Abstract
In this paper, we focus on the existence of normalized solutions to the following Schrödinger equation with the Stein–Weiss potential
where \(2\alpha +\mu \le 2\), \(0<\mu <2\), \(I_{\mu }\) denotes the Riesz potential and \(f:{{\mathbb {R}}} \rightarrow {{\mathbb {R}}} \) has critical exponential growth which behaves like \(e^{\alpha u^2}\). The solutions correspond to critical points of the underlying energy functional subject to the \(L^2\)-norm constraint, namely, \(\int _{{{\mathbb {R}}} ^2} |u|^{2}\textrm{d}x = a^2\) for \(a>0\) given. Under some weak assumptions, we prove the existence of the normalized solution for the equation by developing refined variational methods. In particular, we shall establish two new approaches to estimate precisely the minimax level of the underlying energy functional. As far as we know, our result is the first one in seeking normalized solutions of nonlinear equations involving the nonlocal Stein–Weiss reaction.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
In this paper, we are concerned with the existence of normalized solutions to the following nonlinear Schrödinger equation with the Stein–Weiss reaction under the exponential critical growth case
where \(a>0\), \(0<\mu <2\), \(2\alpha +\mu \le 2\). \(I_{\mu }\) denotes the Riesz potential defined by
where \(\Gamma \) represents the gamma function, \(*\) indicates the convolution operator, F(s) is the primitive of f(s) with that f(s) has exponential critical growth in \({{\mathbb {R}}} ^{2}\). Next, we shall introduce three typical features of this problem to set the tone for the rest of the paper.
1.1 Introduction of Three Typical Features
1.1.1 \(L^2\)-Constraint
We aim to search for solutions to \(({\mathcal {P}}_{a})\) having prescribed mass, the normalized stationary states, whose existence can be formulated as follows: given \(a>0\), we aim to find \((\lambda ,u)\in {{\mathbb {R}}} \times H^{1}({{\mathbb {R}}} ^{2})\) solving \(({\mathcal {P}}_{a})\) together with the normalization condition
and in this case \(\lambda \in {{\mathbb {R}}} \) cannot be prescribed but appear as Lagrange multipliers in the variational approach. This type of problem has important physical significance in Bose–Einstein condensates and the nonlinear optics framework, and the \(L^{2}\)-norm of such solutions is a preserved quantity of the evolution and the corresponding variational feature contributes to analyzing the orbital stability or instability. Naturally, such problems have attracted much attention in the fields of nonlinear PDEs in the last decades. We give brief introduction of the relative progress with the most general Schrödinger equation
which has been investigated extensively via the variational methods. One can search for the existence of the normalized solutions of (1.1) by considering the critical points of the corresponding energy functional \({\mathcal {J}}: H^1({{\mathbb {R}}} ^N) \rightarrow {{\mathbb {R}}} \) defined by
under the constraint
Generally, the types of such problems are divided into three parts: \(L^{2}\)-subcritical case, \(L^{2}\)-critical case, \(L^{2}\)-supercritical case, after a simple stretching consideration presented by
during which one can find there exists a new \(L^{2}\)-critical exponent \(q^{*}=2+4/N\). Generalizing to more general nonlinear term f, one can conclude that if f admits a \(L^{2}\)-subcritical growth, i.e., f has a growth \(u^{p-1}\) with \(p<q^{*}\) at infinity, then \({\mathcal {J}}|_{{\mathcal {S}}_{a,N}}\) is bound below and in this occasion, minimization method is the conventional approach to find normalized solutions, we refer to [12, 31] and the references therein in this aspect. If f admits a \(L^{2}\)-supercritical growth, i.e., f has a growth \(u^{p-1}\) with \(p>q^{*}\) at infinity, then \({\mathcal {J}}\) is unbound below on \({\mathcal {S}}_{a,N}\), which implies the traditional minimization method does not work and more efforts are always required in the study of the \(L^2\)-supercritical case.
One of the groundbreaking pieces of work in the \(L^2\)-supercritical case is accomplished by Jeanjean [21]. Jeanjean [21] obtained the normalized solution at the mountain pass level of its energy functional under the following conditions:
-
(H0)
f is odd;
-
(H1)
\(f\in {\mathcal {C}}({{\mathbb {R}}} , {{\mathbb {R}}} )\) and there exist \(\alpha ,\beta \in {{\mathbb {R}}} \) satisfying \((2N+4)/N<\alpha \le \beta <2^*=2N/(N-2)\) such that
$$\begin{aligned} 0<\alpha F(t)\le f(t)t\le \beta F(t), \ \ \ \ \forall \ t\in {{\mathbb {R}}} \setminus \{0\}, \end{aligned}$$
and the existence of ground state solutions was proved if f also satisfies
-
(H2)
the function \({\tilde{F}}(t):=f(t)t-2F(t)\) is of class \({\mathcal {C}}^1\) and
$$\begin{aligned} {\tilde{F}}'(t)t>\frac{2N+4}{N}F(t), \ \ \ \ \forall \ t\in {{\mathbb {R}}} \setminus \{0\}. \end{aligned}$$
Condition \(\mathrm{(H1)}\) is the general Ambrosetti–Rabinowitz condition, which is benefit for us to obtain bounded Palais–Smale sequences for \({\mathcal {J}}\) constrained on \({\mathcal {S}}_{a,N}\), and Jeanjean developed a novel argument related to the mountain pass geometry for the scaled functional \({\tilde{{\mathcal {J}}}}(u,s):=J(u,s)\) with \(s*u(\cdot ):=e^{\frac{Ns}{2}}(e^{s}\cdot )\), which is widely used to find normalized solutions in the \(L^{2}\)-supercritical case. Based on the above conditions and by applying the fountain theorem to the scaled functional \({\mathcal {J}}\), Bartsch and de Valerioda [5] obtained infinitely normalized solutions. Another variational approach is presented by Bartsch and Soave [6, 7], and it is based on Ghoussoub minimax principle [19]. Applying this abstract minimax theorem, Barstch and Soave established the existence and multiplicity of normalized solutions of equation (1.1) with general nonlinearities f. We also refer to Bieganowski and Mederski [8], which considered the normalized solutions under the \(L^{2}\)-supercritical case but Sobolev subcritical, they provided nearly optimal conditions in some ways. For more information, please see [8] and its references.
We conclude this part with a brief progress about the critical Schrödinger equation in the sense of Sobolev embedding when \(N\ge 3\). In 2020, Soave [28] first considered the Schrödinger equation with Sobolev critical growth:
This type of problem is more delicate since they need to analyze how the lower-order term \(|u|^{q-2}u\) affects the structure of the corresponding energy functional and to solve the lack of compactness caused by the Sobolev critical growth. According to the findings in [28], Eq. (1.2) has ground state solutions in the \(L^2\)-subcritical perturbation case \(2<q<2+N/4\) and \(L^2\)-supercritical perturbation case \(2+N/4<q<2^*\), respectively, for \(\mu a(1-\gamma _q)q<\alpha \), where \(\alpha =\alpha (N,q)\) is a specific constant depending on N, q and \(\gamma _q=N(q-2)\). We also refer to Wei and Wu [33], Jeanjean and Le [22], Jeanjean et al. [24] which settled several open questions proposed by Soave [28].
