ( x ) p s ϒ ˆ q n ( x ) p s d x 0 ,

where the function ϒ ˆ C 0 ( R 3 ) . In fact, let

Θ n ( x ) R 3 ϒ ˆ q n ( y ) p s x y 3 2 s d y ϒ ˆ q n ( x ) p s R 3 q n ( y ) p s x y 3 2 s d y ϒ ˆ ( x ) p s ϒ ˆ q n ( x ) p s .

Since ϒ ˆ C 0 ( R 3 ) , we obtain for any δ 0 > 0 , there exists L > 0 such that

(3.25) x L Θ n ( x ) d x < δ 0 for all n 1 .

Since the Riesz potential defines a linear operator and q n ( x ) 0 a.e. in R 3 . Hence, we know that

R 3 q n ( y ) p s x y 3 2 s d y 0 a.e. in R 3 .

So, we have Θ n ( x ) 0 a.e. in R 3 . Note that

Θ n ( x ) = R 3 ( ϒ ˆ ( y ) p s ϒ ˆ ( x ) p s ) q n ( y ) p s x y 3 2 s d y ϒ ˆ q n ( x ) p s R 3 M ( x , y ) q n ( y ) p s d y ϒ ˆ q n ( x ) p s ,

where M ( x , y ) ϒ ˆ ( y ) p s ϒ ˆ ( x ) p s x y 3 2 s . Furthermore, for all x R 3 , there is some R > 0 large enough such that

R 3 M ( x , y ) q n ( y ) p s d y = y R M ( x , y ) q n ( y ) p s d y ϒ ˆ ( x ) p s y R q n ( y ) p s x y 3 2 s d y .

In Lions [30], we note that M ( x , y ) L r ( B R ) for any x B R . By Young’s inequality, there is t > 6 3 2 s such that

B L ( 0 ) B R ( 0 ) M ( x , y ) q n ( y ) p s d y t d x 1 t C θ M ( x , y ) r q n p s 6 3 + 2 s C θ ,

where L is given in (3.25). Additionally, it is clear that for R > 0 large enough

B L ( 0 ) ϒ ˆ ( x ) p s y R q n ( y ) p s x y 3 2 s d y t d x 1 t C

and so we obtain

B L ( 0 ) R 3 M ( x , y ) q n ( y ) p s d y t d x 1 t C θ .

Then, for s > 0 small enough, we obtain

B L ( 0 ) Θ n ( x ) 1 + s d x B L ( 0 ) R 3 M ( x , y ) q n ( y ) p s d y t d x 1 t B L ( 0 ) ϒ ˆ q n p s d x 3 + 2 s 6 C ϒ ˆ .

Using the above formula and Θ n ( x ) 0 a.e. in R 3 , we obtain that

B L ( 0 ) Θ n ( x ) d x 0 as n .

Combining this and (3.25), we have

R 3 Θ n ( x ) d x 0 as n .

So, (3.24) is demonstrated.

Now, for any ϒ ˆ C 0 ( R 3 ) , using the Hardy-Littlewood-Sobolev inequality, we deduce

R 3 R 3 ϒ ˆ q n ( y ) p s x y 3 2 s d y ϒ ˆ q n ( x ) p s d x C ( t , r ) ϒ q n p s * 2 p s .

From (3.24), we have

R 3 ϒ ˆ ( x ) 2 p s R 3 q n ( y ) p s x y 3 2 s d y q n ( x ) p s d x C ( t , r ) ϒ ˆ q n p s * 2 p s + o n ( 1 ) .

Passing to the limit as n , we obtain

(3.26) R 3 ϒ ˆ ( x ) 2 p s d τ ˆ 3 C ( t , r ) R 3 ϒ ˆ p s * d τ ˆ 2 2 p s p s * .

Applying Lemma 1.2 in Lions [30], we know (3.20) holds. Besides, let θ = χ { z i } , i I , and using this fact in (3.26), we obtain

φ i p s * 2 p s C ( t , r ) p s * 2 p s ν i for all i I .

By the definition of S H , L , we obtain

R 3 R 3 ϒ ˆ q n ( y ) p s x y 3 2 s d y ϒ ˆ q n ( x ) p s d x p 2 p s S H , L R 3 R 3 ϒ ˆ q n ( x ) ϒ ˆ q n ( y ) p x y 3 + s p d x d y .

By (3.24) and q n 0 in L loc p ( R 3 ) , it follows that

R 3 ϒ ˆ ( x ) 2 p s R 3 q n ( y ) p s x y 3 2 s d y q n ( x ) p s d x p 2 p s S H , L R 3 R 3 ϒ ˆ p q n ( x ) q n ( y ) p x y 3 + s p d x d y + o n ( 1 ) .

Passing to the limit as n , we have

(3.27) R 3 ϒ ˆ ( x ) 2 p s d τ ˆ 3 p 2 p s S H , L R 3 ϒ ˆ p d τ ˆ 1 .

Applying Lemma 1.2 in Lions [30], we know (3.22) holds. Let η ˆ = χ { z i } , i I , and applying it in (3.27), we obtain

S H , L φ i p 2 p s κ i , i I .

This completes the proof of (3.21).

Then, we are going to demonstrate the possible loss of mass at infinity. For R > 1 , let θ R C ( R 3 ) be such that θ R ( x ) = 0 for x < R , θ R = 1 for x > R + 1 , and 0 θ R ( x ) 1 on R 3 . For every R > 1 , we obtain

limsup n R 3 R 3 v n ( y ) p s v n ( x ) p s x y 3 2 s d y d x = limsup n R 3 R 3 u n ( y ) p s u n ( x ) p s x y 3 2 s d y d x R 3 R 3 u ( y ) p s u ( x ) p s x y 3 2 s d y d x = limsup n R 3 R 3 u n ( y ) p s u n ( x ) p s θ R ( x ) x y 3 2 s d y d x + R 3 R 3 u n ( y ) p s u n ( x ) p s ( 1 θ R ( x ) ) x y 3 2 s d y d x R 3 R 3 u ( y ) p s u ( x ) p s x y 3 2 s d y d x = limsup n R 3 R 3 u n ( y ) p s u n ( x ) p s θ R ( x ) x y 3 2 s d y d x + R 3 ( 1 θ R ) d ζ + R 3 R 3 u ( y ) p s u ( x ) p s ( 1 θ R ( x ) ) x y 3 2 s d y d x R 3 R 3 u ( y ) p s u ( x ) p s x y 3 2 s d y d x ,

as R . From Lebesgue’s theorem, we infer that

limsup n R 3 R 3 u n ( y ) p s u n ( x ) p s x y 3 2 s d y d x = φ + R 3 d φ .

Using the Hardy-Littlewood-Sobolev inequality, we infer that

φ = lim R limsup n R 3 R 3 u n ( y ) p s x y 3 2 s d y θ R u n ( x ) p s d x C ( r , t ) lim R limsup n R 3 u n p s * d x R 3 θ R u n p s * d x p s p s * = C ( r , t ) ν + R 3 d ν ν p s p s *

which means

C ( r , t ) p s * p s φ p s * p s ν ν + R 3 d ν .

According to the definition of S H , L and φ , we obtain

φ = lim R limsup n R 3 R 3 u n ( x ) p s x y 3 2 s d y θ R u n ( x ) p s d x C ( r , t ) lim R limsup n R 3 u n p s * d x R 3 θ R u n p s * d x p s p s * C ( r , t ) S * p s lim R limsup n R 3 R 3 u n ( x ) u n ( y ) p x y 3 + s p d x d y R 3 R 3 θ R u n ( x ) θ R u n ( y ) p x y 3 + s p d x d y p s p = C ( r , t ) S * p s ( κ + R 3 d κ ) κ p s p ,

which means

C ( r , t ) p p s S * p φ p p s κ κ + R 3 d κ .

Besides, if u = 0 then τ ˆ 1 = κ and τ ˆ 2 = φ . Then, the Hölder inequality and (3.26) imply that, for ς 1 C 0 ( R N ) ,

R 3 ς 1 ( x ) 2 p s d ν 1 3 + 2 s S H , L R 3 d κ 2 + 2 s 3 + 2 s R 3 ς 1 2 p s d κ 1 3 + 2 s .

Thus, we can deduce that φ = S H , L p s p R 3 d κ 2 + 2 s κ . It follows from (3.26) that, for ς 1 C 0 ( R 3 ) ,

R 3 ς 1 ( x ) 2 p s d φ 1 3 + 2 s R 3 d φ 2 + 2 s 3 + 2 s R 3 ς 1 2 d φ .

And so, for each open set Ω , one has

φ ( Ω ) 1 3 + 2 s ν ( R 3 ) 2 + 2 s 3 + 2 s φ ( Ω ) .

It follows that φ is concentrated at a single point.□

Next, we give the proof of the compactness result.

Proposition 3.2

J ε satisfies the ( PS ) c condition at the level 0 < c < c * , where

c * = min 1 θ 1 p s * S * p s * p + S H , L 2 p s p p 2 p s p , 1 θ 1 p s * S * p s * p + S H , L 2 p s p p p s * p , × 1 θ 1 p s * S * p s * p + S * p s K p s p C ( r , t ) p p s * p , 1 θ 1 p s * S * p s * p + S * p s K p s p C ( r , t ) p p s p ,

S * is given by formula (2.1), S H , L is given by formula (2.5), K is given from formula (3.15), and C ( r , t ) comes from formula (3.23).

