Singlepeaks for a magnetic Schrodinger

equation with critical growth

Sara Barile
Dipartimento di Matematica, Universita di Bari
via Orabona 4, I-70125 Bari
Silvia Cingolani
Politecnico di Bari
via Amendola 126/B, I-70126 Bari
Simone Secchi
Dipartimento di Matematica, Universita di Milano
via Saldini 50, I-20133 Milano.
June 30, 2006

Abstract
We prove existence results of complex-valued solutions for a semilinear Schrodinger
equation with critical growth under the perturbation of an external electromagnetic
field. Solutions are found via an abstract perturbation result in critical point theory,
developed in [1, 2, 5].

AMS Subject classification: 35J10, 35J20, 35Q55

1 Introduction

This paper deals with some classes of elliptic equations which are perturbation of the time-
dependent nonlinear Schrodinger equation

= ~2 ||p1 (1)
t
Email: s.barile@dm.uniba.it. Supported by MIUR, national project Variational and topological meth-

ods in the study of nonlinear phenomena.

Email: cingolan@poliba.it. Supported by MIUR, national project Variational and topological methods

in the study of nonlinear phenomena.

Email: secchi@mat.unimi.it. Supported by MIUR, national project Variational methods and nonlinear

differential equations.

1
under the effect of a magnetic field B and an electric field E whose sources are small in
L sense. Precisely we will study the existence of wave functions : RN R C satisfying
the nonlinear Schrodinger equation
 2
~
= A (x) + W (x) ||p1 (2)
t i
where A (x) and W (x) are respectively a magnetic potential and an electric one, depending
on a positive small parameter > 0. In the work, we assume that A (x) = A(x),
W (x) = V0 + V (x), being A : RN RN and V0 R, V : RN R, [1, 2].

2
On the right hand side of (2) the operator ~i A denotes the formal scalar product
of the operator ~i A by itself, i.e.
 2
~ 2~ ~
A (x) := ~2 A + |A |2 div A
i i i

being i2 = 1, ~ the Planck constant.

This model arises in several branches of physics, e.g. in the description of the Bose
Einstein condensates and in nonlinear optics (see [7, 11, 23, 25]).
If A is seen as the 1form
XN
A= Aj dxi ,
j=1

then X
B = dA = Bjk dxj dxk , where Bjk = j Ak k Aj ,
j<k

represents the external magnetic field having source in A (cf. [30]), while E = V (x)
is the electric field. The fixed ~ > 0 the spectral theory of the operator has been studied in
detail, particularly by Avron, Herbt, Simon [7] and Helffer [21, 22].
1
The search of standing waves of the type (t, x) = eiV0 ~ t u (x) leads to find a
complex-valued solution u : RN C of the semilinear Schrodinger equation
 2
~
A(x) u + V (x)u = |u|p1 u in RN . (3)
i
From a mathematical viewpoint, this equation has been studied in several papers in the
subcritical case 1 < p < (N + 2)/(N 2). In the pioneering paper [20], M. Esteban
and P.L. Lions proved the existence of standing wave solutions to (2) in the case V = 1
identically, > 0 fixed, by a constrained minimization. Recently variational techniques
are been employed to study equation (3) in the semiclassical limit (~ 0+ ). We refer to
[15, 17, 24, 27]. Recent results on multi-bumps solutions are obtained in [12] for bounded
vector potentials and in [19] without any L restriction on |A|.
In the critical case p = (N + 2)/(N 2), we mention the paper [6] by Arioli and Szulkin
where the potentials A and V are  assumed to be periodic,
 > 0 fixed. The existence of a


2
solution is proved whenever 0 / i A + V . We also cite the recent paper [13] by
Chabrowski and Skulzin, dealing with entire solutions of (3).

2
 In the present paper we are concerned with the critical case p = (N + 2)/(N 2), but
V and A are not in general periodic potentials.
When the problem is nonmagnetic and static, i.e. A = 0, V = 0, and ~ = 1 then problem
(3) reduces to the equation
N +2
u = u N 2 , u D1,2 (RN , C). (4)
In Section 2 we prove that the least energy solutions to (4) are given by the functions
z = ei z, (x), where
N
( 2 1) N 2
z, (x) = N N 2 , N = (N (N 2)) 4
(5)
(2 + |x |2 ) 2

and they correspond to the extremals of the Sobolev imbedding D1,2 (RN , C) L2 (RN , C)
(cf. Lemma 2.1).
The perturbation of (4) due to the action of an external magnetic potential A leads us
to seek for complexvalued solutions. In general, the lack of compactness due to the critical
growth of the nonlinear term produces several difficulties in facing the problem by global
variational methods. We will attack (3) by means of a perturbation method in Critical Point
Theory, see [1, 4, 5], and we prove the existence of a solution u to (3) that is close for
small enough to a solution of (4). After an appropriate finite dimensional reduction, we find
N +1
that stable
 critical points on ]0, +[R of a suitable functional correspond to points
on Z = ei z, : S 1 , > 0, RN from which there bifurcate solutions to (3) for
6= 0. If V changes its sign, we find at least two solutions to (3). The main result of the
paper is Theorem 5.2, stated in Section 5.
We quote the papers [3, 14, 16], dealing with perturbed semilinear equations with critical
growth without magnetic potential A.
Remark 1.1. It is apparent that the compact group S 1 acts on the space of solutions to
(3). For simplicity, we will talk about solutions, rather than orbits of solutions.
Notation. The complex conjugate of any number z C will be denoted by z. The real
part of a number z C will be denoted by Re z. The ordinary inner product between
two vectors a, b RN will be denoted by a b. We use the Landau symbols. For example
O() is a generic function such that lim sup0 O() < , and o() is a function such that
 2 N
lim o() 1,2 N

(R , C) | RN |u|2 dx < , with a
R
= 0. We will denote D (R , C) = u L
0
similar definition for D1,2 (RN , R).

2 The limiting problem

Before proceeding, we recall some known facts about a couple of auxiliary problems. Recall
that 2 = 2N/(N 2).
() The problem (
u = |u|2 2 u in RN
(6)
u D1,2 (RN , R).

3
possesses a smooth manifold of least-energy solutions
n (N 2)
o
Ze = z, = 2 z0 ( x ) | > 0, RN (7)

where
1 N 2
z0 (x) = N N 2 , N = (N (N 2)) 4
. (8)
(1 + |x|2 ) 2

Explicitly,
N
(N 2) 1 ( 2 1)
z, (x) = N 2 = N . (9)
   N 2 (2 + |x |2 )
N 2
 x 2 2 2

1+ 

These solutions are critical points of the Euler functional

Z Z
1 2 1
f0 (u) =
e |u| |u+ |2 dx, (10)
2 RN 2 RN

defined on D1,2 (RN , R) E, and the following nondegeneracy property holds:

ker fe000 (z, ) = Tz, Z

e for all > 0, RN . (11)

() Similarly, f0 C 2 (D1,2 (RN , C)) possesses a finitedimensional manifold Z of least-

energy critical points, given by

Z = ei z, : S 1 , > 0, RN = S 1 (0, +) RN .

