This is a preliminary version; the nal version appeared in

Conferenze del Seminario di Matematica dellUniversit` a di Bari 259 (1995), 27-60.

LAPLACE TRANSFORM METHODS FOR
EVOLUTION EQUATIONS
BORIS B

AUMER and FRANK NEUBRANDER *

Louisiana State University, Baton Rouge, USA
The Laplace transform theory of Banach space valued functions and the eld of evolution
equations presently oer many research problems which are of general interest in mathe-
matical analysis. Although there is a longstanding relationship between the two elds, it
was not until recently that the basic transform principles for Banach space valued functions
could be formulated in a satisfactory way. In Section 1 we will give an introduction to
some aspects of the mathematical theory of the Banach space valued Laplace transform.
In contrast to semigroup methods, where the spectral and regularity assumptions on
the characteristic operator are rather restrictive, the Laplace transform method applies to
wellposed and illposed linear evolution problems alike. No a priori spectral assumptions
have to be made, the characteristic operators might not be closed or closable, might not
be densely dened and could be multivalued. The only assumptions needed are linearity
and relative closedness. This class of operators, which includes any sum, product or limit
of closed operators, will be discussed in Section 2.
Besides new theoretical concepts and a unied approach to many aspects of semigroup
theory, Laplace transform theory provides analytic tools for many of the basic problems
in the theory of evolution equations. We will demonstrate this in Section 3. There we will
investigate existence, uniqueness and regularity of solutions of linear evolution equations.
The Laplace transform part of this paper builds on results of W. Arendt (1987, 1991),
W. Arendt, M. Hieber and F. Neubrander (1994), B. Hennig and F. Neubrander (1993), F.
Neubrander (1994, 1994b) and M. Sova (1979, 1980). In applying the Laplace transform
methods to evolution equations we follow ideas developed in papers by Yu. I. Lyubich
(1966), M. Hieber, A. Holderrieth, and F. Neubrander (1991), B. B aumer and F. Neubran-
der (1994), and F. Neubrander (1994).
* supported by Louisiana Education Quality Support Fund
1
1. THE LAPLACE TRANSFORM
The Laplace transform has a long history, dating back to L. Eulers paper De Con-
structione Aequationum from 1737. Since then it has been widely used in mathematics,
in particular in ordinary dierential, dierence and functional equations. An informative
description of the contributions of mathematicians like Euler, Lagrange, Laplace, Fourier,
Poisson, Cauchy, Abel, Liouville, Boole, Riemann, Pincherle, Amaldi, Tricomi, Picard,
Mellin, Borel, Heaviside, Bateman, Titchmarsh, Bernstein, Doetsch, Widder and many
others can be found in two historical surveys by M. Deakin (1981, 1982).
The article Studio della trasformazione di Laplace e della sua inversa dal punto
di vista dei funzionali analitici by the Italian mathematician Sylvia Martis in Biddau
(1933) provides an interesting overall account of the theory prior to D.V. Widders and G.
Doetschs contributions. D.V. Widders books The Laplace Transform (1941) and An
Introduction to Transform Theory (1971) as well as G. Doetschs Theorie und Anwen-
dung der Laplace-Transformation (1937) and his monumental, three volumed Handbuch
der Laplace-Transformation (1950, 1955, 1956) remain useful and modern works. They
are still among the best introductions to the subject.
A rst look at Laplace transform theory for functions with values in a Banach space
X is contained in E. Hilles monograph Functional Analysis and Semi-Groups from 1948.
In trying to extend the results of the classical numerical theory of the Laplace transform
(L) r() =
_

0
e
t
f(t) dt ( > 0)
to Banach space valued functions, E. Hille remarks on several occasions that this can be
done if X is reexive, but not in general. In fact, it was shown by S. Zaidman (1960)
(see also W. Arendt (1987) or Theorem 1.6 below) that some important results of classical
Laplace transform theory extend to a Banach space X if and only if X has the Radon-
Nikodym property (such as reexive spaces). This negative result led to the development
of special Laplace transform theories on arbitrary Banach spaces for functions with ad-
ditional algebraic properties; the most prominent one being the theory of C
0
-semigroups.
There, the link between the generator A and the semigroup T is given via the Laplace
transform
(I A)
1
x =
_

0
e
t
T(t)x dt (x X).
The crucial algebraic property which made it possible to extend classical transform results
to this abstract setting is the semigroup law T(t +s) = T(t)T(s), (t, s 0). In the revised
version of E. Hilles monograph from 1957, E. Hille and R.S. Phillips comment as follows
on the algebraic methods used to solve evolution equations:
2
Thus in keeping in spirit of the times the algebraic tools now play a major role and
are introduced early in the book; they lead to a more satisfactory operational calculus and
spectral theory... On the other hand, the Laplace-Stieltjes transform methods, used by Hille
for such purposes, have not been replaced but rather supplemented by the new tools.
In more recent monographs on semigroup theory the role of the Laplace transform is
reduced to a footnote at best and almost every proof is based on the semigroup property
T(t +s) = T(t)T(s). The major disadvantage of the algebraic approach to linear evolu-
tion equations becomes obvious if one compares the mathematical theories associated with
them (for example, semigroup theories, cosine families, the theory of integro-dierential
equations, etc.). It is striking how similar the results and techniques are. Still, without a
Laplace transform theory for arbitrary Banach space valued functions at hand, every type
of linear evolution equation required its own theory because the algebraic properties of the
operator families involved change. In recent years, in search of a general analytic principle
behind all these theories, the Laplace transform has been reconsidered for that reason.
The rst major result in this direction is due to M. Sova (1979). His analytic Laplace
representation theorem (see Theorem 1.16 below) is behind every generation result for
analytic solution families of linear evolution equations. The real breakthrough came in
1987 with W. Arendts paper on Vector valued Laplace transforms and Cauchy problems.
He showed that Widders Theorem, one of the main theoretical results of the classical
theory, extends to arbitrary Banach spaces if the Laplace transform is replaced by the
Laplace-Stieltjes transform
(L
S
) r() =
_