1.1.2 Critical Exponential Case
Besides the \(L^2\)-constraint, another novel feature of equation (\({\mathcal {P}}_{a}\)) is that functions f(u) have critical exponential growth that is the maximal growth that allows us to treat (\({\mathcal {P}}_{a}\)) variationally in \(H^1({{\mathbb {R}}} ^2)\), which was shown by Trudinger [32] and Moser [26], and it is motivated by the following Trudinger–Moser inequality [11].
Lemma 1.1
(i) If \(\alpha >0\) and \(u\in H^1({{\mathbb {R}}} ^2)\), then
(ii) if \(u\in H^1({{\mathbb {R}}} ^2), \Vert \nabla u\Vert _2^2\le 1, \Vert u\Vert _2 \le M < \infty \), and \(\alpha < 4\pi \), then there exists a constant \({\mathcal {C}}(M,\alpha )\), which depends only on M and \(\alpha \), such that
Inspired by the Trudinger–Moser-type inequality, we can say that a function \(f\in {\mathcal {C}}({{\mathbb {R}}} ,{{\mathbb {R}}} )\) possesses critical exponential growth if there exists a constant \(\alpha _{0}>0\) such that
Here, we refer Adimurthi–Yadava [36] and de Figueiredo–Miyagaki–Ruf [17] to the readers for more information. Based on the Trudinger–Moser inequalities, many authors considered the existence and multiplicity of weak solutions for the nonlinear Schrödinger equations. We refer to [2, 13, 14] and the references therein for more details about the relative progress in this direction.
1.1.3 Stein–Weiss Convolution Reaction
We briefly recall the related background and some pioneering contributions in this field, and we start with the weighted \(L^{p}\) estimates for the fractional integral
which is a fundamental problem in the field of harmonic analysis and such weighted \(L^{p}\) estimates are generated from quite natural phenomena and have practical significance in the large wide of mathematical fields, which can be summarized as that the appearance of some suitable symmetry hypotheses, notably radial symmetry, contributes to improving the classical estimates and some embedding properties of function spaces.
Many mathematicians have studied on the weighted \(L^{p}\) estimates for the fractional integral \(T_{\mu }\). Historically, Hardy and Littlewood [20] first considered the weighted \(L^{p}\) estimates for the one-dimensional fractional integral operator \(T_{\mu }\), then Sobolev [29] extended it to the N-dimensional case. Later, Stein and Weiss [30] obtained the following two-weight extension of the Hardy–Littlewood–Sobolev inequality, which is known as the Stein–Weiss inequality.
Proposition 1.2
(Doubly weighted Hardy–Littlewood–Sobolev inequality) Let \(t,s>1\) and \(0<\mu <N\) with \(\vartheta +\beta >0\), \(1/t+(\mu +\vartheta +\beta )/N+1/s=2\), \(\vartheta <\frac{N}{t'}\), \(\beta <\frac{N}{s'}\), \(g_{1}\in L^{t}({{\mathbb {R}}} ^{N})\) and \(g_{2}\in L^{s}({{\mathbb {R}}} ^{N})\), where \(t'\) and \(s'\) denote the Hölder conjugate of t and s, respectively. Then there exists a constant \(C(N,\mu ,\vartheta ,\beta ,t,s)\), independent of \(g_{1}\), \(g_{2}\) such that
For \(\vartheta =\beta =0\), it is reduced to the Hartree type (also called the Choquard type) nonlinearity, which is driven by the classical Hardy–Littlewood–Sobolev inequality.
Integrability for integral operators can be quantified using the Stein–Weiss inequality, which is fundamentally determined by the dilation nature of integral operators. Due to its significance in applications to issues in harmonic analysis and partial differential equations, the study of and comprehension of the Stein–Weiss inequality has sparked a growing amount of attention among scholars. We now look at the applications that the Stein–Weiss term has in relation to them. The polyharmonic Kirchhoff equations involving the critical Choquard-type exponential nonlinearity with singular weights were explored by Giacomoni et al. [4]. It is important to note the beautiful work of Du et al. [18], where they investigated the equation as below,
where \(2_{\alpha ,\mu }^{*}=(2N-2\alpha -\mu )/(N-2)\). In order to analyze the existence of solutions, study the regularity, and symmetry of positive solutions by moving plane arguments under the critical situation, as well as the results under the subcritical situation, the authors created a nonlocal version of the concentration—compactness principle. Yang et al. [38] achieved the symmetry, regularity, and asymptotic features of the weighted nonlocal system with critical exponents associated with the Stein–Weiss inequality by employing the moving plane arguments in integral form. For other results, we refer to [9, 37, 39, 40] and the references therein.
1.2 Introduction of Our Goal and Main Results
Among the investigations into normalized solutions of the nonlinear Schrödinger equation with critical growth, an new emerging interest is seeking the normalized solutions under the critical exponential growth in the sense of the Trudinger–Moser inequality, which was recently constructed by Alves et al. [3]. Under following hypothesis,
- (F2\('\)):
-
\(\lim _{|t|\rightarrow 0}|f(t)|/|t|^{l}=0\) for some constant \(l>3\);
- (F3\('\)):
-
there exists a constant \(\mu _0>4\) such that \(f(t)t\ge \mu _0 F(t)>0\) for all \(t\in {{\mathbb {R}}} \setminus \{0\}\);
- (F4\('\)):
-
there exist constants \(p>4\) and \(\gamma >0\) such that \(F(t)\ge \gamma |t|^p\) for all \(t\in {{\mathbb {R}}} \);
- (F5\('\)):
-
the function \({\tilde{F}}(t):=f(t)t-2F(t)\) is of class \({\mathcal {C}}^1\) and satisfies
$$\begin{aligned} {\tilde{F}}'(t)t\ge 4F(t), \ \ \ \ \forall \ t\in {{\mathbb {R}}} , \end{aligned}$$
they established the existence of normalized solutions to equation:
Their result reads as follows in this topic.
Theorem 1.3
([3, Theorem 1.2]) Assume that f possesses critical exponential growth and satisfies (F2\('\))-(F4\('\)). If \(a\in (0,1)\), then there exists \( \gamma ^*(a)>0\) such that (1.3) has a radial solution for all \(\gamma \ge \gamma ^*(a)\), where \(\gamma \) is given by (F4\('\)), moreover, this solution can be chosen as a positive ground state solution if f also satisfies (H0) and (F5\('\)).
Based on the idea introduced by Jeanjean [21], for every \(a\in (0,1)\), they constructed a special (PS) sequence \(\{u_n\}\subset {\mathcal {S}}_a^r:= {\mathcal {S}}_a\cap H_r^1({{\mathbb {R}}} ^2)\) such that
where the mountain pass level \(c_{\gamma }^{\infty }(a)\) depends on \(\gamma \) and \(\varphi : H^1({{\mathbb {R}}} ^2)\rightarrow {{\mathbb {R}}} \) is given by
on the constraint \({\mathcal {S}}_a\) which is defined by
As a consequence of the Pohozaev identity (see [21, Lemma 2.7]), any solution u of \(\varphi \) exists in the Pohozaev manifold given by
where \(\vartheta \) is called the Pohozaev functional defined by
Next, let us discuss a few key components of their proofs in the work [3]. By establishing the crucial estimation
they could overcome the influence caused by the exponential critical growth. Indeed, as long as \(\limsup _{n\rightarrow \infty }\Vert \nabla u_n\Vert _2^2<1-a\), then the following property
is a natural conclusion, since we have the Trudinger–Moser inequality and the compact embedding \(H_r^1({{\mathbb {R}}} ^2)\hookrightarrow L^s({{\mathbb {R}}} ^2)\) for any \(s>2\). Then they were able to derive the Brezis–Lieb property
and also the convergence
if \(u_n\rightharpoonup {\bar{u}}\) in \(H^1({{\mathbb {R}}} ^2)\). In this case, the dealing process of their problem is similar to the one of nonlinearities like \(f(u)\sim |u|^{q-2}u\) with \(q>2\). After this work, a series of subsequent studies have been done on the existence of different types of normalized solutions to the nonlinear equations in the planar case. Here, we refer to the work by Jeanjean and Lu [23], and Chen and Tang [15] and the references therein for more details.
