메인 Abstract and Applied Analysis On the location of the peaks of least-energy solutions to semilinear Dirichlet problems with...

# On the location of the peaks of least-energy solutions to semilinear Dirichlet problems with critical growth

이 책이 얼마나 마음에 드셨습니까?
파일의 품질이 어떻습니까?
책의 품질을 평가하시려면 책을 다운로드하시기 바랍니다
다운로드된 파일들의 품질이 어떻습니까?
권:
7
년:
2002
언어:
english
기록:
Abstract and Applied Analysis
DOI:
10.1155/s1085337502206028
파일:
PDF, 1.88 MB
Conversion to is in progress
Conversion to is failed

리뷰를 게시하기에 로그인 또는 가입 하세요
책에 대한 리뷰를 작성하거나 귀하의 독서 경험을 공유할 수 있습니다. 다른 독자들이 귀하가 읽은 책에 대한 의견에 귀를 기울일 것입니다. 개인적으로 책이 마음에 들었거나 그렇지 않았거나 정직하고 상세한 호평은 다른 독자들이 자신에게 적합한 책을 찾는데 도움이 됩니다.
1

년:
2003
언어:
english
파일:
PDF, 2.13 MB
2

### Continuity and general perturbation of the Drazin inverse for closed linear operators

년:
2002
언어:
english
파일:
PDF, 1.86 MB
```ON THE LOCATION OF THE PEAKS OF LEAST-ENERGY
SOLUTIONS TO SEMILINEAR DIRICHLET PROBLEMS
WITH CRITICAL GROWTH
MARCO A. S. SOUTO

We study the location of the peaks of solution for the critical growth problem
∗
−ε2 ∆u + u = f (u) + u2 −1 , u > 0 in Ω, u = 0 on ∂Ω, where Ω is a bounded domain; 2∗ = 2N/(N − 2), N ≥ 3, is the critical Sobolev exponent and f has a behavior like u p , 1 < p < 2∗ − 1.
1. Introduction
In this paper, we will study the location of the peaks of least-energy solution for
the problem
∗

−ε2 ∆u + u = f (u) + u2

−1

in Ω,

u > 0 in Ω,

(1.1)

u = 0 on ∂Ω,
where Ω is a bounded domain in RN , ε > 0, and f is a function satisfying some
subcritical conditions. Here 2∗ = 2N/(N − 2), N ≥ 3, is the critical Sobolev exponent.
By least-energy solution for problem (1.1) we mean a critical point at the
Mountain-Pass level of the associated energy functional
Jε (u) =

1
2




Ω



ε2 |∇u|2 + u2 dz −

 
Ω



F(u) +

1  + 2∗
u
dz,
2∗

(1.2)

(where u+ = max{u,0}), defined on the Hilbert space Ho1 (Ω) endowed with the
norm

u2ε =


Ω



ε2 |∇u|2 + u2 dz.

Abstract and Applied Analysis 7:10 (2002) 547–561
2000 Mathematics Subject Classification: 35A05, 35A15, 35J20
URL: http://dx.doi.org/10.1155/S1085337502206028

(1.3)

548

Location of the peaks of solutions to critical growth problems

The Mountain-Pass level of Jε is defined by




cε = inf max Jε g(t) ,
g ∈Γ 0≤t ≤1

(1.4)

where Γ is the set of all continuous paths joining the origin and a fixed nonzero
element e in Ho1 (Ω), such that e = 0 and Jε (e) ≤ 0. Under suitable hypothesis
(e.g., (f1 ), (f4 ), (f5 ) below), it is not hard to check that cε > 0 does not depend on
the element 0 = v ∈ Ho1 (Ω) and u is a least-energy solution if and only if Jε (u) = c
and Jε (u) = 0, and Jε (u) ≤ Jε (v) for all v = 0 such that Jε (v) = 0.
The existence of least-energy solution of problem (1.1) was given in Brézis
and Nirenberg in [3, Theorem 2.1] (se; e Lemma 2.4 in this paper).
In this paper, we will study some properties of the least-energy solution uε of
problem (1.1) when ε is small. In order to describe these properties, we introduce
the hypotheses on the function f .
