Calc. Var.

18, 117–143 (2003)

Calculus of Variations
DOI: 10.1007/s00526-002-0182-9

Filippo Gazzola · Hans-Christoph Grunau · Marco Squassina

Existence and nonexistence results

for critical growth biharmonic elliptic equations
Received: 18 April 2002 / Accepted: 3 September 2002/
Published online: 17 December 2002 –
c Springer-Verlag 2002

Abstract. We prove existence of nontrivial solutions to semilinear fourth order problems

at critical growth in some contractible domains which are perturbations of small capacity of
domains having nontrivial topology. Compared with the second order case, some difficulties
arise which are overcome by a decomposition method with respect to pairs of dual cones.
In the case of Navier boundary conditions, further technical problems have to be solved by
means of a careful application of concentration compactness lemmas. The required general-
ization of a Struwe type compactness lemma needs a somehow involved discussion of certain
limit procedures. Also nonexistence results for positive solutions in the ball are obtained,
extending a result of Pucci and Serrin on so-called critical dimensions to Navier boundary
conditions. A Sobolev inequality with optimal constant and remainder term is proved, which
is closely related to the critical dimension phenomenon. Here, this inequality serves as a tool
in the proof of the existence results and in particular in the discussion of certain relevant
energy levels.
Mathematics Subject Classification (2000): 35J65; 35J40, 58E05

1. Introduction

We consider the following biharmonic critical growth problem

∆2 u = λu + |u|2∗ −2 u in Ω (1)
either with Navier boundary conditions
u = ∆u = 0 on ∂Ω (2)
or with Dirichlet boundary conditions
u = ∇u = 0 on ∂Ω. (3)

F. Gazzola: Dipartimento di Scienze e Tecnologie Avanzate, Corso Borsalino 54,

15100 Alessandria, Italy (e-mail: [email protected])
H.-Ch. Grunau: Fakultät für Mathematik der Universität, Postfach 4120,
39016 Magdeburg, Germany (e-mail: [email protected])
M. Squassina: Dipartimento di Matematica e Fisica, Via Musei 41,
25121 Brescia, Italy (e-mail: [email protected])
The first author was supported by MURST project “Metodi Variazionali ed Equazioni Dif-
ferenziali non Lineari”.
118 F. Gazzola et al.

2n
Here Ω ⊂ Rn (n ≥ 5) is a bounded smooth domain, λ ≥ 0 and 2∗ = n−4 is
2 n 2∗ n
the critical Sobolev exponent for the embedding H (R ) → L (R ). We are
interested in existence and nonexistence of nontrivial solutions of (1) with one of
the above boundary conditions.
We consider first the case where λ = 0, namely the equation
∆2 u = |u|8/(n−4) u in Ω. (4)
It is well known that (4)-(2) and (4)-(3) admit no positive solutions if Ω is star shaped
(see [27, Theorem 3.3] and [30, Corollary 1]). Clearly, these are not exhaustive
nonexistence results as they only deal with positive solutions, see also Sect. 3 below
for further comments. On the other hand, by combining the Pohožaev identity [34]
with the unique continuation property [22, Section 3], it is known that one indeed has
nonexistence results of any nontrivial solution for the correponding second order
equation −∆u = |u|4/(n−2) u in star shaped domains. This suggests that, in order
to obtain existence results for (4), one should either add subcritical perturbations
like the linear term λ u in (1) or modify the topology or the geometry of the domain
Ω. For general subcritical perturbations we refer to [4, 13] and references therein.
Domains with nontrivial topology are studied in [3] for Dirichlet and in [11] for
Navier boundary conditions. Here we are interested in topologically simple but
geometrically complicated domains. As far as we are aware, the case of “strange”
contractible domains has not been tackled before and this is precisely the first aim
of the present paper. We show that (4)-(2) admits a positive solution in a suitable
contractible domain Ω. This result is the exact counterpart of [27, Theorem 3.3]
where it is shown that (4)-(2) admits no positive solution if Ω is star shaped. For
the Dirichlet boundary condition we obtain a weaker result: we show that (4)-
(3) admits a nontrivial solution in the same kind of contractible domain. Clearly,
this still leaves open some problems concerning Dirichlet boundary conditions,
see Sect. 3. On one hand our proof is inspired by strong arguments developed by
Passaseo [32], on the other hand we have to face several hard difficulties, especially
(and somehow unexpectedly) under Navier boundary conditions. One of the crucial
steps in the approach by Passaseo is to prove that sign changing solutions of (4)
“double the energy” of the associated functional. For the second order problem
this may be shown by the usual technique of testing the equation with the positive
and negative parts of the solution. Of course, this technique fails for (4) where
higher order derivatives are involved and we overcome this difficulty thanks to
the decomposition method in dual cones developed in [14]. This method enables
us to bypass the lack of nonexistence results for nodal solutions of (4) in star
shaped domains. Moreover, when dealing with Navier boundary conditions, the
required generalization of the Struwe compactness lemma [37] turns out to be very
delicate because of the second boundary datum and the lack of a uniform extension
operator for H 2 ∩ H01 -functions in families of domains. See Lemma 5 and its proof
in Sect. 7. The same problem arises in Lemma 7 where a uniform lower bound
for an enlarged optimal Sobolev constant has to be found in a suitable class of
domains. An important tool for this Lemma is a Sobolev inequality with optimal
constant and remainder term, see Theorem 5. This inequality is closely related to
nonexistence results, which will be summarized in what follows.
Biharmonic equations at critical growth 119

Next, we restrict our attention to the case where λ > 0 and Ω = B, the unit
ball in Rn . We are interested in positive radially symmetric solutions. Let λ1 > 0
be the first eigenvalue of ∆2 with homogeneous Dirichlet boundary conditions.
A celebrated result by Pucci-Serrin [36] shows that (1)-(3) admits a nontrivial
radially symmetric solution for all λ ∈ (0, λ1 ) if and only if n ≥ 8 (n is the space
dimension). If n = 5, 6, 7 the range 0 < λ < λ1 is no longer correct since there
exists λ∗ ∈ (0, λ1 ) such that (1)-(3) admits no nontrivial radial solution whenever
λ ∈ (0, λ∗ ]; moreover, there exists λ∗ ∈ [λ∗ , λ1 ) (presumably λ∗ = λ∗ ) such that
(1)-(3) admits even a positive radially symmetric solution for all λ ∈ (λ∗ , λ1 ).
In other words, the well known [6] nonexistence result for radially symmetric
solutions of the second order problem for small λ in dimension n = 3 carries over
to (1)-(3) in dimensions n = 5, 6, 7: Pucci-Serrin call these dimensions critical.
Furthermore, the space dimensions n for which one has nonexistence of positive
radially symmetric solutions of (1)-(3) for λ in some (right) neighbourhood of 0
are called weakly critical [17]. In Theorem 3 below we prove that the dimensions
n = 5, 6, 7 are weakly critical also for Navier boundary conditions. Therefore,
critical dimensions seem not to depend on the boundary conditions. This result is
perhaps related to the fact that critical dimensions may have an explanation in terms
of the summability of the fundamental solution corresponding to the differential
operator ∆2 , see [20, 26].
Finally, when Ω = B we show that (4)-(3) admits no nontrivial radially sym-
metric solution; note that no positivity assumptions on the solution are made. This
result was already obtained in [16, Theorem 3.11]; we give here a different (and
simpler) proof.

2. Existence results
By D2,2 (Rn ) we
denote the completion of the space Cc∞ (Rn ) with respect to the
norm f 22,2 = Rn |∆f |2 ; this is a Hilbert space when endowed with the scalar
product 
(f, g)2,2 = ∆f ∆g dx.
Rn
By solution of (1) we mean here a function u ∈ H 2 ∩ H01 (Ω) (when dealing with
(2)) or a function u ∈ H02 (Ω) (when dealing with (3)) which satisfies
   
∆u∆ϕ dx = λuϕ + |u|8/(n−4) uϕ dx
Ω Ω
2
for all ϕ ∈ H ∩ H01 (Ω) (resp. H02 (Ω)). For Dirichlet boundary conditions this
variational formulation is standard, while for Navier boundary data we refer to [4,
p.221]. By well known regularity results (see [25, Theorem 1] and [43, Lemma
B3]), if ∂Ω ∈ C 4,α , then the solutions u of (1) satisfy u ∈ C 4,α (Ω) and solve (1)
in the classical sense.
Definition 1. Let K ⊂ Rn be bounded. We say that u ∈ D2,2 (Rn ) satisfies u ≥ 1
on K in the sense of D2,2 (Rn ) if there exists a sequence (uh ) in Cc2 (Rn ) such that
uh ≥ 1 on K for each h ∈ N and uh → u in D2,2 (Rn ). Analogously, u ≤ 1 and
u = 1 on K are defined.
120 F. Gazzola et al.