0
e
t
dF(t) = lim

_

0
e
t
dF(t) ( > 0)
where F Lip
0
([0, ); X); i.e., F(0) = 0 and |F|
Lip
:= sup
t,s0
F(t)F(s)
|ts|
< .
In order to see how the Laplace-Stieltjes transform relates to the Laplace trans-
form, let f L

([0, ); X). Then F(t) :=

_
t
0
f(s) ds is in Lip
0
([0, ); X). In this case
the Laplace-Stieltjes transform reduces to the Laplace transform; i.e.,
_

0
e
t
dF(t) =
_

0
e
t
f(t) dt. We recall that a space X is said to have the Radon-Nikodym property if
the fundamental theorem of calculus holds; i.e., if any absolutely continuous function is
the antiderivative of a Bochner integrable function. Hence, if a space X has the Radon-
Nikodym property, then the Laplace-Stieltjes transform reduces to the Laplace transform.
Otherwise it is a proper generalization. As it turned out, it is the generalization needed in
dealing with Banach space valued functions.
3
Clearly, the Laplace transform can also be considered for f L
p
((0, ); X) and the
Laplace-Stieltjes transform for functions F of bounded p-variation or semivariation (see,
for example, W. Arendt (1993), J. Pr uss (1993), P. Vieten and L. Weis (1993), and L.
Weis (1993)). However, we will restrict the discussion of L and L
S
to the case where
f L

([0, ); X) and F Lip

0
([0, ); X). These special cases are relatively easy to
deal with and have immediate and important applications to evolution equations.
The key to Laplace transform theory is the Riesz-Stieltjes representation of bounded
linear operators from L
1
(0, ) into X. The proof of this well known observation is taken
from F. Neubrander (1994). We will denote by
[0,t]
the characteristic function of the
interval [0, t].
Theorem 1.1 (Riesz-Stieltjes Representation). There exists an isometric iso-
morphism 1
S
: Lip
0
([0, ); X) L(L
1
(0, ), X) given by 1
S
(F) = T where Tg :=
_

0
g(t) dF(t) for all continuous functions g L
1
(0, ) and T
[0,t]
= F(t), for all t 0.
PROOF. Let F Lip
0
([0, ); X) and let D be the linear span of the functions with
compact support which are either continuous or of bounded variation. For such functions
we can dene an operator T
F
on the subspace D by T
F
g :=
_

0
g(s) dF(s) and it is easy
to see that for such g we have that |T
F
g| |F|
Lip
|g|
1
. It follows from F(0) = 0 that
T
F

[0,t]
=
_
t
0
dF(s) = F(t) for all t 0. Since D is dense in L
1
(0, ), there exists a unique
extension of T
F
in L(L
1
(0, ), X), denoted by the same symbol. Clearly, |T
F
| |F|
Lip
and, if T
F
= 0, then T
F

[0,t]
= F(t) = 0 for all t 0. This shows that the linear mapping
1
S
: F T
F
is one-to-one. Moreover, if g is a continuous function in L
1
(0, ), then
g = lim
t
g
[0,t]
in L
1
(0, ). Thus,
T
F
g = lim
t
T
F
(g
[0,t]
) = lim
t
_
t
0
g(s) dF(s) =
_

0
g(s) dF(s).
To prove that 1
S
is onto, let T L(L
1
(0, ), X). Dene F(t) := T
[0,t]
. Then |F(t)
F(s)| = |T
[s,t]
| |T||
[s,t]
|
1
= |T|[ts[. So F Lip
0
([0, ); X) , |F|
Lip
|T|, and
T
[0,t]
= F(t) =
_
t
0
dF(s) = T
F

[0,t]
. Since the characteristic functions
[0,t]
, t 0 are
total in L
1
(0, ), we conclude that T = T
F
and |T
F
| = |F|
Lip
for all F Lip
0
([0, ); X).

4
The Riesz-Stieltjes Representation Theorem is crucial for the following reason. Con-
sidering the Laplace-Stieltjes transform L
S
we want to know how properties of F and its
transform r aect each other. Observe that
r() = T
F
e

and F(t) = T
F

[0,t]
for all , t > 0, where e

denotes the exponential function t e

t
. Thus, if one knows
the function F, then the operator T
F
is determined on the set of characteristic functions,
which is total in L
1
(0, ). Hence T
F
, and in particular T
F
e

= r(), is completely
determined. Conversely, any information on r() for > 0 translates into information on
T
F
on the set of exponential functions, which is also total in L
1
(0, ) (see below). Thus,
r determines the properties of T
F
and, in particular, of T
F

[0,t]
= F(t), (t 0).
Lemma 1.2. Let
n
(n IN) be a sequence of distinct complex numbers with Re
n

> 0 for some > 0. If
(1.1)
N

n=1
1
[
n
1[
[
n
+ 1[

as N , then the exponential functions e

n
are total in L
1
(0, ).
PROOF. a) Let < n
0
IN. Clearly, the monomials t
1
n
0
t
n/n
0
(n IN
0
) are
total in C[0, 1] and thus in L
1
(0, 1). Since : L
1
(0, 1) L
1
(0, ) dened by (g)(t) :=
n
0
e
n
0
t
g(e
n
0
t
) is an isometric isomorphism which maps the monomials onto the set c of
exponential functions e
n
(n n
0
), it follows that c is total in L
1
(0, ).
b) Let H be the closure of the linear span of the functions e

n
(n IN) in L
1
(0, 1).
If H contains the exponential functions e
n
(n n
0
), the claim follows from a). Assume
that there exists m n
0
such that e
m
, H. By the Hahn-Banach theorem there exists
L