Proof

Let { u n } H ε be a ( PS ) c sequence for J ε . From the study of Chabrowski [13], we know that { u n } is bounded in E . Hence, by Lions [30] and Lemma 3.4, there are three positive measures κ , ν , φ ( R 3 ) so that

(3.28) [ u n ] s , p p κ , u n p s * ν , and R 3 u n ( y ) p s x y 3 2 s d y u n ( x ) p s φ .

For proving the Proposition 3.2, the proof will be divided into the following three steps.

Step I. We claim that κ i ( ν i + φ i ) .

In fact, define σ C 0 ( R 3 ) as a truncation function with σ [ 0 , 1 ] , σ 1 in B 1 2 ( 0 ) , and σ 0 in R 3 \ B 1 ( 0 ) . For any ε > 0 , set

σ ε ( x ) σ x x i ε = 1 if x x i ε 2 , 0 if x x i ε .

Due to the boundedness of { u n } in H ε , we know that { σ ε u n } is also bounded in H ε , thus J ε ( u n ) , u n σ ε = o n ( 1 ) ,

(3.29) R 3 R 3 u n ( x ) u n ( y ) p 1 ( u n ( x ) u n ( y ) ) ( σ ε ( x ) σ ε ( y ) ) x y 3 + s p d x d y + R 3 V ε ( x ) u n p 1 u n σ ε d x = R 3 ϕ u n u n p s 1 u n σ ε d x + R 3 ( f ( u n ) u n σ ε + u n p s * σ ε ) d x + o n ( 1 ) .

Then, by (3.28), the definition of σ ε , and Chabrowski [13], it follows that

(3.30) lim ε 0 lim n R 3 σ ε ϕ u n u n p s d x = lim ε 0 R 3 σ ε d φ = φ i ,

(3.31) lim ε 0 lim n R 3 u n ( x ) u n ( y ) p ( σ ε ( x ) σ ε ( y ) ) x y 3 + s p d x = lim ε 0 R 3 σ ε d κ = κ i ,

and

(3.32) lim ε 0 lim n R 3 σ ε u n p s * d x = lim ε 0 R 3 σ ε d ν = ν i .

By the absolute continuity of the Lebesgue integral and the definition of σ ε , one has

(3.33) lim ε 0 lim n R 3 f ( u n ) u n σ ε d x = lim ε 0 R 3 f ( u ) u σ ε d x = lim ε 0 x x j < ε f ( u ) u σ ε d x = 0 .

Summing up, from (3.29) to (3.33), taking the limit as ε 0 and then the limit as n , we gain

R 3 ϕ u n u n p s σ ε d x + R 3 f ( u n ) u n σ ε d x + R 3 u n p s * σ ε d x = R 3 u n ( x ) u n ( y ) p ( σ ε ( x ) σ ε ( y ) ) x y 3 + s p d x + R 3 V ( x ) u n p σ ρ d x R 3 u n ( x ) u n ( y ) p ( σ ε ( x ) σ ε ( y ) ) x y 3 + s p d x .

So, we have κ i φ i + ν i .

Step II. We claim that κ φ + ν .

In fact, from Lemmas 3.3 and 3.4, let σ R C 0 ( R 3 ) be a truncation function with σ R [ 0 , 1 ] , σ R 0 in B 1 2 ( 0 ) , and σ R 1 in R 3 \ B 1 ( 0 ) . For any R > 0 , set

σ ρ ( x ) σ R ( x R ) = 0 if x R 2 , 1 if x R .

Analogously, by the boundedness of { u n } in H ε , we know that { σ ε u n } is also bounded in H ε , thus J ε ( u n ) , u n σ R = o n ( 1 ) ,

(3.34) R 3 R 3 u n ( x ) u n ( y ) p 1 ( u n ( x ) u n ( y ) ) ( σ R ( x ) σ R ( y ) ) x y 3 + s p d x d y + R 3 V ε ( x ) u n p 1 u n σ R d x = R 3 ϕ u n u n p s 1 u n σ R d x + R 3 ( f ( u n ) u n σ R + u n p s * σ R ) d x + o n ( 1 ) .

By the definition of σ R , we have

{ x R 3 : x > R } u n ( x ) u n ( y ) p x y 3 + s p d x R 3 u n ( x ) u n ( y ) p ( σ R ( x ) σ R ( y ) ) x y 3 + s p d x { x R 3 : x R 2 } u n ( x ) u n ( y ) p x y 3 + s p d x .

Thus, given that Lemma 3.4, we obtain

(3.35) lim R lim n R 3 u n ( x ) u n ( y ) p ( σ R ( x ) σ R ( y ) ) x y 3 + s p d x = κ .

Similarly, we obtain that

(3.36) lim R lim n R 3 u n p s 1 u n ϕ u n σ R d x = φ ,

(3.37) lim R lim n R 3 u n p s * 1 u n σ R d x = ν ,

and

(3.38) lim R lim n R 3 f ( u n ) u n σ R d x = lim R R 3 f ( u ) u σ R d x = lim R { x > R 2 } f ( u ) u σ R d x = 0 .

Summing up, from (3.34)–(3.38), similar to Step II, taking the limit as R , and then the limit as n , we obtain κ φ + ν .

Step III. We claim that φ i + ν i = 0 for all i I and φ + ν = 0 .

In fact, suppose by contradiction that there exists i 0 I so that φ i 0 + ν i 0 > 0 . Step II and Lemmas 3.3, Chabrowski [13], and Lemma 3.4 imply that

(3.39) ν i 0 ( κ i 0 S * 1 ) p s * p S * p s * p ( φ i 0 + ν i 0 ) p s * p

and

(3.40) φ i 0 ( κ i 0 S H , L 1 ) 2 p s p S H , L 2 p s p ( φ i 0 + ν i 0 ) 2 p s p .

From (3.39) and (3.40), we know that

φ i 0 + ν i 0 S * p s * p ( φ i 0 + ν i 0 ) p s * p + S H , L 2 p s p ( φ i 0 + ν i 0 ) 2 p s p .

If φ i 0 + ν i 0 1 , we have

ν i 0 + φ i 0 S * p s * p ( φ i 0 + ν i 0 ) p s * p + S H , L 2 p s p ( φ i 0 + ν i 0 ) 2 p s p S * p s * p ( ν i 0 + φ i 0 ) 2 p s p + S H , L 2 p s p ( ν i 0 + φ i 0 ) 2 p s p = ( ν i 0 + φ i 0 ) 2 p s p S * p s * p + S H , L 2 p s p .

This fact implies that

(3.41) ν i 0 + φ i 0 S * p s * p + S H , L 2 p s p p 2 p s p .

If 0 < φ i 0 + ν i 0 1 , we have

ν i 0 + φ i 0 S * p s * p ( ν i 0 + φ i 0 ) p s * p + S H , L 2 p s p ( ν i 0 + φ i 0 ) 2 p s p S * p s * p ( ν i 0 + φ i 0 ) p s p + S H , L 2 p s p ( ν i 0 + φ i 0 ) p s p = ( ν i 0 + φ i 0 ) p s p S * p s * p + S H , L 2 p s p .

From this fact, one has

(3.42) ν i 0 + φ i 0 S * p s * p + S H , L 2 p * p p p s * p .

By using the selection of c , combining (3.41) and (3.42), we have

c = lim n J ε ( u n ) 1 θ J ε ( u n ) , u n lim n 1 θ 1 2 p * R 3 ϕ u n u n p * σ ε d x + 1 θ 1 p * R 3 u n p * σ ε d x 1 θ 1 p * ( ν i 0 + φ i 0 ) c .

This obviously contradicts with c < c .

On the other hand, if φ + ν > 0 , from Step II and Lemmas 3.33.4, we have

(3.43) ν ( κ S * 1 ) p s * p S p s * p ( φ + ν ) p s * p

and

(3.44) φ ω p * p C ( r , t ) S * p * R 3 d ω + ω p * p S * p * K p * p C ( r , t ) ( φ + ν ) p * p .

By (3.43) and (3.44), we obtain

φ + ν S * p s * p ( φ + ν ) p s * p + S * p s K p s p C ( r , t ) ( φ + ν ) p * p .

If φ + ν 1 , we have

(3.45) φ + ν S * p s * p ( φ + ν ) p s * p + S * p s K p s p C ( r , t ) ( φ + ν ) p * p S * p s * p ( ν + φ ) p s * p + S * p s K p s p C ( r , t ) ( ν + φ ) p s * p = ( ν + φ ) p s * p S * p s * p + S * p s K p s p C ( r , t ) .

This implies that

(3.46) ν + φ S * p s * p + S * p s K p s p C ( r , t ) p p s * p .

If 0 < φ + ν 1 , we have

(3.47) ν + φ S * p s * p ( ν + φ ) p s * p + S * p s K p s p C ( r , t ) ( ν + φ ) p s p S * p s * p ( ν + φ ) p s p + S * p s K p s p C ( r , t ) ( ν + φ ) p s p = S * p s * p + S * p s K p s p C ( r , t ) ( ν + φ ) p s p .

This means that

(3.48) ν + φ S * p s * p + S * p s K p s p C ( r , t ) p p s p .

Moreover, we have

c = lim ε lim n J ε ( u n ) 1 θ J ε ( u n ) , u n lim ε lim n 1 θ 1 2 p * R 3 ϕ u n u n p * σ ε d x + 1 θ 1 p s * R 3 u n p s * σ ε d x 1 θ 1 p s * ( ν + φ ) c .

Similarly, this contradicts the selection of c . Thus, we have φ i 0 + ν i 0 = 0 and φ + ν = 0 for all i I . Hence,

(3.49) R 3 u n u p s * d x 0 as n ,

(3.50) R 3 ϕ ( u n u ) u n u p s d x 0 as n ,

and

(3.51) R 3 f ( u n u ) u n u d x 0 as n .