Definition 2. We define the (2, 2)−capacity of K (cap K) as

 
2 2,2
cap K = inf |∆u| dx : u = 1 on K in the D (R ) sense .
n
Rn

We set cap ∅ := 0.
Since the nonempty set
 
u ∈ D2,2 (Rn ) : u = 1 on K in the D2,2 (Rn ) sense
is closed and convex, there exists a unique function zK ∈ D2,2 (Rn ) such that
zK = 1 on K and 
|∆zK |2 dx = cap K.
Rn
Finally we make precise what we mean by set deformations:
Definition 3. Let Ω ⊂ Rn and let H, Ω ⊂ Ω. We say that H can be deformed in
Ω into a subset of Ω if there exists a continuous function
H : H × [0, 1] → Ω
such that H (x, 0) = x and H (x, 1) ∈ Ω for all x ∈ H.
Our first result states the existence of positive solutions for the critical growth
equation (4) with Navier boundary conditions. Combined with the already men-
tioned nonexistence result in star shaped domains [27], this shows that the existence
of positive solutions strongly depends on the geometry of the domain.
Theorem 1. Let Ω be a smooth bounded domain of Rn (n ≥ 5) and let H be a
closed subset contained in Ω. Then there exists ε > 0 such that if Ω ⊂ Ω is a
smooth domain with cap(Ω \ Ω) < ε and such that H cannot be deformed in Ω
into a subset of Ω then there exists a positive solution of

∆2 u = u(n+4)/(n−4) in Ω
(5)
u = ∆u = 0 on ∂Ω.
We will prove this result in Sect. 5.
Next we turn to Dirichlet boundary conditions. In this case, we merely show the
existence of nontrivial solutions to (4). The lack of information about the sign of the
solution is due to the lack of information about the sign of the corresponding Green’s
function. Moreover, in general domains sign change has even to be expected. See
[18] for a survey on this feature and for further references.
Theorem 2. Let Ω be a smooth bounded domain of Rn (n ≥ 5) and let H be a
closed subset contained in Ω. Then there exists ε > 0 such that if Ω ⊂ Ω is a
smooth domain with cap(Ω \ Ω) < ε and such that H cannot be deformed in Ω
into a subset of Ω then there exists a nontrivial solution of

∆2 u = |u|8/(n−4) u in Ω
(6)
u = ∇u = 0 on ∂Ω.
Biharmonic equations at critical growth 121

The proof of this result is simpler than the one of Theorem 1. As we will
explain below, for (5), one has to study very carefully the behaviour of suitable
sequences uh ∈ H 2 ∩ H01 (Ωh ) for varying domains Ωh . In contrast with the spaces
H02 (Ωh ), there is no “trivial” extension operator into H 2 (Rn ). Only as positivity of
solutions is concerned, the situation with respect to Dirichlet boundary conditions
is more involved than with respect to Navier boundary conditions. Here we have
no positivity statement of the solution because Lemma 3 below does not seem to
hold.
If Ω is an annulus and if one restricts to radial functions, the problem is no
longer critical and then both (5) and (6) admit a nontrivial solution. For the second
order problem −∆u = |u|4/(n−2) u in Ω, u = 0 on ∂Ω, it was shown with help of
very subtle methods by Coron [7] and Bahri-Coron [2] that this result extends to
domains with nontrivial topology. Recently, corresponding results have been found
also in the higher order case: see [11] for problem (5) and [3] for the polyharmonic
analogue of (6) under Dirichlet boundary conditions. However, the solutions which
are constructed in these papers and here are not related. To explain this, assume that
Ω satisfies the assumptions of [11] or [3]. If uΩ is the solution of (5) or (6), then
by exploiting the proof of Theorem 1 one has that uΩ converges weakly to zero as
cap(Ω \ Ω) → 0. Hence, any nontrivial solution in Ω may not be obtained as limit
of the solutions uΩ in Ω. On the other hand one expects the nontrivial solutions in
Ω to be stable and to remain under “small” perturbations Ω ⊂ Ω. These solutions
in Ω will be different from ours. The latter situation was studied in [9] for the
second order problem.
To conclude the section, we observe that there are also contractible domains Ω,
which satisfy the assumptions of the preceding theorems. In the following example
we describe precisely such a situation. Further examples may be adapted to the
biharmonic setting from [32, pp. 39-41].

Example 1. We consider an annular shaped domain as Ω ⊂ Rn with n > 5, where

we drill a sufficiently “thin” cylindrical hole along a straight line in order to obtain
the smooth contractible subdomain Ω.
To be more precise: we assume that for ε small enough, Ω \ Ωε is contained
in a cylinder with basis Bε  ⊂ Rn−1and fixed height. Then by simple scaling
arguments one finds that cap Ω \ Ωε = O(εn−5 ) → 0 as ε → 0, provided that
the dimension satisfies n > 5.
We choose H to be a spherical hypersurface in Ω, which cannot be de-
formed into a subset of Ωε . This can be seen by looking at the degree of map-
ping d(H ( . , t), H, 0) for t ∈ [0, 1], where H is assumed to exist according to
Definition 3.
Instead of the above mentioned segment of fixed length, one may consider any
bounded piece of a fixed generalized plane, provided that its codimension is at
least 5.
122 F. Gazzola et al.

3. Nonexistence results and a Sobolev inequality

with optimal constant and remainder term

We collect here a number of known as well as of new nonexistence results for the
equation
∆2 u = λu + |u|8/(n−4) u in Ω (7)
under either Navier or Dirichlet boundary conditions. Here, we include the linear
term λu. It turns out that the discussion of the borderline case λ = 0, where the
purely critical equation has to be considered, is somehow involved and still not
exhaustive. This lack of nonexistence results complicates the formulation of the
compactness result Lemma 5 below.
The functional analytic counterpart of nonexistence results for (7) for small λ >
0 in so-called critical dimensions are Sobolev inequalities with optimal constant and
remainder term. (This becomes clear in the proof of Theorem 5. See also [14].) Such
an inequality in H02 (Ω) – i.e. under Dirichlet boundary conditions – was proved
in [14] while the result in H 2 ∩ H01 (Ω) – i.e. under Navier boundary conditions –
is provided in Sect. 3.3 below. In the present paper, this refined Sobolev inequality
is applied to show that the optimal Sobolev constant can be increased uniformly
for all subdomains Ω ⊂ Ω with help of a condition on the “barycenter” of the
functions u ∈ H 2 ∩ H01 (Ω). See Lemma 7 below.

3.1. Nonexistence under Navier boundary conditions

It turns out that nonexistence questions in this case are relatively hard and that only
restricted results are available.
The case λ = 0
In this case, no positive solution to (5) may exist, provided Ω is star shaped. See
[42, Theorem 3.10] and also [27, Theorem 3.3]. The key ingredient is a Pohožaev
type identity [35]. By means of techniques developed by Pucci and Serrin in [36],
such a nonexistence result may be also shown for nontrivial radial solutions in the
ball.
More difficult seems this nonexistence result for any nontrivial solution of (5).
In this case the Pohožaev type identity is not powerful enough to extend u outside
Ω by zero as a solution of (4) in Rn and then to apply the unique continuation
property [33].
The case λ > 0
In Sect. 6 we prove the following result.
Theorem 3. Let n ∈ {5, 6, 7} and B ⊂ Rn be the unit ball. Then there exists a
number λ∗ > 0 such that the problem

∆2 u = λu + u(n+4)/(n−4) in B
(8)
u = ∆u = 0 on ∂B

admits no positive solution if λ ∈ (0, λ∗ ].

Biharmonic equations at critical growth 123

By [41], positive solutions to (8) are radially symmetric.

Theorem 3 answers a question left open by van der Vorst in [44, Theorem
3]. Moreover, in [44, Theorem 1] complementary existence results were shown
implying that such a nonexistence result can hold at most in the dimensions 5, 6, 7.
These are the critical dimensions for problem (8), at least, when restricting to
positive solutions in the ball. Theorem 3 can also be shown just for nontrivial radial
solutions: here one has to combine methods of [36] with some techniques, which
appear in the proof of Theorem 4 below.
Again, nonexistence of any nontrivial solution, although expected, has to be
left open.
The case λ < 0
In the preceding nonexistence results, the positivity of a solution, the nonexistence
of which had to be shown, is intensively exploited. For λ ≥ 0 one can conclude
from u ≥ 0 that even −∆u ≥ 0 in Ω.
For λ negative, we can only argue as before, when λ is close enough to 0, see
[28, 21]. If λ << 0, the framework of positive solutions is no longer adequate
for (7), and the argument in [42, pp. 390/391] and [44, Theorem 3] breaks down.
Instead, one has to look for nonexistence of any nontrivial solution. But such a
result seems to be still unknown, even for nontrivial radial solutions in the ball.