(0, ) such that (H) 0 and (e

m
) ,= 0. Let 0 < < . The function
() := (e

) =
_

0
e
t
(t) dt is analytic and bounded for Re and the
function : z
1+z
1z
+ is a conformal map between the unit disk and the halfplane
Re > . Dene h(z) := ((z)) and
n
:=

n
1

n
+1
. Then h is analytic and bounded
on the unit disk and h(
n
) = (
n
) = (e

n
) = 0 for all n IN. Let a
n
:= 1 [
n
[ and
b
n
:= 1
|
n
1|
|
n
+1|
. Then
a
n
b
n
=
[
n
+ 1[
[
n
+ 1[

[
n
+ 1[
2
[
n
1[
2
[
n
+ 1[ +[
n
1[

[
n
+ 1[ +[
n
1[
[
n
+ 1[
2
[
n
1[
2
=
[
n
+ 1[
[
n
+ 1[

Re
n

Re
n

[
n
+ 1[ +[
n
1[
[
n
+ 1[ +[
n
1[
.
5
Since [
n
+ 1[ [
n
1[ and [
n
1[ [
n
+ 1[ 2 it follows that
a
n
b
n

Re
n

Re
n

[
n
+ 1[ +[
n
1[
[
n
+ 1[ +[
n
1[

Re
n

Re
n

2[
n
+ 1[ 2
2[
n
+ 1[
(1

)(1
1
1 +
) > 0.
Since

N
n=1
b
n
as N , we obtain that

N
n=1
a
n
=

N
n=1
1 [
n
[ as
N . Thus h() = 0 for [[ < 1 (see, for example, W. Rudin (1987), Theorem 15.23).
But this implies that () = 0 for Re > contradicting (m) ,= 0. Thus H = L
1
(0, ).

Typical examples of sequences satisfying (1.1) are equidistant sequences

n
= a +
nb (a, b > 0) or convergent sequences
n
, (Re > 0). Examples of sequences not
satisfying (1.1) are given by
n
= n

, ( > 1) and
n
= 1 +in.
Combining Lemma 1.2 with the Riesz-Stieltjes Representation 1.1 we obtain immedi-
ately the following uniqueness theorem for the Laplace-Stieltjes transform (see Yu-Cheng
Shen (1947) or G. Doetsch (1950), Satz 2.9.6).
Corollary 1.3 (Uniqueness). Let F Lip
0
([0, ); X). If
_

0
e

n
t
dF(t) = 0 for a
sequence of complex numbers
n
satisfying (1.1), then F 0.
Next we determine the range of the Laplace-Stieltjes transform. Let F Lip
0
([0, ); X).
Then r() =
_

0
e
t
dF(t) is analytic and
(1.2) r
(n)
() =
_

0
e
t
(t)
n
dF(t)
for n IN
0
and Re > 0. For a proof, see for example, G. Doetsch (1950), Satz 3.2.1 or
W. Arendt, M. Hieber and F. Neubrander (1995). To obtain an estimate for the Taylor
coecients of r let x

and dene f
x
(t) := F(t), x

). Then f
x
Lip
0
([0, )) and
|f
x
|
Lip
|F|
Lip
|x

|. It follows that f
x
is dierentiable a.e. and f
x
(t) =
_
t
0
f

x
(s) ds
for all t 0. Clearly, |f

x
(t)| |F|
Lip
|x

| a.e. and, by (1.2),

k+1
1
k!
r
(k)
(), x

) =
k+1
1
k!
_

0
e
t
(t)
k
f

x
(t) dt.
6
Since
n+1
_

0
e
t t
n
n!
dt = 1 ( > 0, n IN
0
) we obtain [
k+1 1
k!
r
(k)
(), x

)[ |F|
Lip
|x

|
for all x

and > 0. Thus

(1.3) |r|
W
:= sup
kIN
0
sup
>0
|
k+1
1
k!
r
(k)
()| |F|
Lip
.
Let C

W
((0, ); X) := r C

((0, ); X) : |r|
W
< . Then the Widder space
C

W
((0, ); X) is a Banach space and we obtain the following crucial result due to D.V.
Widder (1936) for the numerical case and due to W. Arendt (1987) for the vector-valued
case. The following proof is a modernized version of Widders classical proof. It is taken
from F. Neubrander (1994).
Widders Theorem 1.4. The Laplace-Stieltjes transform L
S
: Lip
0
([0, ); X)
C

W
((0, ); X) is an isometric isomorphism.
PROOF. It follows from (1.3) that L
S
maps Lip
0
([0, ); X) into C

W
((0, ); X) and
that |L
S
(F)|
W
= |r|
W
|F|
Lip
. It follows from Corollary 1.3 that L
S
is one-to-one. It
remains to be shown that L
S
is onto. Let r C

W
((0, ); X). Dene T
k
L(L
1
(0, ), X)
by
T
k
f :=
_

0
f(t)(1)
k
1
k!
_
k
t
_
k+1
r
(k)
_
k
t
_
dt.
Then |T
k
| |r|
W
for all k IN
0
and
T
k
e

= (1)
k
1
k!
_

0
e
t
(
k
t
)
k+1
r
(k)
(
k
t
) dt.
Using change of variables and integration by parts one obtains (with some work; for details,
see D.V. Widder (1941) or B. Hennig and F. Neubrander (1993)),
T
k
e

= u
_

0
1
k!
_
k
u
_
k+1
e
kt
u
t
k
h(t) dt
where h(t) :=
1
t
r(
1
t
) and 1/u := . The functions

k
(t) :=
1
k!
_
k
u
_
k+1
e
kt
u
t
k
(t 0)
7
are approximate identities; i.e., they are positive, their L
1
-norm is one and for all > 0
and all open intervals I [0, ) containing u we have
_
t/ I

k
(t) dt < for all suciently
large k. Since h is continuous in u > 0 and |h|

|r|
W
, it follows from
(1.4)
|
_

0

k
(t)h(t) dt h(u)| = |
_

0

k
(t) (h(t) h(u)) dt|
2|h|

_
t/ I

k
(t) dt + sup
tI
|h(t) h(u)|
that
_

0

k
(t)h(t) dt h(u) as k . Hence,
T
k
e

uh(u) = r(
1
u
) = r()
as k . By Lemma 1.2, the set of exponential functions e

: > 0 is total in
L
1
(0, ). Thus, by the Theorem of Banach-Steinhaus, there exists a bounded operator
T L(L
1
(0, ), X) with |T| |r|
W
such that T
k
f Tf for all f L
1
(0, ). In
particular,
r() = lim
k
T
k
e