(12)

More precisely, following the ideas of [24] and [27], we give the following characterization.
Lemma 2.1. Any least-energy solution to the problem
(
u = |u|2 2 u in RN
(13)
u D1,2 (RN , C)

is of the form u = ei z, for some suitable [0, 2], > 0 and RN .

Proof. It is convenient to divide the proof into two steps.
Step 1: Let z0 = U the least energy solution associated to the energy functional (10) on
the manifold
 Z Z 
2 2
M0,r = v D1,2 (RN , R) \ {0} | |v| dx = |v| dx .
RN RN

It is well-known that z0 = U is radially symmetric and unique (up to translation and

dilation) positive solution to the equation (6). Let b0,r = br = fe0 (U ) = fe0 (z0 ). In a similar
way, we define the class
 Z Z 
2 2
M0,c = v E \ {0} | |v| dx = |v| dx
RN RN

4
and denote by b0,c = bc = f0 (v) on M0,c . Let R, RN , > 0 ,e
v (x) = z, (x) positive
i i e = ei z, is a
solution to (6) and U = e ve = e z, (i.e. z, = |U (x)|). It results that U
e e
non-trivial least energy solution for b0,c = f0 (v) with v M0,c .

Step 2: The following facts hold:

(i) b0,c = b0,r ;

(ii) If Uc = U
e is a least energy solution of problem (13), then

for a.e. x RN .


||Uc |(x)| = |Uc (x)| and Re iUc (x)Uc (x) = 0

(iii) There exist R and a least energy solution ur : RN R of problem (6) with

Uc (x) = ei ur (x) for a.e. x RN

or, equivalently, the least energy solution Uc for b0,c is the following

Uc (x) = ei ur (x) = ei z, (x) for a.e. x RN .

Observe that
b0,r = min fe0 (v) and b0,c = min f0 (v)
v M0,r v M0,c

where M0,r and M0,c are the real and complex Nehari manifolds for fe0 and f0 ,
n o
M0,r = v D1,2 (RN , R) \ {0} | fe00 (v)[v] = 0
 Z Z 
2 2
= v D1,2 (RN , R) \ {0} | |v| dx = |v| dx
RN RN

and

M0,c = {v E \ {0} | f00 (v)[v] = 0}

 Z Z 
2 2
= v E \ {0} | |v| dx = |v| dx
RN RN

So (i) is equivalent to

b0,r = min fe0 (v) = fe0 (ur )

v M0,r

b0,c = min f0 (v) = f0 (Uc )

v M0,c

Proof of (i)(iii). Let u E be given. For the sake of convenience, we introduce the
functionals
Z
2
T (u) = |u| dx
RN
Z
1 2
P (u) = |u| dx
2 RN

5
(resp. Te(u) and Pe(u) as u D1,2 (RN , R)) such that f0 (u) = 12 T (u) P (u) as u E (resp.
fe0 (u) = 12 Te(u) Pe(u) as u D1,2 (RN , R)).
Consider the following minimization problems
n o
r = min Te(u) | u D1,2 (RN , R), Pe(u) = 1
c = min {T (u) | u E, P (u) = 1}

Note that, obviously, there holds c r . If we denote by u the Schwarz symmetric rear-
rangement (see [8]) of the positive real valued function |u| D1,2 (RN , R), then, Cavalieris
principle yields Z Z
2 2
|u | dx = |u| dx
RN RN

which entails Pe(u ) = P (|u|). Moreover, by the Polya-Szego inequality, we have

Z Z Z
2 2 2
Te(u ) = |u | dx ||u|| dx |u| dx = T (u)
RN RN RN

where the second inequality follows from the following diamagnetic inequality
Z Z
2 2
||u|| dx |D |u|| dx for all u HA,V

RN RN

with D = i A and A = 0. Therefore, one can compute c by minimizing over the

subclass of positive, radially symmetric and radially decreasing functions u D1,2 (RN , R).
As a consequence, we have r c . In conclusion, r = c . Observe now that
n o
b0,r = min fe0 (u) | u D1,2 (RN , R) \ {0} is a solution to (6) ,
b0,c = min {f0 (u) | u E \ {0} is a solution to (13)} .

The above inequalities hold since any nontrivial real (resp. complex) solution of (6) (resp.
(13)) belongs to M0,r (resp. M0,c ) and, conversely, any solution of b0,r (resp. b0,c ) produces
a nontrivial solution of (6) (resp. (13)). Moreover, it follows from an easy adaptation of [8,
Th. 3] that b0,r = r as well as b0,c = c . In conclusion, there holds

b0,r = r = b0,c = c

which proves (i).

To prove (ii), let Uc : RN C be a least energy solution to problem (13) and assume by
contradiction that
LN x RN : ||Uc || < |Uc |


> 0
where LN is the Lebesgue measure in RN . Then, we would get Pe(|Uc |) = P (Uc ) and
Z Z
1 2 1 2
P (|Uc |) =
e |Uc | dx = |Uc | dx = P (Uc )
2 RN 2 RN
and Z Z
2 2
r ||Uc || dx < |Uc | dx = c
RN RN

6
which is a contradiction. The second assertion in (ii) follows by direct computations. Indeed,
a.e. in RN , we have

||Uc || = |Uc | if and only if Re Uc ( Im Uc ) = Im Uc (Re Uc ) .

If this last condition holds, in turn, a.e. in RN , we have

Uc Uc = Re Uc (Re Uc ) + Im Uc (Im Uc )

which implies the desired assertion.

Finally, the representation formula of (iii) Uc (x) = ei ur (x) is an immediate consequence
of (ii), since one obtains Uc = ei |Uc | for some R.
Remark 2.2. For the readers convenience, we write here the second derivative of f0 at any
z Z:
Z Z

hf000 (z)v, wiE = Re v w dx Re |z|2 2 vw dx
RN RN
Z


Re(2 2) |z|2 4 Re(zv)zw dx. (14)
RN

In particular, f000 (z) can be identified with a compact perturbation of the identity operator.
We now come to the most delicate requirement of the perturbation method.
Lemma 2.3. For each z = ei z, Z, there holds

Tz Z = ker f000 (z) for all z Z,

where
ei z, ei z, ei z, ei z,
 
Tei z, Z = spanR ,..., , , = iei z, . (15)
1 N
Proof. The inclusion Tz Z ker f000 (z) is always true, see [1]. Conversely, we prove that for
any ker f000 (z) there exist numbers a1 , . . . , aN , b, d R such that
N
X ei z, ei z,
= aj +b + diei z, . (16)
j=1
j