3.2. Nonexistence under Dirichlet boundary conditions

In this situation, existence results for positive solutions are more involved than
under Navier boundary conditions (see [15]) and cannot even be expected in general
domains. On the other hand, we know much more about nonexistence for relatively
large classes of solutions.
The case λ < 0
Nonexistence of any nontrivial solution is shown in [35], provided the domain Ω
is star shaped. This proof is based on a generalized Pohožaev identity and covers
also the general polyharmonic problem with (−∆)K as linear principal part.
The case λ > 0
In [36] we already find the following result, which gives a stronger statement than
the analogous Theorem 3 above for Navier boundary conditions:
Let n ∈ {5, 6, 7} and assume that Ω is the unit ball B. Then there exists
a number λ∗ > 0 such that a necessary condition for a nontrivial radial
solution to (6) to exist is λ > λ∗ .
As this special behaviour of the critical growth Dirichlet problem (6) may be ob-
served only in space dimensions 5, 6 and 7, Pucci and Serrin call these dimensions
critical. For the critical dimension phenomenon for general polyharmonic problems
with Dirichlet boundary conditions we refer to [17].
The case λ = 0
Also in this case, the nonexistence of any nontrivial solution to (6) has to be left open,
only more restricted results are available. Oswald [30] proved the nonexistence of
124 F. Gazzola et al.

positive solutions in strictly star shaped domains by means of a Pohožaev type

identity (see e.g. [27, (2.6)]) and strong maximum principles for superharmonic
functions.
Moreover, we can also exclude nontrivial radial solutions:

Theorem 4. Let B ⊂ Rn (n ≥ 5) be the unit ball. Then problem

∆2 u = |u|8/(n−4) u in B,
(9)
u = ∇u = 0 on ∂B

admits no nontrivial radial solution.

This result can be found in [16, Theorem 3.11]. The proof there is based on a
refined integral identity of Pucci and Serrin [36] and some Hardy type embedding
inequalities. In Sect. 6 we will give a more elementary and geometric proof.
Unfortunately neither this proof nor the one given in [16, Theorem 3.11] carry
over to the general polyharmonic problem, where the discussion of nonexistence
of nontrivial radial solutions in the borderline case λ = 0 remains essentially open.

3.3. An optimal Sobolev inequality with remainder term

Sobolev embeddings with critical exponents and the corresponding optimal Sobolev
constants will play a crucial role in the proof of our existence results.
Let us set

∆u22
S= infn . (10)
u∈D 2,2
(R )\{0} u22∗

When working in the whole Rn , this optimal Sobolev constant is attained.

According to Lemma 2 below, minimizers are necessarily of one sign. Positive
solutions of the corresponding Euler Lagrange equations are known to be of the
following form.

Lemma 1. The constant S in (10) is a minimum and (up to nontrivial real multiples)
it is attained only by the functions

[(n − 4)(n − 2)n(n + 2)](n−4)/8 ε(n−4)/2

uε,x0 (x) = , (11)
(ε2 + |x − x0 |2 )(n−4)/2

for any x0 ∈ Rn and each ε > 0. Moreover, the functions uε,x0 are the only positive
solutions of the equation

∆2 u = u(n+4)/(n−4) in Rn . (12)

Proof. From Lemma 2 below, we take that any minimizer of (10) may be assumed
to be positive. Then, the result follows from [12, Theorem 2.1], [39, Theorem 4]
and [23, Theorem 1.3].
Biharmonic equations at critical growth 125

Similarly one can define S(Ω) for any Ω ⊂ Rn . It is well known that S(Ω) is not
attained if Ω is bounded [12] and that it does not depend neither on the domain nor on
the boundary conditions [43]. The following extension of the embedding inequality
corresponding to (10) with optimal constant S gives a quantitative formulation for
the fact that S is not attained. Here, it will be used to show Lemma 7, which in turn
is an important step in the proof of our existence results.
Theorem 5. Let Ω ⊂ Rn be open and of finite measure and p ∈ [1, n−4 n
). Then
there exists a constant C = C(n, p, |Ω|) > 0, such that for every u ∈ H ∩H01 (Ω),
2

one has 
1
|∆u|2 dx ≥ S u22∗ + u2p . (13)
Ω C
Here, |Ω| denotes the n-dimensional Lebesgue measure of Ω. Moreover, if ω > 0,
then (13) holds with a constant C = C(n, p, ω) for all Ω with |Ω| ≤ ω.
For the smaller class H02 (Ω) a more refined result can be found in [14]. The
proof there is based on an extension and decomposition method (see also Sect. 4
below), by means of which analogues to (13) for the spaces H0m (Ω) of any order
can be also obtained. This proof, however, does not seem to carry over directly to
the space H 2 ∩ H01 due to the lack of information at the boundary. In Sect. 6 we
sketch a technical proof of Theorem 5, which is based on nonexistence results like
in Theorem 3 and on Talenti’s symmetrization result [40].

4. Preliminary results
In the previous section we have briefly discussed energy minimizing solutions of
(12). These are necessarily of one sign (this is obtained as a byproduct of the
following lemma) and of the form given in Lemma 1. The next results show that
nodal solutions of (4) double the energy level of the corresponding “free” energy
functional (observe that here, we are working with a constrained functional). This
will allow us to bypass the lack of nonexistence results for nodal solutions of (4)
in star shaped domains.
Lemma 2. Let u ∈ D2,2 (Rn ) be a nodal solution of the equation
∆2 u = |u|8/(n−4) u in Rn . (14)

Then ∆u22 ≥ 24/n Su22∗ .

Proof. Consider the convex closed cone
 
K = u ∈ D2,2 (Rn ) : u ≥ 0 a.e. in Rn ,
and its dual cone
 
K  = u ∈ D2,2 (Rn ) : (u, v)2,2 ≤ 0 ∀v ∈ K .
Let us show that K  ⊆ −K . For each h ∈ Cc∞ (Rn ) ∩ K consider the solution
uh of the problem
∆2 uh = h in Rn .
126 F. Gazzola et al.

Then by the positivity of the fundamental solution of ∆2 in Rn , it follows uh ∈ K

and thus if v ∈ K  ,
 
∞
∀h ∈ Cc (R ) ∩ K :
n
hv dx = ∆2 uh v dx = (uh , v)2,2 ≤ 0.
Rn Rn

By a density argument one obtains v ≤ 0 a.e. in Rn , i.e. v ∈ −K . Now, by a

result of Moreau [29], for each u ∈ D2,2 (Rn ) there exists a unique pair (u1 , u2 )
in K × K  such that

u = u1 + u2 , (u1 , u2 )2,2 = 0. (15)

Let u be a nodal solution of (14) and let u1 ∈ K and u2 ∈ K  be the components

of u according to this decomposition. We obtain that ui ≡ 0 and

|u(x)|2∗ −2 u(x)ui (x) ≤ |ui (x)|2∗ , i = 1, 2, (16)

for a.e. x ∈ Rn . Indeed, if i = 1 and u(x) ≤ 0 then (16) is trivial, while if u(x) ≥ 0,
since u2 ∈ −K one has u(x) = u1 (x) + u2 (x) ≤ u1 (x) and again (16) holds.
The case i = 2 is similar. By combining the Sobolev inequality with (15) and (16),
we get for i = 1, 2
 
Sui 22∗ ≤ ui 22,2 = ∆u∆ui dx = ∆2 uui dx =
Rn Rn
 
= |u|2∗ −2
uui dx ≤ |ui |2∗ dx = ui 22∗∗ ,
Rn Rn

which implies ui 22∗ ≥ S (n−4)/4 for i = 1, 2. Hence, again by (15), one obtains
  −2/2∗  4/n
∆u22 2 2∗
= |∆u| dx |u| dx = |∆u|2 dx =
u22∗ Rn Rn Rn
  4/n
2
= |∆u1 | dx + |∆u2 |2 dx ≥
Rn Rn
4/n
≥ Su1 22∗ + Su2 22∗ ≥ 24/n S ,

which concludes the proof.

In a completely similar fashion, one may extend the previous result to any
domain in case of Navier boundary conditions. For Dirichlet boundary conditions
this seems not possible due to the lack of information about the positivity of the
corresponding Green’s function.

Lemma 3. Let Ω be a bounded domain of Rn and assume that u solves (5) and
changes sign. Then ∆u22 ≥ 24/n Su22∗ .

In the case of the half space, by exploiting nonexistence results for positive
solutions, we have a stronger result.
Biharmonic equations at critical growth 127

Lemma 4. Let Ω = {xn > 0} be the half space and assume that u ∈ D2,2 (Ω)
solves the equation

∆2 u = |u|8/(n−4) u in {xn > 0} (17)

with boundary data either (2) or (3). Then ∆u22 ≥ 24/n Su22∗ .