= Te

.
It follows from the Riesz-Stieltjes Representation Theorem 1.1 that there exists F
Lip
0
([0, ); X) with |F|
Lip
= |T| |r|
W
such that Tg =
_

0
g(t) dF(t) for all con-
tinuous g L
1
(0, ). Hence, for all > 0,
r() = Te

=
_

0
e
t
dF(t).
Closely related to Widders Theorem is the following inversion of the Laplace transform
which has been widely used in applications to dierential equations.
Theorem 1.5 (Post-Widder Inversion). Let f L
1
loc
([0, ); X). Assume that there
exists IR such that r() =
_

0
e
t
f(t) dt exists for all > . Then for all continuity
points t > 0 of f,
f(t) = lim
k
(1)
k
1
k!
_
k
t
_
k+1
r
(k)
(
k
t
).
8
PROOF. Since we will make no use of this result in the applications below, we will verify it
only for bounded continuous functions f. For a proof of the general case see, for example,
G. Doetsch (1950), Satz 8.2.1 or W. Arendt, M. Hieber and F. Neubrander (1995). If f is
bounded and continuous, then it follows from (1.4) that
f(t) = lim
k
1
k!
_
k
t
_
k+1
_

0
e
ks
t
s
k
f(s)ds.
Now the statement follows from (1.2).
If F : [a, b] X is Lipschitz continuous and g : [a, b] C continuous, then g and F
are Riemann-Stieltjes integrable with respect to each other and
_
b
a
g(t) dF(t) = F(b)g(b)
F(a)g(a)
_
b
a
F(t) dg(t). Thus, for F Lip
0
([0, ); X) and > 0 the Laplace-Stieltjes
transform equals the Laplace-Carson transform, or -multiplied Laplace transform; i.e.,
r() =
_

0
e
t
dF(t) =
_

0
e
t
F(t) dt.
Combining this with the Post-Widder inversion formula we obtain that the inverse Laplace-
Stieltjes transform L
1
S
: C

W
((0, ); X) Lip
0
([0, ); X) is given by
F(t) = lim
k
(1)
k
1
k!
_
k
t
_
k+1
_
r()

_
(k)
=k/t
= lim
k
k

j=0
(1)
j
1
j!
_
k
t
_
j
r
(j)
_
k
t
_
.
Studying whether or not Widders original proof of the surjectivity of L
S
extends to Banach
space valued functions, E. Hille (1948) remarks that Widders
... argument seems to hold for any abstract space in which bounded sets are weakly
compact, but it is not clear in the present writing whether or not his conditions are always
sucient for the existence of the representation.
This statement was made precise by S. Zaidman (1960) and W. Arendt (1987). The
result is the following characterization of Banach spaces with the Radon-Nikodym property.
9
Theorem 1.6 (Riesz Representation). Let X be a Banach space. The following are
equivalent.
(i) The antiderivative operator I : f F, F(t) :=
_
t
0
f(s) ds is an isometric isomorphism
between L

([0, ); X) and Lip

0
([0, ); X).
(ii) The Riesz operator 1 : f T
f
, T
f
g :=
_

0
g(t)f(t) dt is an isometric isomorphism
between L

([0, ); X) and L(L

1
(0, ), X).
(iii) The Laplace transform L : f r, r() :=
_

0
e
t
f(t) dt is an isometric isomorphism
between L

([0, ); X) and C

W
((0, ); X).
PROOF. As shown above, the Riesz-Stieltjes operator 1
S
: F T
F
, T
F
g =
_

0
g(t)dF(t)
is an isometric isomorphism between Lip
0
([0, ); X) and L(L
1
(0, ), X) and the Laplace-
Stieltjes transform L
S
: F r, r() =
_

0
e
t
dF(t) is an isometric isomorphism between
Lip
0
([0, ); X) and C

W
((0, ); X). Now the statement follows from the fact that L =
L
S
I and 1 = 1
S
I on L

([0, ); X).
Let X = L

[0, ) and f(t) :=

[0,t]
(t 0). Since |f(t)f(s)| = 1 for all t ,= s, the
range of f does not contain a countable dense set. Thus, by Pettis Theorem (see E. Hille
and R.S. Phillips (1957), Theorem 3.5.3), f is not measurable. However, f is Riemann
integrable. In fact,
F(t) := (R)
_
t
0
f(s) ds = lim
||0
n

i=1

[0,
i
]
(s
i
s
i1
) = (t )
[0,t]
for all t 0. It is easy to see that F is Lipschitz continuous. However, although F
is Lipschitz continuous and an indenite Riemann integral, F is nowhere dierentiable.
Again, it is easy to check that
1
h
(F(t + h) F(t)) does not converge in X as h 0 for
any t 0. Since F C
0
[0, ) L