Suppose that f : R+ → R+ is a C 1,α function such that
(f1 ) f (0) = f  (0) = 0;
(f2 ) there is q1 ∈ (1,(N + 2)/(N − 2)) such that
lim

s→∞

f (s)
= 0;
sq1

(1.5)

(f3 ) there are q2 ∈ (1,(N + 2)/(N − 2)) and λ > 0 such that
f (s) ≥ λsq2 ,

∀s > 0

(1.6)

(when N = 3, we need q2 > 2, otherwise we require a suﬃciently large λ);
s
(f4 ) if F(s) = o f (t)dt, for some θ ∈ (2, q1 + 1) we have
0 < θF(s) ≤ f (s)s,

∀s > 0;

(1.7)

(f5 ) the function f (s)/s is increasing for s > 0.
Since our interest is on positive solutions we define f (s) = 0, in s ≤ 0.
Now we will state our main result.
Theorem 1.1. Suppose that Ω is a bounded domain in RN ; f satisfies (f1 ), (f2 ),
(f3 ), (f4 ), (f5 ); and let uε be the least-energy solution of (1.1). Then, there is a εo > 0
such that
(i) uε attains only one local maximum at some zε ∈ Ω (hence global maximum), for all ε ∈ (0,εo ];
(ii) uε converges uniformly to zero over compact subsets of Ω \ {zε } as ε → 0;
(iii) dist(zε ,∂Ω) → maxz∈Ω dist(z,∂Ω) .

Marco A. S. Souto 549
This statement is analogous to the one given by Ni and Wei in , in the
subcritical case
−ε2 ∆u + u = h(u),

u = 0,

in Ω,

on ∂Ω,

(1.8)

where h satisfies the following hypothesis:
(i) (f1 ), (f2 ), (f4 ), and (f5 ) hold;
(ii) the global problem
−∆u + u = h(u),

in RN

(1.9)

has a unique positive solution in H 1 (RN );
(iii) this solution is nondegenerate in the sense that
−∆v + v = h (u)v,

in RN

(1.10)

has no nontrivial spherically symmetric solution in L2 (RN ).
In , Ni and Wei also have described the asymptotic profile (in ε) of uε ,
giving a detailed description for ε small. Here in the critical case, the solutions
have the same profile.
In this work we will show that a ground state solution of the critical problem
(1.1) is also solution of a subcritical problem (1.8) by showing that for small ε
we have a uniform bound for the L∞ norm of uε .
The diﬃculty here lies in finding an upper bound for uε L∞ (Ω) by obtaining
a bound for uε in L p (Ω) norm, for all p ≥ 2. In the subcritical case this boundedness is obtained since the family uε is bounded in H 1 (Ω) but this argument
does not work in the critical case. Here, we obtain an L∞ -bound for uε through
the estimate below, which is based on Moser’s iteration technique (see ) and
is essentially due to Brézis and Kato .
Proposition 1.2. Let Λ be an open subset and q ∈ LN/2 (Λ). Suppose that g : Λ ×
R → R is a Caratheodory function satisfying

 

g(x,s) ≤ q(x) + Cg |s|,

∀s ∈ R, x ∈ Λ and for some Cg > 0.

(1.11)

Then, if v ∈ Ho1 (Λ) is such that
−∆v = g(x,v),

in Λ

(1.12)

we have v ∈ L p (Λ) for all 2 ≤ p < ∞. Moreover, there is a positive constant C p =
C(p,Cg , q) such that
v L2∗ (p+1) (Λ) ≤ C p v L2(p+1) (Λ) .

(1.13)

Remark 1.3. The dependence on q of C p can be given uniformly on a family of
functions {qε }ε>0 such that qε converges in LN/2 (see the appendix).

550

Location of the peaks of solutions to critical growth problems

We have organized this paper as follows: the next section contains the proof
of Theorem 1.1. This proof consists in a series of lemmas which show the L∞ bound for uε , where these functions are solutions of a class of subcritical problems (1.8). The third section is an appendix proving Proposition 1.2, for the sake
of completeness.
2. Proof of Theorem 1.1
Before proving Theorem 1.1, let us fix some notation and preliminaries.
Remark 2.1. Throughout this section, we use the equivalent characterization
of cε , which is more adequate to our purposes, given by
cε =

max Jε (tv).

inf

(2.1)

v∈Ho1 (Ω)\{0} t ≥0

(see Willem [13, Theorem 4.2]).
We denote by J : H 1 (RN ) → R the functional given by
1
J(u) = u2 −
2




RN



F(u) +

1  2∗
u+
dx,
2∗

(2.2)

where

2

u  =

RN



|∇u|2 + u2 dx,

(2.3)

associated with the problem
∗

−∆u + u = f (u) + |u|2

−2

u,

in RN .