So, we have

(3.52) J ε ( u n u ) , u n u = R 3 R 3 ( u n u ) ( x ) ( u n u ) ( y ) p x y 3 + s p d x d y + R 3 V ε ( x ) u n u p d x R 3 ϕ ( u n u ) u n u p s d x R 3 u n u p s * d x R 3 f ( u n u ) ( u n u ) d x .

Now, from (3.49) to (3.52), we have

R 3 R 3 ( u n u ) ( x ) ( u n u ) ( y ) p x y 3 + s p d x d y + R 3 V ε ( x ) ( u n u ) p d x 0 as n .

This completes the proof of Proposition 3.2.□

According to Propositions 3.1 and 3.2, we obtain the following result.

Corollary 3.1

ϕ ε satisfies the ( PS ) c condition if 0 < c < c * .

4 The autonomous problem

In this section, we mainly consider the following equation (4.1):

(4.1) ( Δ ) p s u + V 0 ( x ) u p 2 u ϕ u u p s 2 u = f ( u ) + u + p s * 2 u in R 3 .

According to equation (4.1), we give the energy functional is

J ε ( u ) = 1 p R 3 R 3 u ( x ) u ( y ) p x y 3 + s p d x d y + 1 p R 3 V ε ( x ) u p d x 1 2 p s R 3 ϕ u u p s d x R 3 F ( u ) d x 1 p s * R 3 u + p s * d x ,

which is well defined on the Sobolev space H V 0 ,

H V 0 = u W s , p ( R 3 ) : R N V 0 u p d x <

and norm in this space is given

u V 0 p = R 3 R 3 u ( x ) u ( y ) p x y 3 + s p d x d y + R 3 V 0 u p d x .

From ( f 1 ) , ( f 2 ) , and Proposition 2.1, we deduce that J V 0 C 1 ( H V 0 , R ) and we give the following derivative:

J V 0 ( u ) , φ = R 6 u ( x ) u ( y ) p 2 ( u ( x ) u ( y ) ) ( φ ( x ) φ ( y ) ) x y 3 + s p d x d y + R 3 V 0 ( x ) u p 2 u φ d x R 3 ϕ u u p s 2 u φ d x R 3 f ( u ) φ d x R 3 u + p s * 2 u + φ d x

for any u , φ H V 0 . Then, we define the following set:

H V 0 + = { u H V 0 : supp ( u + ) > 0 }

and

S V 0 + = S V 0 H V 0 + ,

where S V 0 is the unit sphere in H V 0 . Moreover, S V 0 = T u S V 0 + R u . Then, we define the Nehari manifold associated with J V 0 by

N V 0 = { u H V 0 { 0 } : J V 0 ( u ) , u = 0 } .

Next, we establish the following mountain pass geometry.

Lemma 4.1

One of the following holds:

  1. there exist α , β > 0 satisfying J V 0 ( u ) β with u V 0 = σ ;

  2. there exist e H V 0 satisfying J V 0 ( e ) < 0 .

Similar to the results of Lemma 3.1 and Proposition 3.1 in Section 3, we introduce the following lemma.

Lemma 4.2

Assume that ( V 1 ) ( V 2 ) and ( f 1 )–( f 5 ) hold. Then

  1. For each u H V 0 + , let h : R + R be defined by h u ( t ) = J V 0 ( t u ) . Then, there is a unique t u > 0 such that

    h u ( t ) > 0 i n ( 0 , t u ) , h u ( t ) < 0 i n ( t u , ) .

  2. There exists τ > 0 independent of u such that t u τ for any u S V 0 + . In addition, for each compact set K S V 0 + , there exists a positive constant C K such that t u C K for any u K ;

  3. The map m ˇ V 0 : H V 0 N V 0 given by m ˇ V 0 ( u ) = t u u is continuous and m V 0 m ˇ V 0 S V 0 + is a homeomorphism between S V 0 + and N V 0 . Moreover, m V 0 1 ( u ) = u u V 0 ;

  4. If there is a sequence { u n } S V 0 + + such that d i s t ( u n , S V 0 + ) 0 , then m V 0 ( u n ) V 0 and J V 0 ( m V 0 ( u n ) ) .

Now, we consider the maps

ψ ˇ V 0 : H V 0 R and ψ V 0 : S V 0 + R ,

where ψ ˇ V 0 ( u ) J V 0 ( m ˇ V 0 ( u ) ) and ψ V 0 ψ ˇ V 0 S V 0 + .

Proposition 4.1

Suppose that the hypotheses ( V 1 ) ( V 2 ) and ( f 1 )–( f 5 ) hold. Then

  1. ψ ˇ V 0 C 1 ( H V 0 + , R ) and

    ψ ˇ V 0 ( u ) , v = m ˇ V 0 ( u ) V 0 u V 0 J V 0 ( m ˇ V 0 ( u ) ) , v for a l l v H V 0 + a n d u H V 0 + ;

  2. ψ V 0 C 1 ( S V 0 + , R ) and

    ψ V 0 ( u ) , v = m V 0 ( u ) V 0 J V 0 ( m V 0 ( u ) ) , v , f o r e v e r y v T u S V 0 + { v H V 0 : u , v V 0 = 0 } ,

    where

    R 6 u ( x ) u ( y ) p 2 ( u ( x ) u ( y ) ) ( v ( x ) v ( y ) ) x y 3 + p s d x d y + V 0 R 3 u p 2 u v d x ;

  3. Assume that { u n } is a ( PS ) c sequence of ψ V 0 , which means { m V 0 ( u n ) } is also a ( PS ) c sequence of J V 0 . Assume that { u n } N V 0 is a bounded ( PS ) c sequence of J V 0 , thus { m V 0 1 ( u n ) } is also a ( PS ) c sequence of the functional ψ V 0 ;

  4. u is a critical point of ψ V 0 , if and only if, m V 0 ( u ) is a nontrivial critical point for J V 0 . Moreover, the corresponding critical values coincide and

    inf u S V 0 + ψ V 0 ( u ) = inf u N V 0 J V 0 ( t u ) .

Remark 4.1

As in Szulkin and Weth [50], we can know that the following equalities hold:

c ε inf u N V 0 J V 0 ( u ) = inf u H V 0 + max t > 0 J V 0 ( t u ) = inf u S V 0 + max t > 0 J V 0 ( t u ) = inf γ Γ max t [ 0 , 1 ] J V 0 ( γ ( t ) ) ,

where Γ = { γ C ( [ 0 , 1 ] , H V 0 ) : γ ( 0 ) = 0 a n d J V 0 ( γ ( 1 ) ) < 0 } .

Lemma 4.3

There exists μ 0 > 0 such that d V 0 < c * , for any μ μ 0 .

Proof

Let e H V 0 be given in Lemma 4.1. There exists t μ > 0 such that

J V 0 ( t μ e ) = max t 0 J V 0 ( t e ) .

By J V 0 ( u ) , u = 0 and ( f 2 ) , we have

(4.2) t μ e V 0 p = R 3 ϕ t μ e t μ e p s d x + R 3 f ( t μ e ) t μ e d x + R 3 t μ e + p s * d x = t μ 2 p s R 3 ϕ e e p s d x + R 3 f ( t μ e ) t μ e d x + t μ p s * R 3 e + p s * d x μ t μ q 2 R 3 e q 2 d x

which means that { t μ } μ > 0 is bounded. Then, there exists a sequence { μ n } R satisfying

μ n as n and t μ n t 0 as n .

Combining with (4.2), we can know that t μ n 0 as n . We fix γ ( t ) = t e for t [ 0 , 1 ] , then γ Γ and

d V 0 max t [ 0 , 1 ] J V 0 ( γ ( t ) ) = J V 0 ( t μ e ) 1 p t μ e V 0 p .

Then, as μ , we know that d V 0 0 , thus we completed the proof.□

Now, we use the following lemma to verify the weak limit of the ( PS ) c V 0 sequence whether it is nontrivial.

Lemma 4.4

Assume that { u n } H V 0 be a ( PS ) c V 0 sequence for J V 0 such that J V 0 ( u n ) d V 0 and J V 0 ( u n ) 0 . Then, one of the following results will be held:

  1. if u 0 , then u n u in H V 0 ;

  2. if u = 0 , there exists a sequence { y n } R 3 such that u n ( + y n ) has a strongly convergent subsequence in H V 0 .

Proof

According to the above arguments, we can obtain { u n } is bounded in H V 0 . First, we suppose that u n u in H V 0 . Then, J V 0 ( u ) = 0 holds. If we assume that u 0 . By using Fatou’s lemma, we have the following that

d V 0 = liminf n J V 0 ( u n ) 1 2 p s J ε ( u n ) , u n = liminf n 1 p 1 2 p s u n V 0 p + R 3 1 2 p s f ( u n ) u n F ( u n ) d x + 1 2 p s 1 p s * R 3 u n + p s * d x 1 p 1 2 p s u V 0 p + R 3 1 2 p s f ( u ) u F ( u ) d x + 1 2 p s 1 p s * R 3 u + p s * d x = J V 0 ( u ) 1 2 p s J ε ( u ) , u J V 0 ( u ) d V 0 .

Thus, we know that u n u in H V 0 . For the case u 0 in H V 0 , there exists a sequence { y n } R 3 and constants R , β > 0 such that

lim n sup y R 3 B R ( y ) u n p d x β for all R > 0 ,

we assume that if there exists a contradiction, such that

(4.3) lim n sup y R 3 B R ( y ) u n p d x = 0 for all R > 0 .