If we can prove the following representation formul, then (16) will follow.
N
X z, z,
Re(ei ) = aj +b (17)
j=1
j

Im(ei ) = dz, . (18)

We will use a well-known result for the scalar case:

 
00 z, z, z,
ker f0 (z, ) Tz, Z = spanR
e e ,..., ,
1 N

7
Step 1: proof of (17). We wish to prove that Re(ei ) ker fe000 (z, ). Recall that
ker f000 (ei z, ), so
hf000 (ei z, ), i = 0 for all E. (19)
Select = ei v, with v C0 (RN , R).
Z
0 = hf000 (ei z, ), vei i = Re (ei )v
Z Z
2 2
(2 2) |z, | i
Re(e )v |z, |2 2 Re(ei )v
R N R N
Z Z

2
= (Re(ei )v (2 1) |z, |2
Re(ei )v = hfe000 (z, ) Re(ei ), vi.
RN RN

This implies that

Re(ei ) ker ff00
0 (z, ) Tz, Z
e

from which it follows

N
X z, z,
Re(ei ) = aj +b
j=1
j

for some real constants a1 , . . . , aN and b.

Step 2: proof of (18). Test (19) on = iei w E with w : RN R. We get
Z Z

0 = hf000 (ei z, ), iei wi = Re (iei ) w Re |z, |2 2 (iei )w
RN RN
i i
[being Re(ie ) = Im(e )]
Z Z

i
= (Im(e )) w |z, |2 2 Im(ei )w
N N
ZR Z R h i

= (Im(ei )) w |z, |2 2 Im(ei ) w. (20)
RN RN +

We can take = 1 and = 0, otherwise we perform the change of variable x 7 x + .

From (20) we get that u := Im(ei ) satisfies the equation

N (N 2)
u = u in D1,2 (RN , R). (21)
(1 + |x|2 )2

We will study this linear equation by an inverse stereographic projections onto the sphere
S N . Precisely, for each point S N , denote by x its corresponding point under the
stereographic projection from S N to RN , sending the north pole on S N to . That is,
2xi
suppose = (1 , 2 , . . . , N +1 ) is a point in S N , x = (x1 , . . . , xN ), then i = 1+|x| 2 for
2
1 i N ; N +1 = |x| 1
|x|2 +1 .
Recall that, on a Riemannian manifold (M, g), the conformal Laplacian is defined by

N 2
Lg = g + Sg ,
4(N 1)

8
where g is the LaplaceBeltrami operator on M and Sg is the scalar curvature of (M, g).
It is known that N +2
Lg ((u)) = N 2 L (u),
 (N 2)/2
2
where is the euclidean metric of RN , (x) = 1+|x| 2 and

u((x))
: D1,2 (RN ) H 1 (S n ), (u)(x) =
((x))
is an isomorphism between H 1 (S n ) and E := D1,2 (RN ). Therefore, if U = (u), then (21)
changes into the equation

N 2 N (N 2)
g0 U + Sg0 U = U, (22)
4(N 1) 4

where g0 is the standard riemannian metric on S N , and

Sg0 = N (N 1)

is the constant scalar curvature of (S N , g0 ). As a consequence, (22) implies that

g0 U = 0,

i.e. U is an eigenfunction of g0 corresponding to the eigenvalue = 0. But the point

spectrum of g0 is completely known (see [9, 10]), consisting of the numbers

k = k(k + N 1), k = 0, 1, 2, . . .

with associated eigenspaces of dimension

(N + k 2)! (N + 2k 1)
.
k! (N 1)!

Hence we deduce that k = 0, and U belongs to an eigenspace of dimension 1. Since z, is

a solution to (21), we conclude that there exists d R such that

Im(ei ) = dz, .

This completes the proof.

3 The functional framework

In the variational framework of the problem, solutions to (3) can be found as critical points
of the energy functional f : E R defined by
 2

Z 

Z Z
1 2 1
|u|2 dx,

f (u) =  A(x) u dx +
 V (x)|u| dx (23)
2 RN  i 2 RN 2 RN

9
on the real Hilbert space
 Z 
1,2 N 2 N 2
E=D (R , C) = v L (R , C) | |v| dx < (24)
RN

endowed with the inner product

Z
hu, viE = Re u v dx. (25)
RN

We shall assume throughout the paper that

(N) N > 4,
(A1) A C 1 (RN , RN ) L (RN , RN ) Lr (RN , RN ) with 1 < r < N
(A2) div A LN/2 (RN , R),
(V) V C(RN , RN ) L (RN , R) Ls (RN , R), with 1 < s < N/2.
The functional f is well defined on E. Indeed,
Z   2 Z
 u
Z Z
2 2 2 2

 A(x) u  = |u| + |A| |u| Re A u,
RN
 i 
RN RN RN i

and all the integrals are finite by virtue of (A1). Moreover, f C 2 (E, R).
In this section, we perform a finitedimensional reduction on f according to the methods
of [1, 5]. Roughly speaking, since the unperturbed problem (i.e. (3) with = 0) has a whole
C 2 manifold of critical points, we can deform this manifold is a suitable manner and get a
finitedimensional natural constraint for the EulerLagrange functional associated to (3).
As a consequence, we can find solutions to (3) in correspondence to (stable) critical points
of an auxiliary map called the Melnikov function in finite dimension.
Now we focus on the case = 2, as in the other cases [1, 2[ the magnetic potential
A no longer affects the finite-dimensional reduction (see Remark (5.3)).
So that we can write the functional f as

f (u) = f0 (u) + G1 (u) + 2 G2 (u) (26)

where Z Z
1 1
f0 (u) = |u|2

|u|2 , (27)
2RN 2 RN
Z Z Z
1 1 1
G1 (u) = Re u A u, G2 (u) = |A|2 |u|2 + V (x)|u|2 . (28)
i RN 2 RN 2 RN
We can now use the arguments of [1, 5] to build a natural constraint for the functional f .
Theorem 3.1. Given R > 0 and BR = {u E : ||u|| R}, there exist 0 and a smooth
function w = w(z, ) = w(ei z, , ) = w(, , , ), w(z, ) : M = Z BR (0 , 0 ) E
such that
1. w(z, 0) = 0 for all z Z BR

10
 2. w(z, ) is orthogonal to Tz Z, for all (z, ) M . Equivalently w(z, ) (Tz Z)

3. the manifold Z = {z + w(z, ) : (z, ) M } is a natural constraint for f0 : if u Z

and f0 |Z = 0, then f0 (u) = 0.

For future reference let us recall that w satisfies 2. above and Df (z + w) Tz Z,

namely f000 (z)[w] + G01 (z) + o() Tz Z. As a consequence, if G01 (z)Tz Z (to be proved as
Lemma 3.2), one finds
w(, z) = Lz G01 (z) + o(), (29)
where Lz denotes the inverse of the restriction to (Tz Z) of f000 (z).
Lemma 3.2. G1 (z) = 0 for all z Z.