Motivated by the above works, especially [3], in this paper, we discuss some refined analysis for the existence of normalized solutions to nonlinear Schrödinger equation with Stein–Weiss convolution term. As far as we know, our result is the first one in the fields of seeking normalized solutions involving the nonlocal Stein–Weiss reaction.
To state our results, besides (F1), we make the assumptions on f:
-
(F2)
\(\lim _{|t|\rightarrow 0}f(t)/t^{\frac{4-\mu -2\alpha }{2} }=0\);
-
(F3)
\(f(t)t\ge (6-\mu -2\alpha )F(t)/2\) for all \(t\in {{\mathbb {R}}} \setminus \{0\}\);
-
(F4)
\(\liminf _{|t|\rightarrow \infty }\frac{ f(t)}{e^{\alpha _0 t^2}}>0\);
-
(F5)
there exist constants \(M_0>0\) and \(\beta _0>0\) such that
$$\begin{aligned} F(t)\le M_0 |f(t)|, \ \ \ \ \forall \ |t|\ge \beta _0. \end{aligned}$$
Theorem 1.4
Assume that \(a>0\), \(\alpha <\mu \) and f satisfies (F1)–(F5). Then equation \(({\mathcal {P}}_{a}) \) has a radial normalized solution. Moreover, for any solution the associated Lagrange multiplier \(\lambda \) is positive.
Remark 1.5
We would like to point out that the purpose of both Theorem 1.4 and following Theorem 1.6 is to demonstrate that radial solutions exist for \(({\mathcal {P}}_{a})\). As a result, we focus on the space \(H_r^1({{\mathbb {R}}} ^2)\) since it compactly embeds in \(L^s({{\mathbb {R}}} ^2)\) for all \(s>2\) and aids in the recovery of compactness. The solutions in \(H_r^1({{\mathbb {R}}} ^2)\) are in reality solutions in whole \(H^1({{\mathbb {R}}} ^2)\) according to Palais’ principle of symmetric criticality [35].
Set
where A is defined in (3.10).
Theorem 1.6
Assume that \(a>0\), \(\alpha <\mu \) and f satisfies (F1)–(F3), (F4\('\)) with \(\gamma >\gamma ^*(a)\) and (F5). Then conclusions of Theorem 1.4 hold.
Let us now outline the main strategy to prove Theorems 1.4 and 1.6. Our arguments are based on variational approaches and refined analysis techniques in order to complete the proofs of main results. It is easily seen that solutions of problem \(({\mathcal {P}}_{a})\) can be found by looking for critical points of the energy functional \(\Phi :H^{1}({{\mathbb {R}}} ^{2})\rightarrow {{\mathbb {R}}} \) given by
By proposition 1.2, we know the latter term of the right-hand side of equation \(\Phi (u)\) is well defined if \(F(u)\in L^{t}({{\mathbb {R}}} ^{2} )\) for \(t>1\) given by \(2/t+(\mu +2\alpha )/2=2\). This means that we must require \(F(u)\in L^{\frac{4}{4-\mu -2\alpha }}({{\mathbb {R}}} ^{2})\), which can be guaranteed by (F1), (F2) and the continuous Sobolev embedding \(H^1({{\mathbb {R}}} ^2)\hookrightarrow L^p({{\mathbb {R}}} ^2)\), where \(p\ge 2\). Furthermore, it is standard to show that \(\Phi \in {\mathcal {C}}^1(H^1({{\mathbb {R}}} ^2), {{\mathbb {R}}} )\). More precisely, since the exponential critical growth in \({{\mathbb {R}}} ^{2}\) implies that the energy functional \(\Phi \) is no more bounded from below on \({\mathcal {S}}_{a}\), and we shall look for a critical point satisfying a minimax characterization. Here, we introduce the following definition.
Definition 1.7
For given \(a>0\), we say that \(\Phi \) possesses a mountain pass geometry on \({\mathcal {S}}_a\) if there exists \(\rho _a>0\) such that
where \(\Gamma _a:=\{g\in {\mathcal {C}}([0,1],{\mathcal {S}}_a):\Vert \nabla g(0)\Vert _2^2\le \rho _a, \Phi (g(1))<0\}\).
We intend to make use of the above definition to verify the mountain pass geometry of our problem. Besides, we highlight how the Stein–Weiss convolution term, in conjunction with the critical exponential growth, presents some new difficulties to our approach, as well as the accomplishments in our paper, which can be seen below.
-
(i)
One of the main challenges comes after the appearance of the Stein–Weiss convolution term. We must confirm that the weak limit of the (PS) sequence of energy functional \(\Phi \) is, in fact, the solution to the equation \( ({\mathcal {P}}_{a})\). However, in the case of exponential growth, the appearance of the Trudinger–Moser inequality requires that the critical exponent \(\alpha _{0}\) be less than \(2\pi \). We notice that Alves and Shen [1] provided a version of proof of such property; here, we would like to show another version of proof based on the condition that \(\alpha <\mu \), which is motivated by the method explored in Qin and Tang [27, Lemma 4.8].
-
(ii)
The work by Alves et al.[3] is significant in the field of normalized solutions in the two-dimensional critical case, and the technique previously described demonstrates that it is possible to control the energy level of the corresponding energy functional arbitrary small by simply taking a large enough value for the parameter \(\gamma \). This estimation is one of the most crucial components that cannot be overlooked because we are working with the exponential critical case. The first technique we developed in this instance is to provide an exact lower bound \(\gamma \), which is defined by (1.4). We shall provide a precise range of values for \(\gamma \) to fulfill the energy estimation in this method.
-
(iii)
We would like to point out that, in the ordinary methods, one usually could take advantage of the Moser-type functions to pull down the critical value to a particular threshold value. However, it seems like there is no such estimation in seeking normalized solutions of the nonlinear equations except the very recent work by Zhang et al. [41], where they used the traditional Moser-type function, then to perform a stretch to satisfy the constraint mass and then gave the estimation. Here, motivated by Chen et al. [16], we will directly improve the traditional Moser-type function. By the suitable extension of the traditional Moser-type function, we can obtain the test functions in \(H^1({{\mathbb {R}}} ^2)\) on the \(L^2\)-constraint \({\mathcal {S}}_a\). Another highlight in this progress is our assumption with the growth on f at infinity is relatively weak compared to the existing works about the Stein–Weiss convolution term.
-
(iv)
Our results are superior than those of [3] in that we additionally change the constraint condition utilized in [3] from \(a\in (0,1)\) to \(a>0\).
Remark 1.8
We emphasize that the ground state normalized solution can be obtained using our work with some additional assumptions. However, in some ways, finding a ground state normalized solution is similar to the process explored as in our paper.