According to the boundedness of { u n } in H V 0 and combining with Lemma 2.1, we can obtain

(4.4) u n 0 in L r ( R 3 ) for all r ( p , p s * ) .

Then, from (4.4) and taking J V 0 ( u n ) , u n = o n ( 1 ) , it follows that

(4.5) u n V 0 p = R 3 ϕ u n u n p s d x + u n L p s * ( R 3 ) p s * + o n ( 1 ) .

We assume that

(4.6) R 3 R 3 u n ( x ) u n ( y ) p x y 3 + p s d x d y A .

Combining (4.5) and Lemma 2.1, we can obtain A C c * . Due to d V 0 = I V 0 ( u n ) + o n ( 1 ) and (4.4), it means

(4.7) d V 0 = I V 0 ( u n ) = 1 p u n V 0 p 1 2 p s R 3 ϕ u n u n p s d x 1 p s * u n + L p s * ( R 3 ) p s * + o n ( 1 ) .

Due to (4.5)–(4.7), one has

(4.8) d V 0 = 1 p 1 p s * u n V 0 p + o n ( 1 ) 1 p 1 p s * A + o n ( 1 ) .

Letting n , then d V 0 C c * . Therefore, there creates a contradiction with Lemma 4.3, thus we can obtain

lim n sup y R 3 B R ( y ) u n p d x β for all R > 0 .

Meanwhile, v n = u n ( + y n ) has a strongly convergent subsequence in H V 0 .□

Lemma 4.5

Equation (3.1) has a positive ground-state solution.

Proof

Let { v n } be a minimizing sequence for ψ V 0 on S V 0 + . Applying Ekeland’s variational method (see Ekeland [16]) and Lemma 4.2- ( i v ) , we can deduce that

ψ V 0 ( v n ) d V 0 and ψ V 0 ( v n ) 0 in ( T v n ( S V 0 + ) ) * .

Now let u n = m V 0 ( v n ) . According to Proposition 4.1, we have J V 0 ( u n ) d V 0 and J V 0 ( u n ) 0 . Together with Lemma 4.4, we can suppose that u n u 0 in H V 0 . That means u is the ground-state solution of equation (3.1). In addition, due to ( f 1 ) , we have u 0 . Combining Li and Yan [35] with the Harnack-type inequalities in Trudinger [53], one has that u C loc 1 , α ( R 3 ) for some α > 0 and u ( x ) > 0 in R 3 .□

Lemma 4.6

limsup ε 0 c ε d V 0 < c * .

The proof of this part will be given in the next chapter and will be omitted here.

5 Multiplicity of solution to equation (3.1)

In this section, we will consider the multiplicity of solutions to equation (3.1). First, we introduce the following lemma.

Lemma 5.1

Assume that ε n 0 + and { u n } N ε n is a sequence such that J ε n ( u n ) = d V 0 . Then, there exists { y ˜ n } R 3 , such that v n = u n ( x + { y ˜ n } ) has a convergent subsequence in H V 0 . Moreover, up to a subsequence, y n = ε n y ˜ n y 0 Π .

Proof

According to J ε n ( u n ) , u n = 0 and J ε n ( u n ) d V 0 , in the aforementioned, we know that { u n } is bounded in H ε . And suppose that there exists a sequence { y ˜ n } R 3 , R V 0 > 0 , and β > 0 such that

(5.1) lim n inf B R ( y ˜ n ) u n ( x ) p d x = β > 0 .

If we assume that (5.1) is false. Thus, for all R V 0 > 0 , we can obtain

lim n sup y R 3 B R ( y ) u n p d x = 0 .

In L s ( R 3 ) , s ( p , p s * ) , we know that u n 0 . But due to J ε n ( u n ) d V 0 > 0 , obviously un → 0 is impossible. Next, we set v n ( x ) = u n ( x + y ˜ n ) , V n = V ( ε n x + ε n y ˜ n ) and suppose that v n v ̸x≠ 0 as n . Let t n ( 0 , + ) such that v ˜ n = t n v n N V 0 , and set y n ε n y ˜ n . Then, by using g ( x , t ) f ( t ) , we have

(5.2) d V 0 J V 0 ( v ˜ n ) = J ε n ( t n u n ) J ε n ( u n ) d V 0 + o n ( 1 ) .

So, lim n J V 0 ( v ˜ n ) = d V 0 . Therefore, v ˜ n is bounded in H V 0 , and we may also suppose that v ˜ n v ˜ . Obviously, { t n } is bounded and t n t V 0 0 . The first case, we assume that t V 0 = 0 , according to the boundedness of { u ˜ n } n N , we can obtain v ˜ n V 0 = t n v ˜ n V 0 0 , there exists a contradiction between J V 0 ( v ˜ n ) 0 and d V 0 > 0 . The second case, we assume that t V 0 > 0 . From the uniqueness of the weak limit, we can obtain v ˜ n = t V 0 v n and v ˜ 0 . From Lemma 4.3, we can deduce that v ˜ n v ˜ , this means that

v n = v ˜ n t n v ˜ t V 0 = v ,

J V 0 ( v ˜ ) = d V 0 and J V 0 ( v ˜ ) , v ˜ = 0 .

Next, we consider the boundedness of { ε n y ˜ n } , arguing by contradiction, thus we find a subsequence denoted by { ε n y ˜ n } , such that { ε n y ˜ n } . First, due to V is bounded, up to a subsequence, we may suppose that V ( ε n y ˜ n ) V * V V 0 . We know that V is a uniformly continuous function on B R ( 0 ) for any R > 0 , then we have

V n ( x ) V * V ( ε n ( x + y ˜ n ) ) V ( ε n y ˜ n ) + V ( ε n y ˜ n ) V * 0

as n on B R ( 0 ) .

According to Lemma 4 of Thin [52], for any φ W s , p ( R 3 ) , we have

(5.3) R 6 v n ( x ) v n ( y ) p 2 ( v n ( x ) v n ( y ) ) ( φ ( x ) φ ( y ) ) x y 3 + s p d x d y R 3 v ( x ) v ( y ) p 2 ( v ( x ) v ( y ) ) ( φ ( x ) φ ( y ) ) x y 3 + s p d x d y

as n . And, we also know that

(5.4) lim n R B V n ( x ) v n p 2 v n φ d x = V * R 3 v p 2 v φ d x .

Combining (3.49), (3.50), (3.51), (5.3), and (5.4), we can obtain v is a solution of the following equation in R 3 :

( Δ ) p s v + V * ( x ) v p 2 v ϕ v v p s 2 v = f ( v ) + v p s * 2 v .

Via Fatou lemma, (5.2), and condition ( V ) , we know that

d V 0 d V * J V * ( v ) = J V * ( v ) 1 θ J V * ( v ) , v = 1 p 1 θ v V * p + 1 θ 1 2 p s R 3 ϕ v v p s d x + R 3 1 θ f ( x , v ) v F ( x , v ) d x + 1 θ 1 p s * R 3 v p s * d x liminf n J ε n ( v n ) liminf n J ε n ( u n ) d V 0

which is a contradiction, so we prove the boundness of { ε n y ˜ n } . Then, up to a subsequence, there exists y R 3 such that ε n y ˜ n y as n . Now, we prove y 0 Π . We suppose that y 0 Π and by V 2 , we obtain V 0 < V ( y ) , together with

d V 0 d V y J V y ( v ) = J V y ( v ) 1 θ J V y ( v ) , v = 1 p 1 θ v V y p + ( 1 θ 1 2 p s ) R 3 ϕ v v p s d x + R 3 1 θ f ( x , v ) v F ( x , v ) d x + 1 θ 1 p s * R 3 v p s * d x liminf n J ε n ( v n ) d V 0 .

Obviously, this is impossible, hence we know y 0 Π .

Next, for proving V ( y 0 ) = V 0 , we suppose that if V ( y 0 ) > V 0 , according to the v ˜ n v ˜ , Fatou’s lemma and the invariance of R 3 by translations (similar calculation process as above), we can obtain

d V 0 J V 0 ( v ˜ ) < liminf n 1 p [ v ˜ n ] s , p p + 1 p R 3 V ( ε n x + y n ) v ˜ n p d x 1 2 p s R 3 ϕ v ˜ n v ˜ n p s d x R 3 F ( v ˜ n ) d x 1 p s * R 3 v ˜ n p s * d x = liminf n t n p p [ u n ] s , p p + t n p p R 3 V ( ε n z ) t n p d z t n 2 p s 2 p s R 3 ϕ u n u n p s d z R 3 F ( t n u n ) d z t n p s * p s * R 3 u n p s * d z = liminf n J ε ( t n u n ) liminf n J ε n ( u n ) d V 0 ,

which is impossible. So, V ( y 0 ) = V 0 and y 0 Π hold; thus, the proof of Lemma 5.1 is complete.□

Next, we will combine the topology of the set Π and positive solutions of equation (3.1) to obtain the main results in this article. So, we take δ > 0 such that

Π δ = { x R 3 : dist ( x , Π ) δ } Λ .

We choose ψ C 0 ( [ 0 , ) , [ 0 , 1 ] ) as a smooth nonincreasing cut-off function such that ψ ( t ) = 1 if 0 s δ 2 and ψ ( s ) = 0 if s δ . For any ε > 0 and y Π ε , we define the following function:

Ψ ε , y ( x ) = ψ ( ε x y ) w ε x y ε ,

where w is a positive ground-state solution to the autonomous equation (3.1).

There is a unique t ε > 0 such that

max t 0 J ε ( t Ψ ε , y ) = J ε ( t ε Ψ ε , y ) .