(2.4)

It is known that under assumptions (f1 ), (f2 ), (f3 ), (f4 ), (f5 ), and (2.4) possesses a ground state solution ω in the level
c = J(ω) =

max J(tv),

inf

(2.5)

v∈H 1 (RN )\{0} t ≥0

(see ).
Remark 2.2. It is easy to check that for each nonzero v in H 1 (RN ), there is
a unique to = t(v) such that




J to v = max J(tv).

(2.6)

t ≥0

Indeed, since
J(tv) =

t2
v 2 −
2


RN



∗

F(tv) −



t 2  + 2∗
v
dx,
2∗

for t ≥ 0,

(2.7)

Marco A. S. Souto 551
the maximum point to of J(tv) is given by

v 2 =



RN





∗

to−1 v f to v + to2

−2



v+

2∗

dx.

(2.8)

We assume, without loss of generality that 0 ∈ Ω. Set Ωε = {x ∈ RN ; εx ∈ Ω}.
The restriction of J to Ho1 (Ωε ) is the energy functional,
J(u) =

1
2


Ωε



|∇u|2 + u2 dx −





Ωε







1 2∗
u dx,
2∗ +

F u+ +





u ∈ Ho1 Ωε , (2.9)

associated with the problem
∗

−∆u + u = f (u) + u2

−1

in Ωε ,

u = 0 on ∂Ωε .

(2.10)

If uε is a critical point of Jε , the family
vε (x) = uε (z) = uε (εx),

z = εx

(2.11)

is such that each vε is a critical point of functional J restricted to Ho1 (Ωε ) at the
level
 

bε = J vε =

inf

max J(tv).

v∈Ho1 (Ωε )\{0} t ≥0

(2.12)

It is easy to check that bε = ε−N cε and from the definition of c it follows that
bε ≥ c for all ε > 0.
Lemma 2.3. For {bε }ε>0 , limε→0 bε = c.
Proof. Fix ω a ground state solution of problem (2.4) and let ψε (x) = ϕ(εx)ω(x),
where ϕ is a C 1 -function such that

1

if x ∈ B1 ,
ϕ(x) = 
0 if x ∈
/ B2 ,

(2.13)

B1 = Bρ (0), B2 = B2ρ (0) ⊂ Ω. Observe that ψε → ω in H 1 (RN ) and the support
of ψε is in Ωε . By definition of bε , we have tε > 0 such that








bε ≤ max J tψε = J tε ψε .
t>0

(2.14)

552

Location of the peaks of solutions to critical growth problems

From (2.8) and condition (f3 ) it follows that
ψε

2


=
≥




RN





∗

tε−1 ψε f tε ψε + tε2



q −1

q +1

λtε 2 ψε 2

RN

∗

+ tε2

−2

−2

∗

ψε2
∗

ψε2

dx
(2.15)

dx,

so that, tε is bounded. Equality (2.15) and Remark 2.2 show that tε → t(ω) = 1,
as ε → 0. Then we have tε ψε → ω in H 1 (RN ) and




lim J tε ψε = J(ω) = c.

(2.16)

ε→0

Combining (2.14), (2.16), and the inequality bε ≥ c, for all ε > 0, we have proved

this lemma.
Lemma 2.4. The inequality c < (1/N)SN/2 holds, where S is the best Sobolev con∗
stant for the embedding D1,2 (RN )  L2 (RN ).
Proof. For each h > 0, consider the function


N(N − 2)h

(N −2)/4

φh (x) = 
(N −2)/2 .
h + |x|2

(2.17)

We recall that φh satisfies the problem
∗

−∆u = u2 −1


u(x) > 0,

RN



∇φh 2 dx =



RN

in RN ,
|∇u|2 dx < ∞,

∗

RN

φh2 dx = SN/2

(2.18)

(see Talenti ).

