Abstract
In this paper, we consider the following nonlinear Choquard equation with magnetic field
where \(\varepsilon >0\) is a small parameter, \(N\ge 3\), \(0<\mu <N\), \(2_{\mu }^{*}=\frac{2N-\mu }{N-2}\), \(V(x):{\mathbb {R}}^{N}\rightarrow {\mathbb {R}}^{N}\) and \(A(x):{\mathbb {R}}^{N}\rightarrow {\mathbb {R}}^{N}\) is a continuous potential, f is a continuous subcritical term, and F is the primitive function of f. Under a local assumption on the potential V, by the variational methods, the penalization techniques and the Ljusternik–Schnirelmann theory, we prove the multiplicity and concentration properties of nontrivial solutions of the above problem for \(\varepsilon >0\) small enough.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction and main results
In this paper, we study the following nonlinear Choquard equation with magnetic fields
where \(x\in {\mathbb {R}}^{N}\), \(\varepsilon >0\) is a parameter, \(N\ge 3\), \(0<\mu <N\), \(2_{\mu }^{*}=\frac{2N-\mu }{N-2}\), \(V(x)\in C({\mathbb {R}}^{N},{\mathbb {R}})\) is an electric potential, and \(A(x)\in C({\mathbb {R}}^{N},{\mathbb {R}}^{N})\) is a magnetic potential. The operator \(\left( \frac{1}{i}\nabla -A(x)\right) ^{2}\), called magnetic Laplacian, is defined by
For problem (1.1), there are vast literature concerning the existence and multiplicity of nonlinear Choquard equation without magnetic field, namely \(A \equiv 0\). Then, (1.1) turns to the Choquard equation
When \(\varepsilon =1\), Eq. (1.2) reduces to the generalized Choquard equation:
The case of (1.3) \(N=3, p=2\), \(V(x)=1\) and \(\mu =1\) describes the quantum mechanics of a polaron at rest by Pekar [36]. In 1976, Choquard [26] described an approximation to Hartree–Fock theory of a one component plasma by (1.3). Thus, Eq. (1.3) is also called the nonlinear Schrödinger–Newton equation. By using critical point theory, Lions [28] obtained the existence of infinitely many radially symmetric solutions in \(H^{1}\left( {\mathbb {R}}^{N}\right) \) and Ackermann [1] proved the existence of infinitely many geometrically distinct weak solutions for a general case.
Moroz and Van Schaftingen [31, 33] eliminated this restriction and showed the regularity, positivity and radial symmetry of the ground states for the optimal range of parameters, and also derived that these solutions decay asymptotically at infinity.
On the other hand, the semiclassical limit problem attracts people’s attention, namely \(\varepsilon \rightarrow 0\) in (1.2). The question of the existence of semiclassical solutions for the nonlocal problem (1.2) has been posed in [6]. Moroz and Van Schaftingen [32] develop a novel nonlocal penalization by variational methods to show that equation (1.2) with \(G(u)=|u|^{q}\) has a family of solutions concentrated at the local minimum of V, with V satisfying some additional assumptions at infinity. Other results see [5, 12, 25, 30, 43] and references therein.
The magnetic nonlinear Schrödinger equation (1.1) has been extensively investigated by many authors by applying suitable variational and topological methods (see [3, 9, 13, 14, 16, 17, 20, 23, 40, 41] and references therein). To the best of our knowledge, the first result involving the magnetic field was obtained by Esteban and Lions [20]. They used the concentration–compactness principle and minimization arguments to obtain solutions for \(\varepsilon >0\) fixed and \(N=2,3\). For the nonlinear magnetic Schrödinger equation
Alves [3] used the method of the Nehari manifold, the penalization method and the Ljusternik–Schnirelmann category theory to relate the number of solutions with the topology of the set for a subcritical nonlinearity \(f \in C^{1}\). Ji and R\(\breve{a}\)dulescu [23] showed that the arguments developed in [3] fail if f is only continuous, and they improved the methods to study the multiplicity and concentration results for magnetic Schrödinger equation in which the subcritical nonlinearity f is only continuous. Later, Ji and R\(\breve{a}\)dulescu [24] used the same methods to study the multiplicity and concentration results for magnetic Schrödinger equation with critical growth.
About the Choquard equations with magnetic potential like
For \(\varepsilon =1\), \(\mu \in (0,N)\), \(\frac{2N-\mu }{N}<p<\frac{2N-\mu }{N-2}\), \(f(u)=|u|^{p}\), Cingolani et al. [14] established the existence of multiple complex-valued solutions. Alves et al. [4] proved the existence and multiplicity of solutions by using variational methods, penalization techniques and Ljusternik–Schnirelmann theory under the suitable conditions of V and f.
It is quite natural to consider the multiplicity and concentration phenomena of nontrivial solutions for problem (1.1) with critical growth. Inspired by [4, 24], the main purpose of this paper is to investigate multiplicity and concentration of nontrivial solutions for problem (1.1) by proposing a local assumption on V(x) and adapting the penalization technique and Ljusternik–Schnirelmann category theory.
To go on studying the problem (1.1), we recall an important inequality, namely:
Lemma 1.1
(Hardy–Littlewood–Sobolev inequality) [27] Let \(t,r>1\) and \(0<\mu <N\) with \(1/t+\mu /N+1/r=2\), \(f\in L^{t}({\mathbb {R}}^N)\) and \(h\in L^{r}({\mathbb {R}}^N)\). There exists a sharp constant \(C(t,N,\mu ,r)\), independent of f, h, such that
If \(t=r=2N/(2N-\mu )\), then there is equality in (1.6) if and only if \(f\equiv Ch\) and
for some \(A\in {\mathbb {C}}\), \(0\ne \gamma \in {\mathbb {R}}\) and \(a\in {\mathbb {R}}^{N}\).
Notice that, by the Hardy–Littlewood–Sobolev inequality, the integral
is well defined if \(|u|^{q}\in L^{t}({\mathbb {R}}^N)\) for some \(t>1\) satisfying
Thus, for \(u\in H^{1}({\mathbb {R}}^N)\), by Sobolev embedding Theorems, we know
that is
Thus, \(\frac{2N-\mu }{N}\) is called the lower critical exponent and \(2_{\mu }^{*}=\frac{2N-\mu }{N-2}\) is the upper critical exponent in the sense of the Hardy–Littlewood–Sobolev inequality and the Laplace operator.
Before giving our main result, we present the following hypothesis on the potential \(V(x)\in C({\mathbb {R}}^{N},{\mathbb {R}})\):
\((V_{1})\) there exists \(V_{1}>0\) such that \(V_{1}=\inf _{x\in {\mathbb {R}}^{N}}V(x)\);
\((V_{2})\) there exists a bounded open set \(\Lambda \subset {\mathbb {R}}^{N}\)such that
Observe that
Moreover, the nonlinearity \(f \in C({\mathbb {R}},{\mathbb {R}})\) be a function satisfying:
- \((f_{1})\):
-
\(f(t)=0\) if \(t \le 0\);
- \((f_{2})\):
-
there exists \(p,q \in \left( \frac{2N-\mu }{N},2_{\mu }^{*}\right) \) and \(\lambda >0\) such that
$$\begin{aligned} f(t)\ge \lambda t^{(p-2)/2} \quad \forall t>0, \quad \lim _{t\rightarrow +\infty } \frac{f(t^{2}) t}{t^{q-1}}=0; \end{aligned}$$ - \((f_{3})\):
-
f(t) is strictly increasing in \((0,\infty )\).
The main result of this paper is the following:
Theorem 5.5
Suppose that condictions \((V_{1})\)–\((V_{2})\) and \((f_{1})\)–\((f_{3})\) are satisfied. Then, for any \(\delta >0\) such that
there exists \(\varepsilon _{\delta }>0\) such that, for any \(0<\varepsilon <\varepsilon _{\delta }\), problem (1.1) has at least \({\text {cat}}_{M_{\delta }}(M)\) nontrivial solutions. Moreover, for every sequence \(\{\varepsilon _{n}\}\) such that \(\varepsilon _{n}\rightarrow 0^{+}\) as \(n \rightarrow \infty \), if we denote by \(u_{\varepsilon _{n}}\) one of these solutions of (1.1) for \(\varepsilon =\varepsilon _{n}\) and \(\eta _{\varepsilon _{n}}\in {\mathbb {R}}^{N}\) the global maximum point of \(|u_{\varepsilon _{n}}|\), then
We shall use the variational methods, penalization techniques and the Ljusternik–Schnirelmann theory to prove Theorem 1.2. There are three main difficulties to overcome. The first difficulty is the presence of the magnetic field A(x), problem (1.1) became a complex-valued problem. We use the diamagnetic inequality to transform complex values into real values, while corresponding embedding theorem also holds. The second one is that the nonlocal terms appearing in the equation, the commonly used methods and techniques may be ineffective. We can use the Hardy–Littlewood–Sobolev inequality to overcome this difficulty. The last one is that critical exponent occurs the right of problem (1.1), we need more accurate estimate to overcome the lack of compactness.
The paper is organized as follows. In Sect. 2, we give some preliminaries. In Sect. 3, we introduce a modified problem. In Sect. 4, we study the associated autonomous problem. In Sect. 5, we obtain the multiplicity of solutions for problem (1.1). In Sect. 6, we complete the proof Theorem 1.2.
2 Preliminaries
In this section, we collect some notations and some useful preliminary lemmas. Recall that the Sobolev space with magnetic potential \(H_{A}^{1}({\mathbb {R}}^{N},{\mathbb {C}})\) is defined by
where \(\nabla _{A}u=(i^{-1}\nabla -A)u\). The space \(H_{A}^{1}({\mathbb {R}}^{N},{\mathbb {C}})\) is an Hilbert space endowed with the scalar product
where \(\Re \) and the bar denote the real part of a complex number and the complex conjugation, respectively.
Using the change of variable \(x\rightarrow \varepsilon x\), problem (1.1) is equivalent to the following one
where \(A_{\varepsilon }(x)=A(\varepsilon x)\) and \(V_{\varepsilon }(x)=V(\varepsilon x)\).
Owing to the presence of potential \(V_{\varepsilon }(x)\), we introduce the subspace
The space \(H_{\varepsilon }\) is an Hilbert space endowed with the scalar product
where \(\nabla _{\varepsilon }u=(D_{\varepsilon }^{1}u,\ldots ,D_{\varepsilon }^{N}u)\) and \(D_{\varepsilon }^{j}u=i^{-1}\partial _{j}u-A_{j}(\varepsilon x)u\) for \(j=1,\ldots ,N\). The norm induced by this inner product is given by
For readers’ convenience, we recall some useful lemmas.
Lemma 2.1
(Diamagnetic Inequality) [20] For any \(u\in H_{\varepsilon }\), we get \(|u|\in H^{1}({\mathbb {R}}^{N},{\mathbb {R}})\) and it holds
Lemma 2.2
[20] The space \(H_{\varepsilon }\) is continuously embedded into \(L^{r}\left( {\mathbb {R}}^{N},{\mathbb {C}}\right) \) for any \(r \in \left[ 2,2^{*}\right] \) and compactly embedded into \(L^{r}_{loc}\left( {\mathbb {R}}^{N}, {\mathbb {C}}\right) \) for any \(r \in \left[ 1,2^{*}\right) \).
From Gao and Yang [21], we denote by
where \(D_{A}^{1,2}({\mathbb {R}}^{N})=\{u \in L^{2^{*}} ({\mathbb {R}}^{N},{\mathbb {C}}):\nabla _{A}u \in L^{2}({\mathbb {R}}^{N},{\mathbb {C}})\}\). The equality between \(S_{H,L}\) and \(S_{A}\) was proved in Mukherjee and Sreenadh [34]. We remark that \(S_{A}\) is attained if and only if \({\text {rot}} A=0\), see [34, Theorem 4.1].