We denote that

Φ ε ( y ) = t ε Ψ ε , y ,

where Φ ε ( y ) : Π N ε and Φ ε ( y ) has a compact support for any y Π ε .

Proceeding line by line as in Qu and He [44] Lemma 5.2, we can obtain the following result:

Lemma 5.2

It follows that

lim ε 0 J ε ( Φ ε ( y ) ) = d V 0 u n i f o r m l y i n y Π ε .

Next, we define the barycenter map, for any δ > 0 , let γ = γ ( δ ) > 0 be such that Γ δ B γ ( 0 ) .

Define χ : R 3 R 3 as follows:

χ ( x ) = x , x < γ ; γ x x , x γ .

In the following, we define the map β ε : N ε R 3

β ε ( u ) = R 3 χ ( ε x ) ( u p ) R 3 ( u p ) d x .

Lemma 5.3

We have the following limit:

lim ε 0 β ε ( Φ ε ( y ) ) = y u n i f o r m l y i n y Π ε .

Proof

We assume that it is false, then there exist δ 0 > 0 , { y n } Π and ε n 0 + such that

(5.5) β ε n ( Φ ε n ( y n ) ) y n δ 0 > 0 for all n R .

Combining with β ε n , Φ ε n ( y n ) , and z = ( ε n x y n ) ε n , we obtain

β ε n ( Ψ ε n ( y n ) ) = y n + R 3 [ χ ( ε n z + y n ) y n ] ( ψ ( ε n z ) ω ( z ) p ) d z R 3 ( ψ ( ε n z ) ω ( z ) p ) d z .

It is easy to obtain

δ 0 β ε n ( Φ ε n ( y n ) ) y n 0

which contradicts relation (5.5).□

Let

f ( ε ) max y Γ J ε ( Φ ε ( y ) ) d V 0

and lim ε 0 + f ( ε ) = 0 . Thus, we give the definition in the following

N ε ˜ = { u N ε : J ε ( u ) d V 0 + f ( ε ) } .

For each y Γ , Lemma 5.2, and ε > 0 , we obtain Φ ε ( y ) N ε ˜ . So N ε ˜ . Meanwhile, we have the following results.

Lemma 5.4

For any δ > 0 , the following limit holds

lim ε 0 sup u N ˜ inf y Γ β ε ( u ) y = 0 .

Proof

Let ε 0 . By definition, there exists { u n } N ε ˜ such that

inf y Γ β ε n ( u n ) y = sup u N ˜ inf y Γ β ε n ( u ) y + o n ( 1 ) .

Thus, it is sufficient to prove a sequence { u n } Γ δ such that

lim n β ( u n ) z n = 0 .

In fact, from { u n } N ε ˜ , we have

d V 0 c ε n J ε n ( u n ) d V 0 + f ( ε n ) .

According to Lemma 5.1, there exists { y ˜ n } R 3 such that v n = u n ( x + y ˜ n ) and ε n y ˜ n Γ δ for n sufficiently large. Then,

β ε n ( u n ) = R 3 χ ( ε n x ) ( u n p ) d x R 3 ( u n p ) d x .

Using the ε n x + y n y , we have

lim n β ε n ( u n ) y n = 0 .

Therefore, the proof is completed.□

Now, we give the multiplicity result for equation (3.4).

Theorem 5.1

Suppose that ( V 1 ) ( V 2 ) and ( f 1 )–( f 5 ) hold. Then, for all δ > 0 , there exists ε ¯ δ > 0 , such that, for any ε ( 0 , ε ¯ δ ) , problem (3.4) has at least cat Π δ ( Π ) positive solutions.

Proof

For any ε > 0 , let us observe the map υ ε : Π S ε + and define as υ ε ( y ) = m ε 1 ( Φ ε ( y ) ) . From Lemma 5.2, we can obtain

(5.6) lim ε 0 Ψ ε ( υ ε ( y ) ) = lim ε 0 J ε ( Φ ε ( y ) ) = d V 0 uniformly in y Γ .

Set

(5.7) S ˘ ε + = { w S ε + : Ψ ε ( w ) d V 0 + f ( ε ) } ,

where f ( ε ) 0 as ε 0 + . Thus, we can obtain

f ( ε ) = Ψ ε ( υ ε ( y ) ) d V 0 0 as ε 0 + .

Moreover, there is ε ¯ > 0 such that Ψ ε ( υ ε ( y ) ) S ε + and S ε + . According to (5.7), Lemma 5.2, we can find ε ¯ = ε ¯ δ > 0 , such that

Π Φ ε N ε ˜ m ε 1 υ ε ( Π ) m ε N ε ˜ β ε Π δ

is well defined for any ε ( 0 , ε ¯ ) . Therefore, we can prove

(5.8) cat υ ε ( Π ) ( υ ε ( Π ) ) cat Π δ ( Π ) .

Therefore, for ε > 0 small enough, J ε satisfies the ( PS ) c condition for c ( d V 0 , d V 0 + f ( ε ) ) , by Corollary 4.2 and Corollary 2.8 in Szulkin and Weth [50]. Therefore, J ε has at least cat Θ δ ( Θ ) critical points on S ˘ ε + . Taking into account Proposition 3.1-(d) and (5.8), we can deduce system (1.1) has at least cat Π δ ( Π ) positive solutions.□

Next, we will consider equation (3.4) via a variant of the Moser iteration argument in Moser [39].

Lemma 5.5

Let u ˜ n be a solution to the following:

(5.9) ( Δ ) p s v n + V ( ε n x + ε n y ˜ n ) v n p 2 v n ϕ v n v n p s 2 v n = f ( v n ) + v n p s * 2 v n in R 3 .

Then, up to a subsequence, v n = u n ( + y ˜ n ) L ( R 3 ) , and there exists C > 0 , such that v n L ( R 3 ) C for all n N .

Proof

We take β > 1 and K > 0 , such that

( t ) = 0 t 0 , t β 0 t K , β K β 1 t ( β 1 ) K β t > K .

Then,

( t ) = β t β 1 t K ; β K β 1 t > K .

According to v n 0 , ( t ) , and ( t ) , we can obtain

(5.10) ( v n ) 0 , ( v n ) 0 , and v n ( x ) ( v n ) β ( v n ) β ( ( v n ) ) p s 1 .

Due to the function ( t ) is continuous, thus ( t ) D s , p ( R 3 ) , and thanks to He and Zou [23], we have

(5.11) ( Δ ) p s ( v n ) ( v n ) ( Δ ) p s v n .

By the Sobolev inequality and (5.11), we infer that

(5.12) S * ( v n ) p s * p R 3 ( v n ) ( Δ ) p s ( v n ) d x

and

(5.13) 0 R 3 ( v n ) ( Δ ) p s ( v n ) d x R 3 ( v n ) ( v n ) ( Δ ) p s v n d x .

Via the definition of f and ( f 5 ) , there exist ζ > 0 and C ζ > 0 , such that

f ( t ) ζ t p 1 + C ζ t σ 1 ,

where σ ( p , p s * ) . We take ζ ( 0 , V 0 ) , by (5.9), (5.10), (5.12), and (5.13), we have

(5.14) S * ( v n ) p s * p R 3 ( v n ) ( Δ ) p s ( v n ) d x R 3 ( v n ) ( v n ) ( Δ ) p s v n d x = R 3 ϕ v n v n p s 2 v n ( v n ) ( v n ) d x R 3 V n ( x ) v n ( v n ) ( v n ) d x + R 3 v n p s * 2 v n ( v n ) ( v n ) d x + R 3 f ( v n ) ( v n ) ( v n ) d x β R 3 ϕ v n v n p s 2 v n ( ( v n ) ) p s d x R 3 V n ( x ) v n ( v n ) ( v n ) d x + β R 3 v n p s * 2 v n 2 ( v n ) d x + ζ R 3 v n p 1 ( v n ) ( v n ) d x + C ζ R 3 v n σ 1 ( v n ) ( v n ) d x β R 3 ϕ v n v n p s 2 ( ( v n ) ) p s d x + C 5 β R 3 v n p s * 2 2 ( v n ) d x .

Next, we estimate the first term on the right side of (5.14), by using the Hardy-Littlewood-Sobolev inequality, we can obtain

(5.15) R 3 ϕ v n v n p s 2 ( ( v n ) ) p s d x = R 3 R 3 v n ( y ) p s 2 ( v n ( y ) ( v n ) ) p s x y 3 s p d y d x C v n ( y ) p s p s 2 R 3 v n ( y ) p s * ( v n ) p s * d x p s p s * .

There exists m > 0 , we use the following inequality:

( a + b ) j C ( a j + b j )

for any a , b 0 and j ( 0 , 1 ) , C was a positive constant, we can obtain

(5.16) R 3 v n ( y ) p s * ( v n ) p s * d x p s p s * = v n m v n ( y ) p s * ( v n ) p s * d x + v n > m v n ( y ) p s * ( v n ) p s * d x p s p s * C v n m v n ( y ) p s * ( v n ) p s * d x p s p s * + C v n > m v n ( y ) p s * ( v n ) p s * d x p s p s * C m p s * v n m ( v n ) p s * d x p s p s * + C v n > m v n p s * d x p s p s * v n > m ( v n ) p s * d x p s p s * .

Due to v n v in H V 0 , we can take m > 0 large enough. Thus, we can obtain

(5.17) v n > m v n p s * d x p s p s * 1 4 β C .

Combining (5.16) and (5.17), we have

(5.18) R 3 v n ( y ) p s * ( v n ) p s * d x p s p s * C m p s * v n m ( v n ) p s * d x p s p s * + 1 2 β C v n > m ( v n ) p s * d x p s p s * .