The organization of the remainder of this paper is as follows. In Sect. 2, we shall introduce some preliminary results and establish the mountain pass geometry of the associated energy functional. In Sect. 3, we shall apply two different approaches to give a precise estimation for the mountain pass energy level. In Sect. 4, we shall restore the compactness and prove the existence of normalized solutions of the equation \({\mathcal {P}}_{a}\).
Finally, we introduce some notations that will clarify what follows.
\(\bullet \) \(C,C_{i},c_{i}\ (i=1,2,...)\) denote positive constants which may vary from line to line.
\(\bullet \) For any exponent \(p>1\), \(p'\) denotes the conjugate of p and is given as \(p'=p/(p-1)\).
\(\bullet \) \(B_{r}(x)\) denotes the ball of radius r centered at \(x\in {{\mathbb {R}}} ^{2}\).
\(\bullet \) The arrows \(\rightharpoonup \) and \(\rightarrow \) denote the weak convergence and strong convergence, respectively.
\(\bullet \) \(L^{s}({{\mathbb {R}}} ^{2})(1\le s<+\infty )\) denotes the Lebesgue space with the norm \(\Vert u\Vert _{s}=(\int _{{{\mathbb {R}}} ^{2}}|u|^{s}\textrm{d}x)^{1/s}\).
\(\bullet \) \( \beta : H:=H^1({{\mathbb {R}}} ^2)\times {{\mathbb {R}}} \rightarrow H^1({{\mathbb {R}}} ^2)\) is a continuous map defined by
where H is a Banach space equipped with the scalar product
and corresponding norm \(\Vert (v,t)\Vert _{H}:=\left( \Vert v\Vert ^2+|t|^2\right) ^{1/2}\) for all \((v,s)\in H\).
2 Preliminary Results
In this section, we give some preliminary results which will be useful throughout the rest of the paper.
Lemma 2.1
(Gagliardo–Nirenberg inequality [34]) Let \(q>2\). Then there exists a sharp constant \({\mathcal {S}}_{q}>0\) such that
where \({\mathcal {S}}_{q}=\frac{q}{2\Vert U_{q}\Vert _{2}^{q-2}}\), and \(U_q\) is the ground state solution of the following equation:
To deal with the nonlocal type problem (\({\mathcal {P}}_{a}\)), we also need the following inequality.
Lemma 2.2
(Cauchy–Schwarz type inequality [25, Sect. 5]) For f, \(h\in L_{\textrm{loc}}^{1}({{\mathbb {R}}} ^{2})\), there holds
Finding the bounded (PS) sequence relies on the mountain pass geometry based on Definition 1.7. Especially, verifying the boundedness of such a sequence is not trivial, which needs the information of \(L^2\)-Pohozaev inequality. Here, we introduce the following:
and
Following that, we show that \(\Phi \) has a mountain pass geometry on the constraint \({\mathcal {S}}_a\), which is as follows.
Lemma 2.3
Assume that (F1)–(F3) and (F5) hold. Then
-
(i)
there exists \(K(a)>0\) small enough such that \( \Phi (u)> 0\) and \(J(u)>0\) if \(u\in A_{2K}\) and
$$\begin{aligned} 0<\sup _{u\in A_{K}}\Phi (u)<\inf \left\{ \Phi (u): u\in {\mathcal {S}}_a, \Vert \nabla u\Vert _2^2= 2K(a) \right\} , \end{aligned}$$(2.3)where \(A_{K}=\left\{ u\in {\mathcal {S}}_a: \Vert \nabla u\Vert _2^2\le K(a)\right\} \) and \(A_{2K}=\left\{ u\in {\mathcal {S}}_a: \Vert \nabla u\Vert _2^2\le 2K(a)\right\} \).
-
(ii)
\(\Gamma _a=\{g\in {\mathcal {C}}([0,1],{\mathcal {S}}_a):\Vert \nabla g(0)\Vert _2^2\le K(a), \Phi (g(1))<0\}\ne \emptyset \) and
$$\begin{aligned} {c}(a) =\inf _{g\in \Gamma _a}\max _{t\in [0,1]}\Phi (g(t))&\ge \inf \left\{ \Phi (u): u\in {\mathcal {S}}_a, \Vert \nabla u\Vert _2^2= 2K(a) \right\} \\&>\max _{g\in \Gamma _a}\max \{\Phi (g(0)),\Phi (g(1))\}. \end{aligned}$$
Proof
(i) Fixing \(\alpha >\alpha _0\), by (F1) and (F2), we know that for any \(\varepsilon >0\) and any \(q\ge 1\), there exists \(C_{\alpha ,\varepsilon ,q}>0\) such that
moreover, using (2.4), we deduce that for any \(\varepsilon >0\), there exists \(C_{\alpha ,\varepsilon }>0\) such that
Thus, by Proposition 1.2, Lemma 2.1, and (2.5) and by selecting \(\varepsilon \) small enough, we have
Now, let \(0<K\) be arbitrary but fixed and suppose \(u,u_0,v,v_0\in {\mathcal {S}}_a\) satisfy that \(\Vert \nabla u\Vert _2^2\le K\), \(\Vert \nabla v\Vert _2^2\le 2K\), and \(\Vert \nabla v_0\Vert _2^2=2K\). From above, when \(K>0\) is small enough, we can say that \(\Phi (v)>0\). Similarly, we can obtain that \(J(v)>0\) and again by selecting \(\varepsilon \), K small enough, we have
From above, there exists \(K=K(a)>0\) sufficiently small such that \(\Phi (u)>0\) and \(J(u)>0\) if \(u\in A_{2k}\), and (2.3) holds.
(ii) We first prove that \(\Gamma _a\ne \emptyset \). Using (F3) and (F5), it is easy to see that
For any given \(w\in {\mathcal {S}}_a\), we have \(\Vert tw_t\Vert _2=\Vert w\Vert _2\), and so \(tw_t\in {\mathcal {S}}_a\) for every \(t>0\). Then (2.1) and (2.6) yield
Thus we can deduce that there exist \(t_1>0\) small enough and \(t_2>0\) large enough such that
Let \(g_0(t):=(t_1+(t_2-t_1)t)w_{t_1+(t_2-t_1)t}\). Then \(g_0\in \Gamma _a\), and so \(\Gamma _a\ne \emptyset \). Now using the intermediate value theorem, for any \(g\in \Gamma _a\), there exists \(t_0\in (0,1)\), depending on g, such that \(\Vert \nabla g(t_0)\Vert _2^2=2K(a)\) and
which, together with the arbitrariness of \(g\in \Gamma _a\), implies
Hence, (ii) follows directly from (2.3) and (2.8), and the proof is completed. \(\square \)
We recall that any solution of \(({\mathcal {P}}_{a})\) lives in the \(L^2\)-Pohozaev manifold given by
Remark 2.4
From \(J(v)>0\) when \(v\in {\mathcal {S}}_{c}\) and \(\Vert \nabla v\Vert _{2}^{2}\le 2K\), we can deduce that for any \(a>0\), there exists a constant \(\rho (a)>0\), just depending on \(a>0\), such that \(\Vert \nabla u\Vert _2\ge \rho (a)\) for all \(u\in {\mathcal {M}}_a\).
Next, inspired by [21], we consider the following auxiliary functional:
We shall show that \({\tilde{\Phi }}\) also possesses a kind of mountain pass geometrical structure on \({\mathcal {S}}_a\times {{\mathbb {R}}} \). Since the proof is standard, we omit it here.