Lemma 2.3
[21] The constant \(S_{H,L}\) defined in (2.3) is achieved if and only if
where \(C>0\) is a fixed constant, \(a\in {\mathbb {R}}^{N}\) and \(b\in (0,\infty )\) are parameters. Furthermore,
where S is the best Sobolev constant of the immersion \(D^{1,2}({\mathbb {R}}^{N}) \hookrightarrow L^{2^{*}}({\mathbb {R}}^{N})\) and \(C(N, \mu )\) depends on N and \(\mu \).
If we consider the minimizer for S given by \(U(x):=\frac{[N(N-2)]^{\frac{N-2}{4}}}{\left( |1+|x|^{2} \right) ^{\frac{N-2}{2}}}\) (see [42, Theorem 1.42]), then
is the unique minimizer for \(S_{H,L}\) that satisfies
with
Lemma 2.4
[29] Let \(N\ge 3\) and \(r\in [2,2^{*})\). If \(\left\{ u_{n}\right\} _{n\in {\mathbb {N}}}\) is a bounded sequence in \(H^{1}({\mathbb {R}}^{N},{\mathbb {R}})\) and if
where \(R>0\), then \(u_{n}\rightarrow 0\) in \(L^{t}({\mathbb {R}}^{N},{\mathbb {R}})\) for all \(t\in (2,2^{*})\).
3 The modified problem
We will adapt an argument explored by the penalization method, which introduced by del Pino and Felmer [19] to overcome the lack of compactness. More precisely, let \(K>0\) determined later, there exists a unique number \(a>0\) satisfying
by \((f_{3})\), where \(V_{0}\) is given in \((V_{2})\). Suppose that
and introduce the penalized nonlinearity
where \(\chi _{\Lambda }\) is the characteristic function on \(\Lambda \) and \(G(x,t):=\int \limits _{0}^{t}f(s)\text {d}s\).
From assumptions \((f_{1})\)–\((f_{3})\), it follows that g satisfies the following properties:
- \((g_{1})\):
-
\(g(x,t)=0\) for each \(t\le 0\) and \(\lim \limits _{t\rightarrow 0^{+}}g(x,t)=0\) uniformly in \(x\in {\mathbb {R}}^{N}\);
- \((g_{2})\):
-
\(g(x,t)\le f(t)+(t^{+})^{2_{\mu }^{*}}\) for all \(t\ge 0\) uniformly in \(x\in {\mathbb {R}}^{N}\);
- \((g_{3})\):
-
\(0<G(t)\le g(x,t)t\), for each \(x\in \Lambda , t>0\);
- \((g_{4})\):
-
\(0<G(t)\le g(x,t)t\le V_{0}t/K\), for each \(x\in \Lambda ^{c}, t>0\);
- \((g_{5})\):
-
\(t\mapsto g(x,t)\) and \(t\mapsto \frac{G(x,t)}{t}\) are increasing for all \(x\in {\mathbb {R}}^{N}\) and \(t\in (0,\infty )\). Then, we consider the following modified problem
$$\begin{aligned} \bigg (\frac{1}{i}\nabla -A_{\varepsilon }(x)\bigg )^{2}u+V_{\varepsilon }(x)u =\left( \frac{1}{|x|^{\mu }}*G(\varepsilon x,|u|^{2})\right) g(\varepsilon x,|u|^{2})u, \hspace{2mm}x\in {\mathbb {R}}^{N}. \end{aligned}$$(3.1)Observe that if u is a nontrivial solution of problem (3.1) with
$$\begin{aligned} |u(x)|^{2}\le a,~~~~~\forall x\in \Lambda ^{c}_{\varepsilon }, \end{aligned}$$where \(\Lambda _{\varepsilon }=\{x\in {\mathbb {R}}^{N}:\varepsilon x\in \Lambda \}\), then u is a nontrivial solution of problem (2.1).
The energy function associated to problem (3.1) is
In the sequel, set \(p_{0}=\frac{2N}{2N-\mu }\), then for any \(u\in H_{\varepsilon }\), we have
and
Therefore, the Hardy–Littlewood–Sobolev inequality implies that
and
As the argument in the proof of [5, Theorem 1.3], we have the following lemma.
Lemma 3.1
Suppose that \(\left( V_1\right) \) and \(\left( f_1\right) -\left( f_3\right) \) are satisfied, then \(J_{\varepsilon }^{\prime }\) is weakly sequentially continuous. Namely, if \(u_n \rightarrow u\) in \(H_{\varepsilon }\), then \(J_{\varepsilon }^{\prime }\left( u_n\right) \rightarrow J_{\varepsilon }^{\prime }(u)\) in \(\left( H_{\varepsilon }\right) ^*\).
Proof
Observe that
Since f(t) has subcritical growth in the sense of the Hardy–Littlewood–Sobolev inequality and the definition of g(x, t), to prove that \(J_{\varepsilon }^{\prime }\) is weakly sequentially continuous, we only need to check that if \(u_n \rightarrow u\) in \(H_{\varepsilon }\), then
for any \(v \in H_{\varepsilon }\) as \(n \rightarrow \infty \). Indeed, by the Hardy–Littlewood–Sobolev inequality, the Riesz potential defines a linear continuous map from \(L^{\frac{2 N}{2 N-\mu }}\left( {\mathbb {R}}^N\right) \) to \(L^{\frac{2 N}{\mu }}\left( {\mathbb {R}}^N\right) \), so we know
Since \(\left| u_n\right| ^{2_\mu ^*-2} u_n \rightarrow |u|^{2_\mu ^*-2} u\) in \(L^{\frac{2 N}{N-\mu +2}}\left( {\mathbb {R}}^N\right) \) as \(n \rightarrow \infty \), we have
as \(n \rightarrow \infty \). Hence (3.6) holds. \(\square \)
By a similar argument to Proposition 3.2 in [33], we have the following results.
Lemma 3.2
Let \(N\ge 3\) and \(\mu \in (0,N)\). If \(H,K\in L^{\frac{2 N}{N-\mu +2}}({\mathbb {R}}^{N})+L^{\frac{2N}{N-\mu }}({\mathbb {R}}^{N})\), and \(u\in H_{\varepsilon }\) satisfies
Then, \(u \in L^{r}({\mathbb {R}}^{N})\) for some \(r \in \left[ 2,\frac{N}{N-\mu }\frac{2N}{N-2}\right) \). Moreover, there exists a positive constant C(r) independent of u such that
Now, let us define
where \(d>0\) is a constant, and we set
The next lemma implies that we can treat the convolution term as a bounded term.
Lemma 3.3
Suppose that \((f_{1})-(f_{3})\) hold. Then, there exists \(C_{0}>0\) such that
Proof
Observe that
Thereby,
From [4, Lemma 2.5], we know there exist a constant C independent of x and \(\varepsilon \) such that
Let \(H(u)=K(u)=|u|^{\frac{N-\mu +2}{N-2}}\), then it is easy to verify that
Using Lemma 3.2, we obtain \(u\in L^{r}({\mathbb {R}}^{N})\) for \(r \in \left[ 2,\frac{N}{N-\mu }\frac{2N}{N-2}\right) \). Moreover,
for some constant \(C>0\) independent of \(\varepsilon \).
It follows from \(\mu <N\) that
Choosing \(t>\frac{N}{\mu }\) and \(\frac{2N^{2}}{3N\mu -\mu ^{2}}<s<\frac{N}{\mu }\), then
By the Hölder inequality, we have
and
where the constant C is independent of x and \(\varepsilon \).
The above equalities with (3.8) imply (3.7). Then, we can find \(C_{0}>0\) such that
\(\square \)
From Lemma 3.3, we know there exists \(k_{0}\) such that
hence we can choose \(K\ge k_{0}\).
We denote by \({\mathcal {N}}_{\varepsilon }\) the Nehari manifold of \(J_{\varepsilon }\), that is,
and define the number \(c_{\varepsilon }\) by
Let \(H_{\varepsilon }^{+}\) be the open subset of \(H_{\varepsilon }\) given by
and \(S_{\varepsilon }^{+}=S_{\varepsilon }\cap H_{\varepsilon }^{+}\), where \(S_{\varepsilon }\) is the unit sphere of \(H_{\varepsilon }\). Observe that \(S_{\varepsilon }^{+}\) is a non-complete \(C^{1,1}\)-manifold of codimension 1, modeled on \(H_{\varepsilon }\) and contained in \(H_{\varepsilon }^{+}\). Hence, \(H_{\varepsilon }=T_{u}S_{\varepsilon }^{+}\oplus {\mathbb {R}}u\) for each \(u\in S_{\varepsilon }^{+}\), where \(T_{u}S_{\varepsilon }^{+}=\{v\in H_{\varepsilon }:\langle u,v\rangle _{\varepsilon }=0\}\).
The functional \(J_{\varepsilon }\) satisfies the mountain pass geometry [42].
Lemma 3.4
For any fixed \(\varepsilon >0\), the functional \(J_{\varepsilon }\) satisfies the following properties:
(a) there exist \(\alpha ,\rho >0\) such that \(J_{\varepsilon }(u)\ge \alpha \) with \(\Vert u\Vert _{\varepsilon }=\rho \);
(b) there exists \(e\in H_{\varepsilon }\) such that \(\Vert e\Vert _{\varepsilon }\ge \rho \) and \(J_{\varepsilon }(e)<0.\)
Proof
-
(a)
For any \(u\in H_{\varepsilon }\setminus \{0\}\), it follows from (3.4) and Lemma 2.2 that
$$\begin{aligned} \begin{aligned} J_{\varepsilon }(u)&=\frac{1}{2}\Vert u\Vert _{\varepsilon }^{2}-\frac{1}{2\cdot 2_{\mu }^{*}}\int \limits _{{\mathbb {R}}^{N}}\left( \frac{1 }{|x|^{\mu }} *G(\varepsilon x, |u|^{2})\right) G(\varepsilon x, |u|^{2})\text {d}x\\&\ge \frac{1}{2}\Vert u\Vert _{\varepsilon }^{2} -C\left( \Vert u\Vert _{\varepsilon }^{4} -\Vert u\Vert _{\varepsilon }^{2q} -\Vert u\Vert _{\varepsilon }^{2\cdot 2_{\mu }^{*}}\right) . \end{aligned} \end{aligned}$$Hence, we can find \(\alpha ,\rho >0\) such that \(J_{\varepsilon }\ge \alpha \) with \(\Vert u\Vert _{\varepsilon }=\rho \).
-
(b)
Fix a positive function \(u_{0} \in H_{\varepsilon }^{+}\) with \({\text {supp}}\left( u_{0}\right) \subset \Lambda _{\varepsilon }\), and observe
$$\begin{aligned} G(\varepsilon x,|u_{0}|^{2})=F(|u_{0}|^{2}). \end{aligned}$$
Set
where
By assumption \((f_{3})\), we have
Therefore, we deduce that
Integrating (3.9) on \(\left[ 1, t\left\| u_{0}\right\| _{\varepsilon }\right] \) with \(t>\frac{1}{\left\| u_{0}\right\| _{\varepsilon }}\), we have
Hence, we have
Taking \(e=t u_{0}\) with t sufficiently large, we can see (b) holds. \(\square \)
Since f is only continuous, we can apply the next two conclusions to overcome the non-differentiability of \({\mathcal {N}}_{\varepsilon }\) and the incompleteness of \(S_{\varepsilon }^{+}\).