Proof.
z
Z
A(x) z dx = z = ei z,
 
G1 (z) = Re
N i
ZR
z,
= Re ei A(x) ei z, dx =
N i
ZR
z,
= Re A(x) z, dx = 0.
RN i

Hence we cannot hope to apply directly the tools contained in [1], since the Melnikov
function would vanish identically. However, following [4], we can find a slightly implicit
Melnikov function whose stable critical points produce critical points of f .

Lemma 3.3. Let : Z R be defined by setting

1
(z) = G2 (z) (Lz G01 (z), G01 (z)) . (30)
2
Then we have
f (z + w(, z)) = f0 (z) + 2 (z) + o(2 ). (31)

Proof. Since G1 |Z 0, then G01 (z) (Tz Z) . Then one finds

f (z + w(, z)) = f0 (z + w(, z)) + G1 (z + w(, z)) + 2 G2 (z + w(, z))

1
= f0 (z) + f000 (z)[w, w] + G1 (z) + G01 (z)[w] + 2 G2 (z) + o(2 ).
2
Using (29) and Lemma 3.2 the lemma follows.

Remark 3.4. We notice that = G2 (z) + 1

2 (G01 (z), ), where z stands for ei z, and
= lim0 w .

11
Remark 3.5. By the definition of z Z, it results: (z) = (ei z, ) = (, , ). In the
sequel, we will write freely (, , ) (, ) since is -invariant. Indeed, it is easy to
check that G2 is -invariant. In fact, by the definition of G2 (z) and z = ei z, , it results:
Z Z
i 1 2 2 1
G2 (, , ) = G2 (e z, ) = |A(x)| |z, | dx + |V (x)||z, |2 dx G2 (, ).
2 2
It remains to prove that hG01 (z), i is -invariant. We will show that = ei (, ) with
(, ) C independent on which immediately gives
Z

0 i 1 i
e z, A(x)ei (, ) dx

G1 (e z, ), = Re
i
Z
1
Re , A(x)z, dx = hG01 (z, ), (, )i .
i
w(,z)
We begin to recall that = lim0+ , where w(, z) is such that

f0 (ei z, + w(, , )) Tei z, Z.

By (15), this condition means that

N
X z, z,
f0 (ei z, + w(, , )) = ai ei + bei + dei iz, , (32)
i=1
i

with a1 , . . . , aN , b, d, R.
Let w(, , ) = ei w e D1,2 (RN , C). Testing (32) by ei v(x) with v
e with w
1,2 N
D (R , C), we derive that z, + w e is a solution of an equation independently on . Thus,
also we is independent on and it can be denoted as w(, e ). Set (, ) = lim0+ w(,)
e
,
i
we deduce that = e (, ).

4 Asymptotic study of
In order to find critical points of it is convenient to study the behavior of as 0 and
as + || . Our goal is to show:

Proposition 4.1. can be extended smoothly to the hyperplane (0, ) R RN by set-
ting
(0, ) = 0. (33)
Moreover there results
(, ) 0, as + || +. (34)
The proof of this Proposition is rather technical, so we split it into several lemmas in
which we will use the formulation of = G2 (z) + 12 (G01 (z), ), where = lim0 w .
Lemma 4.2. Under assumption (A1) there holds
Z
1
lim+ |A(x)|2 |z, |2 dx = 0. (35)
0 2 RN

12
Proof. Let z = ei z, Z. Then
Z
1
H2 (z) = |A(x)|2 |z, |2 dx (36)
2 RN
2
2 ! 2N

 2

Z
1 (N 2) x
|A(x)|2 N 2
 
= 1 +   dx
2 RN 

2N |A(x)|2
Z
= dx
2(N 2) RN 1 +  x 2
N 2
 

x
Using the change of variable y = , or x = y + , we can write

2N
Z
1
H2 (z) = |A(y + )|2 N dy
2(N 2) RN (1 + |y|2 )N 2
2N 2 |A(y + )|2
Z
= 2 N 2
dy
2 RN (1 + |y| )

and using the hypothesis (A1)

Z
2 1
H2 (z) 2 CN kAk dy, (37)
RN (1 + |y|2 )N 2

the lemma follows.

The proof of the following Lemma is similar and thus omitted.

Lemma 4.3. Under assumption (V) there holds

Z
1
lim+ V (x)|z, |2 dx = 0. (38)
0 2 RN

Lemma 4.4. There holds

lim hG01 (z), i = 0. (39)
0+

Proof. We write
hG01 (z), iE = 1 + 2 ,
where
z
Z
1 = Re A(x) dx (40)
N i
ZR

2 = Re A(x)z dx. (41)
RN i

It is convenient to introduce (y) by setting

N
(y) = , (y) = 2 1 (y + )

13
 (N 2)
Using the expression of z = ei 2 z0 ( x ) and the change of variable x = y + we
can write:
 
x
Z
1 (N 2)
1 = Re x ei 2 z0 A(x)(x) dx
RN i
Z
1 i N
= Re e y z0 (y) 2 A(y + )(y + ) dy
RN i
Z
1 i
= Re e y z0 (y) A(y + ) (y) dy
RN i

and
 
x
Z
1 (N 2)
2 = Re x (x) A(x)ei 2 z0 dx
RN i
Z
1 (N 2)
= Re y (y + )1 A(y + )ei 2 N z0 (y) dy
N i
ZR
1 N
= Re (y + )1 A(y + )ei 2 +1 z0 (y) dy
R N i
Z
1
= Re (y) A(y + )ei z0 (y) dy.
R N i
Now the conclusion follows easily from the next lemma.
Lemma 4.5. As 0+ ,
, 0 strongly in E. (42)
Proof. For all v E, due to the divergence theorem, we have
z v
Z Z
0
hG1 (z), viE = Re A(x)v dx Re A(x)z dx
R N i R N i
N
z
Z Z
1 X v
= Re A(x)v dx Re Aj (x)z dx
RN i RN i j=1 xj

N
z
Z Z
1X
= Re A(x)v dx + Re v (Aj z) dx
RN i RN i j=1 xj

z
Z Z Z
1 1
= Re A(x)v dx + Re v div A z dx + Re vA z dx
R N i R N i R N i
z
Z Z
1
= 2 Re A(x)v dx Re div A zv dx
RN i RN i

where the last integral is finite by assumption (A2) and

Z Z

(f000 (z)w, , v) = Re w, v dx Re |z|2 2 w, v dx
RN RN
Z

Re (2 2)|z|2 4 Re(zw, )zv dx. (43)
RN

14
We know that w, = Lei z, G01 (ei z, ) + o(), and hence

hf000 (z), , viE = hG01 (z), viE , v E (44)

w,
where , = lim0  . This implies that , solves
Z Z Z
2 2
Re , vdx Re |z| , v dx Re (2 2)|z|2 4 Re(z, )zv dx
RN RN R N
Z Z
1 1
= 2 Re z A(x)vdx + Re div A zv dx.
RN i RN i
N (N 2)
Multiplying by 2 1 and using the expression of z = ei 2 z0 ( x
), we get