Proof. Notice first that by [27, Theorem 3.3], equation (17)-(2) does not admit
positive solutions. A similar nonexistence result holds for boundary conditions (3),
see [30, Corollary 1] which may be easily extended to unbounded domains.
Therefore, any nontrivial solution of (17) (with boundary conditions (2) or (3))
necessarily changes sign. We obtain the result if we repeat the argument of the proof
of Lemma 2. With Navier boundary conditions this is straightforward, while with
Dirichlet boundary data one may invoke Boggio’s principle [5] in the half space.

As pointed out in the introduction, the H 2 ∩ H01 -framework is more involved

than the H02 -case. Therefore, from now on we restrict our attention to the first
situation.
Consider the functional f : V (Ω) → R

f (u) = |∆u|2 dx (18)
Ω

constrained on the manifold

  
V (Ω) = u ∈ H 2 ∩ H01 (Ω) : |u|2∗ dx = 1 . (19)
Ω

We say that (uh ) ⊂ V (Ω) is a Palais-Smale sequence for f at level c if

lim f (uh ) = c, lim f  (uh )(Tu ∗

V (Ω)) →0
h h h

where Tuh V (Ω) denotes the tangent space to the manifold V (Ω) at uh .
We can now state a global compactness result for the biharmonic operator,
in the spirit of [37]. Even if the proof is on the lines of that of Struwe, some
difficulties arise. In the appendix we give an outline of the proof emphasizing the
main differences with respect to second order equations.

Lemma 5. Let (uh ) ⊂ V (Ω) be a Palais-Smale sequence for f at level c ∈ R.

Then, for a suitable subsequence, one has the following alternative: either (uh ) is
relatively compact in H 2 ∩ H01 (Ω) or there exist k nonzero functions u
1 , ..., u
k ∈
D2,2 , solving either (14) or (17) with boundary condition (2) and a solution u 0 ∈
H 2 ∩ H01 (Ω) of (5) such that
 −1/2∗
k 
uh 0 
u uj |2∗ dx
| in H 2 ∩ H01 (Ω) (20)
j=0
128 F. Gazzola et al.

and
  −2/2∗
k  k 
lim f (uh ) =  uj |2 dx 
|∆ uj |2∗ dx
| . (21)
h
j=0 j=0

The domain of integration for u 1 , ..., u

0 is just Ω, while for u k , it is either a half
space or the whole Rn .

When working in H02 (Ω) a similar compactness result holds true with Navier
boundary conditions replaced with Dirichlet boundary conditions.
We believe that nontrivial solutions of (17) either with Navier (2) or Dirichlet (3)
boundary conditions do not exist and that the previous lemma may be strengthened
(see Sect. 3.1 and 3.2). Nevertheless, Lemma 5 is sufficient to prove the following
compactness property in the second critical energy range.

Lemma 6. Let (uh ) ⊂ V (Ω) be a Palais-Smale sequence for f at level c ∈

(S, 24/n S). Then, up to a subsequence, (uh ) strongly converges in H 2 ∩ H01 (Ω).

Proof. Assume by contradiction that the sequence (uh ) is not relatively compact
in H 2 ∩ H01 (Ω). Then, by Lemma 5 one finds functions u 0 ∈ H 2 ∩ H01 (Ω) and
u k ∈ D2,2 satisfying (20) and (21). Assume first that all u
1 , ..., u 1 , . . . , u
k are
positive solutions of (14). Then, by Lemma 1 each u j is of type (11) and attains
the best Sobolev constant, i.e.
   2/2∗
2∗ 2
|
uj | dx = |∆
uj | dx = S uj |2∗ dx
| , j = 1, . . . k, (22)
Rn Rn Rn


which implies Rn
uj |2∗ dx = S n/4 for j = 1, ..., k. In particular, one obtains
|
 4/n
lim f (uh ) = u0 |2∗ dx + kS n/4
| .
h Ω

0 ≡ 0, we get f (uh ) → k 4/n S, while if u

If u 0 ≡ 0 for each k ≥ 1 one has
 4/n
u0 |2∗ dx + kS n/4
| > (k + 1)4/n S.
Ω

In any case we have a contradiction with S < lim f (uh ) < 24/n S.
h
j is sign changing or a solution
We now consider the case in which at least one u
in the half space {xn > 0}. By Lemmas 2 and 4 we then have for these u j :
   2/2∗
uj |2∗ dx =
| uj |2 dx ≥ 24/n S
|∆ uj |2∗ dx
| , (23)
Rn Rn Rn


and hence uj |2∗ dx ≥ 2S n/4 , while for the remaining u
| Rn
j , (22) holds true. In
 n/4 4/n 4/n
any case, we have lim f (uh ) ≥ 2 S = 2 S, again a contradiction.
h
Biharmonic equations at critical growth 129

5. Proof of Theorems 1 and 2

The proof of Theorem 1 is by far more involved than the proof of Theorem 2,
because the trivial extension of any function u ∈ H 2 ∩ H01 (Ω) by 0 does not yield
a function in H 2 (Rn ). In particular, for the space H02 (Ω) Lemma 7 easily follows
by this extension argument. A further difference is that the positivity conclusion
of Lemma 3 does not hold in the Dirichlet boundary value case. But in the latter
case, one simply has to drop this argument. Therefore we only deal with the proof
of Theorem 1.
For all smooth Ω ⊆ Ω let β : V (Ω) → Rn be the “barycenter” map

β(u) = x|u(x)|2∗ dx.
Ω

Since Ω is smooth one finds r > 0 such that Ω is a deformation retract of

 
Ω+ = x ∈ Rn : d(x, Ω) < r .

First we show that the energy or, equivalently, the optimal Sobolev constants
will remain relatively large if we prevent the functions from concentrating “too
close” to their domain of definition.

Lemma 7.
   
γ = inf inf f (u) : u ∈ V (Ω), β(u) ∈ Ω+ , Ω smooth subset of Ω > S.

Proof. Assume by contradiction that for each ε > 0 there exists a smooth Ωε ⊆ Ω
and uε ∈ H 2 ∩ H01 ∩ C ∞ (Ωε ) such that

|uε |2∗ dx = 1 (24)
Ωε

|∆uε |2 dx ≤ S + ε (25)
Ωε

β(uε ) ∈ Ω+ . (26)
2
Let Uε ∈ H (R ) be any entire extension of uε . Since Ωε is smooth, the existence
n

of such an extension is well known. We emphasize that the quantitative properties

of Uε outside Ωε (which are expected to blow up for ε  0) will not be used.
Further, let 1Ωε be the characteristic function of Ωε .

Step I. We claim that 1Ωε Uε → 0 in L2 (Rn ) as ε → 0.

By Theorem 5 there exists C > 0 independent of ε such that
1
∀u ∈ H 2 ∩ H01 (Ωε ) : ∆u22 ≥ Su22∗ + u21 . (27)
C
130 F. Gazzola et al.

Putting together (24), (25) and (27), for each ε > 0 we have

1 1
S+ε≥ |∆uε |2 dx ≥ Suε 22∗ + uε 21 = S + 1Ωε Uε 21
Ωε C C

so that 1Ωε Uε 1 → 0 as ε → 0. By (24) and classical Lp -interpolation, we also

get 1Ωε Uε 2 → 0 as ε → 0.

Step II. The weak* limit (in the sense of measures) of 1Ωε |Uε |2∗ is a Dirac mass.
Integration by part shows that ∇uε 2L2 (Ωε ) ≤ ∆uε L2 (Ωε ) uε L2 (Ωε ) . Then by
Step I and (25), we infer that

1Ωε ∇Uε L2 (Rn ) = ∇uε L2 (Ωε ) → 0.

Therefore, taking any ϕ ∈ Cc∞ (Rn ), as ε → 0 we find

 
1Ωε |∆(ϕUε )|2 dx = 1Ωε |ϕ|2 |∆Uε |2 dx + o(1) . (28)
Rn Rn

By the L1 (Rn ) boundedness of the sequences (1Ωε |Uε |2∗ ) and (1Ωε |∆Uε |2 ) there
exist two bounded non negative measures ν, µ on Rn such that

1Ωε |Uε |2∗ ∗

ν, 1Ωε |∆Uε |2 ∗
µ (29)

in the sense of measures. By the Sobolev inequality in Ωε we have

 2/2∗ 
2∗ 2∗
∀ε > 0 : S 1Ωε |Uε | |ϕ| dx ≤ 1Ωε |∆(ϕUε )|2 dx
Rn Rn

and therefore, by (28) and (29), letting ε → 0 yields

 2/2∗ 
2∗
S |ϕ| dν ≤ |ϕ|2 dµ.
Rn Rn

By (24) and (25) we also know that

 2/2∗ 
S dν = dµ.
Rn Rn

Hence by [24, Lemma I.2] there exist x ∈ Ω and σ > 0 such that ν = σδx . From
(24) we see that σ = 1.
The contradiction follows by Step II, since β(uε ) → x, against (26). This
completes the proof.