Lemma 2.5
Assume that (F1)–(F3) and (F5) hold. Let \(v\in {\mathcal {S}}_a\) be arbitrary but fixed. Then we have
-
(i)
\(\Vert \nabla \beta (v, t)\Vert _2\rightarrow 0\) and \(\Phi (\beta (v,t))\rightarrow 0\) as \(t\rightarrow -\infty ;\)
-
(ii)
\(\Vert \nabla \beta (v, t)\Vert _2\rightarrow +\infty \) and \(\Phi (\beta (v,t))\rightarrow -\infty \) as \(t\rightarrow +\infty ;\)
-
(iii)
there exist \(s_1<0\) and \(s_2>0\), depending on a and v, such that the functions \({\tilde{v}}_1=\beta (v,s_1)\) and \({\tilde{v}}_2=\beta (v,s_2)\) satisfy
$$\begin{aligned} \Vert \nabla {\tilde{v}}_1\Vert _2^2\le K(a),\ \ \Vert \nabla {\tilde{v}}_2\Vert _2^2>2K(a) \ \ \hbox {and}\ \ \Phi ({\tilde{v}}_2)<0. \end{aligned}$$
Lemma 2.6
Assume that (F1)–(F3) and (F5) hold. Then
where
By the argument explored as in [35], we know that for any \(a>0\), \({\mathcal {S}}_a\) is a submanifold of \(H^1({{\mathbb {R}}} ^2)\) with codimension 1 and the tangent space at \({\mathcal {S}}_a\) is defined as
The norm of the \({\mathcal {C}}^1\) restriction functional \(\Phi |_{{\mathcal {S}}_a}'(u)\) is defined by
As in Jeanjean [21], for every \((u,t)\in {\mathcal {S}}_a\times {{\mathbb {R}}} \), we define the following linear space
We see that \({\tilde{\Phi }} (v,t)\) is of class \({\mathcal {C}}^1\) and for any \((w,s)\in H\),
The norm of the derivative of the \({\mathcal {C}}^1\) restriction functional \({\tilde{\Phi }}|_{{\mathcal {S}}_a\times {{\mathbb {R}}} }\) is defined by
In the same way as [21, Proposition 2.2], we have the following proposition.
Proposition 2.7
Assume that \({\tilde{\Phi }}\) has a mountain pass geometry on the constraint \({\mathcal {S}}_a\times {{\mathbb {R}}} \). Let \(a>0\) and \(\{{\tilde{g}}_n\} \subset {\tilde{\Gamma }}_a\) be such that
Then there exists a sequence \(\{(v_n,t_n)\}\subset {\mathcal {S}}_a\times {{\mathbb {R}}} \) such that
-
1.
\({\tilde{\Phi }}(v_n,t_n)\in \left[ {\tilde{c}}(a)-\frac{1}{n},{\tilde{c}}(a)+\frac{1}{n}\right] \);
-
2.
\(\min _{\tau \in [0,1]}\Vert (v_n,t_n)-{\tilde{g}}_n(\tau )\Vert _H\le \frac{1}{\sqrt{n}}\);
-
3.
\(\Vert {\tilde{\Phi }}|_{{\mathcal {S}}_a\times {{\mathbb {R}}} }'(v_n,t_n)\Vert \le \frac{2}{\sqrt{n}}\), i.e.,
$$\begin{aligned} |\langle {\tilde{\Phi }}'(v_n,t_n),(v,s)\rangle |\le \frac{2}{\sqrt{n}}\Vert (v,s)\Vert _{H}, \ \ \ \ \forall \ (v,s)\in {\tilde{T}}_{v_n,t_n}. \end{aligned}$$
Note that
With the aforementioned lemmas, we can get the desired sequence as follows.
Lemma 2.8
Assume that (F1)–(F3) and (F5) hold. Then there exists a bounded sequence \(\{u_n\}\subset {\mathcal {S}}_a\) such that
Proof
Let
where \(\beta \) is defined by (1.6), \(v_n, t_n\), and \({\tilde{g}}_n\) are given in Proposition 2.7. Then \(u_n\in {\mathcal {S}}_a\) and \(g_n\in \Gamma _a\) by (ii) of Lemma 2.3. Moreover, by (2.9), (2.10), Lemma 2.6, and Proposition 2.7, we have
and
By (F3), we have
which implies that \(\int _{{{\mathbb {R}}} ^{2}}\left( I_{\mu }\times \frac{F(u)}{|x|^{\alpha }}\right) \frac{F(u)}{|x|^{\alpha }}\textrm{d}x\) is bounded. Using (F3) and (F4), we know that for any \(\delta >0\) there exists \(R_{\delta }>0\) such that
Then we have
which implies that \(\{u_n\}\) is bounded in \(H^1({{\mathbb {R}}} ^2)\). To finish the proof, it remains to prove that \(\Phi |_{{\mathcal {S}}_a}'(u_n) \rightarrow 0\), i.e., \(\left\langle \Phi '(u_n),w\right\rangle \rightarrow 0\) for all \(w\in T_{u_n}\). For this, we just need to show that \(\{(\beta (w,-t_n),0)\}\subset T_{v_n,t_n}\) and \(\{(\beta (w,-t_n),0)\}\) is bounded in H since
Indeed, for any \(w\in T_{u_n}\), i.e.,
we have
which implies
Moreover, by (ii) of Proposition 2.7, we have
which leads to
This shows that \(\{(\beta (w,-t_n),0)\}\subset T_{v_n,t_n}\) is bounded in H. Jointly with (2.15), we get \(\Phi |_{{\mathcal {S}}_a}'(u_n) \rightarrow 0\). From this, (2.13) and (2.14), we conclude that \(\{u_n\}\), defined by (2.12), is bounded, and satisfies (2.11). The proof is completed. \(\square \)
3 Energy Estimates for Minimax Level
In this subsection, we give a precise estimation for the energy level c(a) given by (2.8), which helps us to restore the compactness in the critical exponential case in next subsection.
Let \(\kappa :=\liminf _{|t|\rightarrow \infty }\frac{f(t)}{e^{\alpha _0 t^2}}\). By (F4), we know that \(\kappa >0\). Then we can choose \(d>0\) such that
For large \(n\in {\mathcal {N}}\), let \(R_n\ge d\) be such that
Then one has
Now we define the following new Moser-type functions \(w_n(x)\) supported in \(B_{d}:=B_{d}(0)\)
Computing directly, we get that for large \(n\in {\mathcal {N}}\),
and
We also give the observation on the estimation of convolution term.
Lemma 3.1
Assume that (F1)–(F4) hold. Then there exists \({\bar{n}}\in {\mathbb {N}}\) such that
Proof
By (F4), we may choose \(\varepsilon >0\) small and \(t_{\varepsilon }>0\) such that
Using (3.3), we have
There are three cases to distinguish. Without mentioning, all inequalities hold for large \(n\in {\mathbb {N}}\) in the rest of the Lemma.
Case i\(t\in \left[ 0,\sqrt{(4-2\alpha -\mu )\pi /2\alpha _0}\right] \). Then by (F3), we have
which yields the existence of \({\bar{n}}\in {\mathbb {N}}\) satisfying (3.4).