Now, consider ψh (x) = ϕφh (x)/ ϕφh L2∗ (RN ) , where ϕ is the function defined in
the proof of Lemma 2.3. From condition (f3 ) we have




J tψh ≤

t2
2

 
∗

q2 +1 


t2
q +1
∇ψh 2 + ψ 2 dx − λt
ψh 2 dx − ∗ .
h

q2 + 1

B2

2

B2

(2.19)

Using arguments as in , there exists h > 0 such that


t2
max
t ≥0
2


 

q2 +1 
2∗
2

1
λt
t
q
+1
2
2
∇ψh  + ψ dx −
ψh dx − ∗ < SN/2 .
h
B2

q2 + 1

B2

2

N

(2.20)

Marco A. S. Souto 553
Therefore, from (2.19) and (2.20) we have that




1 N/2
S ,
N

max J tψh <
t ≥0

(2.21)


and the proof of the lemma is completed.

Notice that the same proof of Lemma 2.4 can be used to show that bε <
(1/N)SN/2 , for all ε > 0. Using [3, Theorem 2.1], this inequality implies the existence of vε and then the existence of uε .
Lemma 2.5. There are εo > 0; a family { yε }{0<ε≤εo } ⊂ RN , yε ∈ Ωε ; constants R > 0
and β > 0 such that

BR (yε )

vε2 dx ≥ β > 0,


∀0 < ε ≤ εo ,


lim d yε ,∂Ωε = ∞.

(2.22)
(2.23)

ε→o

Proof. Start by showing that there is a family satisfying inequality (2.22). Arguing to the contrary, there is εn 0 such that for all R > 0


lim sup

n→∞ x∈RN BR (x)

vε2n dx = 0.

(2.24)

Using (Lions [6, Lemma I.1]) we have


q

RN

as n −→ ∞, ∀2 < q < 2∗ ,

vεn dx = on (1),

(2.25)

and, from (f1 ) and (f2 ),




RN





F vεn dx =





vεn f vεn dx = on (1).

RN

(2.26)

Since J  (vεn ) · vεn = 0, we conclude from (2.26) that
vεn

2


=

∗

RN

vε2n dx + on (1).

(2.27)

Let ≥ 0 be such that vεn 2 → . Passing to the limit in J(vεn ) = bεn and using
(2.26) we have
= Nc

(2.28)

and hence > 0. Now, using the definition of the constant S, we have
vεn

2


≥S

RN

∗

vε2n dx

2/2∗

.

(2.29)

554

Location of the peaks of solutions to critical growth problems

Taking the limit in the above inequalities, as n → ∞, we achieve that
≥S

2/2∗

,

(2.30)

1 N/2
S
N

(2.31)

and by (2.28), that
c≥

which contradicts Lemma 2.4 and then (2.22) holds.
Finally, to establish (2.23), suppose the contrary. That is, there exist εn → 0
and R > 0 such that dist(yεn ,∂Ωεn ) ≤ R, hence dist(εn yεn ,∂Ω) ≤ εn R. Without loss
of generality, we have εn yεn → yo for some yo ∈ ∂Ω. The arguments that follow
can be found in .
Let ν be the unit interior normal to ∂Ω at yo , and δ > 0 such that Bδ (yo +
δν) ⊂ Ω and Bδ (yo − δν) ∩ Ω = ∅. Let Ωn = {x ∈ RN : yo + εn x ∈ Ω} and wn (x)
= uεn (yo + εn x). This sequence wn is bounded in H 1 (RN ), −∆wn + wn = f (wn ) +
∗
wn2 −1 in Ωn ,




B2R (0)

wn2 dx

≥

BR (yεn )

vε2n dx ≥ β > 0,

∀n,

(2.32)

and we have that wn converges weakly to some w in H 1 (RN ).
Let RN+,ν be the half space {x ∈ RN : x · ν > 0}. Notice that Bεn−1 δ (εn−1 δν) ⊂ Ωn
and Bεn−1 δ (−εn−1 δν) ∩ Ωn = ∅ and then we can prove that for all compacts K+ ⊂
N
N
N
RN
+,ν and K− ⊂ R−,ν = R \ R+,ν , we have K+ ⊂ Ωn and K− ∩ Ωn = ∅, for n large.
∞
N
Then for each φ ∈ Co (R+,ν ) such that suppφ ⊂ Ωn , we have

RN
+,ν



∇wn ∇φ + wn φ dx =


RN
+,ν

 



∗

f wn + wn2

−1



φ dx.