According to the previous lemma, we may choose µ such that S < µ <
min{24/n S, γ}. Let ϕ ∈ Cc∞ (B1 (0)) be such that
 
|ϕ|2∗ dx = 1, |∆ϕ|2 dx < µ. (30)
B1 (0) B1 (0)
Biharmonic equations at critical growth 131

Define for each σ > 0 and y ∈ Rn the function ϕσ,y : Rn → R by setting

 
x−y
ϕσ,y (x) = ϕ
σ

where ϕ is set to zero outside B1 (0) . It is readily seen that there exists σ > 0 such
that Bσ (y) ⊂ Ω and hence ϕσ,y ∈ Cc∞ (Ω) for each y ∈ H . For all Ω
Ω let
zΩ ∈ D2,2 (Rn ) be such that zΩ = 1 in Ω\Ω and

|∆zΩ |2 dx = cap(Ω\Ω),
Rn

see what follows Definition 1. Note that one has

(1 − zΩ )Ty (ϕ) ∈ H 2 ∩ H01 (Ω) ∀y ∈ H,

ϕσ,y
where Ty (ϕ) := ϕσ,y 2∗ . Moreover, for each δ > 0 there exists ε > 0 with

sup zΩ Ty (ϕ)H 2 ∩H01 (Ω) < δ (31)

y∈H

whenever cap(Ω\Ω) < ε. Then one gets

(1 − zΩ )Ty (ϕ)L2∗ (Ω) = 0 ∀y ∈ H,

if ε is sufficiently small. So, we may define the map ΦΩ : H → V (Ω) by

(1 − zΩ ) Ty (ϕ)
ΦΩ (y) = .
(1 − zΩ ) Ty (ϕ)2∗
Taking into account (30) and (31), we find ε > 0 such that

sup{f (ΦΩ (y)) : y ∈ H} < µ (32)

provided that cap(Ω\Ω) < ε. From now on ε is fixed subject to the previous
restrictions. Let Ω ⊂ Ω be such that cap(Ω\Ω) < ε and r ∈ (0, r) be such that Ω
is a deformation retract of
 
Ω+ = x ∈ Rn : d(x, Ω) < r .

As in Lemma 7, one obtains

inf {f (u) : u ∈ V (Ω), β(u) ∈ Ω+ } > S. (33)

Notice that

inf {f (u) : u ∈ V (Ω), β(u) ∈ Ω+ } ≤ sup {f (ΦΩ (y)) : y ∈ H} (34)

otherwise the map R : H × [0, 1] → Ω given by

R(y, t) = (1 − t)y + tβ(ΦΩ (y)) ∀y ∈ H, ∀t ∈ [0, 1]

132 F. Gazzola et al.

would deform H in Ω into a subset of Ω+ and then into a subset of Ω (Ω is a

deformation retract of Ω+ ) contradicting the assumptions. Here, in order to see
R(y, t) ∈ Ω, we used the fact that supp ΦΩ (y) ⊂ Bσ (y) and that, since ϕ ≥ 0,
β (ΦΩ (y)) ∈ Bσ (y) ⊂ Ω.
Therefore, by combining (32), (33) and (34) one gets
S < inf {f (u) : u ∈ V (Ω), β(u) ∈ Ω+ } ≤ sup {f (ΦΩ (y)) : y ∈ H} (35)
 
< µ < γ ≤ inf f (u) : u ∈ V (Ω), β(u) ∈ Ω+ .
In what follows we need to find two appropriate levels, such that the corresponding
sublevel sets
f c = {u ∈ V (Ω) : f (u) ≤ c}
cannot be deformed into each other. For this purpose let c1 , c2 > S be such that
c1 < inf{f (u) : u ∈ V (Ω), β(u) ∈ Ω+ } ,
c2 = sup{f (ΦΩ (y)) : y ∈ H}.
Assume by contradiction that there exists a deformation ϑ of f c2 into f c1 , i.e.
ϑ : f c2 × [0, 1] → f c2 , ϑ( . , 0) = Idf c2 and ϑ(f c2 , 1) ⊆ f c1 .
We define H : H × [0, 1] → Ω by setting for each x ∈ H


(1 − 3t)x + 3tβ(ΦΩ (x)) if t ∈ [0, 1/3]
H (x, t) = (β(ϑ(ΦΩ (x), 3t − 1))) if t ∈ [1/3, 2/3]


 ( (β (ϑ(ΦΩ (x), 1)) , 3t − 2)) if t ∈ [2/3, 1]
where  : Ω+ → Ω is a retraction and  : Ω+ × [0, 1] → Ω+ is a continuous
map with (x, 0) = x and (x, 1) ∈ Ω for all x ∈ Ω+ . In order to see that
H (x, t) ∈ Ω, one should observe that c2 < γ and ϑ(ΦΩ (x), 3t − 1) ∈ f c2 , hence
β(ϑ(ΦΩ (x), 3t − 1)) ∈ Ω+ .
As ϑ(ΦΩ (x), 1) ∈ f c1 for each x ∈ H and
c1 < inf {f (u) : u ∈ V (Ω), β(u) ∈ Ω+ } ,
then for each x ∈ H
β(ϑ(ΦΩ (x), 1)) ∈ Ω+
and H is a deformation of H in Ω into a subset of Ω, in contradiction with our
assumptions. Then the sublevel set
f c2 = {u ∈ V (Ω) : f (u) ≤ c2 }
cannot be deformed into
f c1 = {u ∈ V (Ω) : f (u) ≤ c1 } .
Hence, by combining Lemma 6 with the standard deformation lemma (see [31,
Theorem 4.6] and also [10, Lemma 27.2], [38, Theorem 3.11]) one obtains a con-
strained critical point uΩ such that
S < f (uΩ ) ≤ sup{f (ΦΩ (y)) : y ∈ H} < µ < 24/n S.
Finally, uΩ does not change sign by Lemma 3.
Biharmonic equations at critical growth 133

6. Proof of Theorems 3, 4 and 5

Proof of Theorem 3. We assume that Ω = B = B1 (0) is the unit ball and that
(8) with λ > 0 has a positive solution u. As we have already mentioned above,
the result by Troy [41, Theorem 1] shows that u is radially symmetric. As the right
hand side in (8) is positive, an iterated application of the maximum principle for
−∆ yields that −∆u and u are strictly positive and strictly radially decreasing. We
will exploit the following Pohožaev type identity, which may be taken e.g. from
[42, Lemma 3.9]:
    
2 ∂u ∂
2λ u dx = (−∆u) dS. (36)
B ∂B ∂r ∂r
As u is radially symmetric we may proceed as follows with the term on the right
hand side; here, en denotes the n-dimensional volume of the unit ball B and c(n)
denotes a generic positive constant which may vary from line to line:
   
∂u ∂ ∂u ∂
(−∆u) dS = −n en (1) · (∆u)(1)
∂B ∂r ∂r ∂r ∂r
       
1 ∂u ∂
= − dS (∆u) dS
n en ∂r ∂r
∂B   ∂B

1
= (−∆u) dx ∆2 u dx . (37)
n en B B

The first integral needs some further consideration. First of all, by making use of
homogeneous Navier boundary conditions, we find:
 n  
 
1 − |x|2 ∆2 u dx = − (−2xj )(∆u)xj dx = 2 n (−∆u) dx.
B j=1 B B

Furthermore, as we have already mentioned, u and by (8) also ∆2 u are positive

and strictly radially decreasing. This fact can be used to get rid of the degenerate
weight factor 1 − |x|2 in the previous term:
  
2 2
∆ u dx ≤ ∆ u dx + ∆2 u dx
B |x|≤1/2 1/2≤|x|≤1
  
1
≤ ∆2 u dx + c(n)∆2 u
|x|≤1/2 2
 
≤ ∆2 u dx + c(n) ∆2 u dx
|x|≤1/2 |x|≤1/2

 
≤ c(n) 1 − |x|2 ∆2 u dx.
B

Combining this estimate with the previous identity, we obtain from (36) and (37):
  2
2
λ u dx ≥ c(n) ∆ u dx = c(n)∆2 u21 .
2
(38)
B B
134 F. Gazzola et al.

By a duality argument and using elliptic estimates, we find

u2 ≤ c(n)∆2 u1 , provided n < 8. (39)
Observe that L2 −estimates for the biharmonic operator under homogeneous Navier
boundary conditions follow immediately from the L2 −estimates for the Laplacian.
Combining (38) and (39), we see that for any positive radial solution u of (8),
we have  
λ u2 dx ≥ c(n) u2 dx;
B B
where the constant c(n) is in particular independent of u. As u > 0, we have
necessarily λ ≥ c(n).