Lemma 3.5
Suppose that the function V satisfies \((V_{1})\)–\((V_{2})\) and f satisfies \((f_{1})\)–\((f_{3})\), then the following properties hold:
- \((A_{1})\):
-
For each \(u\in H_{\varepsilon }^{+}\), let \(\varphi _{u}:{\mathbb {R}}^{+}\rightarrow {\mathbb {R}}\) be given by \(\varphi _{u}(t)=J_{\varepsilon }(tu)\). Then, there exists a unique \(t_{u}>0\) such that \(\varphi '_{u}(t)>0\) in \((0, t_{u})\) and \(\varphi '_{u}(t)<0\) in \((t_{u},\infty )\).
- \((A_{2})\):
-
There is a \(\sigma >0\) independent on u such that \(t_{u}>\sigma \) for all \(u\in S_{\varepsilon }^{+}\). Furthermore, for each compact set \({\mathcal {W}}\subset S_{\varepsilon }^{+}\), there is \(C_{{\mathcal {W}}}>0\) such that \(t_{u}\le C_{{\mathcal {W}}}\) for all \(u\in {\mathcal {W}}\).
- \((A_{3})\):
-
The map \({\widehat{m}}_{\varepsilon }:H_{\varepsilon }^{+}\rightarrow {\mathcal {N}}_{\varepsilon }\) given by \({\widehat{m}}_{\varepsilon }(u)=t_{u}u\) is continuous and \(m_{\varepsilon }={\widehat{m}}_{\varepsilon }|_{S_{\varepsilon }^{+}}\) is a homeomorphism between \(S_{\varepsilon }^{+}\) and \({\mathcal {N}}_{\varepsilon }\). Moreover, \(m_{\varepsilon }^{-1}=\frac{u}{\Vert u\Vert _{\varepsilon }}\).
- \((A_{4})\):
-
If there is a sequence \(\{u_{n}\}\subset S_{\varepsilon }^{+}\) such that \({\text {dist}}(u_{n}, \partial S_{\varepsilon }^{+})\rightarrow 0\), then \(\Vert m_{\varepsilon }(u_{n})\Vert _{\varepsilon }\rightarrow \infty \) and \(J_{\varepsilon }(m_{\varepsilon }(u_{n}))\rightarrow \infty \).
Proof
- \((A_{1})\):
-
We notice from the proof of Lemma 3.4 that \(\varphi _{u}(0)=0\), \(\varphi _{u}(t)>0\) for \(t>0\) small and \(\varphi _{u}(t)<0\) for \(t>0\) large. Therefore, \(\max _{t\ge 0}\varphi _{u}(tu)\) is achieved at a global maximum point \(t=t_{u}\) verifying \(\varphi '_{u}(t_{u})=0\) and \(t_{u}u\in {\mathcal {N}}_{\varepsilon }\). From \((f_{3})\), the definition of \(\varphi \) and \(|{\text {supp}}(u)\cap \Lambda _{\varepsilon }|>0\), we may obtain the uniqueness of \(t_{u}\). Therefore, \(\max _{t\ge 0}\varphi _{u}(tu)\) is achieved at a unique \(t=t_{u}\) such that \(\varphi '_{u}(t_{u})=0\) and \(t_{u}u\in {\mathcal {N}}_{\varepsilon }\).
- \((A_{2})\):
-
Suppose \(u\in S_{\varepsilon }^{+}\), then from (3.5) we have
$$\begin{aligned} \begin{aligned} t_{u}^{2}=\frac{1}{2^{*}_{\mu }}\int \limits _{{\mathbb {R}}^{N}} \left( \frac{1 }{|x|^{\mu }}*G(\varepsilon x, |u|^{2})\right) g(\varepsilon x, |t_{u}u|^{2})|t_{u}u|^{2}\text {d}x&\le C\left( t_{u}^{4}+t_{u}^{2q} +t_{u}^{2\cdot 2^{*}_{\mu }}\right) . \end{aligned} \end{aligned}$$which implies that \(t_{u}>\sigma \) for some \(\sigma >0\).
If \({\mathcal {W}}\subset S_{\varepsilon }^{+}\) is compact, and suppose by contradiction that there is \(\{u_{n}\}\subset {\mathcal {W}}\) such that \(t_{n}:=t_{u_{n}}\rightarrow \infty \). Since \({\mathcal {W}}\) is compact, there is \(u\in {\mathcal {W}}\) such that \( u_{n}\rightarrow u\) in \(H_{\varepsilon }\). Then, \(u\in {\mathcal {W}}\subset S_{\varepsilon }^{+}\). Moreover, using the proof of Lemma 3.4 (b), we have \(J_{\varepsilon }(t_{n}u_{n})\rightarrow -\infty \).
On the other hand, using the proof of lemma 3.4 (a), there exists \(\rho >0\) such that \(\inf _{\Vert u\Vert _{\varepsilon }=\rho }J_{\varepsilon }(u)>0\). Then combining this with \((A_{1})\), we have \(c_{\varepsilon }=\inf _{{\mathcal {N}}_{\varepsilon }}J_{\varepsilon }(u)\ge \inf _{\Vert u\Vert _{\varepsilon }=\rho }J_{\varepsilon }(u)>0\), which yields a contradiction. Hence \((A_{2})\) is true.
- \((A_{3})\):
-
Firstly, we observe that \(m_{\varepsilon }\), \({\widehat{m}}_{\varepsilon }\) and \(m_{\varepsilon }^{-1}\) are well defined. In fact, by \((A_{1})\), for each \(u\in H_{\varepsilon }^{+}\), there exists a unique \(t_{u}>0\) such that \(t_{u}u\in {\mathcal {N}}_{\varepsilon }\), hence there is a unique \({\widehat{m}}_{\varepsilon }(u)=t_{u}u\in {\mathcal {N}}_{\varepsilon }\). On the other hand, if \(u\in {\mathcal {N}}_{\varepsilon }\), then \(u\in H_{\varepsilon }^{+}\). Therefore, \(m_{\varepsilon }^{-1}=\frac{u}{\Vert u\Vert _{\varepsilon }}\in S_{\varepsilon }^{+}\), is well defined and it is a continuous function. Since
$$\begin{aligned} m_{\varepsilon }^{-1}(m_{\varepsilon }(u)) =m_{\varepsilon }^{-1}(t_{u}u) =\frac{t_{u}u}{\Vert t_{u}u\Vert _{\varepsilon }}=u, \hspace{2mm}\forall u\in S_{\varepsilon }^{+}, \end{aligned}$$we conclude that \(m_{\varepsilon }\) is a bijection. To prove that \({\widehat{m}}_{\varepsilon }:H_{\varepsilon }^{+}\rightarrow {\mathcal {N}}_{\varepsilon }\) is continuous, let \(\{u_{n}\}\subset H_{\varepsilon }^{+}\) and \(u\in H_{\varepsilon }^{+}\) satisfy \(u_{n}\rightarrow u\) in \(H_{\varepsilon }^{+}\). By \((A_{2})\), there is a \(t_{0}>0\) such that \(t_{n}:=t_{u_{n}}\rightarrow t_{0}\). From \(t_{n}u_{n}\in {\mathcal {N}}_{\varepsilon }\), we obtain
$$\begin{aligned} \begin{aligned} t_{n}^{2}\Vert u_{n}\Vert _{\varepsilon }^{2}=\frac{1}{2^{*}_{\mu }}\int \limits _{{\mathbb {R}}^{N}} \left( \frac{1 }{|x|^{\mu }}*G(\varepsilon x, |t_{n}u_{n}|^{2})\right) g(\varepsilon x, |t_{n}u_{n}|^{2})|t_{u}u_{n}|^{2}\text {d}x. \end{aligned} \end{aligned}$$By Lemma 3.1 and passing to the limit as \(n\rightarrow \infty \), it follows that
$$\begin{aligned} \begin{aligned} t_{0}^{2}\Vert u\Vert _{\varepsilon }^{2}=\frac{1}{2^{*}_{\mu }} \int \limits _{{\mathbb {R}}^{N}} \left( \frac{1 }{|x|^{\mu }}*G(\varepsilon x, |t_{0}u|^{2})\right) g(\varepsilon x, |t_{0}u|^{2})|t_{0}u|^{2}\text {d}x\\ \end{aligned} \end{aligned}$$which means that \(t_{0}u\in {\mathcal {N}}_{\varepsilon }\) and \(t_{u}=t_{0}\). This proves \({\widehat{m}}_{\varepsilon }(u_{n})\rightarrow {\widehat{m}}_{\varepsilon }(u)\) in \(H_{\varepsilon }^{+}\). So, \({\widehat{m}}_{\varepsilon }\), \(m_{\varepsilon }\) are continuous functions and \((A_{3})\) is proved.
- \((A_{4})\):
-
Let \(\{u_{n}\}\subset S_{\varepsilon }^{+}\) be a subsequence such that \({\text {dist}}(u_{n}, \partial S_{\varepsilon }^{+}) \rightarrow 0\), then for each \(v \in \partial S_{\varepsilon }^{+}\)and \(n\in {\mathbb {N}}\), we have \(\left| u_{n}\right| =\left| u_{n}-v\right| \) a.e. in \(\Lambda _{\varepsilon }\). Therefore, by \((V_{1}),(V_{2})\) and the Sobolev embedding, for any \(t \in \left[ 2,2^{*}\right] \), there exists a constant \(C_{t}>0\) such that
$$\begin{aligned} \begin{aligned} \Vert u_{n}\Vert _{L^{t}(\Lambda _{\varepsilon })}&\le \inf _{v\in \partial S_{\varepsilon }^{+}} \Vert u_{n}-v\Vert _{L^{t}(\Lambda _{\varepsilon })} \\&\le C_{t}\left( \inf _{v\in \partial S_{\varepsilon }^{+}} \int \limits _{\Lambda _{\varepsilon }}\left( |\nabla _{\varepsilon } (u_{n}-v)|^{2}+V_{\varepsilon }(x)\left| u_{n}-v\right| ^{2}\right) \text {d} x\right) ^{\frac{1}{2}} \\&\le C_{t} {\text {dist}}\left( u_{n}, \partial S_{\varepsilon }^{+}\right) \end{aligned} \end{aligned}$$for all \(n\in {\mathbb {N}}\). By \((g_{2})\) and \((g_{4})\), for each \(t>0\), we have
$$\begin{aligned} \begin{aligned} \int \limits _{{\mathbb {R}}^{N}}&\left( \frac{1}{|x|^{\mu }}*G(\varepsilon x,|tu_{n}|^{2})\right) G(\varepsilon x,|tu_{n}|^{2})\text {d}x\\&=\int \limits _{\Lambda _{\varepsilon }}\left( \frac{1}{|x|^{\mu }}*G(\varepsilon x,|tu_{n}|^{2})\right) G(\varepsilon x,|tu_{n}|^{2})\text {d}x +\int \limits _{\Lambda _{\varepsilon }^{c}}\left( \frac{1}{|x|^{\mu }}*G(\varepsilon x,|tu_{n}|^{2})\right) G(\varepsilon x,|tu_{n}|^{2})\text {d}x\\&\le C\left( t^{4}|u_{n}|_{L^{2p_{0}}(\Lambda _{\varepsilon })}^{4} +t^{2q}|u_{n}|_{L^{qp_{0}}(\Lambda _{\varepsilon })}^{2q} +t^{2\cdot 2^{*}_{\mu }}|u_{n}|_{L^{2^{*}}(\Lambda _{\varepsilon })}^{2\cdot 2^{*}_{\mu }}\right) +\frac{k_{0}t^{2}}{2K} \int \limits _{\Lambda _{\varepsilon }^{c}} V(\varepsilon x)\left| u_{n}\right| ^{2} \text {d} x \\&\le C\left( t^{4}|u_{n}|_{L^{2p_{0}}(\Lambda _{\varepsilon })}^{4} +t^{qp_{0}}|u_{n}|_{L^{2q}(\Lambda _{\varepsilon })}^{2q} +t^{2\cdot 2^{*}_{\mu }}|u_{n}|_{L^{2^{*}}(\Lambda _{\varepsilon })}^{2\cdot 2^{*}_{\mu }}\right) +\frac{k_{0}t^{2}}{2K}\Vert u_{n}\Vert _{\varepsilon }^{2} \\&\le C\left( t^{4}{\text {dist}}(u_{n}, \partial S_{\varepsilon }^{+})^{4}+t^{2q}{\text {dist}}(u_{n},\partial S_{\varepsilon }^{+})^{2q}+t^{2\cdot 2^{*}_{\mu }} {\text {dist}}(u_{n},\partial S_{\varepsilon }^{+})^{2\cdot 2^{*}_{\mu }}\right) +\frac{k_{0}t^{2}}{2K}. \end{aligned} \end{aligned}$$Therefore,
$$\begin{aligned} \limsup _{n} \int \limits _{{\mathbb {R}}^{N}}\left( \frac{1}{|x|^{\mu }}*G(\varepsilon x,|tu_{n}|^{2})\right) G(\varepsilon x,|tu_{n}|^{2})\text {d}x \le \frac{k_{0}t^{2}}{2K}, \quad \forall t>0. \end{aligned}$$On the other hand, from the definition of \(m_{\varepsilon }\) and the last inequality, for all \(t>0\), we obtain
$$\begin{aligned} \begin{aligned} \liminf _{n}J_{\varepsilon }(m_{\varepsilon }\left( u_{n}\right) )&\ge \liminf _{n}J_{\varepsilon }(t u_{n}) \\&\ge \liminf _{n} \frac{t^{2}}{2}\left\| u_{n}\right\| _{\varepsilon }^{2} -\frac{k_{0}t^{2}}{2K}\\&=\frac{K-k_{0}}{2K}t^{2}. \end{aligned} \end{aligned}$$This implies that
$$\begin{aligned} \liminf _{n} \frac{1}{2}\Vert m_{\varepsilon }(u_{n})\Vert _{\varepsilon }^{2} \ge \liminf _{n}J_{\varepsilon }\left( m_{\varepsilon }(u_{n})\right) \ge \frac{K-k_{0}}{2 K} t^{2}, \quad \forall t>0. \end{aligned}$$From the arbitrariness of \(t>0\), we see that
$$\begin{aligned} \Vert m_{\varepsilon }(u_{n})\Vert _{\varepsilon }\rightarrow \infty \quad \text{ and } \quad J_{\varepsilon }\left( m_{\varepsilon }(u_{n})\right) \rightarrow \infty \quad \text{ as } n\rightarrow \infty . \end{aligned}$$This completes the proof of Lemma 3.5.