[0, ), it follows that C

0
and L

-spaces do not have

the Radon-Nikodym property. Since L

[0, ) L(L
p
[0, )) for 1 p it follows
that L(L
p
[0, )) does not have the Radon-Nikodym property. In fact, there seems to be
no known innite dimensional Banach space X for which L(X) has the Radon-Nikodym
property.
10
The function F : [0, ) c
0
, F(t) := (
1
n
sin(nt))
nIN
is another example of a Lips-
chitz continuous function which is nowhere dierentiable. Thus c
0
and any Banach space
containing it does not have the Radon-Nikodym property.
Examples of spaces with the Radon-Nikodym property are reexive Banach spaces and
spaces with a boundedly complete Schauder basis (like
1
). For a proof of this statement
and a thorough discussion of the Radon-Nikodym property, see J. Diestel and J.J. Uhl
(1977).
As we will see in Section 3, the inversion formulas for the Laplace and Laplace-
Stieltjes transforms play a fundamental role in applications. Compared to the Post-Widder
inversion, it is remarkable that in the following inversion formula (due to G. Doetsch (1950),
Satz 8.1.1) only the values of r(n) for large n IN are needed and that the convergence is
uniform for all t 0.
Theorem 1.7 (Phragmen-Doetsch Inversion). Let F Lip
0
([0, ); X) and r() =
_

0
e
t
dF(t). Then
|F(t)

j=1
(1)
j+1
j!
e
tnj
r(nj)|
c
n
|r|
W
for all t 0 and n IN, where c 1.0159....
PROOF. By the Riesz-Stieltjes Representation Theorem 1.1 and Widders Theorem
1.3, there exists T L(L
1
(0, ), X) such that r() =
_

0
e
t
dF(t) = Te

( > 0),
T
[0,t]
= F(t) (t 0) and |T| = |r|
W
= |F|
Lip
. Thus
|F(t)

j=1
(1)
j+1
j!
e
tnj
r(nj)| |T| |
[0,t]

j=1
(1)
j+1
j!
e
tnj
e
nj
|
1
.
Dene p
n,t
:= 1 e
e
nt
e
n()
=

j=1
(1)
j+1
j!
e
tnj
e
nj
. Then
|
(0,t]
p
n,t
|
1
=
_
t
0
[p
n,t
(s)1[ ds+
_

t
[p
n,t
(s)[ ds =
_
t
0
e
e
n(ts)
ds+
_

t
1e
e
n(ts)
ds
=
1
n
_
e
nt
1
1
u
e
u
du +
1
n
_
1
0
1 e
u
u
du
1
n
__

1
1
u
e
u
du +
_
1
0
1 e
u
u
du
_
for all t 0 and n IN. Now the claim follows from the fact that
_

1
1
u
e
u
du +
_
1
0
1e
u
u
du = 2Ei(1) + 1.0159... where Ei(z) is the exponential integral and
is Eulers constant; see N.N. Lebedev (1972), 3.1.
11
The Phragmen-Doetsch inversion of the Laplace-Stieltjes transform shows that any r
C

W
((0, ); X) (and thus also the unique F Lip
0
([0, ); X) with r() =
_

0
e
t
dF(t)
( > 0)) is uniquely determined by the sequence x
n
:= r(n) (n IN). We ask now the
converse question: i.e. to characterize those sequences (x
n
)
nIN
in a Banach space X for
which there exists r C

W
((0, ); X) (or equivalently F Lip
0
([0, ); X)) such that
x
n
= r(n) =
_

0
e
t
dF(t) =
_
1
0
t
n
dH(t)
(where H(t) := F(lnt)). This special Hausdor moment problem can be solved,
among others, by adapting H. Hahns (1927) method to solve the moment problem (see,
for example, IV.5 in K. Yosida (1971) or Theorem 2.7.6 in E. Hille and R.S. Phillips
(1957)).
Theorem 1.8 (Moment Theorem). Let M > 0 and x
n
X (n IN). The following
are equivalent.
(i) There exists T L(L
1
(0, ), X) with |T| M and x
n
= T(e
n
) for all n IN.
(ii) There exists F Lip
0
([0, ); X) with |F|
Lip
M and x
n
=
_

0
e
nt
dF(t) for all
n IN.
(iii) There exists r C

W
((0, ); X) with |r|
W
M and x
n
= r(n) for all n IN.
(iv) |

n
i=1

i
x
i
| M|

n
i=1

i
e
i
|
L
1 for all n IN and
i
C.
PROOF. The equivalence of (i),(ii) and (iii) follows from the Riesz-Stieltjes represen-
tation and Widders Theorem. The implication (i)(iv) follows from |

n
i=1

i
x
i
|
|T| |

n
i=1

i
e
i
|
L
1.
To prove (iv)(i), let z =

n
i=1

i
e
i
spane
i
, i IN =: c. If z =

n
i=1

i
e
i
=

m
j=1

j
e
j
, then |

n
i=1

i
x
i


m
j=1

j
x
j
| = |

l
k=1

k
x
k
| M|

l
k=1

k
e
k
|
L
1 =
M|

n
i=1

i
e
i

m
j=1

j
e
j
|
L
1 = 0. Thus we can dene on c an operator T by T(z) :=

n
i=1

i
x
i
, where z =

n
i=1

i
e
i
. Now (i) follows from |T(z)| M|z|
L
1 and the density
of c.
We remark that any r C

W
((0, ); X) has a Laplace-Stieltjes representation and thus
an analytic extension for Re > 0 which we denote by the same symbol. Widders growth
conditions sup
>0
|
k+1 1
k!
r
(k)
()| M translate via Cauchys Integral Theorem to the
following complex conditions (see also M. Sova (1980)).
12
Theorem 1.9 (Complex Widder Conditions). Let M > 0 and let r : Re > 0 X
be an analytic function with lim

|r()| = 0 and sup

Re>
|r()| < for all > 0.
The following are equivalent.
(i) r C

W
((0, ); X), with |r|
W
M.
(ii) For some, or equivalently, for all k
0
IN
0
,
sup
>0
_
_
_
_

k+1
1
k!
r
(k)
()
_
_
_
_
M for all k k
0
.
(iii) For some, or equivalently, for all k
0
IN
0
,
sup
>0
sup
s>0
_
_
_
_
1
2
_

r( +it)
(1 ist)
k+2
dt
_
_
_
_
M for all k k
0
.
PROOF. Clearly, (i) implies (ii). First we show that (ii) implies (i). This follows from
the observation that
sup
>0
|
k
0
+1
1
k
0
!
r
(k
0
)
()| M
for some k
0
IN implies that
sup
>0
|
k+1
1
k!
r
(k)
()| M
for all 0 k k
0
1. In order to prove this statement dene
r
k
() :=
(1)
k+1
k!
_