Next, we calculate another term in (5.14)

(5.19) R 3 v n p s * 2 2 ( v n ) d x = v n m v n p s * 2 2 ( v n ) d x + v n > m v n p s * 2 2 ( v n ) d x m p s * v n m v n 2 ( v n ) d x + v n > m v n p s * d x p s * 2 p s * v n > m ( v n ) p s * d x 2 p s * m p s * v n m v n p s * d x 1 p s * v n m ( v n ) 2 p s * p s d x p s p s * + v n > m v n p s * d x p s * 2 p s * v n > m ( v n ) p s * d x 2 p s * .

We take m > 0 large enough again. Thus, we can obtain

(5.20) v n > m v n p s * d x p s * 2 p s * 1 4 β C 6 .

Combining (5.16) and (5.17), we have

(5.21) R 3 v n p s * 2 2 ( v n ) d x m p s * C R 3 ( v n ) 2 p s * p s d x p s p s * + 1 4 β C 6 R 3 ( v n ) p s * d x 2 p s * .

Together with (5.14)–(5.16), (5.18)–(5.20), and (5.21), we deduce that

(5.22) ( v n ) p s * P β C S * m p s * R 3 ( v n ) 2 p s * p s d x p s p s * .

Via the famous Fatou’s lemma and we take K , we can obtain that

(5.23) R 3 v n p s * β p β p s * β β C S * m p s * R 3 v n 2 p s * p s β d x p s p s * .

That means

(5.24) v n p s * β C S * m p s * 1 p β β 1 p β v n p β p s * p s * 1 .

Letting β = p s * 1 p s * > 1 , we have

(5.25) v n p s * β C S * m p s * 1 p β β 1 p β v n p s * .

Note that β p = p s * 1 p β p 1 , we obtain

(5.26) v n p s * β p C S * m p s * 1 p β p β 1 β p v n p s * β p 1 .

Substituting (5.25) into (5.26) and repeating the process, we have

(5.27) v n p s * β t C S * m p s * i = 1 t 1 p β i β i = 1 t i p β i v n p s * β p 1 .

We can easily obtain

i = 1 1 β i = 1 β 1 .

Using the D’Alembert principle, we have

lim i i + 1 p β i + 1 × p β i i = lim i i + 1 p β = 1 β < 1 ,

which means that i = 1 j p β i = K for some K R + . Taking the limit in (5.27) as t , we can obtain

v n L ( R 3 ) C S * m p s * 1 p ( β 1 ) β L v n p s * C uniformly in  n N .

Therefore, equation (5.9) can be rewritten in the following form:

( Δ ) p s v n + v n = ϖ n ,

where ϖ n = v n V ( ε n x + ε n y ˜ n ) v n p 2 v n + ϕ v n v n p s 2 v n + f ( v n ) + v n p s * 2 v n . The fact is that ϖ n L ( R 3 ) and is uniformly bounded. By interpolation inequality and v n v in H V 0 , we can obtain ϖ n ϖ in L p ( R 3 ) for any p [ 2 , + ) , where ϖ = v V 0 v p 2 v + ϕ v v p s 2 v + f ( v ) + v p s * 2 v . Thus, we have

v n ( x ) = ( Ξ * ϖ n ) ( x ) = R 3 ( Ξ ( x y ) Λ ϖ n ( y ) ) d y ,

where Ξ is a Bessel potential, which possesses the following properties (see Felmer [19]):

  1. Ξ is positive, radially symmetric, and smooth in R 3 \ { 0 } ;

  2. there exists C > 0 , such that Ξ ( x ) C 3 + s p ;

  3. Ξ L r ( R 3 ) for any r [ 1 , 3 3 s p ) .

Therefore, arguing as in the proof of Lemma 2.6 in Alves and Miyagaki [2], we have that

v n ( x ) 0 as x + ,

uniformly in n N .□

Proof of Theorem 1.1

We suppose that there exist some sequences ε n 0 and consider the corresponding solution u ε n of the energy function, then v ε n ( x ) = u ε n ( x + y ˜ n ) is a solution of equation (5.9). Choose δ > 0 such that Π δ Λ . First, there exists ε ˜ δ > 0 such that, for every ε ( 0 , ε ˜ δ ) , we have

(5.28) u ε L ( Λ ε c ) < a .

Assume that there exist some sequences ε n 0 , we can find u n = u ε n N ε ˜ such that J ε n ( u n ) = 0 and

(5.29) u ε L ( Λ ε c ) a .

From the above fact v ε n ( x ) = u ε n ( x + y ˜ n ) , we can obtain u ε n ( x ) = v ε n ( x y ˜ n ) , it follows from Lemma 5.14 that there exists R > 0 such that

v ε n ( x y ˜ n ) a for any  x y ˜ n R .

We know that

u ε n ( x ) < a for any  x B R c ( y ˜ n ) .

From Lemma 5.1, there exists y 0 Π such that ε n y ˜ n y 0 Π . Thus, there exists r > 0 such that B r ε n y ˜ n Λ , we have

Λ ε n c B r ε n ( y ˜ n ) B R c ( y ˜ n ) for any n n 0 .

So, we deduce that for any x Λ ε n c ,

u ε n ( x ) < a .

Therefore, we reach a contradiction with u ε L ( Λ ε c ) a . Thus, (5.28) holds. So, there exists least cat Π δ ( Π ) positive solutions.

According to u n N ˜ ε , u ε L ( Λ ε c ) < a and the definition of g , we have g ( ε x , u ε ) = f ( u ε ) + u ε p s * 1 . Thus, it follows that u ε is a solution of system (1.1). Then, we take ε n 0 and a sequence { u n } n N of solutions to system (1.1). Arguing as before, there exists R > 0 such that

(5.30) u ε L ( B R c ( y ˜ n ) ) < γ .

And up to a subsequence, we can suppose that

(5.31) u ε L ( B R c ( y ˜ n ) ) γ .

We can easily prove that this is impossible. So (5.30) holds. Taking into account (5.30) and Theorem 5.1, we can deduce that the maximum points p ε n R 3 of u n belong to B R c ( y ˜ n ) . So, p ε n = y ˜ n + p ε n for some q n B R ( 0 ) . Since ε n y ˜ n y 0 and q n < R for any n N , we have ε n p ε n y 0 and combined with the continuity of V yields

lim n V ( ε n p ε n ) = V ( y 0 ) = V 0 .

Therefore, the proof of Theorem 1.1 is completed.□

  1. Funding information: Shaoyun Shi was supported by the NSFC (Grant No. 12271203). Shuaishuai Liang acknowledges the financial support provided by the China Scholarship Council (No. 202306170138). Yueqiang Song was supported by the Science and Technology Development Plan Project of Jilin Province, China (No. 20230101287JC), the National Natural Science Foundation of China (No. 12001061), and the Young outstanding talents project of Scientific Innovation and entrepreneurship in Jilin (No. 20240601048RC).

  2. Author contributions: All authors contributed equally to the writing of this article. All authors read and approved the final manuscript.

  3. Conflict of interest: The authors state no conflict of interest.

  4. Competing interests: The authors declare that they have no competing interests.

  5. Data availability statement: Date sharing is not applicable to this article as no new data were created or analyzed in this study.

References

[1] R. A. Adams, Sobolev Spaces, Pure and Applied Mathematics, Vol. 65, Academic Press, New York-London, 1975. Search in Google Scholar

[2] C. O. Alves and O. H. Miyagaki, Existence and concentration of solution for a class of fractional elliptic equation in RN via penalization method, Calc. Var. Partial Differential Equations 55 (2016), 47, https://doi.org/10.1007/s00526-016-0983-x. Search in Google Scholar

[3] A. Ambrosetti, On Schrödinger-Poisson systems, Milan J. Math. 76 (2008), 257–274, https://doi.org/10.1007/s00032-008-0094-z. Search in Google Scholar

[4] A. Ambrosetti and A. Malchiodi, Perturbation Methods and Semilinear Elliptic Problems on RN, Birkhäuser, Verlag, 2006, https://doi.org/10.1007/3-7643-7396-2. Search in Google Scholar

[5] A. Ambrosetti and D. Ruiz, Multiple bound states for the Schrödinger-Poisson problem, Commun. Contemp. Math. 10 (2008), 391–404, https://doi.org/10.1142/S021919970800282X. Search in Google Scholar

[6] A. Azzollini and A. Pomponio, Ground state solutions for the nonlinear Schrödinger-Maxwell equations, J. Math. Anal. Appl. 345 (2008), 90–108, https://doi.org/10.1016/j.jmaa.2008.03.057. Search in Google Scholar

[7] A. Azzollini, P. d’Avenia, and A. Pomponio, On the Schrödinger-Maxwell equations under the effect of a general nonlinear term, Ann. Inst. Henri Poincaré, Anal. Non Linéire. 27 (2010), 791–799, https://doi.org/10.1016/j.anihpc.2009.11.012. Search in Google Scholar

[8] V. Benci and D. Fortunato, An eigenvalue problem for the Schrödinger-Maxwell equations, Top. Meth. Nonlinear Anal. 11 (1998), 283–293, https://doi.org/10.12775/TMNA.1998.019. Search in Google Scholar

[9] M. Bhakta, K. Perera, and F. Sk, A system of equations involving the fractional p-Laplacian and doubly critical nonlinearities, Adv. Nonlinear Stud. 23 (2023), 20230103, https://doi.org/10.1515/ans-2023-0103. Search in Google Scholar