\(\square \)
Now we define the function
by \({\widehat{\Psi }}_{\varepsilon }(u)=J_{\varepsilon }({\widehat{m}}_{\varepsilon }(u))\) and denote by \(\Psi _{\varepsilon }(u)=({\widehat{\Psi }}_{\varepsilon }(u))|_{S_{\varepsilon }^{+}}\).
From Lemma 3.2, by a similar argument to that of [39, Corollary 10], we have the following results.
Lemma 3.6
Assume that \((V_{1})\)–\((V_{2})\) and \((f_{1})\)–\((f_{3})\) are satisfied, then
- \((B_{1})\):
-
\({\widehat{\Psi }}_{\varepsilon }\in C^{1}(H_{\varepsilon }^{+},{\mathbb {R}})\) and
$$\begin{aligned} \langle {\widehat{\Psi }}'_{\varepsilon }(u),v\rangle =\frac{\Vert {\widehat{m}}_{\varepsilon }(u)\Vert _{\varepsilon }}{\Vert u\Vert _{\varepsilon }}\langle J'_{\varepsilon }({\widehat{m}}_{\varepsilon }(u)),v \rangle ,\hspace{5mm}\forall u\in H_{\varepsilon }^{+} \text { and }\forall v\in H_{\varepsilon }. \end{aligned}$$ - \((B_{2})\):
-
\(\Psi _{\varepsilon }\in C^{1}(S_{\varepsilon }^{+},{\mathbb {R}})\) and
$$\begin{aligned} \langle \Psi '_{\varepsilon }(u),v\rangle =\Vert m_{\varepsilon }(u)\Vert _{\varepsilon }\langle J'_{\varepsilon }({\widehat{m}}_{\varepsilon }(u)),v\rangle , \hspace{5mm} \forall v\in T_{u}S_{\varepsilon }^{+}. \end{aligned}$$ - \((B_{3})\):
-
If \(\{u_{n}\}\) is a \((PS)_{c}\) sequence of \(\Psi _{\varepsilon }\), then \(\{m_{\varepsilon }(u_{n})\}\) is a \((PS)_{c}\) sequence of \(J_{\varepsilon }\). If \(\{u_{n}\}\subset {\mathcal {N}}_{\varepsilon }\) is a bounded \((PS)_{c}\) sequence of \(J_{\varepsilon }\), then \(\{m_{\varepsilon }^{-1}(u_{n})\}\) is a \((PS)_{c}\) sequence of \(\Psi _{\varepsilon }\).
- \((B_{4})\):
-
u is a critical point of \(\Psi _{\varepsilon }\) if and only if \(m_{\varepsilon }(u)\) is a critical point of \(J_{\varepsilon }\). Moreover, the corresponding critical values coincide and
$$\begin{aligned} \inf _{u\in S_{\varepsilon }^{+}}\Psi _{\varepsilon }(u)= \inf _{u\in {\mathcal {N}}_{\varepsilon }}J_{\varepsilon }(u). \end{aligned}$$
As in [39], we have the following variational characterization of the infimum of \(J_{\varepsilon }\) over \({\mathcal {N}}_{\varepsilon }\):
Lemma 3.7
Suppose that \((f_{1})\)–\((f_{3})\) hold. Assume that \(\{u_{n}\}\subset {\mathcal {N}}_{\varepsilon }\) is a \((PS)_{c}\) sequence with
then \(\{u_{n}\}\) is bounded in \(H_{\varepsilon }\). Moreover, \(\{u_{n}\}\) cannot be vanishing, namely there exist \(r, \delta >0\) and a sequence \(\{y_{n}\}\subset {\mathbb {R}}^{N}\) such that
Proof
Firstly, we prove the boundedness of \(\{u_{n}\}\). Arguing by contradiction, we suppose that \(\{u_{n}\}\) is unbounded in \(H_{\varepsilon }\). Without loss of generality, we assume that \(\Vert u_{n}\Vert _{\varepsilon }\rightarrow \infty \). Let \(v_{n}:=\frac{u_{n}}{\Vert u_{n}\Vert _{\varepsilon }}\), up to a subsequence, then there exists \(v\in H_{\varepsilon }\) such that
If \(v_{n}\) is vanishing, i.e.,
then Lemma 2.4 implies that \(v_{n}\rightarrow 0\) in \(L^{t}({\mathbb {R}}^{N},{\mathbb {R}})\) for all \(t\in (2,2^{*})\). By (3.2) and (3.3), we have
Hence for sufficiently large n, we have
which is a contradiction. Therefore, \(\left\{ v_{n}\right\} \) is nonvanishing, namely there exists \(y_{n} \in {\mathbb {R}}^{N}\) and \(\delta >0\) such that
Denote \({\tilde{v}}_{n}(\cdot )=v_{n}\left( \cdot +y_{n}\right) \), then we can suppose that
By (3.12), we have \({\tilde{v}}\ne 0\). Hence, there exists a measure set E such that \({\tilde{v}}(x) \ne 0\) for \(x\in E\). Let \(|{\hat{u}}_{n}|:=|{\tilde{v}}_{n}|\Vert u_{n}\Vert _{\varepsilon }\). Then, \(|{\hat{u}}_{n}(x)|\rightarrow +\infty \) for \(x\in E\). By \((f_{3})\), we have
Hence, we know
namely
Then, we have
which is a contradiction. Therefore, \(\{u_{n}\}\) is bounded in \(H_{\varepsilon }\).
Next we prove the second conclusion. By contradiction, if \(\left\{ u_n\right\} \) is vanishing, then similar to (3.12), we have
and
If \(\left\| u_n\right\| _{\varepsilon } \rightarrow 0\), then it follows from (3.13) and (3.14) that \(c_{\varepsilon }=0\), which is impossible. Then, \(\left\| u_n\right\| _{\varepsilon } \nrightarrow 0\) and by virtue of (3.14) we get
Hence,
From (3.13) and (3.14), we deduce that
which yields a contradiction. Therefore, \(\left\{ u_n\right\} \) is nonvanishing. \(\square \)
Lemma 3.8
Assume \((V_{1})\)–\((V_{2})\) and \((f_{1})\)–\((f_{3})\) hold, the functional \(J_{\varepsilon }\) satisfies the \((PS)_{c}\) condiction for all \(c \in \) \(\left[ c_{\varepsilon }, \frac{N+2-\mu }{2(2N-\mu )} S_{H,L}^{\frac{2N-\mu }{N+2-\mu }}\right) \).
Proof
Let \(\left\{ u_{n}\right\} \) be a \((PS)_{c}\) sequence for \(J_{\varepsilon }\). By Lemma 3.7, \(\left\{ u_{n}\right\} \) is bounded in \(H_{\varepsilon }\). Therefore, up to a subsequence, \(u_{n}\rightharpoonup u\) in \(H_{\varepsilon }\) and \(u_{n} \rightarrow u\) in \(L_{loc}^{r}({\mathbb {R}}^{N})\) for any \(r \in [2,2^{*})\).
Let \(R>0\) be such that \(\Lambda _{\varepsilon }\subset B_{R}(0)\). We show that for each \(\xi >0\) and R large enough, it holds that
Let us consider a cut-off function \(\eta _{R} \in C^{\infty }({\mathbb {R}}^{N},{\mathbb {R}})\) defined as
and \(|\nabla \eta _{R}|_{\infty }\le C/R\), where \(C>0\) is a constant independent of R. Since \(\{u_{n}\eta _{R}\}\) is bounded in \(H_{\varepsilon }\), it follows that
therefore,
Since \(\overline{\nabla _{\varepsilon }(u_{n}\eta _{R})} =i{\overline{u}}_{n}\nabla \eta _{R}+\eta _{R}\overline{\nabla _{\varepsilon }u_{n}}\), using \((g_{4})\), we obtain
By the Hölder inequality and the boundedness of \(\left\{ u_{n}\right\} \) in \(H_{\varepsilon }\), we deduce that
and so (3.15) holds.
Next we prove that \(u_{n} \rightarrow u\) in \(H_{\varepsilon }\) as \(n \rightarrow \infty \). Setting \(\omega _{n}=\left\| u_{n}-u\right\| _{\varepsilon }^{2}\), we have
Observe that \(\left\langle J_{\varepsilon }^{\prime }(u_{n}), u_{n}\right\rangle =\left\langle J_{\varepsilon }^{\prime }\left( u_{n}\right) , u\right\rangle =o_{n}(1)\), so we only need to prove that
By the [21, lemma 2.4], we have
Now, by \((g_{4})\) and (3.15) we obtain
Therefore, we have
\(\square \)
Since f is only assumed to be continuous, the following Lemma is required for the multiplicity result in the next section.
Lemma 3.9
The functional \(\Psi _{\varepsilon }\) satisfies the \((PS)_{c}\) condiction in \(S_{\varepsilon }^{+}\) for all \(c \in \) \(\left[ c_{\varepsilon }, \frac{N+2-\mu }{2(2N-\mu )} S_{H,L}^{\frac{2N-\mu }{N+2-\mu }}\right) \).