( )
k
r
(k
0
)
() d ( > 0).
These integrals exist absolutely and
|r
k
()|
M k
0
!
k!
_

_
_
_
_
( )
k

k
0
+1
_
_
_
_
d =
M(k
0
k 1)!

k
0
k
.
Since r
(k
0
)
k
0
1
= r
(k
0
)
it follows that (r
k
0
1
r)
(k
0
)
= 0. Thus r
k
0
1
r is a polynomial with
lim

r
k
0
1
() r() = 0. Hence r = r
k
0
1
and
|r
(k)
()| = |r
k
0
k1
()|
M k!

k+1
for all k < k
0
.
Next we show the equivalence of (ii) and (iii). Let > > 0, n IN and let
n
be the
counterclockwise oriented path consisting of the line segments
n,1
:= +i[n, n],
n,2
:=
13
+ n + i[n, n],
n,3
:= [, + n] + in,
n,4
:= [, + n] in. By Cauchys Integral
Theorem we obtain
r
(k+1)
() =
(k + 1)!
2i
_

n
r(z)
(z )
k+2
dz (k IN
0
).
For n the boundedness of r for Re implies that
r
(k+1)
() =
(k + 1)!
2i
_
+iIR
r(z)
(z )
k+2
dz.
Hence
( )
k+2
1
(k + 1)!
r
(k+1)
() = (1)
k+1
( )
k+2
2i
_
+iIR
r(z)
(z )
k+2
dz
= (1)
k+1
( )
k+2
2
_

r( +it)
( +it )
k+2
dt.
=
(1)
k+1
2
_

r( +it)
(1 ist)
k+2
dt,
where s :=
1

. Thus (ii) implies (iii) and (iii) implies that |()

k+2 1
(k+1)!
r
(k+1)
()|
M for all > 0 and all k k
0
. Taking 0 we obtain that (iii) implies (ii).
The following inversion theorem uses the fact that any r C

W
((w, ); X) is a
Laplace-Stieltjes transform and therefore has an analytic extension to Re > 0 which
will be denoted by the same symbol.
Theorem 1.10 (Complex Inversion). The inverse Laplace-Stieltjes transform L
1
S
:
C

W
((0, ); X) Lip
0
([0, ); X) is given by
L
1
S
(r)(t) = F(t) = lim
n
_
1+in
1in
e
t
r()

d,
where the limit is uniform on compact intervals.
14
PROOF. By the Riesz-Stieltjes Representation Theorem 1.1 and Widders Theorem 1.3,
there exist F Lip
0
([0, ); X) and T L(L
1
(0, ), X) such that r() =
_

0
e
t
dF(t) =
Te

( > 0) and T
[0,t]
= F(t) (t 0). Then
|F(t)
_
1+in
1in
e
t
r()

d| |T||
[0,t]

_
1+in
1in
e
t
e

d|
1
.
Now it is an exercise to show that the functions
_
1+in
1in
e
t
e

d converge towards the

characteristic function
[0,t]
in L
1
(0, ) as n uniformly (in t) on compact intervals
(see F. Neubrander (1994b)).
Other consequences of the Riesz-Stieltjes Representation Theorem are Laplace trans-
form versions of Trotter-type approximation theorems of semigroup theory (for further the-
orems of this type, see B. Hennig and F. Neubrander (1993) and F. Neubrander (1994a)).
Theorem 1.11 (Approximation Theorem). Let F
n
Lip
0
([0, ); X) with |F
n
|
Lip

M for all n IN and r
n
() =
_

0
e
t
dF
n
(t), ( > 0). The following are equivalent.
(i) There exist a, b > 0 such that lim
n
r
n
(a +kb) exists for all k IN
0
.
(ii) There exists F Lip
0
([0, ); X) with |F|
Lip
M such that lim
n
r
n
() =
_

0
e
t
dF(t) uniformly (in > 0) on compact sets.
(iii) lim
n
F
n
(t) exists for all t 0.
(iv) There exists F Lip
0
([0, ); X) with |F|
Lip
M such that F
n
(t) F(t) uniformly
(in t 0) on compact sets.
PROOF. By the Riesz-Stieltjes Representation Theorem there exist T
n
L(L
1
(0, ), X)
with |T
n
| M such that T
n
e

= r
n
() and T
n

[0,t]
= F
n
(t) for all n IN, all t 0
and all > 0. Each of the statements imply that the operators T
n
converge on a total
subset of L
1
(0, ). By the theorem of Banach-Steinhaus there exists T L(L
1
(0, ), X)
such that T
n
T as n . Let b > 0. Then the set of characteristic functions

b
:=
[0,t]
: 0 t b and the set of exponential functions E
b
:= e

:
1
b
b
are compact in L
1
(0, ). Hence, the uniformly bounded sequence T
n
converges uniformly
on
b
and E
b
(H.H. Schaefer (1980), Theorem III.4.5). Now the statement follows from
Theorem 1.1.
15
Applying vector-valued Laplace-Stieltjes transform theory to evolution equations, it
is desirable to allow for Lipschitz functions with arbitrary exponential growth. This can
be done by using the following shifting procedure.
Remark 1.12 (The Laplace Transform of Functions with Exponential Growth).
For w IR let Lip
w
([0, ); X) be the space of all functions G : [0, ) X with
G(0) = 0 and |G(t) G(s)| M
_
t
s
e
wr
dr for all 0 s t and some constant M. Then
Lip
w
([0, ), X) is a Banach space with norm
|G|
Lip(w)
:= infM : |G(t) G(s)| M
_
t
s
e
wr
dr for all 0 s t,
and I
w
dened by I
w
G(t) :=
_
t
0
e
ws
dG(s) is an isometric isomorphism between the spaces
Lip
w
([0, ); X) and Lip
0
([0, ); X).
For w IR let C