[10] D. Cao, Nontrivial solution of semilinear elliptic equation with critical exponent in R2, Comm. Partial Differential Equations 17 (1992), 407–435, https://doi.org/10.1080/03605309208820848. Search in Google Scholar

[11] S. Chen, V. D. Rădulescu, and X. Tang, Multiple normalized solutions for the planar Schrödinger-Poisson system with critical exponential growth, Math. Z. 306 (2024), no. 3, Paper No. 50, 32 pp. https://doi.org/10.1007/s00209-024-03432-9. Search in Google Scholar

[12] G. M. Coclite, A multiplicity result for the nonlinear Schrödinger-Maxwell equations, Commun. Appl. Anal. 7 (2003), 417–423. Search in Google Scholar

[13] J. Chabrowski, Concentration-compactness principle at infinity and semilinear elliptic equations involving critical and subcritical Sobolev exponents, Calc. Var. Partial Differential Equations 3 (1995), 493–512, https://doi.org/10.1007/BF01187898. Search in Google Scholar

[14] M. Del Pino and P. L. Felmer, Local Mountain Pass for semilinear elliptic problems in unbounded domains, Calc. Var. Partial Differential Equations 4 (1996), 121–137, https://doi.org/10.1007/BF01189950. Search in Google Scholar

[15] Y. Du, J. B. Su, and C. Wang, On a quasilinear Schrödinger-Poisson system, J. Math. Anal. Appl. 1 (2022), 125446, https://doi.org/10.1016/j.jmaa.2021.125446. Search in Google Scholar

[16] I. Ekeland, On the variational principle, J. Math. Anal. Appl. 47 (1974), 324–353, https://doi.org/10.1016/0022-247X(74)90025-0. Search in Google Scholar

[17] E. Di Nezza, G. Palatucci, and E. Valdinoci, Hitchhiker’s guide to the fractional Sobolev spaces, Bull. Sci. Math. 136 (2012), 521–573, https://doi.org/10.1016/j.bulsci.2011.12.004. Search in Google Scholar

[18] P. D’Avenia, Non-radially symmetric solutions of nonlinear Schrödinger equation coupled with Maxwell equations, Adv. Nonlinear Stud. 2 (2002), 177–192, https://doi.org/10.1515/ans-2002-0205. Search in Google Scholar

[19] P. Felmer, A. Quaas, and J. Tan, Positive solutions of the nonlinear Schrödinger equation with the fractional Laplacian, Proc. R. Soc. Edinb. Sect. A. 142 (2012), 1237–1262, https://doi.org/10.1017/S0308210511000746. Search in Google Scholar

[20] X. J. Feng, Ground state solution for a class of Schrödinger-Poisson-type systems with partial potential, Z. Angew. Math. Phys. 71 (2020), 1–16, https://doi.org/10.1007/s00033-020-1254-4. Search in Google Scholar

[21] X. He, Positive solutions for fractional Schrödinger-Poisson system with doubly critical exponents, Appl. Math. Lett. 120 (2021), 107190. Search in Google Scholar

[22] X. He and D. B. Wang, Multiple bound state solutions for fractional Schrödinger-Poisson systems with critical nonlocal terms, J. Geom. Anal. 33 (2023), 194, https://doi.org/10.1007/s12220-023-01247-4. Search in Google Scholar

[23] X. He and W. Zou, Existence and concentration result for the fractional Schrödinger equations with critical nonlinearities, Calc. Var. Partial Differential Equations 55 (2016), 91, https://doi.org/10.1007/s00526-016-1045-0. Search in Google Scholar

[24] X. He, X. Zhao, and W. Zou, The Benci-Cerami problem for the fractional Choquard equation with critical exponent, Manuscripta Math. 170 (2023), 193–242, https://doi.org/10.1007/s00229-021-01362-y. Search in Google Scholar

[25] L. Jeanjean, On the existence of bounded Palais-Smale sequences and application to a Landesman-Lazer-type problem set on RN, Proc. R. Soc. Edinb. Sect. A. 129 (1999), 787–809, https://doi.org/10.1017/S0308210500013147. Search in Google Scholar

[26] K. Jin and L. Wang, Infinitely many solutions for the nonlinear Schrödinger-Poisson system, J. Dyn. Control Syst. 29 (2023), 1299–1322, https://doi.org/10.1007/s10883-022-09636-8. Search in Google Scholar

[27] S. Liang and V. D. Rădulescu, Least-energy nodal solutions of critical Kirchhoff problems with logarithmic nonlinearity, Anal. Math. Phys. 10 (2020), no. 45, 1–31, https://doi.org/10.1007/s13324-020-00386-z. Search in Google Scholar

[28] C. Lei, V. D. Rădulescu, and B. L. Zhang, Ground states of the Schrödinger-Poisson-Slater equation with critical growth, RACSAM Rev. R. Acad. A. 117 (2023), 128, https://doi.org/10.1007/s13398-023-01457-z. Search in Google Scholar

[29] E. Lieb and M. Loss, Analysis, Graduate Studies in Mathematics, American Mathematical Society, vol. 14. Providence, 2001.Search in Google Scholar

[30] P. L. Lions, The concentration compactness principle in the calculus of variations, The locally compact case, I, II, Ann. Inst. Poincaré Anal. Nonlin. 1 (1984), 109–145, 223–283, https://doi.org/10.1016/S0294-1449(16)30428-0. Search in Google Scholar

[31] P. L. Lions, The Choquard equation and related questions, Nonlinear Anal. 4 (1980), 1063–1072, https://doi.org/10.1016/0362-546X(80)90016-4. Search in Google Scholar

[32] X. Liang and B. L. Zhang, A modified zero energy critical point theory with applications to several nonlocal problems, Electron. J. Qual. Theory Differ. Equ. 6 (2024), 20, https://doi.org/10.14232/ejqtde.2024.1.6. Search in Google Scholar

[33] Y. Li, B. L. Zhang, and X. Han, Existence and concentration behavior of positive solutions to Schrödinger-Poisson-Slater equations, Adv. Nonlinear Anal. 12 (2023), 20220293, https://doi.org/10.1515/anona-2022-0293. Search in Google Scholar

[34] Y. Li, V. D. Rădulescu, and B. Zhang, Critical planar Schrödinger-Poisson equations: existence, multiplicity and concentration, Math. Z. 307 (2024), 43, https://doi.org/10.1007/s00209-024-03520-w. Search in Google Scholar

[35] G. B. Li and S. S. Yan, Eigenvalue problems for quasilinear elliptic equations on RN, Comm. Partial Differential Equations 14 (1989), 1291–1314, https://doi.org/10.1080/03605308908820654. Search in Google Scholar

[36] Z. Liu, Z. Zhang, and S. Huang, Existence and nonexistence of positive solutions for a static Schrödinger-Poisson-Slater equation, J. Differential Equations 266 (2019), 5912–5941, https://doi.org/10.1016/j.jde.2018.10.048. Search in Google Scholar

[37] Z. Liu, L. Tao, D. Zhang, S. Liang, and Y. Song, Critical nonlocal Schrödinger-Poisson system on the Heisenberg group, Adv. Nonlinear Anal. 11 (2022), 482–502, https://doi.org/10.1515/anona-2021-0203. Search in Google Scholar

[38] S. L. Liu and H. B. Chen, Existence of ground-state solutions for p-Choquard equations with singular potential and doubly critical exponents, Math. Nachr. 296 (2023), 2467–2502, https://doi.org/10.1002/mana.202100255. Search in Google Scholar

[39] J. Moser, A new proof of De Giorgias theorem concerning the regularity problem for elliptic differential equations, Comm. Pure Appl. Math. 13 (1960), 457–468, https://doi.org/10.1002/cpa.3160130308. Search in Google Scholar

[40] Y. X. Meng and X. M. He, Normalized solutions for the Schrödinger-Poisson system with doubly critical growth, Topol. Methods Nonlinear Anal. 62 (2023), 509–534, https://doi.org/10.12775/TMNA.2022.075. Search in Google Scholar

[41] N. S. Papageorgiou, J. Zhang, and W. Zhang, Global existence and multiplicity of solutions for nonlinear singular eigenvalue problems, Discrete Contin. Dyn. Syst. Ser. S (2024), https://doi.org/10.3934/dcdss.2024018. Search in Google Scholar

[42] P. Pucci, M. Xiang, and B. L. Zhang, Multiple solutions for nonhomogeneous Schrödinger-Kirchhoff type equations involving the fractional p-Laplacian in RN, Calc. Var. Partial Differential Equations 54 (2015), 2785–2806, https://doi.org/10.1007/s00526-015-0883-5. Search in Google Scholar

[43] P. Pucci, L. Wang, and B. Zhang, Bifurcation and regularity of entire solutions for the planar nonlinear Schrödinger-Poisson system, Math. Ann. 389 (2024), 4265–4300, https://doi.org/10.1007/s00208-023-02752-1. Search in Google Scholar

[44] S. Qu and X. He, On the number of concentrating solutions of a fractional Schrödinger-Poisson system with doubly critical growth, Anal. Math. Phys. 12 (2022), 59, https://doi.org/10.1007/s13324-022-00675-9. Search in Google Scholar

[45] D. Qin, V. D. Rădulescu, and X. Tang, Ground states and geometrically distinct solutions for periodic Choquard-Pekar equations, J. Differential Equations 275 (2021), 652–683, https://doi.org/10.1016/j.jde.2020.11.021. Search in Google Scholar

[46] P. H. Rabinowitz, On a class of nonlinear Schrödinger equations, Z angew Math Phys. 43 (1992), 270–290, https://doi.org/10.1006/jdeq.2000.3946. Search in Google Scholar