Case ii \(t\in \left[ \sqrt{(4-2\beta -\mu )\pi /2\alpha _0},\sqrt{(4-2\alpha -\mu )(1+\varepsilon )\pi /\alpha _0}\right] \). In this case, \(tw_n(x)\ge t_{\varepsilon }\) for \(x\in B_{d/n}\) and \(n\in {\mathbb {N}}\) large. Then it follows from (3.2) and (3.5) that
then
Choosing \(t_n>0\) be such that \(\varphi _n'(t_n)=0\), then we have
Let
then it follows that
and
Using (3.6) and (3.7), we are led to
where from (3.1), we know that
thus we have
Then we deduce that (3.4) holds for some \({\bar{n}}\in {\mathbb {N}}\).
Case iv\(t\in (\sqrt{(4-2\beta -\mu )(1+\varepsilon )\pi /\alpha _0},\infty )\). Since \(tw_n(x)\ge t_{\varepsilon }\) for \(x\in B_{d/n}\) and \({\bar{n}}\in {\mathbb {N}}\) large, we deduce from (3.2) that
where we have used the fact that the function
is decreasing on \(t\in \left( \sqrt{\frac{(4-2\alpha -\mu )\pi }{\alpha _0}(1+\varepsilon )}, +\infty \right) \) for large n. In fact,
Assume that \(s_n\ge \sqrt{\frac{(4-2\alpha -\mu )\pi }{\alpha _0}(1+\varepsilon )}\) such that \(\varphi _n'(s_n)=0\) for large n. Then
which yields
This implies that \(\lim _{n\rightarrow \infty }s_n^2=\frac{(4-2\alpha -\mu )\pi }{\alpha _0}\), a contradiction. So \(\varphi _n(t)\) is decreasing for large n when \(t\in \left( \sqrt{\frac{(4-2\alpha -\mu )\pi }{\alpha _0}(1+\varepsilon )}, +\infty \right) \). Thus (3.4) holds for some \({\bar{n}}\in {\mathbb {N}}\). Till now, we have completed the proof. \(\square \)
Lemma 3.2
Assume that (F1)–(F5) hold. Then \({c}(a)<2\pi /\alpha _0\) for any \(a>0\).
Proof
Let \(w_{{\bar{n}}}\) be given by Lemma 3.1. Since \(\Vert \nabla t(w_{{\bar{n}}})_{t}\Vert _2^2=t^2\Vert \nabla w_{{\bar{n}}}\Vert _2^2\), we know that there exists \(t_w>0\) small enough and \(T_w>0\) large enough such that \(\Vert \nabla t_w(w_{{\bar{n}}})_{t_w}\Vert _2^2\le K(a)\) and \(\Phi (T_w(w_{{\bar{n}}})_{T_w})<0\) by (2.7). Set
Then \(g_0\in \Gamma _a\). Jointly with the definition of c(a), we have \({c}(a)<2\pi /\alpha _0\) for any \(a>0\). \(\square \)
Lemma 3.3
Assume that f satisfies (F1)–(F3) and (F4\('\)) with \(\gamma >\gamma ^*(a)\). Then \({c}(a)<[(4-2\mu -\beta )\pi ]/(2\alpha _0)\), where \(\gamma ^*(a)\) is given by (1.4).
Proof
Since
we can choose \(v_n\in {\mathcal {S}}_a\) such that
Note that
Let \(g_n'(t_n)=0\), then one has
It is easy to see that \(g_n(t)\le g_n(t_n)\) for all \(t>0\). We define that
Then it follows from (3.8) and (3.9) that
Since \(p>\frac{6-2\alpha -\mu }{2}\) and \(\gamma >\gamma ^*(a)\), then there exists \(\epsilon _0>0\) such that
which together with (3.11) imply that
which implies that there exists \({\bar{n}}\in {\mathbb {N}}\) large enough such that
Replacing \(w_{{\bar{n}}}\) by \(v_{{\bar{n}}}\) in the proof of Lemma 3.2, we can get \({c}(a)\le \max _{t>0}\Phi (t(v_{{\bar{n}}})_{t})\) for any \(\gamma >\gamma ^*(a)\). From this and (3.12), we derived the desired conclusion, and so the proof is completed. \(\square \)
4 Restore the Compactness
Let us first establish the following two convergence results which contribute to the final proof.
Lemma 4.1
Assume that \(\alpha <\mu \), \(u_{n}\rightharpoonup u\) in \( H_r^1({{\mathbb {R}}} ^2) \) and
for some constant \({\mathcal {K}}>0\). Then for every \(\varphi \in {\mathcal {C}}_{0}^{\infty }({{\mathbb {R}}} ^{2})\), we have
Proof
By the Fatou’s Lemma we have
Take \(\Omega =\textrm{supp}\ \varphi \), for any given \(\varepsilon >0\), let \(M_{\varepsilon }:={\mathcal {K}}\Vert \varphi \Vert _{\infty }\varepsilon ^{-1}\), then it follows that for n large enough,
and similarly
Since \(|f(u_{n})|\chi _{|u_{n}|\le M_{\varepsilon }}\rightarrow |f(u)|\chi _{|u|\le M_{\varepsilon }}\) a.e. in \(\Omega {\setminus } D_{\varepsilon }\), where \(D_{\varepsilon }=\{x\in \Omega :|u(x)|=M_{\varepsilon }\}\), and
the Lebesgue dominated convergence theorem leads to
Here, we choose \(K_{\varepsilon }>t_{0}\) such that
and
With the help of Lemma 2.2, we have
then from (F5) and Proposition 1.2, one has
For any \(x\in {{\mathbb {R}}} ^{2}\), define \(\zeta _{n}(x)\) and \({\bar{\zeta }}(x)\) as follows:
and
Let us first point out some relationships here. For fixed \(x\in {{\mathbb {R}}} ^{2}\), we consider the term
When \(x\in {{\mathbb {R}}} ^{2}/B_{2R}(0)\), \(y\in B_{R}(x)\), thus \(|x-y|<|y|\). Select \(p_{1}\) such that \((\mu +\alpha )p_{1}<2\), and thus we have,
When \(x\in B_{2R}(0)\), one has
That is
Choosing q such that \(\alpha q<2<\mu q\), one has
Then from (2.4), we have
which implies that for any \(x\in {{\mathbb {R}}} ^{2}\), we have \(\zeta _{n}(x)\rightarrow {\bar{\zeta }}(x)\). Similarly, by choosing suitable \(p_2\) and \(p_3\), then for any \(x\in {{\mathbb {R}}} ^{2}\), we know that
It follows that
By \(\alpha <2\), it is easy to verify that \(\frac{1}{|x|^{\alpha }}\in L_{\textrm{loc}}^{1}({{\mathbb {R}}} ^{2})\). Therefore, together with \(\zeta _{n}(x)\rightarrow {\bar{\zeta }}(x)\) and the Lebesgue dominated convergence theorem, we have
From the arguments above all and by the arbitrariness of \(\varepsilon >0\), we can conclude this Lemma. \(\square \)
Lemma 4.2
Assume that \(\{u_{n}\}\) is bounded in \(H_r^1({{\mathbb {R}}} ^2)\), \(u_{n}\rightharpoonup u\) in \(H_r^1({{\mathbb {R}}} ^2)\) and
Then we have
Proof
In view of \(u_n\rightharpoonup {\bar{u}}\ \ \text{ in }\ H_r^1({{\mathbb {R}}} ^2)\), we know \(u_{n}\rightarrow u\) in \(L^{q}({{\mathbb {R}}} ^{2})\) with \(q> 2\). By [35, Theorem A.1], there exists \(g\in L^{q}({{\mathbb {R}}} ^{2})\) such that
For any given \(\varepsilon \in (0,M_{0}/t_{0})\), it follows from (F5) that
Similarly, one has
Now, we can choose \(R_{\varepsilon }>0\) such that
and
Let C be the constant in (4.1) and choose \(K \ge \max \{CM_{0}/\varepsilon ,t_{0}\}\) such that
By (F5), one has
By (F2), we know that there exist \(C>0\) and \({\tilde{q}}>\frac{(4-2\alpha -\mu ) }{2}\) such that for \(|t|\le K\),
Thus we have
which leads to
On the other hand,
It remains to prove that as \(n\rightarrow \infty \),
Combining (4.2) with (4.3), we can see that
In order to prove (4.5), it remains to verify that as \(n\rightarrow +\infty \) there holds
Indeed, it can be easily verified that as \(n\rightarrow \infty \),
From (4.4), we have
From [10, Theorem 4.9], there exists \({\mathcal {F}}\in L^{1}({{\mathbb {R}}} ^{2})\) such that up to a subsequence, still denoted by \(\{u_{n}\}\), for each \(n\in {\mathbb {N}}\), we have
Using the Lebesgue dominated convergence theorem, we can conclude this Lemma. \(\square \)
Lemma 4.3
Assume that (F1)–(F3) hold. If there exist \(u\in H^1({{\mathbb {R}}} ^2)\) and \(\lambda \in {{\mathbb {R}}} \) such that
then \(J(u)=0\), where J is defined by (2.2).
