Solutions concentrating around the saddle points of the potential for critical Schrödinger equations (Q5963602)

In this paper, the authors consider the standing wave of the nonlinear Schrödinger equation \[ -\varepsilon^2 \Delta u + V(x)u = f(u)\,,\; u\in H^1(\mathbb R^N)\,,\quad N\geq3\,. \eqno{(1.1)} \] The authors are concerned with the existence of spikes of problem (1.1) around the saddle points of function \(V\) which satisfies \begin{itemize} \item[(V0)] \(0 < \alpha_1 \leq V(x) \leq \alpha_2\) for all \(x \in\mathbb R^N\). \end{itemize} Furthermore, they assume that V satisfies one of the following conditions. \begin{itemize} \item[(V1)] \(V(0) = 1\), \(V\) is \(C^1\) in a neighborhood of \(0\) and \(0\) is an isolated local maximum of \(V\); \item[(V2)] \(V(0) = 1\), \(V\) is \(C^2\) in a neighborhood of \(0\) and \(0\) is a non-degenerate saddle critical point of \(V\); \item[(V3)] \(V(0) = 1\), \(V\) is \(C^{N-1}\) in a neighborhood of \(0\) and \(0\) is an isolated critical point of \(V\) and there exists a vector space \(E\) such that: \begin{itemize} \item[(a)] \(V |_E\) has a local maximum at \(0\), \item[(b)] \(V |_{E^\perp}\) has a local minimum at \(0\). \end{itemize} \end{itemize} The following hypotheses are imposed on \(f\). \begin{itemize} \item[(\(f_1\))] \(f \in C^1(\mathbb R^+,\mathbb R^+)\) and \(\lim_{t\to 0} f(t)/t = 0\); \item[(\(f_2\))] \(\lim_{t\to+\infty} \frac{f(t)}{t^{\frac{N+2}{N-2}}} = 1\) where \(N \geq 3\); \item[(\(f_3\))] There exist \(C > 0\) and \(p < 2^*\) such that \(f (t) \geq t^{\frac{N+2}{N-2}} + Ct^{p-1}\) for \(t \geq 0\). \end{itemize} (AR) There exists \(\mu > 2\) such that \(s f (s) \geq \mu F(s) > 0\) for \(s > 0\), where \(F(s) =\int_0^s f (t)\, dt\). The main theorem of this paper reads as {Theorem 1.1} Let \(p > 2\) if \(N\geq 4\) or \(p > 4\) if \(N = 3\). Suppose that (\(f_1\)),(\(f_2\)),(\(f_3\)) and (AR) hold and that \(V\) satisfies (V0) and one of (V1)--(V3). Then for sufficiently small \(\varepsilon > 0\), (1.1) admits a positive solution \(v_\varepsilon\) , which satisfies \begin{itemize} \item[(i)] there exists a maximum point \(x_\varepsilon \in \mathbb{R}^N\) of \(v_\varepsilon\) such that \(w_\varepsilon (x)\equiv v_\varepsilon (\varepsilon x + x_\varepsilon )\) converges (up to a subsequence) uniformly to a least energy solution of \[ - \Delta u + u = f (u)\,,\quad u > 0\,, \quad u \in H^1(\mathbb{R}^N ); \eqno{(1.2)} \] \item[(ii)] \(v_\varepsilon (x) \leq C \text{exp}(-\frac{c}{\varepsilon} |x - x_\varepsilon |)\) for some \(c,C > 0\). \end{itemize} To prove this theorem, the authors use variational techniques and construct a solution \(u_\varepsilon\) which concentrates around the saddle points of \(V\) as \(\varepsilon \to 0\).