Proof
Let \(\{u_{n}\} \subset S_{\varepsilon }^{+}\) be a \((PS)_{c}\) sequence for \(\Psi _{\varepsilon }\). Thus, \(\Psi _{\varepsilon }\left( u_{n}\right) \rightarrow c\) and \(\left\| \Psi _{\varepsilon }^{\prime }\right\| _{*} \rightarrow 0\), where \(\Vert \cdot \Vert _{*}\) is the norm in the dual space \((T_{u_{n}} S_{\varepsilon }^{+})^{*}\). It follows from Lemma 3.6\((B_{3})\) that \(\{m_{\varepsilon }(u_{n})\}\) is a \((PS)_{c}\) sequence for \(J_{\varepsilon }\) in \(H_{\varepsilon }\). From Lemma 3.8, we see that there exists a \(u \in S_{\varepsilon }^{+}\) such that \(m_{\varepsilon }(u_{n}) \rightarrow m_{\varepsilon }(u)\) in \(H_{\varepsilon }\). From Lemma 3.5\((A_{3})\), we conclude that
and the proof is complete. \(\square \)
4 The autonomous problem
For our scope, we start by considering the limit problem associated to (2.1), namely the problem
which has the following associated functional
The functional \(I_{0}\) is well defined on the Hilbert space \(H_{0}=H^{1}({\mathbb {R}}^{N},{\mathbb {R}})\) with the inner product
and the norm
We denote the Nehari manifold associated to \(I_{0}\) by
and the least energy on \( {\mathcal {N}}_{0}\) is defined by
Let \(H_{0}^{+}\) be the open subset of \(H_{0}\) given by
and \(S_{0}^{+}=S_{0}\cap H_{0}^{+}\), where \(S_{0}\) is the unit sphere of \(H_{0}\). Note that \(S_{0}^{+}\) is a non-complete \(C^{1,1}\)-manifold of codimension 1, modeled on \(H_{0}\) and contained in \(H_{0}^{+}\). Therefore, \(H_{0}=T_{u}S_{0}^{+}\oplus {\mathbb {R}}u\) for each \(u\in S_{0}^{+}\), where \(T_{u}S_{0}^{+}=\{v\in H_{0}:\langle u,v\rangle _{V_{0}}=0\}\).
Next we have the following Lemmas and the proofs that follow from a similar argument used in the proofs of Lemmas 3.5 and 3.6.
Lemma 4.1
Let \(V_{0}\) be given in \((V_{1})\) and suppose that f satisfies \((f_{1})\)–\((f_{3})\). Then, the following properties hold:
- \((a_{1})\):
-
For each \(u\in H_{0}\setminus \{0\}\), let \(\phi _{u}:{\mathbb {R}}^{+}\rightarrow {\mathbb {R}}\) be given by \(\phi _{u}(t)=I_{0}(tu)\). Then, there exists a unique \(t_{u}>0\) such that \(\phi '_{u}(t)>0\) in \((0, t_{u})\) and \(\phi '_{u}(t)<0\) in \((t_{u}, \infty )\).
- \((a_{2})\):
-
There is a \(\sigma >0\) independent on u such that \(t_{u}>\sigma \) for all \(u\in S_{0}^{+}\). Moreover, for each compact set \({\mathcal {W}}\subset S_{0}^{+}\), there is \(C_{{\mathcal {W}}}>0\) such that \(t_{u}\le C_{{\mathcal {W}}}\) for all \(u\in {\mathcal {W}}\).
- \((a_{3})\):
-
The map \({\widehat{m}}:H_{0}\setminus \{0\}\rightarrow {\mathcal {N}}_{0}\) given by \({\widehat{m}}(u)=t_{u}u\) is continuous and \(m={\widehat{m}}|_{S_{0}^{+}}\) is a homeomorphism between \(S_{0}^{+}\) and \({\mathcal {N}}_{0}\). Moreover, \(m^{-1}(u)=\frac{u}{\Vert u\Vert _{0}}\).
- \((a_{4})\):
-
If there is a sequence \(\{u_{n}\}\subset S_{\varepsilon }^{+}\) such that \({\text {dist}}(u_{n}, \partial S_{\varepsilon }^{+})\rightarrow 0\), then we have \(\Vert m_{\varepsilon }(u_{n})\Vert _{\varepsilon }\rightarrow \infty \) and \(I_{0}(m_{\varepsilon }(u_{n}))\rightarrow \infty \).
We shall consider the functional defined by
Lemma 4.2
Let \(V_{0}\) be given in \((V_{1})\) and suppose that f satisfies \((f_{1})\)-\((f_{3})\), then
- \((b_{1})\):
-
\({\widehat{\Psi }}_{0}\in C^{1}(H_{0}^{+},{\mathbb {R}})\) and
$$\begin{aligned} \langle {\widehat{\Psi }}'_{0}(u),v\rangle =\frac{\Vert {\widehat{m}}(u)\Vert _{0}}{\Vert u\Vert _{0}}\langle I'_{0}({\widehat{m}}(u)),v\rangle ,\hspace{2mm} \forall u\in H_{0}^{+} \text { and }\hspace{2mm} \forall v\in H_{0}. \end{aligned}$$ - \((b_{2})\):
-
\(\Psi _{0}\in C^{1}(S_{0}^{+},{\mathbb {R}})\) and
$$\begin{aligned} \langle \Psi '_{0}(u),v\rangle =\Vert m(u)\Vert _{0}\langle I'_{0}(m(u)),v\rangle ,\forall v\in TS_{0}^{+}. \end{aligned}$$ - \((b_{3})\):
-
If \(\{u_{n}\}\) is a \((PS)_{c}\) sequence of \(\Psi _{0}\), then \(\{m(u_{n})\}\) is a \((PS)_{c}\) sequence of \(I_{0}\). If \(\{u_{n}\}\subset {\mathcal {N}}_{0}\) is a bounded \((PS)_{c}\) sequence of \(I_{0}\), then \(\{m^{-1}(u_{n})\}\) is a \((PS)_{c}\) sequence of \(\Psi _{0}\).
- \((b_{4})\):
-
u is a critical point of \(\Psi _{0}\) if and only if m(u) is a critical point of \(I_{0}\). Moreover, the corresponding critical values coincide and
$$\begin{aligned} \inf _{u\in S_{0}^{+}}\Psi _{\varepsilon }(u)= \inf _{u\in {\mathcal {N}}_{0}}I_{0}(u). \end{aligned}$$
Similar to the previous argument, we have the following variational characterization of the infimum of \(I_{0}\) over \({\mathcal {N}}_{0}\):
The next result is useful in later arguments.
Lemma 4.3
Let \(\{u_{n}\}\in H_{0}\) be a \((PS)_{c}\) sequence with \(c\in \left[ c_{V_{0}},\frac{N+2-\mu }{2(2N-\mu )} S_{H,L}^{\frac{2N-\mu }{N+2-\mu }}\right) \) for \(I_{0}\) such that \(u_{n}\rightharpoonup 0\). Then, only one of the following conclusions holds.
- (i):
-
\(u_{n}\rightarrow 0\) in \( H_{0}\) as \(n\rightarrow \infty \);
- (ii):
-
there exists \(R, \beta >0\) and \(\{y_{n}\}\subset {\mathbb {R}}^{N}\) such that
$$\begin{aligned} \int \limits _{B_{R}(y_{n})}|u_{n}|^{2}\ge \beta . \end{aligned}$$
Proof
Suppose that (ii) does not hold. Then, for any \(R>0\), we have
Similarly to Lemma 3.7, \(\{u_{n}\}\) is bounded in \(H_{0}\). Then, by Lemma 2.4, we know
Thus, by (f2) we have
From \(\langle I_{0}^{\prime }(u_{n}),u_{n}\rangle \rightarrow 0\), we get
Therefore the conclusion follows. \(\square \)
From Lemma 4.3, we can see that, if u is the weak limit of a \((PS)_{c_{V_{0}}}\) sequence \(\{u_{n}\}\) for the functional \(I_{0}\), then we can assume \(u_{n}\ne 0\). Otherwise we would have \(u_{n}\rightharpoonup 0\) and once it does not occur that \(u_{n} \rightarrow 0\), we conclude from Lemma 4.3 that there exist \(\{y_{n}\}\subset {\mathbb {R}}^{N}\) and \(R, \beta >0\) such that
Then set \(v_{n}(x)=u_{n}(x+y_{n})\), making a change of variable, we can prove that \(\{v_{n}\}\) is also a \((P S)_{c_{V_{0}}}\) sequence for the functional \(I_{0}\), it is bounded in \(H_{0}\) and there exists \(v \in H_{0}\) such that \(v_{n} \rightarrow v\) in \(H_{0}\) with \(v \ne 0\).
Next we devote to estimating the least energy \(c_{V_{0}}\).
Lemma 4.4
There exists \(u_{\varepsilon }\) such that
provided that one of the following conditions holds:
- (i):
-
\(\frac{N+2-\mu }{N-2}<p<2_{\mu }^{*}, N=3,4\) and \(\lambda >0\);
- (ii):
-
\(\frac{2N-\mu }{N}<p\le \frac{N+2-\mu }{N-2}, N=3,4\) and \(\lambda \) sufficiently large;
- (iii):
-
\(\frac{2N+2-\mu }{N-2}<p<2_{\mu }^{*}, N\ge 5\) and \(\lambda >0\);
- (iv):
-
\(\frac{2 N-\mu }{N}<p\le \frac{2N-2-\mu }{N-2}, N\ge 5\) and \(\lambda \) sufficiently large.
Proof
We consider \(\psi \in C_{0}^{\infty }\left( {\mathbb {R}}^{N}\right) \) such that
where \(C>0\) is a constant. For \(\varepsilon >0\), we define
From [21], we have the estimates
and
By direct computation, we have
Case 1. \(\frac{N+2-\mu }{N-2}<p<2_{\mu }^{*}\) and \(N=3,4\) or \(\frac{2N-2-\mu }{N-2}<p<2_{\mu }^{*}\) and \(N\ge 5\).
From Lemma 4.1, there exists unique \(t_{\varepsilon }>0\) such that \(t_{\varepsilon }u_{\varepsilon }\in {\mathcal {N}}_{0}\). Now we estimate \(I_{0}(t_{\varepsilon }u_{\varepsilon })\). Notice that
For \(I_{1}\), we may assume that
and consider the function \(\theta :[0,\infty )\rightarrow {\mathbb {R}}\) defined by
We have that \(t_{0}=\left( \frac{E}{F}\right) ^{\frac{1}{2\cdot 2_{\mu }^{*}-2}}\) is a maximum point of \(\theta \) and
Hence, we have
Observing that for \(\varepsilon >0\) sufficiently small, it holds
Therefore, we conclude that for any \(\varepsilon >0\) sufficiently small, we have
We claim that there exists \(C_{0}>0\) such that for all \(\varepsilon >0\),
In fact, assume that there exists a sequence \(\{\varepsilon _{n}\}\) and \(\varepsilon _{n} \rightarrow 0\) as \(n \rightarrow \infty \), such that \(t_{\varepsilon _{n}} \rightarrow 0\) as \(n \rightarrow \infty \). Thus,
Since \(u_{\varepsilon _{n}}\in H_{0}\) is bounded and \(t_{\varepsilon _{n}} \rightarrow 0\) as \(n \rightarrow \infty \), we have \(t_{\varepsilon _{n}} u_{\varepsilon _{n}} \rightarrow 0\) in \(H_{0}\).
The continuity of \(I_{0}\) implies that \(I_{0}(t_{\varepsilon _{n}} u_{\varepsilon _{n}})\rightarrow I_{0}(0)=0\). Therefore,
which is a contradiction. The claim is holds.