   2 2
x 
Z Z
N N
1 2
2 1 , (x)vdx

Re 2x , (x)vdx Re  z0

RN RN 
   2 4    
x  x
Z
N 4  i N +1 N
1

Re (2 2)  z0 Re e 2 z0 2 , (x)
RN 
 
N x
ei 2 +1 z0 vdx

   
x x
Z Z
1 i 1 i
= 2 Re e x z0 A(x)vdx + Re div A e z0 vdx.
RN i RN i
N
Using the expression of ( x
)=
2 1
, (x), we have

     2 2  
x x  x
Z Z
x 2  z0

Re vdx Re vdx

   2 4    !
x  x x
Z
N 4  i N +1

Re (2 2)
 z0  Re e 2 z0



 
i N +1 x
e 2 z0 vdx

   
x x
Z Z
1 i 1 i
= 2 Re e x z0 A(x)vdx + Re div A(x) e z0 vdx.
i i
then, the change of variable x = y + yields
Z Z
2 2
Re 2 y (y)y v(y + )N dy Re N 2 | z0 (y) | (y)v(y + )dy
Z
2 4
 N

Re (2 2)N 4 | z0 (y)| Re ei 2( 2 +1) z0 (y) (y) ei z0 (y)v(y + )N dy
Z
1 i
= 2 Re e y z0 (y) A(y + ))v(y + )N 1 dy
i
Z
1
+ Re div A(y + ) ei z0 (y)v(y + )N dy.
i

15
Replacing x = y and dividing by N 2 , it results
Z Z
2 2
Re x (x)x v(x + ) dx Re | z0 (x) | (x)v(x + ) dx
R N R N
Z
2 4
(2 2) | z0 (x)| Re ei z0 (x) (x) ei z0 (x)v(x + ) dx


Re
RN
Z
1 i
= 2 Re e x z0 (x) A(x + ))v(x + ) dx
RN i
Z
2 1
+ Re divy A(x + ) ei z0 (x)v(x + ) dx.
R N i
This means that, if we write , (x) = x + ,
Z
f000 (ei z0 ) , v


, = k, v ,

for all test function v, in particular that

f000 (ei z0 ) = k,
where
2 i 1
k, (x) =e x z0 (x) A(x + ) + 2 ei divy A(x + ) z0 (x).
i i
We conclude that is a solution of
(x) = Lei z0 k, (x) (45)
Our assumptions on A (i.e. (A1) and (A2)) imply immediately that
k, 0 in E as 0. (46)
From the continuity of Lei z0 we deduce that
lim = lim Lei z0 k, = 0. (47)
0+ 0+

This completes the proof of the Lemma.

Lemma 4.6. Under assumption (A1), there holds
lim H2 (, ) = 0,
+||+

where H2 is defined in (36).

Proof. Firstly, assume that (0, +) and + || +. We notice that
(N 2)
 
x
Z
H2 (, ) = |A(x)|2 z02 dx
2 RN
(N 2)
 
x
Z
= |A(x)|2 z02 dx
2 ||
|x| 2
(N 2)
 
x
Z
+ |A(x)|2 z02 dx.
2 |x|> 2
||

16
Moreover,

(N 2)
 
x
Z
|A(x)|2 z02 dx
2 |x|
||
2

(N 2) ||N
 
x
||A||2 N N sup z02
2 2 |x| ||
2

(N 2) N 2 2(N 2)
|| kN
= ||A||2 N sup N 2
2 2N |x|
|| [2 + |x |2 ]
2

(N 2) ||N 2
kN
||A||2 N N sup h iN 2
2 2 |x| || 2
2 2 + | |x| || |
(N 2) ||N 2
kN
||A||2 N N h iN 2 ,
2 2 ||2
2 + 4

where N is the measure of S N 1 = x RN : |x| = 1 . Since N > 4, we infer



2
kN ||N
0 as || +.
2 N 2
h i
2 + ||4

Finally, we deduce
(N 2)
 
x
Z
|A(x)|2 z02 dx 0
2 |x|
||
2

as and || +.
On the other hand, we have

(N 2)
 
x
Z
|A(x)|2 z02 dx
2 ||
|x|> 2
(N 2)
 
x
Z
||A||2 z02 dx
2 ||
|x|> 2
N (N 2)
Z
= ||A||2 z02 (x)dx.
2 |x+|> 2
||

Since z02 L1 (RN ), we deduce that

2
Z
||A||2 z02 (x)dx 0
2 |x+|>
||
2

as and || +, and thus

(N 2)
 
x
Z
|A(x)|2 z02 dx 0
2 |x|>
||
2

17
as and || +.
Finally, we can conclude that H2 (, ) 0 as and || +.
Conversely, assume that +. After a suitable change of variable, it results
2
Z
H2 (, ) = |A(y + )|2 |z0 (y)|2 dy.
2 RN

By assumption (A1), we can fix 1 < r < N2 such that A2 Lr (RN ). Moreover, let be
r
s = r1 . It is immediate to check that 2s > 2 and then z02s L1 (RN ). By (A1) and Holder
inequality, we deduce that
Z
|A(y + )|2 |z0 (y)|2 dy
RN
Z  r1 Z  1s
|A(y + )|2r dy |z0 (y)|2s dy
RN RN
Z  r1 Z  1s
N 2r 2s
r |A(y)| dy |z0 (y)| dy .
RN RN

As a consequence, by the above inequality, we infer for small

2
Z
G2 (, ) = |A(y + )|2 |z0 (y)|2 dy
2 RN
Z  r1 Z  1s
2 N 2r 2s
r |A(y)| dy |z0 (y)| dy .
RN RN

Now, we notice that r < N2 implies 2 Nr < 0 and thus by the above inequality we can
conclude that G2 (, ) tends to 0 as +.
Arguing as before we can deduce the following result.
Lemma 4.7. Under assumption (V), there holds
Z
lim V (x)|z, (x)|2 dx = 0.
+||+ RN

In order to describe the behavior of the term hG01 (z), iE as + || +, we need the
following lemma.
Lemma 4.8. There is a constant CN > 0 such that

k kE CN for all > 0 and for all RN . (48)