(2.33)

From (2.33), usual arguments show that w ∈ H 1 (RN ) ∩ C 2 (RN+ ) and satisfies
∗
N
−∆w + w = f (w) + w 2 −1 , in RN
+,ν , and w ≡ 0 in R−,ν . Theorem I.1, due to
Esteban and Lions in , shows that w ≡ 0 which contradicts

B2R (0)∩RN
+,ν

w2 dx ≥ β > 0.

(2.34)


This completes the proof of the lemma.

Now we will consider the translation of vε , defined by ωε (x) = vε (x + yε ) =
 ε = {x ∈ RN ;εyε + εx ∈ Ω} and ωε = 0 outside Ω
 ε . From (2.23),
uε (εyε + εx) in Ω
N
 ε , for ε suﬃciently small.
any compact subset of R is contained in Ω
From Lemma 2.5,

BR (0)

ωε2 dx ≥ β > 0,

∀0 < ε ≤ εo .

(2.35)

Marco A. S. Souto 555
 εn , ωn = ωεn , vn = vεn , yε = yεn .
n = Ω
Consider a sequence εn 0 and set Ω
We will prove that ωn is bounded in the L∞ norm. In that case, uε is also
bounded in L∞ (Ω) norm and the proof of Theorem 1.1 follows from the subcritical case, as Lemma 2.8 will show.
Since the sequence ωn a translation of vn , we have a uniform bound for ωn 
and there is a ωo ∈ H 1 (RN ) which is weak limit of ωn in H 1 (RN ). From (2.35)
we have ωo = 0. We can write limit (2.23) in the following form




 n = ∞.
lim d 0,∂Ω

(2.36)

n→∞

 n , we have
Then for each φ ∈ Co∞ (RN ) and large n such that supp φ ⊂ Ω

RN



∇ωn ∇φ + ωn φ dx =



 
RN



∗

f ωn + ωn2

−1



φ dx,

∀n.

(2.37)

From (2.37), usual arguments show that ωo is a solution of problem (2.4), hence
a critical point of J, and J(ωo ) ≥ c.
Lemma 2.6. The sequence ωn converges to ωo in H 1 (RN ) and J(ωo ) = c.
Proof. This fact comes from Lemma 2.5 and Fatou’s lemma applied in the positive sequence ωn f (ωn ) − θF(ωn ). Observe that
  1  
bεn = J vn − J  vn vn
θ



 ∗


 
 
1
θ−2
2 −θ
∗
2
=
vn +
vn f vn − θF vn +
vn2
(2.38)
N
2θ
θ RN
2∗ θ
R



 ∗


 
 
1
θ−2
2 −θ
∗
2
=
ωn +
ωn f ωn − θF ωn +
ωn2 .
N
2θ
θ RN
2∗ θ
R

From (2.38)
 
  1  
c ≤ J ωo = J ωo − J  ωo ωo
θ


 ∗


 
 
1
θ−2
2 −θ
∗
2
=
ωo +
ωo f ωo − θF ωo +
ωo2
∗
N
N
2θ
θ R
2 θ
R



 ∗


 
 
1
θ−2
2 −θ
∗
2
≤ liminf
ωn +
ωn f ωn − θF ωn +
ωn2
2θ
θ RN
2∗ θ
RN

= lim bεn = c.
n→∞

We have proved that J(ωo ) = c and then (2.39) becomes an equality.

(2.39)

556

Location of the peaks of solutions to critical growth problems

Combining (2.39) with the three following inequalities:
2

2

≤ liminf ωn ,



 
 

 
 
ωo f ωo − θF ωo dx ≤ liminf
ωn f ωn − θF ωn dx,
N
N
R

R
∗
2∗
ωo dx ≤ liminf
ωn2 dx,

ωo

RN

(2.40)

RN

we conclude that ωn  → ωo  and then ωn → ωo in H 1 (RN ).



We are ready to conclude the proof of our main result. From Proposition 1.2
∗
∗
and Remark 1.3 with q(x) = ωn2 −2 ∈ LN/2 ; g(x,s) = f (s) + s2 − s, we have ωn ∈
Lt for all t ≥ 2 and
ωn

≤ Ct ,

Lt

(2.41)

where Ct does not depend on n.