The proof of lemma is standard, so we omit it. Hereafter, we are ready to prove the Theorem 1.4.
Proof of Theorem 1.4:
Let \({\mathcal {S}}_a^r={\mathcal {S}}_a\cap H_r^1({{\mathbb {R}}} ^2)\). In the same way as Lemmas 2.8 and 3.2, we can deduce that for any \(a>0\), there exists a bounded sequence \(\{u_n\}\subset {\mathcal {S}}_a^r\) such that
and
where \(\Gamma _{r,a}=\{g\in {\mathcal {C}}([0,1],{\mathcal {S}}_a^r):\Vert \nabla g(0)\Vert _2^2\le K(a), \Phi (g(1))<0\}\) and K(a) is given in Lemma 2.3. Then there exists \({\bar{u}}\in H_r^1({{\mathbb {R}}} ^2)\) such that, passing to a subsequence,
Arguing similar as Lemma 2.8, we know that
Then, it follows that Lemma 4.2 holds, that is
Next, we claim that \({\bar{u}}\ne 0\). Otherwise, we suppose that \(u_n\rightharpoonup 0\) in \(H_r^1({{\mathbb {R}}} ^2)\). Then one has
Choosing \(q\in (1,2)\) be such that
using (F1), we get
and using (ii) of Lemma 1.1, we get
Thus,
Similarly, we have
Then it follows from (4.9) and (4.9) that
which is a contradiction due to \({c}_r(a)>0\) for any \(a>0\). This shows that \({\bar{u}}\ne 0\) as claimed.
By (4.6) and the boundedness of the sequence \(\{u_n\}\), one can easily verify that there exist a bounded sequence \(\{\lambda _n\}\subset {{\mathbb {R}}} \) and \({\bar{\lambda }}\) such that, up to a subsequence,
and
Again, in view of the conclusion of Lemma 4.1, we can see that
Hereafter, the only thing we need to verify is that \(\Vert {\bar{u}}\Vert _2^2=a\), and next our goal is to prove that \(u_n\rightarrow {\bar{u}}\) in \(H_r^1({{\mathbb {R}}} ^2)\). Note that (4.11) yields
and
By (4.13) minus \(J(u_n)\rightarrow 0\), and using (4.10) and (4.7), we have
which, together with \(F(t)>0\) for \(t\ne 0\), yields \({\bar{\lambda }}> 0\). By (4.13) and (4.14), we have
Next, we claim that \(\int _{{{\mathbb {R}}} ^{2}}\left( \int _{{{\mathbb {R}}} ^2}\frac{F(u_{n}(y))}{|y|^{\alpha }|x-y|^{\mu }}\textrm{d}y\right) \frac{f(u_n)(u_n-{\bar{u}})}{|x|^{\alpha }}\textrm{d}x=o(1)\). By (4.12) and the Lemma 4.3, we have \(J({\bar{u}})=0\). This, jointly with (F3) implies
Thus,
Since \(0<{c}_r(a)<\frac{(4-2\mu -\beta )\pi }{2\alpha _0}\) for any \(a>0\), similarly as in (4.8), it follows that there exists \({\bar{\varepsilon }}>0\) such that
Nothing that \(q/(q-1)>2\), by using the Hölder inequality, we have
where \(\frac{4\alpha _0(1+{\bar{\varepsilon }})^2q^2 \Vert \nabla (u_n- {\bar{u}})\Vert _2^2 }{4-2\alpha -\mu }< 4\pi (1-3{\bar{\varepsilon }})^{2}(1+{\bar{\varepsilon }})^2q^{2}<4\pi \). Moreover, we have
and similarly one has
Till now, we have finished the Claim. Then, one has
which, together with \(u_n\rightharpoonup {\bar{u}}\) in \(H_r^1({{\mathbb {R}}} ^2)\) and \(\lambda \rightarrow {\bar{\lambda }}>0\), implies that \(u_n\rightarrow {\bar{u}}\) in \(H_r^1({{\mathbb {R}}} ^2)\). Next, using Palais’ principle of symmetric criticality [35], the above function \({\bar{u}}\in H_r^1({{\mathbb {R}}} ^2){\setminus }\{0\}\) is in fact a radial solution of \(({\mathcal {P}}_{a})\) in \(H^1({{\mathbb {R}}} ^2)\), and so the proof of Theorem 1.4 is completed.
Remark 4.4
We omit here the proof of Theorem 1.6, since the difference between of Theorems 1.4 and 1.6 has been presented in Sect. 3. To prove Theorem 1.6, we just need to replace Lemma 3.2 by Lemma 3.3.
Data availability
Data sharing is not applicable to this article as no new data were created or analyzed in this study.