From (4.5) and the above claim, we have
We want to obtain that
In order to do that, it suffices to show that
and
Suppose that (4.8) and (4.9) hold, let us continue to complete our proof. Since
from (4.7), we have
for \(\varepsilon >0\) sufficiently small. Thus, (4.6) and (4.10) guarantee
for \(\varepsilon >0\) sufficiently small and fixed. Once (4.8) and (4.9) are verified, the proof of Case 1 is complete.
Now we prove (4.8). By direct computation, we know
where \(\omega _{N}\) denotes the volume of the unit ball in \({\mathbb {R}}^{N}\).
Let
The Case N=3. In this case, we have \(5-\mu<p<2_{\mu }^{*}\), thus \(5-\mu -p<0\). And we know \(\min \left\{ N-2, \frac{2N-\mu }{2}\right\} =N-2=1\) by \(0<\mu <N\). It is easy to show that
Therefore,
Our claim follows.
The Case N=4. In this case, \(\frac{6-\mu }{2}<p<2_{\mu }^{*}\) implies \(6-\mu -2p<0\) and \(\min \left\{ N-2,\frac{2N-\mu }{2}\right\} =N-2=2\) if \(0<\mu <4\). We also have
So,
Our claim follows.
The Case \(N\ge 5\) . We have
It is easy to show that if \(N \ge 5\), then the integral
converges.
There are two cases left to be considered:
-
\(0<\mu <4\) and \(N \ge 5\);
-
\(\mu \ge 4\) and \(N \ge 5\).
Assume that \(0<\mu <4\) and \(N\ge 5\), we have
Also \(\frac{2N-\mu -2}{N-2}<p<\frac{2N-\mu }{N-2}\) implies \(2N-\mu -(N-2)p-2<0.\) Therefore, \(I_{\varepsilon } \rightarrow -\infty \) as \(\varepsilon \rightarrow 0\). Now we consider the case \(\mu \ge 4\) and \(N\ge 5\). We have \(N-2\ge \frac{2N-\mu }{2}\) and therefore
Since
we conclude that \(I_{\varepsilon } \rightarrow -\infty \) as \(\varepsilon \rightarrow 0\). We are done.
Now we prove (4.9).
For \(\delta >0\) sufficiently large, we have \(U_{\varepsilon }^{2}(x) \le \varepsilon ^{1+\eta }\) if \(|x| \ge \delta \). Since
our proof is complete.
Case 2. For \(\lambda \) sufficiently large, \(\frac{2 N-\mu }{N}<p\le \frac{N+2-\mu }{N-2}\) and \(N=3,4\) or \(\frac{2 N-\mu }{N}<p\le \frac{2N-2-\mu }{N-2}\) and \(N\ge 5\).
From lemma 4.1, we know that \(\max _{t\ge 0} I_{0}(tu_{\varepsilon })\) is attained at some \(t_{\lambda }>0\). Since \(I'_{0}(t_{\lambda }u_{\varepsilon })=0\), we have
Thus, \(t_{\lambda }\rightarrow 0\) as \(\lambda \rightarrow +\infty \) and
Since \(t_{\lambda } \rightarrow 0\) as \(\lambda \rightarrow +\infty \) and \(\frac{N+2-\mu }{2(N-\mu )}\left( S_{H,L}\right) ^{\frac{2 N-\mu }{N+2-\mu }}>0\), we conclude that
for \(\lambda >0\) sufficiently large.
Therefore,
for \(\lambda >0\) sufficiently large. \(\square \)
Theorem 4.1
Assume that \((f_{1})\)–\((f_{3})\) hold. Then, autonomous problem (4.1) has a positive ground state solution u with \(I_{0}(u)=c_{V_{0}}\).
Proof
By Lemma 3.4 with \(V(x)=V_{0}\) and the mountain pass theorem without (PS) condition, there exists a \((PS)_{c_{V_{0}}}\) sequence \(\{u_{n}\}\subset H_{0}\) of \(I_{0}\) with
If \(u_{0}\in {\mathcal {N}}_{0}\) satisfies \(I_{0}\left( u_{0}\right) =c_{V_{0}}\), then \(m^{-1}(u_{0})\in S_{0}\) is a minimizer of \(\Psi _{0}\), so that \(u_{0}\) is a critical point of \(I_{0}\) by Lemma 4.2. Now, we show that there exists a minimizer \(u \in {\mathcal {N}}_{0}\) of \(\left. I_{0}\right| _{{\mathcal {N}}_{0}}.\) Since \(\inf _{S_{0}} \Psi _{0}=\inf _{{\mathcal {N}}_{0}} I_{0}=c_{V_{0}}\) and \(S_{0}\) is a \(C^{1}\) manifold, by Ekeland’s variational principle, there exists a sequence \(\omega _{n} \subset S_{0}\) with \(\Psi _{0}\left( \omega _{n}\right) \rightarrow c_{V_{0}}\) and \(\Psi _{0}^{\prime }\left( \omega _{n}\right) \rightarrow 0\) as \(n \rightarrow \infty \). Put \(u_{n}=m\left( \omega _{n}\right) \in {\mathcal {N}}_{0}\) for \(n \in {\mathbb {N}}\). Then by Lemma 4.2\((b_{3})\), we have \(I_{0}\left( u_{n}\right) \rightarrow c_{V_{0}}\) and \(I_{0}^{\prime }\left( u_{n}\right) \rightarrow 0\) as \(n \rightarrow \infty \). Similar to the proof of Lemma 3.7, it is easy to check that \(\{u_{n}\}\) is bounded in \(H_{0}\). Thus, we have \(u_{n} \rightharpoonup u\) in \(H_{0}\), \(u_{n} \rightarrow u\) in \(L_{loc}^{r}({\mathbb {R}}^{N}), 1 \le r<2^{*}\) and \(u_{n} \rightarrow u\) a.e. in \({\mathbb {R}}^{N}\). Similar to the proof of Lemma 3.1, we have \(I_{0}^{\prime }(u)=0\). From Lemma 4.3, we know that \(u \ne 0\). Moreover,
where we used Fatou’s lemma. Therefore, \(I_{0}(u)=c_{V_{0}}\), which means that u is a ground state solution for (4.1). From the assumption of f, by [11, Propositions 6 and 7], we know that \(u(x)>0\) for \(x\in {\mathbb {R}}^{N}\). The proof is complete. \(\square \)
The next result is a compactness result on autonomous problem which we will use later.
Lemma 4.5
Let \((u_{n})\subset {\mathcal {N}}_{0}\) be a sequence such that \(I_{0}\left( u_{n}\right) \rightarrow c_{V_{0}}\). Then, \(\left\{ u_{n}\right\} \) has a convergent subsequence in \(H_{0}\).
Proof
Since \(\left\{ u_{n}\right\} \subset {\mathcal {N}}_{0}\), it follows from Lemma 4.1\((a_{3})\), Lemma 4.2\((b_{4})\) and the definition of \(c_{V_{0}}\) that
and
Although \(S_{0}^{+}\) is not a complete \(C^{1}\) manifold, we still can apply Ekeland’s variational principle to the functional \({\mathcal {E}}_{0}: H \rightarrow {\mathbb {R}}\cup \{\infty \}\) defined by
and
where \(H=\overline{S_{0}^{+}}\) is the complete metric space equipped with the metric
In fact, by Lemma 4.1\((a_{4})\), \({\mathcal {E}}_{0} \in C(H,{\mathbb {R}}\cup \{\infty \})\), and from Lemma 4.2\((b_{4})\), \({\mathcal {E}}_{0}\) is bounded from below. Therefore, there exists a sequence \(\left\{ {\tilde{v}}_{n}\right\} \subset S_{0}^{+}\) such that \(\left\{ {\tilde{v}}_{n}\right\} \) is a \((P S)_{c_{V_{0}}}\) sequence for \(\Psi _{0}\) on \(S_{0}^{+}\) and
Arguing as in Lemma 3.8, we obtain the conclusion of this lemma. \(\square \)
5 Solutions for the penalized problem
This subsection is devoted to proving a multiplicity result for the modified problem (3.1) by applying the Ljusternik–Schnirelmann category theory.
Let \(\delta >0\) be such that \(M_{\delta }\subset \Lambda \), \(\omega \in H^{1}({\mathbb {R}}^{N},{\mathbb {R}})\) be a positive ground state solution of the limit problem (4.1), and \(\eta \in C^{\infty }({\mathbb {R}}^{+},[0,1])\) such that \(\eta =1\) if \(0\le t\le \frac{\delta }{2}\) and \(\eta =0\) if \(t\ge \delta \). We define
for each \(y\in M\), where \(\tau _{y}:=\sum \limits _{1}^{N}A_{i}(y)x_{i}\). Let \(t_{\varepsilon }>0\) be the unique positive number such that
By noticing that \(t_{\varepsilon }\Psi _{\varepsilon ,y}\in {\mathcal {N}}_{\varepsilon }\), we consider the function \(\Phi _{\varepsilon }:=M \rightarrow {\mathcal {N}}_{\varepsilon }\) defined by setting
By construction, \(\Phi _{\varepsilon }(y)\) has compact support for any \(y\in M\).
Lemma 5.1
The limit
holds uniformly in \(y\in M\).
Proof
Arguing by contradiction, we deduce that there exist \(\delta _{0}>0,\{y_{n}\}\subset M\) and \(\varepsilon _{n} \rightarrow 0^{+}\) satisfying
For simplicity, we write \(\Phi _{n}, \Psi _{n}\) and \(t_{n}\) for \(\Phi _{\varepsilon _{n}}, \Psi _{\varepsilon _{n}, y_{n}}\) and \(t_{\varepsilon _{n}}\), respectively.
Arguing as in [13, Lemma 3.2], we see that
On the other hand, since \(\left\langle J_{\varepsilon _{n}}^{\prime }\left( t_{n} \Psi _{n}\right) , t_{n} \Psi _{n}\right\rangle =0\), by the change of variables \(z=(\varepsilon _{n} x-y_{n})/\varepsilon _{n}\), we obtain
If \(z\in B_{\delta /\varepsilon _{n}}(0)\), then \(\varepsilon _{n}z+y_{n}\in B_{\delta }(y_{n})\subset M_{\delta }\subset \Lambda \). Hence, \(g(x,s)=f(s)\) for any \(x\in \Lambda \) and \(\Psi (s)=0\) for \(s\ge \delta \), the last equality leads to
where \(\alpha =\min \{\omega (z):|z|\le \delta /2\}\).
If \(t_{n}\rightarrow \infty \), by \((f_{3})\) we deduce that \(\left\| \Psi _{n}\right\| _{\varepsilon _{n}}^{2} \rightarrow \infty \) which contradicts (5.2). Therefore, up to a subsequence, we may assume that \(t_{n} \rightarrow t_{0}\ge 0\).
Since g has critical growth and \(t_{n} \Psi _{n}\in {\mathcal {N}}_{\varepsilon }\), it follows that \(t_{0}>0\). Thereby, taking the limit in (5.3), we obtain
which implies that \(t_{0} \omega \in {\mathcal {N}}_{V_{0}}\). Since \(\omega \in {\mathcal {N}}_{V_{0}}\), we obtain \(t_{0}=1\). Using the Lebesgue dominated convergence theorem, we get
Hence,
which is a contradiction with (5.1) and the proof is complete. \(\square \)
Now we define the barycenter map.
Let \(\rho >0\) be such that \(M_{\delta } \subset B_{\rho }\) and \(\Upsilon : {\mathbb {R}}^{N} \rightarrow \) \({\mathbb {R}}^{N}\) be defined by
The barycenter map \(\beta _{\varepsilon }: {\mathcal {N}}_{\varepsilon }\rightarrow {\mathbb {R}}^{N}\) is defined by
Arguing as Lemma 4.3 in [8], it is easy to see that the function \(\beta _{\varepsilon }\) verifies the following limit:
Lemma 5.2
holds uniformly in \(y\in M\).