Proof of Theorem 4. Assume by contradiction that u is a nontrivial radial solution

of (9). According to the remarks in Sect. 3.2 on the nonexistence of positive
solutions, we may assume that u is sign changing. Further, u cannot have infinitely
many oscillations near ∂B, because in this case, as u ∈ C 4 (B) we would also have
d
∆u = 0 as well as dr ∆u = 0 on ∂B and could extend u by 0 to the whole of
Rn as a solution of the differential equation. The unique continuation property [33]
would give the desired contradiction.
As one may replace u with −u if necessary, we may thus assume that there
exists a number a ∈ (0, 1) such that
u(x) = 0 for |x| = a, u(x) > 0 for a < |x| < 1;
moreover, by a Pohožaev-type identity,
u(x) = ∇u(x) = ∇2 u(x) = 0 for |x| = 1. (40)
Clearly d
dr u(a) ≥ 0. We define

f (x) := |u(x)|2∗ −2 u(x)

and compare u with the solution v of
∆2 v(x) = f (x) for a < |x| < 1,
v(x) = ∇v(x) = 0 for |x| ∈ {a, 1}.

Obviously v > 0 on a < |x| < 1. We now make use of comparison results, which
hold true for radial solutions of biharmonic inequalities in the annulus {x : a <
|x| < 1} and can e.g. be found in [8, Section 2]. First one obtains
d2
v(1) > 0. (41)
dr2
d
Next, as u and v satisfy the same differential equation, if dr u(a) = 0 it follows
that u ≡ v while if dr u(a) > 0 = dr v(a) we find u > v in a < |x| < 1. In both
d d

cases (41) gives

d2
u(1) > 0
dr2
Biharmonic equations at critical growth 135

and hence a contradiction with (40).

Proof of Theorem 5. We keep p ∈ [1, n−4 n

) fixed as in the theorem. Let B = B1 (0)
denote the unit ball, centered at the origin. By scaling and Talenti’s comparison
principle [40, Theorem 1] it suffices to prove that there exists CB > 0 with

1
∆u22 ≥ Su22∗ + u2p (42)
CB

for each u ∈ H 2 ∩ H01 (B), which is radially symmetric and radially decreasing
with respect to the origin (and hence positive). Let us set

∆u22 − λu2p
Sλ,p = inf 1 .
u∈H 2 ∩H0 (B) u22∗
u radially decreasing

If by contradiction (42) does not hold, one gets

∀λ > 0 : Sλ,p < S. (43)

Let us fix some λ ∈ (0, λ1,p ), where

∆u22
λ1,p := inf1 (44)
u∈H 2 ∩H0 (B)\{0} u2
p

is the first positive eigenvalue of

∆2 u = λu2−p
p |u|p−2 u in B
u = ∆u = 0 on ∂B.

Arguing along the lines of [6] and [15], one finds a smooth positive radial strictly
decreasing solution of

∆2 u = u(n+4)/(n−4) + λu2−p
p |u|p−2 u in B
(45)
u = ∆u = 0 on ∂B.

On the other hand, for any positive radially decreasing solution of (45) one has the
following variant of Pohožaev’s identity:
     
n(p − 2) 2 ∂u ∂
λ 2− up = (−∆u) dS. (46)
2p ∂B ∂r ∂r

As in the proof of Theorem 3, we conclude

λu2p ≥ c(n, p)∆2 u21 ≥ c(n, p) u2p , (47)

as we have assumed that p < n/(n − 4). For λ > 0 sufficiently small, we obtain a
contradiction.
136 F. Gazzola et al.

7. Appendix: Proof of Lemma 5

We prove the corresponding statement for the “free” functional

 
1 1
EΩ (u) = |∆u|2 dx − |u|2∗ dx
2 Ω 2∗ Ω

which is defined on the whole space H 2 ∩ H01 (Ω). More precisely, we have

Lemma 8. Let (uh ) ⊂ H 2 ∩ H01 (Ω) be a Palais-Smale sequence for EΩ at level

c ∈ R. Then either (uh ) is relatively compact in H 2 ∩ H01 (Ω) or there exist k > 0
nonzero functions uj ∈ D2,2 (Ω0,j ), j = 1, . . . , k, with Ω0,j either the whole Rn or
a half space, solving either (14) or (17) with boundary condition (2) and a solution
0 ∈ H 2 ∩ H01 (Ω) of (5) such that, as h → ∞:
u

uh 0 in H 2 ∩ H01 (Ω),
u

k 
k
∆uh 22 → u0 22
∆ + uj 22 ,
∆ EΩ (uh ) → EΩ (
u0 ) + EΩ0,j (
uj ).
j=1 j=1

In the second order situation, the corresponding result is due to Struwe [37].
Related generalisations of the Struwe compactness lemma to higher order prob-
lems can be found in [1, 3, 11, 19]. However, none of these applies directly to our
situation. A particular difficulty here arises from the existence of the boundary ∂Ω
in combination with Navier boundary conditions.
Proof of Lemma 8.
Step I. Reduction to the case uh 0.
By well known arguments [6] we know that there exists u ∈ H 2 ∩ H01 (Ω) such

that uh u (up to a subsequence) and EΩ (u) = 0, namely u solves (5). Moreover,
by the Brezis-Lieb lemma we infer that (uh − u) is again a Palais-Smale sequence
for E. Therefore we may assume that

2 n/4
uh 0, E(uh ) → c ≥ S . (48)
n

Indeed, if c < n2 S n/4 , then we are in the compactness range of E and by arguing as
in [6], uh → 0 up to a subsequence. In this case the statement follows with k = 0.
Assuming (48) and arguing as for the proof of [37, (3.2), p.187] we deduce

|∆uh |2 dx ≥ S n/4 + o(1). (49)
Ω

Let L ∈ N be such that B2 (0) is covered by L balls of radius 1. By continuity of

the maps
 
R → sup |∆uh |2 dx, y → |∆uh |2 dx
y∈Ω BR−1 (y) ∩Ω BR−1 (y) ∩Ω
Biharmonic equations at critical growth 137

and (49), for h large enough one finds Rh > 2/diam(Ω) and xh ∈ Ω such that
 
1 n/4
|∆uh |2 dx = sup |∆uh |2 dx = S . (50)
h
B −1 (x ) ∩Ω y∈Ω B −1 (y) ∩Ω 2L
R R
h h

When passing to a suitable subsequence, three cases may now occur:

Case I. Rh  +∞ and (Rh d(xh , ∂Ω)) is bounded;
Case II. (Rh d(xh , ∂Ω)) → +∞;
Case III. (Rh ) is bounded.
Step II. Preliminaries for Case I.
For every x ∈ Rn let us denote by x its projection onto Rn−1 , so that x =

(x , xn ). Since d(xh , ∂Ω) → 0, up to a subsequence xh → x0 ∈ ∂Ω and h :=
Rh d(xh , ∂Ω) → . Moreover, for sufficiently large h (say h ≥ h) there exists a
unique y h ∈ ∂Ω such that d(xh , ∂Ω) = |y h − xh |.
For all h ≥ h, up to a rotation and a translation, we may assume that y h = 0 and
that the tangent hyperplane H to ∂Ω at 0 has equation xn = 0 so that xh −y h ⊥ H.
Then, by the smoothness of ∂Ω, there exist an (n − 1)-dimensional ball Bσ (0) of
radius σ > 0 (independent of h) and smooth maps ψh : Bσ (0) → R such that in
local orthonormal coordinate systems over the tangent hyperplanes at yh , we have:
   
∂Ω ∩ Bσ (0) × [−σ, σ] = (x , ψh (x )) : x ∈ Bσ (0) . (51)
Furthermore, there exists C > 0 (independent of h) such that
|ψh (x )| ≤ C|x |2 , |∇ψh (x )| ≤ C|x |, |D2 ψh (x )| ≤ C ∀x ∈ Bσ (0).
(52)
We now rescale the domain Ω by setting
 
Ωh = Rh Ω − xh ,

so that xh is mapped into the origin 0 while the origin is mapped into (0, −h ) and
(51) becomes
 
∂Ωh ∩ BσRh (0) × [−σRh − h , σRh − h ] =
    
x
= x , Rh ψh − h : x ∈ BσRh (0) .
Rh
Let us set

ωh = (x , x) ∈ Rn : x ∈ BRh σ (0),

    
x x
Rh ψ h − h < xn < Rh ψh − h + R h δ
Rh Rh

where δ > 0 and sufficiently small to have ωh ⊂ Ωh . Consider the injective map
χh : BRh σ (0) × [−Rh δ, Rh δ] → R
n
  