References
Alves, C.O., Shen, L.J.: Critical Schrödinger equations with Stein–Weiss convolution parts in \({\mathbb{R} }^{2}\). J. Differ. Equ. 344, 352–404 (2023)
Alves, C.O., Souto, M.A.S.: Multiplicity of positive solutions for a class of problems with exponential critical growth in \({\mathbb{R} }^2\). J. Differ. Equ. 244, 1502–1520 (2008)
Alves, C.O., Ji, C., Miyagaki, O.H.: Normalized solutions for a Schrödinger equation with critical growth in \({\cal{R} }^N\). Calc. Var. Part. Differ. Equ. 61, 24 (2022)
Arora, R., Giacomoni, J., Mukherjee, T., Sreenadh, K.: Polyharmonic Kirchhoff problems involving exponential non-linearity of Choquard type with singular weights. Nonlinear Anal. 196, 111779 (2020)
Bartsch, T., De Valeriola, S.: Normalized solutions of nonlinear Schrödinger equations. Arch. Math. 100, 75–83 (2013)
Bartsch, T., Soave, N.: A natural constraint approach to normalized solutions of nonlinear Schrödinger equations and systems. J. Funct. Anal. 272, 4998–5037 (2017)
Bartsch, T., Soave, N.: Correction to A natural constraint approach to normalized solutions on nonlinear Schrödinger equations and systems. J. Funct. Anal. 272, 4998–5037 (2017)
Bieganowski, B., Mederski, J.: Normalized ground states of the nonlinear Schrödinger equation with at least mass critical growth. J. Funct. Anal. 280, 108989 (2021)
Biswas, R., Goyal, S., Sreenadh, K.: Quasilinear Schrödinger equations with Stein–Weiss type convolution and critical exponential nonlinearity in \({\cal{R} }^{N}\). http://arxiv.org/abs/2202.07611 (2022)
Brezis, H.: Functional Analysis, Sobolev Spaces and Partial Differential Equations, Universitext. Springer, New York (2011)
Cao, D.M.: Nontrivial solution of semilinear elliptic equation with critical exponent in \({\mathbb{R} }^2\). Commun. Part. Differ. Equ. 17, 407–435 (1992)
Cazenave, T., Lions, P.L.: Orbital stability of standing waves for some nonlinear Schrödinger equations. Commun. Math. Phys. 85, 549–561 (1982)
Chen, S.T., Tang, X.H.: Axially symmetric solutions for the planar Schrödinger–Poisson system with critical exponential growth. J. Differ. Equ. 269, 9144–9174 (2020)
Chen, S.T., Tang, X.H.: On the planar Schrödinger equation with indefinite linear part and critical growth nonlinearity. Calc. Var. Part. Differ. Equ. 60, 27 (2021)
Chen, S.T., Tang, X.H.: New approaches for Schrödinger equations with prescribed mass: the Sobolev subcritical case and the Sobolev critical case with mixed dispersion. http://arxiv.org/abs/2210.14503 (2022)
Chen, S.T., Rădulescu, V.D., Tang, X.H., Yuan, S.: Normalized solutions for Schrödinger equations with critical exponential growth in \({\cal{R} }^2\) (Under Review)
de Figueiredo, D.G., Miyagaki, O.H., Ruf, B.: Elliptic equations in \({\mathbb{R} }^2\) with nonlinearities in the critical growth range. Calc. Var. Part. Differ. Equ. 3, 139–153 (1995)
Du, L.L., Gao, F.S., Yang, M.B.: On elliptic equations with Stein–Weiss type convolution parts. Math. Z. 301, 2185–2225 (2022)
Ghoussoub, N.: Duality and Perturbation Methods in Critical Point Theory. Cambridge Tracts in Mathematics and Mathematical Physics, vol. 107. Cambridge University Press, Cambridge (1993)
Hardy, G.H., Littlewood, J.E.: Some properties of fractional integrals. I. Math. Z. 27, 565–606 (1928)
Jeanjean, L.: Existence of solutions with prescribed norm for semilinear elliptic equations. Nonlinear Anal. 28, 1633–1659 (1997)
Jeanjean, L., Le, T.T.: Multiple normalized solutions for a Sobolev critical Schrödinger equation. Math. Ann. (2021). https://doi.org/10.1007/s00208-021-02228-0
Jeanjean, L., Lu, S.S.: A mass supercritical problem revisited. Calc. Var. Part. Differ. Equ. 59, 43 (2020)
Jeanjean, L., Jendrej, J., Le, T.T., Visciglia, N.: Orbital stability of ground states for a Sobolev critical Schrödinger equation. J. Math. Pures Appl. 164, 158–179 (2022)
Mattner, L.: Strict definiteness of integrals via complete monotonicity of derivatives. Trans. Am. Math. Soc. 349, 3321–3342 (1997)
Moser, J.: A sharp form of an inequality by N. Trudinger. Indiana Univ. Math. J. 20, 1077–1092 (1970/1971)
Qin, D., Tang, X.: On the planar Choquard equation with indefinite potential and critical exponential growth. J. Differ. Equ. 285, 40–98 (2021)
Soave, N.: Normalized ground states for the NLS equation with combined nonlinearities: the Sobolev critical case. J. Funct. Anal. 279, 108610 (2020)
Sobolev, S.L.: On a theorem of functional analysis. Am. Math. Soc. Transl. Ser. 34, 39–68 (1938)
Stein, E., Weiss, G.: Fractional integrals on n-dimensional Euclidean space. J. Math. Mech. 7, 503–514 (1958)
Stuart, C.A.: Bifurcation in \(L^{p}({\mathbb{R} }^{N})\) for a semilinear elliptic equation. Proc. Lond. Math. Soc. 57, 511–541 (1988)
Trudinger, N.S.: On imbeddings into Orlicz spaces and some applications. J. Math. Mech. 17, 473–483 (1967)
Wei, J., Wu, Y.: Normalized solutions for Schrödinger equations with critical Sobolev exponent and mixed nonlinearities. J. Funct. Anal. 283, 109574 (2022)
Weinstein, M.I.: Nonlinear Schrödinger equations and sharp interpolation estimates. Commun. Math. Phys. 87, 567–576 (1982/1983)
Willem, M.: Minimax Theorems, Progress in Nonlinear Differential Equations and Their Applications. Birkhäuser Inc., Boston (1996)
Yadava, S.L.: Multiplicity results for semilinear elliptic equations in a bounded domain of \({\mathbb{R} }^2\) involving critical exponents. Ann. Scuola Norm. Sup. Pisa Cl. Sci. 17, 481–504 (1990)
Yang, M.B., Zhou, X.M.: On a coupled Schrödinger system with Stein–Weiss type convolution part. J. Geom. Anal. 31, 10263–10303 (2021)
Yang, M.B., Rădulescu, V.D., Zhou, X.M.: Critical Stein–Weiss elliptic systems: symmetry, regularity and asymptotic properties of solutions. Calc. Var. Part. Differ. Equ. 61, 109 (2022)
Zhang, Y.P., Tang, X.H.: Large perturbations of a magnetic system with Stein–Weiss convolution nonlinearity. J. Geom. Anal. 32, 27 (2022)
Zhang, Y.P., Tang, X.H., Rădulescu, V.D.: Anisotropic Choquard problems with Stein–Weiss potential: nonlinear patterns and stationary waves. C. R. Math. Acad. Sci. Paris 359, 959–968 (2021)
Zhang, J., Zhang, J. J., Zhong, X.X.: Normalized solutions to Kirchhoff type equations with a critical growth nonlinearity. http://arxiv.org/abs/2210.12911 (2022)
Funding
This work is partially supported by the National Natural Science Foundation of China (Nos. 11971485 and 12001542) and the Hunan Provincial Natural Science Foundation (No. 2021JJ40703).
Author information
Authors and Affiliations
Contributions
Dr. SY and Prof. SC wrote the main manuscript text. Prof. XT checked the paper. All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.
Consent for Publication
All of the authors agree to the submission of this paper.
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
Yuan, S., Tang, X. & Chen, S. Normalized Solutions for Schrödinger Equations with Stein–Weiss Potential of Critical Exponential Growth. J Geom Anal 33, 341 (2023). https://doi.org/10.1007/s12220-023-01396-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s12220-023-01396-6
Keywords
- Nonlinear Schrödinger equation
- Stein–Weiss potential
- Critical exponential growth
- Trudinger–Moser inequality