[47] D. Ruiz, The Schrödinger-Poisson equation under the effect of a nonlinear local term, J. Funct. Anal. 237 (2006), 655–674, https://doi.org/10.1016/j.jfa.2006.04.005. Search in Google Scholar

[48] Y. Su, Positive solution to Schrödinger equation with singular potential and double critical exponents, Atti Accad. Naz. Lincei Rend. Lincei Mat. Appl. 31 (2020), 667–698, https://doi.org/10.4171/RLM/910. Search in Google Scholar

[49] A. Szulkin, Ljusternik-Schnirelmann theory on C1-manifolds, Ann. Inst. H. Poincaré Anal. NonLinéaire 5 (1988), 119–139, https://doi.org/10.1016/S0294-1449(16)30348-1. Search in Google Scholar

[50] A. Szulkin and T. Weth, The method of Nehari manifold, Handbook of Nonconvex Analysis and Applications, International Press, Boston, 2010, pp. 597–632. Search in Google Scholar

[51] K. Teng, Existence of ground-state solutions for the nonlinear fractional Schrödinger-Poisson system with critical Sobolev exponent, J. Differential Equations 261 (2016), 3061–3106, https://doi.org/10.1016/j.jde.2016.05.022. Search in Google Scholar

[52] N. V. Thin, Multiplicity and concentration of solutions to a fractional p-Laplace problem with exponential growth, Ann. Fenn. Math. 47 (2022), 603–639, https://doi.org/10.54330/afm.115564. Search in Google Scholar

[53] N. S. Trudinger, On Harnack type inequalities and their application to quasilinear elliptic equations, Commun. Pure Appl. Math. 20 (1967), 721–747, https://doi.org/10.1002/cpa.3160200406. Search in Google Scholar

[54] M. Willem, Minimax Theorems, Progress in Nonlinear Differential Equations and their Applications, Birkhäuser Boston, Inc. Boston, MA, 1996. Search in Google Scholar

[55] J. Zhang, J. M. do Ó, and M. Squassina, Fractional Schrödinger-Poisson systems with a general subcritical or critical nonlinearity, Adv. Nonlinear Stud. 16 (2016), 15–30, https://doi.org/10.1515/ans-2015-5024. Search in Google Scholar

[56] J. Zhang, H. Zhou, and H. Mi, Multiplicity of semiclassical solutions for a class of nonlinear Hamiltonian elliptic system, Adv. Nonlinear Anal. 13 (2024), 20230139, https://doi.org/10.1515/anona-2023-0139. Search in Google Scholar

[57] J. Zhang and Y. Zhang, An infinite sequence of localized semiclassical states for nonlinear Maxwell-Dirac system, J. Geometric Analysis 34 (2024), 277, https://doi.org/10.1007/s12220-024-01724-4. Search in Google Scholar

Received: 2024-08-24
Revised: 2024-12-17
Accepted: 2024-12-31
Published Online: 2025-02-01

© 2025 the author(s), published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Research Articles
  2. Incompressible limit for the compressible viscoelastic fluids in critical space
  3. Concentrating solutions for double critical fractional Schrödinger-Poisson system with p-Laplacian in ℝ3
  4. Intervals of bifurcation points for semilinear elliptic problems
  5. On multiplicity of solutions to nonlinear Dirac equation with local super-quadratic growth
  6. Entire radial bounded solutions for Leray-Lions equations of (p, q)-type
  7. Combinatorial pth Calabi flows for total geodesic curvatures in spherical background geometry
  8. Sharp upper bounds for the capacity in the hyperbolic and Euclidean spaces
  9. Positive solutions for asymptotically linear Schrödinger equation on hyperbolic space
  10. Existence and non-degeneracy of the normalized spike solutions to the fractional Schrödinger equations
  11. Existence results for non-coercive problems
  12. Ground states for Schrödinger-Poisson system with zero mass and the Coulomb critical exponent
  13. Geometric characterization of generalized Hajłasz-Sobolev embedding domains
  14. Subharmonic solutions of first-order Hamiltonian systems
  15. Sharp asymptotic expansions of entire large solutions to a class of k-Hessian equations with weights
  16. Stability of pyramidal traveling fronts in time-periodic reaction–diffusion equations with degenerate monostable and ignition nonlinearities
  17. Ground-state solutions for fractional Kirchhoff-Choquard equations with critical growth
  18. Existence results of variable exponent double-phase multivalued elliptic inequalities with logarithmic perturbation and convections
  19. Homoclinic solutions in periodic partial difference equations
  20. Classical solution for compressible Navier-Stokes-Korteweg equations with zero sound speed
  21. Properties of minimizers for L2-subcritical Kirchhoff energy functionals
  22. Asymptotic behavior of global mild solutions to the Keller-Segel-Navier-Stokes system in Lorentz spaces
  23. Blow-up solutions to the Hartree-Fock type Schrödinger equation with the critical rotational speed
  24. Qualitative properties of solutions for dual fractional parabolic equations involving nonlocal Monge-Ampère operator
  25. Regularity for double-phase functionals with nearly linear growth and two modulating coefficients
  26. Uniform boundedness and compactness for the commutator of an extension of Riesz transform on stratified Lie groups
  27. Normalized solutions to nonlinear Schrödinger equations with mixed fractional Laplacians
  28. Existence of positive radial solutions of general quasilinear elliptic systems
  29. Low Mach number and non-resistive limit of magnetohydrodynamic equations with large temperature variations in general bounded domains
  30. Sharp viscous shock waves for relaxation model with degeneracy
  31. Bifurcation and multiplicity results for critical problems involving the p-Grushin operator
  32. Asymptotic behavior of solutions of a free boundary model with seasonal succession and impulsive harvesting
  33. Blowing-up solutions concentrated along minimal submanifolds for some supercritical Hamiltonian systems on Riemannian manifolds
  34. Stability of rarefaction wave for relaxed compressible Navier-Stokes equations with density-dependent viscosity
  35. Singularity for the macroscopic production model with Chaplygin gas
  36. Global strong solution of compressible flow with spherically symmetric data and density-dependent viscosities
  37. Global dynamics of population-toxicant models with nonlocal dispersals
  38. α-Mean curvature flow of non-compact complete convex hypersurfaces and the evolution of level sets
  39. High-energy solutions for coupled Schrödinger systems with critical growth and lack of compactness
  40. On the structure and lifespan of smooth solutions for the two-dimensional hyperbolic geometric flow equation
  41. Well-posedness for physical vacuum free boundary problem of compressible Euler equations with time-dependent damping
  42. On the existence of solutions of infinite systems of Volterra-Hammerstein-Stieltjes integral equations
  43. Remark on the analyticity of the fractional Fokker-Planck equation
  44. Continuous dependence on initial data for damped fourth-order wave equation with strain term
  45. Unilateral problems for quasilinear operators with fractional Riesz gradients
  46. Boundedness of solutions to quasilinear elliptic systems
  47. Existence of positive solutions for critical p-Laplacian equation with critical Neumann boundary condition
  48. Non-local diffusion and pulse intervention in a faecal-oral model with moving infected fronts
  49. Nonsmooth analysis of doubly nonlinear second-order evolution equations with nonconvex energy functionals
  50. Qualitative properties of solutions to the viscoelastic beam equation with damping and logarithmic nonlinear source terms
  51. Shape of extremal functions for weighted Sobolev-type inequalities
  52. One-dimensional boundary blow up problem with a nonlocal term
  53. Doubling measure and regularity to K-quasiminimizers of double-phase energy
  54. General solutions of weakly delayed discrete systems in 3D
  55. Global well-posedness of a nonlinear Boussinesq-fluid-structure interaction system with large initial data
  56. Optimal large time behavior of the 3D rate type viscoelastic fluids
  57. Local well-posedness for the two-component Benjamin-Ono equation
  58. Self-similar blow-up solutions of the four-dimensional Schrödinger-Wave system
  59. Existence and stability of traveling waves in a Keller-Segel system with nonlinear stimulation
  60. Existence and multiplicity of solutions for a class of superlinear p-Laplacian equations
  61. On the global large regular solutions of the 1D degenerate compressible Navier-Stokes equations
  62. Normal forms of piecewise-smooth monodromic systems
  63. Fractional Dirichlet problems with singular and non-locally convective reaction
  64. Sharp forced waves of degenerate diffusion equations in shifting environments
  65. Global boundedness and stability of a quasilinear two-species chemotaxis-competition model with nonlinear sensitivities
  66. Non-existence, radial symmetry, monotonicity, and Liouville theorem of master equations with fractional p-Laplacian
  67. Global existence, asymptotic behavior, and finite time blow up of solutions for a class of generalized thermoelastic system with p-Laplacian
  68. Formation of singularities for a linearly degenerate hyperbolic system arising in magnetohydrodynamics
  69. Linear stability and bifurcation analysis for a free boundary problem arising in a double-layered tumor model
  70. Hopf's lemma, asymptotic radial symmetry, and monotonicity of solutions to the logarithmic Laplacian parabolic system
  71. Generalized quasi-linear fractional Wentzell problems
  72. Existence, symmetry, and regularity of ground states of a nonlinear Choquard equation in the hyperbolic space
  73. Normalized solutions for NLS equations with general nonlinearity on compact metric graphs
  74. Boundedness and global stability in a predator-prey chemotaxis system with indirect pursuit-evasion interaction and nonlocal kinetics
  75. Review Article
  76. Existence and stability of contact discontinuities to piston problems
Downloaded on 29.1.2026 from https://www.degruyterbrill.com/document/doi/10.1515/anona-2024-0063/html
Scroll to top button