Now we will make use of a very particular version of [5, Theorem 8.17], due
to Trudinger.
Proposition 2.7. Suppose that t > N, g ∈ Lt/2 (Λ), and u ∈ Ho1 (Λ) satisfies (in
the weak sense)
−∆u + u ≤ g̃(x),

(2.42)

where Λ is an open subset of RN . Then for any ballB2R (y) ⊂ Λ,


sup u ≤ C u+
BR (y)

L2 (B2R (y)) + g Lt/2 (B2R (y))



,

(2.43)

where C depends on N, t, and R.
We know that each ωn satisfies
∗

−∆ωn + ωn = ωn2

−1





+ f ωn ,

n
in Ω

(2.44)

and this implies that
∗

−∆ωn + ωn ≤ gn (x) = ωn2

−1





+ f ωn ,

in RN

(2.45)

in the weak sense.
Since (2.41) holds, gn Lt is bounded from above for some t > N. Using
Proposition 2.7 in (2.45) we have
sup ωn ≤ C
B1 (y)



ωn

L2 (B

+ gn
2R (y))


Lt (B

2R (y))

(2.46)

Marco A. S. Souto 557
for all y ∈ RN , which implies that there is a constant a > 0, independent of n,
such that
ωn (x) ≤ a,

∀x ∈ RN .

(2.47)

It follows that there is a εo > 0 such that
uε (z) ≤ a,

∀z ∈ Ω, ∀ε < εo .

(2.48)

To conclude the proof observe that uε becomes a solution of the subcritical
case (1.8) with h given by

∗

 f (s) + s2 −1 ,
h(s) = 
(2∗ − 1) 2∗ −θ θ−1 (2∗ − θ) 2∗ −1
 f (s) +
s −
,
a
a
(θ − 1)
(θ − 1)

if s ≤ a,
if s > a,

(2.49)

where θ > 2 is that one fixed in condition ( f4 ). It is easy to check that h is a C 1,α
s
function, h and H(s) = o h(τ)dτ satisfy (f1 ), (f2 ), (f3 ), (f4 ), and (f5 ). Let J˜ε be the
C 1 -functional on Ho1 (Ω) given by
1
J˜ε (u) =
2
∗

Since f (s) + s2

−1





Ω

2

2

2





ε |∇u| + u dz −

Ω

H(u)dz.

(2.50)

≥ h(s) for all s > 0, we have that

Jε (u) ≤ J˜ε (u),

∀u ∈ Ho1 (Ω),

(2.51)

Jε (uε ) = J˜ε (uε ), Jε (uε ) = J˜ε (uε ) = 0. We conclude that uε is a least-energy solution
of the subcritical problem (1.8).
Lemma 2.8. (i) If c̃ε is the minimax level of J˜ε , then c̃ε = cε ;
(ii) each uε is a critical point of J˜ε in the minimax level and satisfies (1.8).
Since global problem (1.9) has a unique nondegenerate positive solution
(cf. [9, 10]), Theorem 1.1 comes from [8, Theorem 2.2] applied to the functional J˜ε , and the asymptotic profile comes from [8, Theorem 2.3].
Appendix
Let Λ be some general domain in RN (bounded or unbounded). We will start
with the following lemma due to Brézis and Kato .
Lemma A.1. Let q ∈ LN/2 (Λ) be a nonnegative function. Then, for every ε > 0,
there is a constant σε = σ(ε, q) > 0 such that



Λ

q(x)u2 dx ≤ ε


Λ

|∇u|2 dx + σε

Λ

u2 dx,

∀u ∈ Ho1 (Λ).

(A.1)

558

Location of the peaks of solutions to critical growth problems

Remark A.2. If qk → q in LN/2 (Λ), we can choose a constant σε independent of k.
That is, σ(ε, qk ) = σε and



Λ

qk (x)u2 dx ≤ ε


Λ

|∇u|2 dx + σε

Λ

u2 dx,

∀u ∈ Ho1 (Λ), k ∈ N.