(x , xn ) → x , Rh ψh Rxh + xn − h
138 F. Gazzola et al.

 
and its inverse χ−1
h . We observe that χ h BRh σ (0) × [0, R h δ] = ωh and that
χh is bijective on these sets. Thanks to (52), after some computations, one sees
2
that (χh ) converges in Cloc ({xn ≥ 0}) to the translation by the vector (0, −).
Therefore,
Ω0 = {(x , xn ) ∈ Rn : xn > −}
is the local uniform limit of (Ωh ). In particular, for every ϕ ∈ Cc∞ (Ω0 ) we also
have that ϕ ∈ Cc∞ (Ωh ) for sufficiently large h. This will be used below. Let us
now set  
4−n x
vh (x) = Rh 2 uh xh + , (53)
Rh
so that vh ∈ H 2 ∩ H01 (Ωh ). By boundedness of (uh ) we infer that there exists
C > 0 such that vh H 2 (Ωh ) ≤ C. Let 1Ωh denote the characteristic function of
Ωh , then the sequence (1Ωh vh ) is bounded in D1,2n/(n−2) (Rn ) (and in L2∗ (Rn ))
so that, up to a subsequence, we have
1Ωh vh v0 in D1,2n/(n−2) ∩ L2∗ (Rn ) (54)
where supp(v0 ) ⊆ Ω0 and v0 |xn =− = 0. Moreover, as (1Ωh D2 vh ) is bounded in
L2 (Rn ), by weak continuity of distributional derivatives, we deduce
 
2 2
Dij vh ϕ dx → Dij v0 ϕ dx
Ωh Ω0

for all ϕ ∈ Cc∞ (Ω0 ) and i, j = 1, . . . , n. In particular, v0 has in Ω0 second order

weak derivatives.
Step III. The limiting function v0 in (54) solves (5) in Ω0 .
Fix ϕ ∈ Cc∞ (Ω0 ), then for h large enough we have supp(ϕ) ⊂ Ωh . Define
ϕh ∈ Cc∞ (Ω) by setting
n−4  
ϕh (x) = Rh 2 ϕ Rh (x − xh ) .
Therefore, (D2 ϕh ) being bounded in L2 (Ω) and taking into account (53), one
obtains


n  
o(1) =EΩ (uh )(ϕh ) = Rh2 ∆uh ∆ϕ Rh (x − xh ) dx
 Ω
n−4
2∗ −2
 
− Rh 2
|uh | uh ϕ Rh (x − xh ) dx
 Ω

= ∆vh ∆ϕ dx − |vh |2∗ −2 vh ϕ dx
Rn Rn
 
= ∆v0 ∆ϕ dx − |v0 |2∗ −2 v0 ϕ dx + o(1).
Rn Rn

Then v0 ∈ D2,2 (Ω0 ) solves (5) in distributional sense. The delicate point is to see
that ∆v0 = 0 on ∂Ω0 . To this end, let ϕ ∈ Cc2 (Rn ) with ϕ = 0 on ∂Ω0 and define
 
ϕh (x) = ϕ χ−1 h (x) − (0, ) .
Biharmonic equations at critical growth 139

Notice that for h large enough, also ϕh ∈ Cc2 (Rn ) with ϕh = 0 on ∂Ωh . Taking
into account that (χ−1
h ) tends to the translation by (0, ), we obtain ϕh → ϕ in
C 2 (Rn ) and
 
2∗ −2
(∆v0 ∆ϕ − |v0 | v0 ϕ) dx = lim (∆vh ∆ϕh − |vh |2∗ −2 vh ϕh ) dx
Ω0 h Ωh
  n−4 
= lim ∆uh ∆ Rh 2 ϕh (Rh (x − xh )) −
h Ω
n−4 !
− |uh |2∗ −2 uh Rh 2 ϕh (Rh (x − xh )) dx
 n−4 

= lim EΩ (uh ) Rh 2 ϕh (Rh (x − xh )) = 0.
h

Therefore, by extending any given ϕ ∈ H01 ∩ H 2 (Ω0 ) oddly with respect to xn − 

as a function of H 2 (Rn ) and then by approximating it by a sequence of C 2 functions
(ϕk ) with ϕk = 0 on ∂Ω0 , we get
 
∀ϕ ∈ H01 ∩ H 2 (Ω0 ) : ∆v0 ∆ϕ dx = |v0 |2∗ −2 v0 ϕ dx,
Ω0 Ω0

which, as pointed out in [4, p. 221] implies that v0 is also a strong solution of (5)
in Ω0 .
Step IV. The limiting function v0 in (54) is nontrivial.
Let ϕ ∈ Cc∞ (Rn ) such that Ω0 ∩ suppϕ = ∅; then for h large enough, we may
define
 
v0,h : Ωh ∩ supp(ϕ) → R, v0,h (x) = v0 χ−1 h (x) − (0, ) . (55)

Since v0 ∈ C 2 (Ω0 ), v0 = ∆v0 = 0 on ∂Ω0 and (χh ) converges to the translation

by the vector (0, −), we get as h → +∞

1Ω0 v0 − 1Ωh v0,h L2 (supp(ϕ)) = o(1),

1Ω0 v0 − 1Ωh v0,h L2∗ (supp(ϕ)) = o(1)
(56)
1Ω0 ∇v0 − 1Ωh ∇v0,h L2 (supp(ϕ)) = o(1),
1Ω0 ∆v0 − 1Ωh ∆v0,h L2 (supp(ϕ)) = o(1).

To symplify the notations, in what follows we omit the 1Ωh in front of vh , ∇vh , ∆vh
and 1Ω0 in front of v0 , ∇v0 , ∆v0 . Then by (56) and some computations, one obtains,
as h → +∞
 
2
|∆(ϕv0 − ϕvh )| dx = |∆(ϕv0,h − ϕvh )|2 dx + o(1)
Rn Ωh ∩supp(ϕ)
" #2/2∗
2∗
≥S |ϕ(v0,h − v0 )| dx + o(1)
Ωh ∩supp(ϕ)
 2/2∗
2∗
≥S |ϕ(v0 − vh )| dx + o(1).
Rn
140 F. Gazzola et al.

By compact embedding one has vh → v0 in L2loc (Rn ) and thanks to an integration

by parts one gets
∇(ϕ(vh − v0 ))2 ≤ ϕ(vh − v0 )2 ∆(ϕ(vh − v0 ))2
and therefore ∇vh → ∇v0 in L2loc (Rn ). Hence the previous inequality yields
  2/2∗
2 2 2∗ 2∗
ϕ |∆(v0 − vh )| dx ≥ S |ϕ| |v0 − vh | dx + o(1),
Rn Rn

which implies
  2/2∗
2 2∗
ϕ dµ ≥ S |ϕ| dν ,
Rn Rn
where dµ and dν denote respectively the weak∗ limits of |∆(v0 − vh )|2 and
|v0 − vh |2∗ in the sense of measures.
The rest of this step – to show v0 ≡ 0 – is now very closely along the lines of [1,
Lemma 2] and [38, p. 173]. We emphasize that here, the normalization condition
(50) is exploited.
Step V. In Case II, a solution of (14) appears.
This follows by arguing as in [1], see also [37]. After rescaling as in (53), (vh )
converges to a solution of (14).
Step VI. Case III cannot occur.
By contradiction, assume that (Rh ) is bounded. Then, being Rh > 2/diam(Ω),
we may assume that, up to a subsequence,
xh → x0 ∈ Ω, Rh → R0 > 0.
Let us set Ω0 = R0 (Ω − x0 ) and
 
4−n x
vh = R0 2 uh x0 + .
R0
As in Case I (with the obvious simplifications), one gets for a subsequence vh
v0 ∈ H 2 ∩ H01 (Ω0 ) and that v0 ≡ 0 solves (5) in Ω0 . But this is absurd since
uh 0.
Step VII. Conclusion.
If (uh ) is a Palais-Smale sequence for EΩ , then by Step I its weak limit u 0
solves (5). By Steps III, IV and V the “remaining part” (uh − u 0 ) suitably rescaled
gives rise to a nontrivial solution v0 of (14) (if Case II occurs) or (17) (if Case I
occurs). With help of this solution, we construct from (uh − u 0 ) a new Palais-Smale
sequence (wh ) for EΩ in H 2 ∩ H01 (Ω) at a strictly lower energy level. In the case
where we find the solution v0 in the whole space, one proceeds precisely as in [37],
cf. also [1]. In the case that v0 is a solution of (17) in a half space Ω0 we again have
to use the locally deformed versions v0,h in Ωh of v0 . These have been defined in
(55). Let ϕ ∈ Cc∞ (Rn ) be any cut-off function with 0 ≤ ϕ ≤ 1, ϕ = 1 in B1 (0)
and ϕ = 0 outside B2 (0). We put
(n−4)/2   $ 
wh (x) := (uh − u 0 )(x) − Rh v0,h Rh (x − xh ) ϕ Rh (x − xh ) .
Biharmonic equations at critical growth 141

√
First we remark that wh is indeed well defined as ϕ = 0 for Rh |x − xh | ≥ 2 Rh
and as the domain of definition of v0,h grows at rate Rh . Further we notice that χh
and χ−1
h converge uniformly to translations even on BRh (0). For this reason we
have
 
·
v0,h (·) ϕ √ → v0 in D2,2 (Ω0 )
Rh

and also in D1,2n/(n−2) (Ω0 ) and in L2∗ (Ω0 ). Hence, we have

EΩ (wh ) = EΩ (uh ) − EΩ ( u0 ) − EΩ0 (v0 ) + o(1)


 ∗
and EΩ (wh ) → 0 strongly in H 2 ∩ H01 (Ω) . Now, the same procedure from
Steps II to VI is applied to (wh ) instead of (uh − u0 ). As v0 ≡ 0 and hence
2 n/4
EΩ0 (v0 ) ≥ S ,
n
this procedure stops after finitely many iterations.