Proof. We know that for all > 0 and all z Z

w(, z) = Lz G01 (z) + o()

so that
w(, z)
= lim = Lz G01 (z)
0

18
and
kkE kLz k kG01 (z)k .
We claim that kLz k is bounded above by a constant independent of and . Indeed:

kLz k = sup kLz k = sup |hLz , i|

kk=1 kk=1
kk=1
Z Z
 2 2
= sup  Re |z, |
kk=1 RN RN
kk=1
Z 
2 4
(2 2)

|z, | Re(z, ) Re(z, )
RN
Z Z Z 
2 2 2 2
||   + (2 2)
     
sup ||   + Re |z, | |z, | ||  
kk=1 RN RN RN
kk=1
Z Z 
2 2
||   + (2 1)
   
sup |z, | ||  
kk=1 RN RN
kk=1
Z 1/2 Z 1/2 Z (2 2)/2
2 2 2
sup || || + (2 1) |z, |
kk=1 RN RN RN
kk=1
Z 1/2 Z 1/2
2 2
|| ||
RN RN

We observe that
Z 1/2   2 !1/2

Z
2 (N 2)
2
 x 
|z, | =  z0 
RN RN
 
Z 1/2

(N 2)
2 N
= 2 |z0 (y)| = kz0 kL2 .
RN

Hence
(2 2)
 
kLz k sup 1 + (2 1)k z0 k L2 kkL2 kkL2
kk=1
kk=1
(2 2)
 
sup 1 + (2 1)CN
0
kz0 kE kkE kkE
kk=1
kk=1
(2 2)
1 + (2 1)CN
0
kz0 kE 1
CN
1
where CN is a constant independent from and . At this point it results:
1
kk CN kG01 (z)k

19
and we have to evaluate kG01 (z)k :

kG01 (z)k = sup |hG01 (z), i|

kk=1
 
z
Z Z
 
= sup  Re A(x) dx Re A(x)z dx 
kk=1 RN i RN i
Z Z 
sup |z, | |A(x)| || dx + || |A(x)| |z, | dx
kk=1 RN RN
00
kAkLN sup (kz0 kE kkE CN )
kk=1
00 2
kAkLN kz0 kE CN CN
2
with CN independent from and .
Finally,
1 2
k k CN CN CN
with CN independent from and and the lemma is proved.

Remark 4.9. It is easy to check that k k = kk.

Lemma 4.10. There holds

lim hG01 (z), iE = 0.
+||+

Proof. Firstly, assume that (0, +) and + || +. We can write

hG01 (z), iE = 1 + 2

where
z
Z
1 = Re A(x) dx (49)
N i
ZR

2 = Re A(x)z dx. (50)
RN i
(N 2)
Using the expression of z = ei 2 z0 ( x ) and by assumption (A1) and the Holder
inequality we have:
 
x
Z
1 (N 2)
1 = Re x ei 2 z0 A(x)(x) dx
RN i
Z   2 !1/2 Z 1/2
(N 2)
2
 x 
2
 x z0

 dx
 |A(x)| || dx
RN RN

Z   2 !1/2
(N 2)
2
 x 
kAkLN (RN ) kkL2 (RN )  x z0

 dx
RN

20
We notice that
  2   2
x z0 x  dx = x z0 x  dx
Z Z
   
RN
 
|x| ||/2
 
  2
x z0 x  dx
Z
 
+
|x| > ||/2
 

and   2 |x |2
x z0 x  = 2(2N ) (2 N )2 2N .
 
N
(2 + |x |2 )
2
Moreover, setting CN := (2 N )2 2N ,
2 2
||N
   
x z0 x  dx  x z0 x 
Z
   
N N sup
|x| ||/2
  2 |x| ||/2
 

||N |x |2
= N sup 2(2N ) (2 N )2 2N N
2N |x| ||/2 (2 + |x |2 )
||N 2
CN |x |2
= 2(2N ) N sup N
2N |x| ||/2 (2 + |x |2 )
2
||N 2
CN ( |x| + || )
2(2N ) N sup N
2N

|x| ||/2 2
2 + | |x| || |
9 ||N CN2
||2
N N
4 2 (2 + ||2 /4)N

where N is the measure of S N 1


= x RN : |x| = 1 . From N > 4, we infer
2
CN ||N +2
N
0 as || +.
(2 + ||2 /4)

Finally, we deduce
  2
x z0 x  dx 0
Z
 
as (0, +), || +.
|x| ||/2
 

On the other hand, we have

  2
x z0 x  dx N 2
Z Z
  2
|x z0 (x)| dx.
|x| > ||/2
 
|x+| > ||/2

Since |x z0 |2 L1 (RN ), we deduce that

Z
2
N 2 |x z0 (x)| dx 0
|x+| > ||/2

21
as (0, +) and || + and thus
  2
x z0 x  dx 0
Z
 
|x| > ||/2
 

and
  2 !1/2

Z

(N 2)  x 
1 2 kAkLN (RN ) kkL2 (RN )  x z0  dx 0
RN
 

as (0, +) and || +.

As regards 2 we know that

 
x
Z
1 (N 2)
2 = Re x (x) A(x)ei 2 z0 dx
RN i
Z 1/ Z    2 !1/2
(N 2)
x
2
 
| x (x) A(x)| dx  z0  dx
RN RN
 
!1/2
Z    2
(N 2)
2
 x 
k kE k A kLN  z0

 dx
RN

with = 2N/(N + 2). We notice that

   2    2
 z0 x  dx =  z0 x  dx
Z Z
   
RN
 
|x| ||/2
 
   2
 z0 x  dx.
Z
 
+
|x| >||/2
 

Moreover,
 2 2
||N
   
 z0 x  dx  z0 x 
Z
   
N sup
|x| ||/2
  2N |x| ||/2
 

N
|| 2N 2N
= N sup 2N 2N N
2N |x| ||/2 (2 + |x |2 )

||N 2N
2N N sup N
2N

|x| ||/2 2
2 + | |x| || |

||N
2N 2N
N N
2 (2 + ||2 /4)N

where N is the measure of S N 1


= x RN : |x| = 1 . From N > 4, we infer

2N ||N
N
0 as || +.
(2 + ||2 /4)

22
Finally, we deduce
   2
 z0 x  dx 0
Z
 
|x| ||/2
 

as (0, +) and || +.
On the other hand, we have
   2
 z0 x  dx N
Z Z
  2
| z0 (x) | dx.
|x| > ||/2
 
|x+| > ||/2

Since | z0 |2 L1 (RN ), we deduce that
Z
2
N | z0 (x) | dx 0
|x+| > ||/2

as (0, +) and || + and thus

   2
 z0 x  dx 0
Z
 
|x| > ||/2
 

and !1/2
   2

Z
(N 2)
2
 x 
2 kAkLN (RN ) kkE  z0  dx 0
RN
 

as (0, +) and || +.