(A.2)

Proof. Let σε = σ(ε, q) > 0 be such that
qLN/2 ({q≥σε }) ≤ εS,

(A.3)
∗

where S is a best constant in the Sobolev immersion Ho1 (Λ)  L2 (Λ), where
2∗ = 2N/(N − 2). For all u ∈ Ho1 (Λ), we have



Λ

q(x)u2 dx =


{q≥σε }

q(x)u2 dx +



≤ σε

≤ σε

{q≤σε }

Λ



u2 dx +

{q≤σε }

{q≥σε }

q(x)u2 dx

q(x)u2 dx

(A.4)

u2 dx + qLN/2 ({q≥σε }) u2L2∗ ({q≥σε }) .

Inequality (A.1) follows from Sobolev estimate and the choice of σε .



Remark 1.3 follows from the proof of Lemma A.1 and the inequality



Λ

qk (x)u2 dx ≤

Λ

q(x)u2 dx + qk − q

2
LN/2 (Λ) uL2∗ (Λ) .

(A.5)

Proof of Proposition 1.2. For any n ∈ N and p > 0, consider An = {x ∈ Λ : |v| p ≤
n}, Bn = Λ \ An , and define vn by
vn = v|v|2p

vn = n2 v

in An ,

in Bn .

(A.6)

Observe that vn ∈ Ho1 (Λ), vn ≤ |v|2p+1 and
∇vn = (2p + 1)|v |2p ∇v

∇vn = n2 ∇v

in An ,

in Bn .

(A.7)

So, using vn as a test function



Λ

∇v ∇vn dx =

Λ

g(x,v)vn dx.

(A.8)

Using (A.7), we have


(2p + 1)


2p

|v |

≤

An

Λ

2

|∇v | dx + n

2

|∇v |2 dx



Bn



g(x,v)vn  dx ≤

Λ

 
q(x) + Cg vvn  dx.

(A.9)

Marco A. S. Souto 559
Now consider
ωn = v|v| p

ωn = nv

in An ,

in Bn .

(A.10)

Notice that ωn2 = vvn ≤ |v|2(p+1) and
∇ωn = (p + 1)|v | p ∇v

∇vn = n∇v

in An ,

in Bn .

(A.11)

|∇v |2 dx.

(A.12)

Therefore,

Λ



∇ωn 2 dx = (p + 1)2




2p

An

2

|v | |∇v | dx + n

2
Bn

Combining (A.9) and (A.12), we obtain


2p + 1
(p + 1)2

Λ



∇ωn 2 dx ≤





Λ



q(x) + Cg ωn2 dx.

(A.13)

Let σ p be given by Lemma A.1 with ε = (2p + 1)/2(p + 1)2 . Then

Λ



∇ωn 2 dx ≤ C̃ p



ωn2 dx,

Λ

(A.14)

where C̃n = (2(p + 1)2 /(2p + 1))(Cg + σ p ). Suppose that v ∈ L2(p+1) (Λ) for some
p ≥ 2. Applying Sobolev immersion in inequality (A.14) we have

An

∗
ωn2 dx



2/2∗
≤

Λ

∗
ωn2 dx



2/2∗
≤ SC̃ p

Λ

|v |2(p+1) dx

(A.15)

that is,

An



2/2∗

2∗ (p+1)

|v |

dx ≤ C p

dx

Λ

|v |2(p+1) dx,

(A.16)

where
Cp =


2(p + 1)2 
S Cg + σ p .
2p + 1
∗

Now, passing to the limit in (A.16) we have v ∈ L2

(p+1) (Λ)

v L2∗ (p+1) (Λ) ≤ C p v L2(p+1) (Λ) .

(A.17)
and
(A.18)

The proof follows from the following iteration argument: let p1 a positive
such that 2(p1 + 1) = 2∗ . It is easy to see that 0 < p1 and v ∈ L2(p1 +1) (Λ). Using
inequality (A.18) we have
∗

v ∈ L2

(p1 +1)

(Λ).

(A.19)

560

Location of the peaks of solutions to critical growth problems

Now choose p2 such that 2(p2 + 1) = 2∗ (p1 + 1). It is easy to see that 0 <
p1 < p2 and v ∈ L p2 +1 (Λ). Using inequality (A.18) we have
∗

v ∈ L2

(p2 +1)

(Λ).