We are next to establish the following useful compactness result.
Lemma 5.3
Let \(\varepsilon _{n} \rightarrow 0^{+}\)and \(\left( u_{n}\right) \subset {\mathcal {N}}_{\varepsilon _{n}}\) be such that \(J_{\varepsilon _{n}}\left( u_{n}\right) \rightarrow c_{V_{0}}\). Then, there exists a sequence \(\{{\tilde{y}}_{n}\}\subset {\mathbb {R}}^{N}\) such that \(v_{n}:=\left| u_{n}\right| \left( \cdot +{\tilde{y}}_{n}\right) \) has a convergent subsequence in \(H^{1}({\mathbb {R}}^{N}, {\mathbb {R}})\). Moreover, up to a subsequence, \(y_{n}:=\varepsilon _{n}{\tilde{y}}_{n} \rightarrow y\in M\) as \(n\rightarrow \infty \).
Proof
Since \(\langle J_{\varepsilon _{n}}^{\prime }(u_{n}),u_{n}\rangle =0\) and \(J_{\varepsilon _{n}}\left( u_{n}\right) \rightarrow c_{V_{0}}\), arguing as in the proof of Lemma 3.4, we can prove that there exists \(C>0\) such that \(\left\| u_{n}\right\| _{\varepsilon _{n}} \le C\) for all \(n \in {\mathbb {N}}\). We claim that there exist a sequence \(\left( {\tilde{y}}_{n}\right) \subset {\mathbb {R}}^{N}\) and constants \(R,\beta >0\) such that
Thereby, for some subsequence,
Let \(t_{n}>0\) be such that \({\tilde{v}}_{n}:=t_{n}v_{n}\in {\mathcal {N}}_{V_{0}}\). By the diamagnetic inequality (2.1), we have
Hence, \(J_{0}\left( {\tilde{v}}_{n}\right) \rightarrow c_{V_{0}}\) as \(n \rightarrow +\infty \).
Since the sequences \(\left\{ v_{n}\right\} \) and \(\left\{ {\tilde{v}}_{n}\right\} \) are bounded in \(H^{1}\left( {\mathbb {R}}^{N}, {\mathbb {R}}\right) \) and \(v_{n}\nrightarrow 0\) in \(H^{1}\left( {\mathbb {R}}^{N}, {\mathbb {R}}\right) \), we deduce \(\left( t_{n}\right) \) is also bounded and up to a subsequence, we may assume that \(t_{n} \rightarrow t_{0}\ge 0\).
If \(t_{0}=0\), in view of the boundedness of \(v_{n}\) in \(H_{0}\), we have \({\tilde{v}}_{n}:=t_{n}v_{n}\rightarrow 0\) in \(H_{0}\). Hence, \(I_{0}({\tilde{v}}_{n})\rightarrow 0\), which contradicts \(c_{V_{0}}>0\). Thus, up to a subsequence, we may assume that \({\tilde{v}}_{n} \rightharpoonup {\tilde{v}}:=t_{0} v \ne 0\) in \(H_{0}\), and, by Lemma 4.5, we can deduce that \({\tilde{v}}_{n} \rightarrow {\tilde{v}}\) in \(H_{0}\), which gives \(v_{n}\rightarrow v\) in \(H_{0}\).
Now we show the final part, namely that \(\left\{ y_{n}\right\} \) has a subsequence such hat \(y_{n} \rightarrow y \in M\). Assume by contradiction that \(\left\{ y_{n}\right\} \) is not bounded and so, up to a subsequence, \(\left| y_{n}\right| \rightarrow +\infty \) as \(n \rightarrow +\infty \). Choose \(R>0\) such that \(\Lambda \subset B_{R}(0)\). Then for n large enough, we have \(\left| y_{n}\right| >2 R\), and for any \(x \in B_{R / \varepsilon _{n}}(0)\),
Since \(u_{n} \in {\mathcal {N}}_{\varepsilon _{n}}\), using \((V_{1})\) and the diamagnetic inequality (2.2), we obtain
Since \(v_{n}\rightarrow v\) in \(H_{0}\), \(v_{n}\in {\mathcal {B}}\) and \({\tilde{f}}(t)\le V_{0}/K\), we obtain
that is to say \(v_{n}\rightarrow 0\) in \(H_{0}\), which contradicts \(v\ne 0\).
Therefore, we may assume that \(y_{n} \rightarrow y_{0} \in {\mathbb {R}}^{N}\). Assume by contradiction that \(y_{0}\notin {\bar{\Lambda }}\). Then, there exists \(r>0\) such that for every n large enough, we have that \(|y_{n}-y_{0}|<r\) and \(B_{2r}(y_{0})\subset {\bar{\Lambda }}^{c}\). Then, if \(x\in B_{r /\varepsilon _{n}}(0)\), we have that \(|\varepsilon _{n}x+y_{n}-y_{0}|<2r\) so that \(\varepsilon _{n}x+y_{n}\in {\bar{\Lambda }}^{c}\), and so, arguing as before, we achieve a contradiction. Thus, \(y_{0} \in {\bar{\Lambda }}\).
To prove that \(V\left( y_{0}\right) =V_{0}\), we suppose by contradiction that \(V\left( y_{0}\right) >V_{0}\). Using Fatou’s lemma, the change of variable \(z=x+{\tilde{y}}_{n}\) and \(\max _{t \ge 0} J_{\varepsilon _{n}}\left( t u_{n}\right) =J_{\varepsilon _{n}}\left( u_{n}\right) \), we obtain
which is impossible and the proof is complete. \(\square \)
Let \(h:{\mathbb {R}}^{+}\rightarrow {\mathbb {R}}^{+}\) be any positive function satisfying \(h(\varepsilon )\rightarrow 0\) as \(\varepsilon \rightarrow 0^{+}\). Consider the following subset of the Nehari manifold
Fixed \(y\in M\), we conclude from Lemma 5.1 that \(|J_{\varepsilon }\left( \Phi _{\varepsilon }(y)\right) -c_{V_{0}}|\rightarrow 0\) as \(\varepsilon \rightarrow 0^{+}\). Thus, \(\Phi _{\varepsilon }(y)\in \widetilde{{\mathcal {N}}}_{\varepsilon }\) and \(\tilde{{\mathcal {N}}}_{\varepsilon }\ne \emptyset \) for \(\varepsilon >0\) small enough.
Arguing as Lemma 4.5 in [8], the next statement is to involve in the relationship between \(\widetilde{{\mathcal {N}}}_{\varepsilon }\) and the barycenter map.
Lemma 5.4
For any \(\delta >0\), we have
Next we prove our multiplicity result by presenting a relation between the topology of M and the number of solutions of the modified problem (3.1).
Theorem 1.2
For any \(\delta >0\) such that \(M_{\delta } \subset \Lambda \), there exists \({\tilde{\varepsilon }}_{\delta }>0\) such that, for any \(\varepsilon \in (0, {\tilde{\varepsilon }}_{\delta })\), problem (3.1) has at least \({\text {cat}}_{M_{\delta }} M\) nontrivial solutions.
Proof
For any given \(\delta >0\) such that \(M_\delta \subset \Lambda \), by Lemmas 5.1, 5.2, and 5.3, we employ an argument as in [15] to deduce the existence of \({\tilde{\varepsilon }}_\delta >0\) such that, for any \(\varepsilon \in \left( 0, \varepsilon _\delta \right) \), the following diagram
is well defined and \(\beta _{\varepsilon } \circ \Phi _{\varepsilon }\) is homotopically equivalent to the embedding \(\iota : M \rightarrow M_\delta \). This fact implies that \({\text {cat}}_{\widetilde{{\mathcal {N}}}_{\varepsilon }}(\widetilde{{\mathcal {N}}}_{\varepsilon }) \ge {\text {cat}}_{M_\delta }(M)\) due to [15, Lemma 2.2]. It follows from Lemma 3.9 and standard Ljusternik–Schnirelmann theory that \(J_{\varepsilon }\) possesses at least cat \(_{\tilde{{\mathcal {N}}}_{\varepsilon }}(\tilde{{\mathcal {N}}}_{\varepsilon })\) critical points on \({\mathcal {N}}_{\varepsilon }\). Consequently, equation (3.1) possess at least \({\text {cat}}_{M_\delta }(M)\) critical points. \(\square \)
6 Proof of Theorem (1.2)
In this section, we prove our main result. The idea is to show that the solutions \(u_{\varepsilon }\) obtained in Theorem 5.5 satisfy
for \(\varepsilon \) small enough. Arguing as in [4, Lemma 4.1], we have the following important result.
Lemma 6.1
Let \(\varepsilon _{n} \rightarrow 0^{+}\)and \(u_{n} \in \widetilde{{\mathcal {N}}}_{\varepsilon _{n}}\) be a solution of (3.1). Then, \(J_{\varepsilon _{n}}\left( u_{n}\right) \rightarrow c_{V_{0}}\) and \(|u_{n}|\in L^{\infty }\left( {\mathbb {R}}^{N}\right) \). Moreover, for any given \(\gamma >0\), there exists \(R>0\) and \(n_{0}\in {\mathbb {N}}\) such that
where \({\tilde{y}}_{n}\) is given by Lemma 5.3.
Now, we are ready to give a proof of Theorem 1.2.
Proof of Theorem 1.2
Let \(\delta >0\) be such that \(M_{\delta } \subset \Lambda \). We first claim that there exists \({\tilde{\varepsilon }}_{\delta }>0\) such that for any \(0<\varepsilon <{\tilde{\varepsilon }}_{\delta }\) and any solution \(u \in \widetilde{{\mathcal {N}}}_{\varepsilon }\) of the problem (3.1), it holds
To prove the above claim, we argue by contradiction. Assume that there are two sequences \(\varepsilon _{n} \rightarrow 0^{+}\)and \(u_{n} \in \widetilde{{\mathcal {N}}}_{\varepsilon _{n}}\) verifying \(J_{\varepsilon _{n}}^{\prime }\left( u_{n}\right) =0\) and
As in Lemma 5.1, we have that \(J_{\varepsilon _{n}}(u_{n}) \rightarrow c_{V_{0}}\). Therefore, by Lemma 5.3, there exists obtain a sequence \(\{{\tilde{y}}_{n}\}\subset {\mathbb {R}}^{N}\) such that \(\varepsilon _{n} {\tilde{y}}_{n} \rightarrow y_{0} \in M\). If we take \(r>0\) such that \(B_{r}\left( y_{0}\right) \subset B_{2 r}\left( y_{0}\right) \subset \Lambda \) we have that
Moreover, for any \(z \in B_{r / \varepsilon _{n}}\left( {\tilde{y}}_{n}\right) \), it holds
for n large. For these values of n, we have that \(B_{r / \varepsilon _{n}}\left( {\tilde{y}}_{n}\right) \subset \Lambda _{\varepsilon _{n}}\), that is, \({\mathbb {R}}^{N} \backslash \Lambda _{\varepsilon _{n}}\subset {\mathbb {R}}^{N} \backslash B_{r / \varepsilon _{n}}\left( {\tilde{y}}_{n}\right) \). On the other hand, it follows from Lemma 6.1 with \(\gamma =a\) that, for any \(n \ge n_{0}\) such that \(r / \varepsilon _{n}>R\), it holds
which contradicts with (6.3). So the claim is correct.