W
((w, ); X) be the Banach space of all functions r C

((w, ); X)
with norm
|r|
W,w
:= sup
kIN
0
sup
>w
|
1
k!
( w)
k+1
r
(k)
()| < .
The shift S
w
r() := r( w) is an isometric isomorphism between C

W
((0, ); X) and
C

W
((w, ); X). Hence, the Laplace-Stieltjes transform S
w
L
S
I
w
is an isometric isomor-
phism between Lip
w
([0, ); X) and C

W
((w, ); X) and all theorems mentioned so far in
this section can be rephrased for these spaces.
In general, it is often impossible to verify the Widder growth conditions
sup
>
|
1
k!
r
(k)
()|
M
( )
k+1
(k IN
0
)
or the equivalent complex conditions (Theorem 1.10), whereas the growth of r in the whole
complex halfplane can be estimated. In these cases one can use the following representation
theorem.
16
Theorem 1.13 (Complex Representation Theorem). Let q : Re > 0 X
be analytic and sup
Re>
|q()| < . Then, for all b > 0 there exists f C([0, ); X)
with sup
t>0
|e
t
f(t)
t
b
| < such that q() =
b
_

0
e
t
f(t) dt for Re > .
PROOF. For > we dene f(t) :=
1
2i
lim
n
_
+in
in
e
t
q()

b
d. Observe that the
integral is absolutely convergent and that it converges uniformly for t [0, m] (m IN).
Hence f C([0, ); X). Let
,R
be the path +i(, R] +Re
i[

2
,

2
]
+i[R, ).
It follows from Cauchys integral theorem that the denition of f is independent of and
that
f(t) =
1
2i
_

,R
e
t
q()

b
d =
1
2
_
R

e
(+ir)t
q( +ir)
( +ir)
b
dr
+
1
2
_
2

2
e
(+Re
i
)t
q( +Re
i
)
( +Re
i
)
b
Re
i
d +
1
2
_

R
e
(+ir)t
q( +ir)
( +ir)
b
dr
for all R > 0. Hence, for all t 0,
|f(t)|
M

e
t
_

R
1
r
b+1
dr +
M
2
_
2

2
1
R
b
e
(+Rcos )t
d

M
b
1
R
b
e
t
+
M

1
R
b
e
t
_
2
0
e
Rt cos
d.
Choosing R =
1
t
, we obtain for all > that |f(t)| Ct
b
e
t
for some C > 0. Hence,
|f(t)| Ct
b
e
t
. For Re > we choose < < Re and let
R
be the path consisting of
the halfcircle +Re
i[

2
,

2
]
and the line segment +i[R, R]. Then the residue theorem
and Cauchys theorem imply that
_

0
e
t
f(t) dt =
_

0
e
t
1
2i
_
+i
i
e
t
q()

b
ddt
=
1
2i
_
+i
i
1

1

b
q() d
= lim
R
1
2i
_

R
1

1

b
q() d
=
1

b
q().

17
The following is an immediate consequence of the Complex Representation Theorem
1.13. We say that a function q is polynomially bounded for if there exists a
polynomial p such that |q()| p([[) for . Similarly, q is said to be exponentially
bounded if there exists IR and M > 0 such that |q()| Me
||
for all .
Corollary 1.14. Let q be a function with values in X. The following are equivalent.
(i) q is analytic and polynomially bounded for some half-plane Re > .
(ii) There exists b 0 and an exponentially bounded f C([0, ); X) such that q() =

b
_

0
e
t
f(t) dt for some half-line >

.
As another consequence of Theorem 1.13 we obtain the following representation result
due to J. Pr uss (1993).
Corollary 1.15. Let q : Re > 0 X be analytic. If there exists M > 0 such
that |q()| M and |
2
q

()| M for Re > 0, then there exists a bounded function

f C((0, ); X) such that q() =
_

0
e
t
f(t) dt for Re > 0.
PROOF. It follows from the Complex Representation Theorem 1.13 that there are func-
tions f
i
C([0, ); X), (i 0, 1) and C > 0 such that |f
i
(t)| Ct for t > 0,
q() =
_

0
e
t
f
0
(t) dt, and q

() =
_

0
e
t
f
1
(t) dt
for Re > 0. Hence
q

() =
_

0
e
t
f
0
(t) dt
_

0
e
t
tf
0
(t) dt =
_

0
e
t
f
1
(t) dt.
Integrating by parts yields

_

0
e
t
_
_
t
0
f
0
(s) ds tf
0
(t)
_
dt =
_

0
e
t
_
t
0
f
1
(s) ds dt.
Since the Laplace transform is one-to-one, it follows that tf
0
(t) =
_
t
0
f
0
(s) ds
_
t
0
f
1
(s) ds.
Thus f
0
C
1
((0, ); X) and tf

0
(t) = f
1
(t). Therefore |f

0
(t)| C for all t > 0 and
q() =
_

0
e
t
f
0
(t) dt =
_

0
e
t
f

0
(t) dt
if Re > 0.
18
We conclude this section with M. Sovas characterization of those functions q : (, ) X
which are Laplace transforms of analytic, exponentially bounded functions dened on a
sector. For a proof see M. Sova (1979), F. Neubrander (1989, 1994b), W. Arendt (1991) or
W. Arendt, M. Hieber, F. Neubrander (1995). For 0 < we denote by

the open
sector

:= re
i
: r > 0, < < and by

its closure.
Theorem 1.16 (Analytic Representation Theorem). Let 0 <

2
, IR and
q : (, ) X. The following are equivalent.
(i) There exists an analytic function f :