Conversely, assume that +. Now it is convenient to write

hG01 (z), iE = 1 + 2

where
ei
Z
1 = Re y z0 (y) A(y + ) (y) dy
RN i
and Z
1
2 = Re y (y) A(y + )ei z0 (y) dy.
RN i
The Holder inequality implies that
Z 1/

1 k kL2 ( y z0 (y) A(y + )) dy
RN

where 1/2 +1/ = 1 so = 2N/(N +2). By assumptions (A1), we can fix r (1, (N +2)/2)
such that A Lr (RN ). Moreover, let s = r/(r 1). It is immediate to check that s > 2

23
and then |y z0 |s L1 (RN ). By (A1) and the Holder inequality, we deduce that:

Z 1/

( y z0 (y) A(y + )) dy
RN
Z 1/s  Z 1/r
s r
( y z0 (y)) dy ( A(y + )) dy
RN RN
Z 1/r
N r
r ky z0 (y)kLs ( A(y + )) dy
RN

As a consequence, by the above inequality, we infer for small:

Z 1/r
N r
1 1 r ky z0 (y)kLs ( A(y + )) dy k kL2
RN
N
0
1 r CN kz0 kE kAkLr k kE

Analogously,
Z 1/  Z 1/2
2
2 ( |y (y)| |A(y + )| dy ) |z0 (y)| dy
RN RN
Z 1/s  Z 1/r
N s r
1 r k z0 kL2 ( y (y)) dy ( A(y)) dy
RN RN
N
1 r 00
CN (k z0 kE k A kLr k kE )
N
Since = 2N/(N + 2), we deduce 1 r < 0. The conclusion follows immediately from
Lemma 4.8.

Proposition 4.11. Assume that there exists RN with V () 6= 0. Then

Z
(, ) 1
lim = V () |z0 |2 . (51)
0+ 2 2

In particular, is a non-constant map.

Proof. If V () 6= 0 for some RN , we can immediately check that (, ) is not identically

zero. More precisely, we prove that for every RN there holds
Z
(, ) 1
lim = V () |z0 |2 . (52)
0+ 2 2 RN

Indeed, after a suitable change of variable,

Z Z
G2 (z, ) 1 2 2 1
lim = lim (|A(y + )| |z0 (y)| + V (y + )|z0 (y)|2 dy
0+ 2 0 + 2 RN 2 RN
Z Z
1 1
= |A()|2 |z0 (y)|2 dy + V () |z0 (y)|2 dy. (53)
2 R N 2 R N

24
To complete the proof of (52), we need to study lim 0+ 21 2 hG01 (z, ), , i.
In Lemma 4.5, we have showed that

0 i
G1 (e z, ), , = hf000 (z, ei ), , , i = hf000 (z0 ei ), , , i


where , ( x
)=
N/21
, (x) and f000 (z0 ei ), = k, , where

2 i 2 i
k, (y) = e y z0 (y) A(y + ) + e divy A(y + )z0 (y).
i i
As 0+ , we have k, k , where

2 i
k (x) := e y z0 (y) A().
i
Lz0 k,
,
Let us define (x) = lim0+ = lim0+ . We have that

2 i
f000 (z0 ei ) = e x z0 (y) A(). (54)
i
Setting g (x) = ei (x), we have that for any D1,2 (RN , R)
Z
2 i
hf000 (z0 ei )ei g , ei i = Re e y z0 (y) A()ei dx = 0.
i

This means that for any D1,2 (RN , R)

0 = hf000 (z0 ei )ei g , ei i

Z Z

= Re (ei g ) (ei ) Re |z0 |2 2 ei g ei
Z

Re(2 2) |z0 |2 4 Re(ei z0 ei g )ei z0 ei
Z Z

= Re (g ) Re |z0 |2 2 g
Z

Re(2 2) |z0 |2 4 Re(z0 g )z0

Z Z

= (Re g ) |z0 |2 2 Re g
Z

(2 2) |z0 |2 4 Re g z0 2

= hf000 (z0 ) Re g , i.


It follows that Re g = 0 as , Tei z, Z . Therefore (x) = iei r (x) with
r D1,2 (RN , R). Now we test (54) against functions of the type iei (x), D1,2 (RN , R).

25
 It results:
Z
2 i
f000 (z0 ei ) , iei w


Re e x z0 (x) A()iei w =
i
Z Z

= Re r w Re |z0 |2 2 r w
Z

Re(2 2) |z0 |2 4 Re(iz0 r )z0 iw

or equivalently
Z Z Z

Re r w Re |z0 |2 2 r w = Re 2x z0 (x) A()w.

We deduce that r satisfies the equation

2
r (x) |z0 |2 r (x) = 2z0 A(). (55)

We notice that the function u(x) = z0 (x)A()x solves the equation (55), as u = z0 A()
x + z0 (A() x) + 2z0 (A() x) = z0 A() x + 2z0 A().

Since iz0 (x)(A()|x)ei belongs to (Tei z0 Z) , we deduce that (x) = iei z0 (x)A() x
and thus
1 hG01 (z, ), , i
Z
1 i
lim+ = Re e y z0 (y) A() iei z0 A() x dx
0 2 2 R N i
Z
= y z0 (y) A()z0 A() x dx.
RN

Since we have
Z Z Z
y z0 (y) A()z0 A() x dx = y z0 (y) A()z0 A() x dx |A()|2 z02 dx,
RN RN RN

we conclude that
1 hG01 (z, ), , i
Z Z
1
lim = y z0 (y) A()z0 A() x dx = |A()|2 z02 dx. (56)
0+ 2 2 RN 2 RN

Therefore we have that

Z
(, ) 1 1 1
lim+ = lim+ 2 (G2 (, ) + hG01 (z, ), , i) = V () |z0 |2 .
0 2 0 2 2 RN

Remark 4.12. The presence of a non-trivial potential V is crucial in the previous Propo-
sition. Otherwise, from (53) and (56) we would simply get that lim0+ (,)
2 = 0, and
might still be a constant function. Hence V is in competition with A. It would be interest-
ing to investigate the case in which V = 0 identically. We conjecture that some additional
assumptions on the shape of A should be made.

26
5 Proof of the main result
We recall the following abstract theorem from [4]. See also [5].

Theorem 5.1. Assume that there exist a set A Z with compact closure and z0 A such
that
(z0 ) < inf (z) (resp. (z0 ) > sup (z)).
z A z A

Then, for small enough, f has at least a critical point u Z such that

f0 (z) + 2 inf + o(2 ) f (u ) f0 (z) + 2 sup + o(2 )

A A

(resp. f0 (z) + 2 inf + o(2 ) f (u ) f0 (z) + 2 sup + o(2 )).

A A

Furthermore, up to a subsequence, there exists z A such that un z in E as n 0.

We can finally prove our main existence result for equation (3). According to Remark
1.1, we will use the term solution rather than the more precise S 1 orbit of solutions.