Let \({\hat{\varepsilon }}_{\delta }>0\) be given by Theorem 5.5 and set \(\varepsilon _{\delta }:=\min \left\{ {\hat{\varepsilon }}_{\delta }, {\tilde{\varepsilon }}_{\delta }\right\} \). We will show the theorem holds for this choice of \(\varepsilon _{\delta }\). Let \(0<\varepsilon <\varepsilon _{\delta }\) be fixed. By Theorem 5.5, there is \({\text {cat}}_{M_{\delta }}(M)\) nontrivial solutions of the problem (3.1). If \(u \in H_{\varepsilon }\) is one of these solutions, we have that \(u \in \widetilde{{\mathcal {N}}}_{\varepsilon }\). Then, by (6.2) \(g(\varepsilon x,|u|^{2})= f(|u|^{2})\), u is a solution of the problem (3.1). An easy calculation shows that \({\hat{u}}(x):=u(x / \varepsilon )\) is a solution of the original problem (2.1). Then, problem (2.1) has at least \({\text {cat}}_{M_{\delta }}(M)\) nontrivial solutions.
Take \(\varepsilon _{n}\rightarrow 0^{+}\) and \(\{u_{n}\}\) a sequence of solutions to (3.1). In order to study the behavior of the maximum points of \(\left| u_{n}\right| \), we first notice that, by (g4), there exists \(\gamma >0\) such that
By Lemma 6.1, there are \(R>0\) and \(\{{\tilde{y}}_{n}\}\subset {\mathbb {R}}^{N}\) such that
Up to a subsequence, we can assume that
Indeed, if this is not the case, we have \(\left\| u_{n}\right\| _{L^{\infty }\left( {\mathbb {R}}^{N}\right) }<\gamma \), and therefore, it follows from \(J_{\varepsilon _{n}}^{\prime }\left( u_{n}\right) =0\), (6.4) and the diamagnetic inequality (2.2) that
The above expression implies that \(\Vert |u_{n}|\Vert _{H^{1}({\mathbb {R}}^{N},{\mathbb {R}})}=0\), which is a contradiction. Thus, (6.6) holds. By (6.5) and (6.6), we can infer the maximum point \(p_{n}\in {\mathbb {R}}^{N}\) of \(|u_{n}|\) belongs to \(B_{R}({\tilde{y}}_{n})\). Hence, \(p_{n}={\tilde{y}}_{n}+q_{n}\) for some \(q_{n}\in B_{R}(0)\). Recalling that the associated solution of (2.1) is of the form \({\hat{u}}_{n}(x)=u_{n}(x /\varepsilon _{n})\), we conclude that the maximum point \(\eta _{n}\) of \(|{\hat{u}}_{n}|\) is \(\eta _{\varepsilon _{n}}:=\varepsilon _{n} {\tilde{y}}_{n}+\varepsilon _{n} q_{n}\). Since \(\left( q_{n}\right) \subset B_{R}(0)\) is bounded and \(\varepsilon _{n} {\tilde{y}}_{n} \rightarrow y_{0} \in M\), we obtain
The proof is complete. \(\square \)
Data availability
Data sharing is not applicable to this article as no new data were created or analyzed in this study.
References
Ackermann, N.: On a periodic Schrödinger equation with nonlocal superlinear part. Math. Z. 248, 423–443 (2004)
Alves, C.O., Figueiredo, G.M.: Multiple solutions for a semilinear elliptic equation with critical growth and magnetic field. Milan J. Math. 82(2), 389–405 (2014)
Alves, C.O., Figueiredo, G.M., Furtado, M.F.: Multiple solutions for a nonlinear Schrödinger equation with magnetic fields. Commun. Part. Diff. Eq. 36(9), 1565–1586 (2011)
Alves, C.O., Figueiredo, G.M., Yang, M.: Multiple semiclassical solutions for a nonlinear Choquard equation with magnetic field. Asymptot. Anal. 96(2), 135–159 (2016)
Alves, C.O., Gao, F., Squassina, M., Yang, M.: Singularly perturbed critical Choquard equations. J. Differ. Equ. 263(7), 3943–3988 (2017)
Ambrosetti A., Malchiodi, A.: Concentration phenomena for nonlinear Schrödinger equations: recent results and new perspectives. In: Perspectives in Nonlinear Partial Differential Equations, Contemp. Math., Amer. Math. Soc., Providence, RI, vol. 446, pp. 19–30 (2007)
Ambrosio, V.: Existence and concentration results for some fractional Schrödinger equations in \(\mathbb{R}^{N}\) with magnetic fields. Comm. Partial Differ. Equ. 44(8), 637–680 (2019)
Ambrosio, V., d’Avenia, P.: Nonlinear fractional magnetic Schrödinger equation: existence and multiplicity. J. Differ. Equ. 264, 3336–3368 (2018)
Arioli, G., Szulkin, A.: A semilinear Schrödinger equation in the presence of a magnetic field. Arch. Ration. Mech. Anal. 170(4), 277–295 (2003)
Bueno, H., Lisboa, N., Vieira, L.: Nonlinear perturbations of a periodic magnetic Choquard equation with Hardy–Littlewood–Sobolev critical exponent. Z. Angew. Math. Phys. 71(4), 26 (2020). (Paper No.143)
Byeon, J., Jeanjean, L., Maris, M.: Symmetric and monotonicity of least energy solutions. Calc. Var. Partial. Differ. Equ. 36, 481–492 (2009)
Chen, S., Rădulescu, V.D., Tang, X., Wen, L.: Planar Kirchhoff equations with critical exponential growth and trapping potential. Math. Z. 302, 1061–1089 (2022)
Cingolani, S.: Semiclassical stationary states of nonlinear Schrödinger equations with an external magnetic field. J. Diff. Eq. 188(1), 52–79 (2003)
Cingolani, S., Clapp, M., Secchi, S.: Multiple solutions to a magnetic nonlinear Choquard equation. Z. Angew. Math. Phys. 63(2), 233–248 (2012)
Cingolani, S., Lazzo, M.: Multiple semiclassical standing waves for a class of nonlinear Schrödinger equations. Topol. Methods Nonlinear Anal. 10, 1–13 (1997)
Cingolani, S., Secchi, S.: Semiclassical states for NLS equations with magnetic potentials having polynomial growths. J. Math. Phys. 46(5), 053503 (2005)
d’Avenia, P., Ji, C.: Multiplicity and concentration results for a magnetic Schrödinger equation with exponential critical growth in \(\mathbb{R}^{2}\). Int. Math. Res. Not. IMRN. 2, 862–897 (2022)
d’Avenia, P., Squassina, M.: Ground states for fractional magnetic operators. ESAIM COCV. 24(1), 1–24 (2018)
del Pino, M., Felmer, P.L.: Local mountain passes for semilinear elliptic problems in unbounded domains. Calc. Var. Partial Differ. Equ. 4(2), 121–137 (1996)
Esteban, M., Lions, P. L.: Stationary solutions of nonlinear Schrödinger equations with an external magnetic field. In: Brezis, H. (ed.) Partial Differential Equations and the Calculus of Variations. Vol. I: Progress in Nonlinear Differential Equations and their Applications. Boston, MA: Birkhauser Boston, pp.401–449 (1989)
Gao, F., Yang, M.: The Brézis-Nirenberg type critical problem for nonlinear Choquard equation. Sci. China Math. 61(7), 1219–1242 (2018)
He, M., Zou, W.: Existence and concentration result for the fractional Schrödinger equations with critical nonlinearities. Calc. Var. Partial Differ. Equ. 55, 1–39 (2016)
Ji, C., Rădulescu, V.: Multiplicity and concentration of solutions to the nonlinear magnetic Schrodinger equation. Calc. Var. Partial Differ. Equ. 59, 115 (2020)
Ji, C., Rădulescu, V.: Concentration phenomena for nonlinear magnetic Schrödinger equations with critical growth. Israel J. Math. 241, 465–500 (2021)
Lan, J., He, X., Meng, Y.: Normalized solutions for a critical fractional Choquard equation with a nonlocal perturbation. Adv. Nonlinear Anal. 12(1), 20230112 (2023)
Lieb, E.H.: Existence and uniqueness of the minimizing solution of Choquard’s nonlinear equation. Stud. Appl. Math. 57(2), 93–105 (1976)
Lieb, E.H., Loss, M.: Analysis. In: Gradute Studies in Mathematics, AMS, Providence, Rhode island (2001)
Lions, P.L.: The Choquard equation and related questions. Nonlinear Anal. 4(6), 1063–1072 (1980)
Lions, P.L.: The concentration-compactness principle in the calculus of variations. The locally compact case. II. Ann. Inst. H. Poincaré Anal. Non Linéaire 4, 223–283 (1984)
Ma, L., Zhao, L.: Classication of positive solitary solutions of the nonlinear Choquard equation. Arch. Ration. Mech. Anal. 195(2), 455–467 (2010)
Moroz, V., Van Schaftingen, J.: Groundstates of nonlinear Choquard equations: existence, qualitative properties and decay asymptotics. J. Funct. Anal. 265(2), 153–184 (2013)
Moroz, V., Van Schaftingen, J.: Semi-classical states for the Choquard equation. Calc. Var. Partial. Differ. Equ. 52, 199–235 (2015)
Moroz, V., Van Schaftingen, J.: Existence of groundstates for a class of nonlinear Choquard equations. Trans. Amer. Math. Soc. 367(9), 6557–6579 (2015)
Mukherjee, T., Sreenadh, K.: On concentration of least energy solutions for magnetic critical Choquard equations. J. Math. Anal. Appl. 464(1), 402–420 (2018)
Kurata, K.: Existence and semi-classical limit of the least energy solution to a nonlinear Schrödinger equation with electromagnetic fields. Nonlinear Anal. 41(5–6), 763–778 (2000)
Pekar, S.: Untersuchung über die Elektronentheorie der Kristalle. Akademie Verlag, Berlin (1954)
Secchi, S.: Ground state solutions for nonlinear fractional Schrödinger equations in \(\mathbb{R} ^{N}\). J. Math. Phys. 54(3), 031501 (2013)
Squasssina, M., Volzone, B.: Bourgain-Br\(\acute{e}\)zis-Mironescu formula for magnetic operators. C. R. Math. 354, 825–831 (2016)
Szulkin, A., Weth, T.: The method of Nehari manifold. In: Gao, D.Y., Motreanu, D. (eds.) Handbook of Nonconvex Analysis and Applications, pp. 2314–2351. International Press, Boston (2010)
Tang, H.: Existence of solution for magnetic Schrodinger equation with the Neumann boundary condition. Complex Var. Elliptic Equ. 68(8), 1313–1331 (2023)
Wen, L., Rădulescu, V.D., Tang, X., Chen, S.: Ground state solutions of magnetic Schrödinger equations with exponential growth. Discret. Contin. Dyn. Syst. 42, 5783–5815 (2022)
Willem, M.: Minimax Theorems, Progress in Nonlinear Differential Equations and their Applications, vol. 24. Birkhäuser Boston Inc, Boston, ISBN 0-8176-3913-6 (1996)
Zhang, W., Yuan, S., Wen, L.: Existence and concentration of ground-states forfractional Choquard equation with indefinite potential. Adv. Nonlinear Anal. 11(1), 1552–1578 (2022)
Acknowledgements
The author is grateful to the referees for their careful reading and valuable comments.
Funding
No funding.
Author information
Authors and Affiliations
Contributions
I have written the main manuscript tex and reviewed the manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no conflict of interest.
Ethical approval and consent to participate
There are no experiments or data related to ethical research in this article.
Consent for publication
All authors agree to publish this article.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Tang, H. Multiplicity and concentration behavior of solutions for magnetic Choquard equation with critical growth. Z. Angew. Math. Phys. 75, 183 (2024). https://doi.org/10.1007/s00033-024-02318-4
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00033-024-02318-4