(A.20)

Continuing with this iteration we obtain an increasing sequence pk given by
2(pk+1 + 1) = 2∗ (pk + 1) such that v ∈ L2(pk+1 +1) (Λ) for all k ∈ N. From
pk+1 + 1 =

N
N −2

(pk + 1),

(A.21)

it follows that


pk+1 + 1 =

N
N −2

k

2∗ .

(A.22)

This shows that pk goes to ∞ and therefore,
v ∈ L p (Λ),

∀ p ≥ 2.

(A.23)


Remark A.3. Proposition 1.2 is valid for positive subsolutions of problem (1.12)
as we can check in its proof.
Acknowledgments
This work was done during the author’s visit to the Department of Mathematics
of Rutgers University. The author thanks Professor C. O. Alves for his suggestions
improving the results. The author is supported by CNPq-Brazil.
References











C. O. Alves, D. C. de Morais Filho, and M. A. S. Souto, Radially symmetric solutions
for a class of critical exponent elliptic problems in RN , Electron. J. Diﬀerential Equations (1996), no. 07, 1–12.
H. Brézis and T. Kato, Remarks on the Schrödinger operator with singular complex
potentials, J. Math. Pures Appl. (9) 58 (1979), no. 2, 137–151.
H. Brézis and L. Nirenberg, Positive solutions of nonlinear elliptic equations involving
critical Sobolev exponents, Comm. Pure Appl. Math. 36 (1983), no. 4, 437–477.
M. J. Esteban and P.-L. Lions, Existence and nonexistence results for semilinear elliptic
problems in unbounded domains, Proc. Roy. Soc. Edinburgh Sect. A 93 (1982/83),
no. 1-2, 1–14.
D. Gilbarg and N. S. Trudinger, Elliptic Partial Diﬀerential Equations of Second Order,
Springer-Verlag, New York, 1998.
P.-L. Lions, The concentration-compactness principle in the calculus of variations. The
locally compact case. II, Ann. Inst. H. Poincaré Anal. Non Linéaire 1 (1984), no. 4,
223–283.
O. H. Miyagaki, On a class of semilinear elliptic problems in RN with critical growth,
Nonlinear Anal. 29 (1997), no. 7, 773–781.

Marco A. S. Souto 561









W.-M. Ni and J. Wei, On the location and profile of spike-layer solutions to singularly perturbed semilinear Dirichlet problems, Comm. Pure Appl. Math. 48 (1995),
no. 7, 731–768.
M. A. S. Souto, Uniqueness of positive radial solutions of semilinear elliptic equations,
preprint, 1999.
, Uniqueness of positive radial solutions of problem −∆u + f (u) = 0, Anais do
49o Seminário Brasileiro de Análise, Universidade Estadual de Campinas, Campinas, 1999, pp. 235–251.
M. Struwe, Variational Methods. Applications to Nonlinear Partial Diﬀerential Equations and Hamiltonian Systems, Springer-Verlag, Berlin, 1990.
G. Talenti, Best constant in Sobolev inequality, Ann. Mat. Pura Appl. (4) 110 (1976),
353–372.
M. Willem, Minimax Theorems, Progress in Nonlinear Diﬀerential Equations and
their Applications, vol. 24, Birkhäuser Boston, Massachusetts, 1996.

Marco A. S. Souto: Universidade Federal de Campina Grande, Departamento
de Matemática e Estatı́stica, Campina Grande-Pb, Cep 58109-970, Brazil

Operations Research
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Decision Sciences
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Journal of

Applied Mathematics

Algebra

Hindawi Publishing Corporation
http://www.hindawi.com

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Journal of

Probability and Statistics
Volume 2014

The Scientific
World Journal
Hindawi Publishing Corporation
http://www.hindawi.com

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

International Journal of

Differential Equations
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Volume 2014

http://www.hindawi.com
International Journal of

Combinatorics
Hindawi Publishing Corporation
http://www.hindawi.com

Mathematical Physics
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Journal of

Complex Analysis
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

International
Journal of
Mathematics and
Mathematical
Sciences

Mathematical Problems
in Engineering

Journal of

Mathematics
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Discrete Mathematics

Journal of

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Discrete Dynamics in
Nature and Society

Journal of

Function Spaces
Hindawi Publishing Corporation
http://www.hindawi.com

Abstract and
Applied Analysis

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

International Journal of

Journal of

Stochastic Analysis

Optimization

Hindawi Publishing Corporation
http://www.hindawi.com

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Volume 2014

```