Theorem 5.2. Retain assumptions (N), (A12), (V). Assume that V () 6= 0 for some
RN . Then, there exists 0 > 0 such that for all (0, 0 ) equation (3) possesses at
least one solution u E. If V is a changing sign function, then there exists two solutions
of equation (3).

Proof. Under our assumptions, the Melnikov function , extended across the hyperplane
{ = 0} by reflection, is not constant and possesses at least a critical point (either a minimum
or a maximum point). We can therefore invoke Theorem 5.1 to conclude that there exists
at least one solution u to (3), provided is small enough. If there exist points i RN ,
i = 1, 2, such that V (1 )V (2 ) < 0, then it follows from the previous Proposition that must
change sign near { = 0}. In particular, it must have both a minimum and a maximum.
Hence there exist two different solutions to (3).

Remark 5.3. Consider equation (3). It is clear that our main theorem still applies for any
[1, 2). Indeed, in the expansion (31), the lowest order term in is
Z

V z 2 dx,
RN

and consequently the magnetic potential A no longer affects the finite-dimensional reduction.
In some sense, we have treated with the more all the details the worst situation in the
range 1 2.

Acknowledgement
The authors would like to thank V. Felli for some useful discussions about the proof of
Lemma 3.2.

27
References
[1] A. Ambrosetti, M. Badiale, Variational perturbative methods and bifurcation of bound
states from the essential spectrum, Proc. Royal Soc. Edinburgh 128 A (1998), 1131
1161.

[2] A. Ambrosetti, M. Badiale, S. Cingolani, Semiclassical states of nonlinear Schrodinger

equations, Arch. Ration. Mech. Anal. 140 (1997), 285300.

[3] A. Ambrosetti, J.G. Azorero, I. Peral, Perturbation of u + u(N +2)/(N 2) = 0, the

Scalar Curvature Problem in RN , and Related Topics, Journal of Functional Analysis
165 (1999), 117149.

[4] A. Ambrosetti, A. Malchiodi, A multiplicity result for the Yamabe problem on S n ,

Journal of Functional Analysis 168 (1999), 529561.

[5] A. Ambrosetti, A. Malchiodi, Perturbation methods and semilinear elliptic problems

on Rn , Progress in Mathematics 240, Birkhauser Verlag, 2006.

[6] G. Arioli, A. Szulkin, A semilinear Schrodinger equations in the presence of a magnetic

field, Arch. Ration. Mech. Anal. 170 (2003), 277295.

[7] J. Avron, I. Herbst, B. Simon, Schrodinger operators with magnetic fields I, Duke Math.
J. 45 (1978), 847883.

[8] H. Berestycki, P. L. Lions, Nonlinear scalar field equations I and II, Arch. Ration. Mech.
Anal. 82 (1983) 313345 and 347375.

[9] F. A. Berezin, M. A. Shubin, The Schrodinger equation, Mathematics and its Appli-
cations (Soviet Series), 66. Kluwer Academic Publishers Group, 1991

[10] M. Berger, P. Gauduchon, E. Mazet, Le spectre dune variete riemannienne, Lecture

Notes in Mathematics, Vol. 194, Springer-Verlag, 1971.

[11] A. Bernoff, P. Stenberg, Onset of superconductivity in decreasing fields or general do-

mains, J. Math. Phys. 39 (1998), 12721284.

[12] D. Cao, Z. Tang, Existence and uniqueness of multibump bound states of nonlinear
Schrodinger equations with electromagnetic fields, J. Differential Equations 222 (2006),
381424.

[13] J. Chabrowski, A. Szulkin, On the Schrodinger equation involving a critical Sobolev

exponent and magnetic field, Topol. Methods Nonlinear Anal., 25 (2005), 321.

[14] S. Cingolani, Positive solutions to perturbed elliptic problems in RN involving critical

Sobolev exponent, Nonlinear Anal. 48 (2002), 11651178.

[15] S. Cingolani, Semiclassical stationary states of Nonlinear Schrodinger equations with

an external magnetic field, J. Differential Equations 188 (2003), 5279.

28
[16] S. Cingolani, A. Pistoia, Nonexistence of single blow-up solutions for a nonlinear
Schrdinger equation involving critical Sobolev exponent, Z. Angew. Math. Phys. 55
(2004), 201215.
[17] S. Cingolani, S. Secchi, Semiclassical limit for nonlinear Schrodinger equations with
electromagnetic fields, J. Math. Anal. Appl. 275 (2002), 108130.
[18] S. Cingolani, S. Secchi, Semiclassical states for NLS equations with magnetic potentials
having polynomial growths, J. Math. Phys. 46 (2005), 119.
[19] S. Cingolani, S. Secchi, Multipeak solutions for NLS equations with magnetic fields in
semiclassical regime, to appear.
[20] M. Esteban, P.L. Lions, Stationary solutions of nonlinear Schrodinger equations with
an external magnetic field, in PDE and Calculus of Variations, in honor of E. De Giorgi,
Birkhauser, 1990.
[21] B. Helffer, On Spectral Theory for Schrodinger Operators with Magnetic Potentials,
Advanced Studies in Pure Mathematics vol. 23, 113141 (1994).
[22] B. Helffer, Semiclassical analysis for Schrodinger operator with magnetic wells, in Qua-
siclassical methods (J. Rauch, B. Simon Eds.). The IMA Volumes in Mathematics and
its applications vol. 95, SpringerVerlag NewYork 1997.
[23] B. Helffer, A. Morame, Magnetic bottles in connection with superconductivity, J. Func-
tional Anal. 93 A (2001), 604680.
[24] K. Kurata, Existence and semi-classical limit of the least energy solution to a nonlinear
Schrodinger equation with electromagnetic fields, Nonlinear Anal. 41 (2000), 763778.
[25] K. Lu, X.-B. Pan, Surface nucleation of superconductivity in 3-dimensions, J. Differen-
tial Equations 168 (2000), 386452.
[26] M. Reed, B. Simon, Methods of Modern Mathematical Physics, vol.II, Academic
Press, 1975.
[27] S. Secchi, M. Squassina, On the location of spikes for the Schrodinger equations with
electromagnetic field, Commun. Contemp. Math. 7 (2005), 251268.
[28] Z. Shen, Eigenvalue asymptotics and exponential decay of the eigenfunctions for
Schrodinger operators with magnetic fields, Trans. Amer. Math. Soc. 348 (1996), 4465
4488.
[29] B. Simon, Maximal and minimal Schrodinger forms, J. Operator Theory 1 (1979),
3747.
[30] C. Sulem, P.L. Sulem, The Nonlinear Schrodinger Equation, Self-Focusing and Wave
Collapse, Springer 1999.
[31] G. Talenti, Best constant in Sobolev inequality, Ann. Mat. Pura Appl., 110 (1976),
353372.

29