Acknowledgement. The authors are grateful to Nicola Garofalo for raising their attention on
the paper [22] and to the referees for a very careful reading of the manuscript.

References
1. Alves, C.O., do Ò, J.M.: Positive solutions of a fourth-order semilinear problem involv-
ing critical growth. Preprint, 2001
2. Bahri, A., Coron, J.M.: On a nonlinear elliptic equation involving the critical Sobolev
exponent: the effect of the topology of the domain. Comm. Pure Appl. Math. 41, 253–
294 (1988)
3. Bartsch, T., Weth, T., Willem, M.: A Sobolev inequality with remainder term and critical
equations on domains with nontrivial topology for the polyharmonic operator. preprint,
2002
4. Bernis, F., Garcı́a Azorero, J., Peral, I.: Existence and multiplicity of nontrivial solutions
in semilinear critical problems of fourth order. Adv. Differential Equations 1, 219–240
(1996)
5. Boggio, T.: Sulle funzioni di Green d’ordine m. Rend. Circ. Mat. Palermo 20, 97–135
(1905)
6. Brezis, H., Nirenberg, L.: Positive solutions of nonlinear elliptic equations involving
critical Sobolev exponents. Comm. Pure Appl. Math. 36, 437–477 (1983)
7. Coron, J.–M.: Topologie et cas limite des injections de Sobolev. C. R. Acad. Sci., Paris,
Sér. I 299, 209–212 (1984)
8. Dalmasso, R.: Elliptic equations of order 2m in annular domains. Trans. Amer. Math.
Soc. 347, 3575–3585 (1995)
9. Dancer, E.N.: A note on an equation with critical exponent. Bull. London Math. Soc.
20, 600–602 (1988)
10. Deimling, K.: Nonlinear Functional Analysis. Springer, Berlin etc. 1985
11. Ebobisse, F., Ahmedou, M.O.: On a nonlinear fourth order elliptic equation involving
the critical Sobolev exponent. Preprint, 2001
12. Edmunds, D.E., Fortunato, D., Jannelli, E.: Critical exponents, critical dimensions and
the biharmonic operator. Arch. Ration. Mech. Anal. 112, 269–289 (1990)
13. Gazzola, F.: Critical growth problems for polyharmonic operators. Proc. Royal Soc.
Edinburgh Sect. A 128, 251–263 (1998)
142 F. Gazzola et al.

14. Gazzola, F., Grunau, H.-Ch.: Critical dimensions and higher order Sobolev inequalities
with remainder terms. NoDEA Nonlinear Differential Equations Appl. 8, 35–44 (2001)
15. Grunau, H.-Ch.: Positive solutions to semilinear polyharmonic Dirichlet problems in-
volving critical Sobolev exponents. Calc. Var. Partial Differential Equations 3, 243–252
(1995)
16. Grunau, H.-Ch.: Polyharmonische Dirichletprobleme: Positivität, kritische Exponenten
und kritische Dimensionen. Habilitationsschrift, Universität Bayreuth 1996
17. Grunau, H.-Ch.: On a conjecture of P. Pucci and J. Serrin. Analysis 16, 399–403 (1996)
18. Grunau, H.-Ch., Sweers, G.: Positivity for perturbations of polyharmonic operators with
Dirichlet boundary conditions in two dimensions. Math. Nachr. 179, 89–102 (1996)
19. Hebey, E., Robert, F.: Coercivity and Struwe’s compactness for Paneitz type operators
with constant coefficients. Calc. Var. Partial Differential Equations 13, 491–517 (2001)
20. Jannelli, E.: The role played by space dimension in elliptic critical problems. J. Differ-
ential Equations 156, 407–426 (1999)
21. Kawohl, B., Sweers, G.: On ’anti’-eigenvalues for elliptic systems and a question of
McKenna and Walter. Indiana U. Math. J., to appear
22. Kenig, C.: Restriction theorems, Carleman estimates, uniform Sobolev inequalities and
unique continuation, in: Harmonic analysis and partial differential equations, (El Esco-
rial, 1987), pp. 69–90, Lect. Notes Math. 1384, Springer 1989
23. Lin, C.S.: A classification of solutions of a conformally invariant fourth order equation
in Rn . Comment. Math. Helv. 73, 206–231 (1998)
24. Lions, P.L.: The concentration-compactness principle in the calculus of variations. The
limit case. I, Rev. Mat. Iberoamericana 1, 145–201 (1985)
25. Luckhaus, S.: Existence and regularity of weak solutions to the Dirichlet problem for
semilinear elliptic systems of higher order. J. Reine Angew. Math. 306, 192–207 (1979)
26. Mitidieri, E.: On the definition of critical dimension. Unpublished manuscript, 1993
27. Mitidieri, E.: A Rellich type identity and applications. Comm. Partial Differential Equa-
tions 18, 125–151 (1993)
28. Mitidieri, E., Sweers, G.: Weakly coupled elliptic systems and positivity. Math. Nachr.
173, 259–286 (1995)
29. Moreau, J.J.: Décomposition orthogonale d’un espace hilbertien selon deux cônes
mutuellement polaires. C. R. Acad. Sci. Paris 255, 238–240 (1962)
30. Oswald, P.: On a priori estimates for positive solutions of a semilinear biharmonic
equation in a ball. Comment. Math. Univ. Carolinae 26, 565–577 (1985)
31. Palais, R.: Critical point theory and the minimax principle. Proc. Symp. Pure Math. 15,
185–212 (1970)
32. Passaseo, D.: The effect of the domain shape on the existence of positive solutions of
∗
the equation ∆u + u2 −1 = 0. Topol. Meth. Nonlinear Anal. 3, 27–54 (1994)
33. Pederson, R.N.: On the unique continuation theorem for certain second and fourth order
elliptic equations. Comm. Pure Appl. Math. 11, 67–80 (1958)
34. Pohožaev, S.I.: Eigenfunctions of the equation ∆u + λf (u) = 0. Soviet. Math. Dokl.
6, 1408–1411 (1965)
35. Pucci, P., Serrin, J.: A general variational identity. Indiana Univ. Math. J. 35, 681–703
(1986)
36. Pucci, P., Serrin, J.: Critical exponents and critical dimensions for polyharmonic oper-
ators. J. Math. Pures Appl. 69, 55–83 (1990)
37. Struwe, M.: A global compactness result for elliptic boundary value problems involving
limiting nonlinearities. Math Z. 187, 511–517 (1984)
38. Struwe, M.: Variational methods, Springer, Berlin etc. 1990
39. Swanson, C.A.: The best Sobolev constant. Applicable Anal. 47, 227–239 (1992)
40. Talenti, G.: Elliptic equations and rearrangements. Ann. Scuola Norm. Sup. Pisa Cl.
Sci. (IV) 3, 697–718 (1976)
41. Troy, W.C.: Symmetry properties in systems of semilinear elliptic equations. J. Differ-
ential Equations 42, 400–413 (1981)
42. van der Vorst, R.C.A.M.: Variational identities and applications to differential systems.
Arch. Ration. Mech. Anal. 116, 375–398 (1991)
Biharmonic equations at critical growth 143

43. van der Vorst, R.C.A.M.: Best constant for the embedding of the space H 2 ∩ H01 (Ω)
into L2N/(N −4) (Ω). Differential Integral Equations 6, 259–276 (1993)
44. van der Vorst, R.C.A.M.: Fourth order elliptic equations with critical growth. C. R.
Acad. Sci. Paris, Série I 320, 295–299 (1995)