X such that sup

z

|e
z
f(z)| < for
all 0 < < and q() =
_

0
e
t
f(t) dt for all > .
(ii) The function q is analytic in the sector +
+

2
and sup
+
+

2
|()q()| <
for all 0 < < .
Moreover, for all 0 < < there exists a constant C

> 0 such that

(1.5) |z
k
f
(k)
(z)| C

e
Rez
([[[z[ + 1)
k
for all z

.
Let f and q be as in the previous theorem and assume that that |( )q()| M
for all +
+

2
. Let

< . It is easy to see that if q is analytic for all Re >

,
then there exists 0 < < such that q is analytic on

+
+

2
and |(

)q()|

M.
Thus, by (1.5), for all k IN
0
, there exists a constant C > 0 such that
|f
(k)
(t)| Ce

t
([

[ +
1
t
)
k
for all t > 0.
This shows that the exponential growth (at innity) of the function f and its derivatives
is determined by the abscissa of analyticity of its Laplace transform q, i.e., inf IR : q
extends analytically to Re > .
19
2. EVOLUTION EQUATIONS, CAUCHY PROBLEMS AND CLOSED OPERATORS
To simplify the demonstration of the Laplace transform method, we think of a linear
evolution equation as a system of equations which can be written as an abstract Cauchy
problem
(ACP) u

(t) Au(t) ; u(0) = x ; 0 t < T (T )

where A is a linear, but in general unbounded, not closed, not densely dened and even
multivalued operator on some Banach space X. It is clear from applications to partial
dierential equations that A is in general unbounded with possibly nondense domain. As
a typical example of a problem where A might be not closed, consider an inhomogeneous
initial boundary value problem for an integro-dierential equation; i.e.,
w

(t) = A
0
w(t) +
_
t
0
b(t s)Bw(s) ds +f(t), w(0) = w
0
, Cw(t) = g(t).
Here A
0
, B are operators on a Banach space X
0
; C has its domain in X
0
and maps in a
Banach space Y ; w
0
X
0
, f T L
1
loc
([0, ); X
0
), and g ( L
1
loc
([0, ); Y ) (where
T and ( are Banach function spaces). This equation can be written as an (ACP) on
X = X
0
T ( with initial value x = (w
0
, f, g), and
A =
_
_
A
0

0
0
b()B D
F
0
0 0 D
G
_
_
,
where D
F
, D
G
denote the rst derivative operators on T and (, and
D(A) = (x
0
, f, g) : x
0
D(A
0
) D(B) D(C) , f D(D
F
) , g D(D
G
) , Cx
0
= g(0).
As for many other abstract systems of equations (i.e., if A is an operator matrix on
products of Banach spaces), the operator A might not be closed. And even if the operator
A is closable, we might not want to switch to its closure A
c
, because much of the basic
information on the underlying evolution problem is contained in the domain of A and would
be lost by considering the larger and often dicult to describe domain of the closure A
c
.
20
Illposed problems connected with evolution equations (backwards equations, observa-
tion and inverse problems) lead to abstract Cauchy problems where the operator A has
bad spectral properties (like the pointspectrum p(A) covering a right or upper half-
plane). For example, consider the heat equation w
t
(r, t) = w
rr
(r, t) in an innitely long
bar. Suppose we observe the temperature and ux of heat at the point r = 0 over one time
unit and that the initial temperature distribution is unknown. We would like to estimate
from our observations the initial temperature distribution w(r, 0) for 0 r < . This leads
to
w
t
(r, t) = w
rr
(r, t) , w(0, t) = f(t) , w
r
(0, t) = g(t) , t [0, 1] , 0 < r < ,
where f, g C[0, 1]. Interchanging t and r we obtain the initial value problem
w
r
(t, r) = w
tt
(t, r) , w(0, r) = f(r) , w
t
(0, r) = g(r) , r [0, 1] , 0 t < ,
where f, g C[0, 1], and where we ask for the value of w(t, 0) for 0 t < . Clearly,
this initial value problem can be written as an (ACP) if we choose X := C[0, 1] C[0, 1],
x := (f, g), and A :=
_
0 I
D 0
_
, where D denotes the rst derivative on C[0, 1]. Since the
point spectrum of D covers the whole complex plane, the same holds for A. For further
examples of illposed Cauchy problems see, for example, S. Agmon and L. Nirenberg (1963).
As an example of a linear evolution equation which leads to a multivalued linear
operator A, consider the degenerate Cauchy problem
(DCP) Bu

(t) = A
0
u(t) , u(0) = x,
where A
0
, B are linear operators with domains in a Banach space X and ranges in a
Banach space Y . Clearly, (DCP) can be written as an abstract Cauchy problem u

(t)
Au(t) , u(0) = x on the space X, where A := B
1
A
0
X X is given by (x, y) :
x D(A
0
), y D(B), A
0
x = By. Then A is a possibly multivalued linear operator (i.e.,
A(x +y) = A(x) +A(y) for x, y D(A), C), whose graph is, in general, not closed
in X X.
It is one of the major advantages of the Laplace transform method that it applies to
all of the problems above. With it we can study abstract Cauchy problems (ACP) for a
wide variety of operators A, including sums, products and limits of closed operators which
are not necessarily closed or closable, operators which are not densely dened, and those
where the point spectrum p(A) covers C.
21
For simplicity, we will include in the following survey only the case of single valued
operators and refer to C. Knuckles and F. Neubrander (1994) for the multivalued case.
For other related approaches to illposed Cauchy problems see I. Cioranescu and G. Lumer
(1994), R. deLaubenfels (1994), and G. Lumer (1994).
In order to demonstrate the Laplace transform method for evolution equations, it is
convenient to consider relatively closed operators (see S. Agmon and L. Nirenberg (1963),
S.R. Caradus (1973), and B. B aumer and F. Neubrander (1994)). This class of linear
operators is quite large. It contains basically all linear operators appearing in applications
and imposes just enough continuity on the operator A such that A commutes with the
Bochner and Stieltjes integral for suciently regular functions.
If a Banach space Z is continuously embedded in a Banach space X, then we will use
the notation Z X.
Denition 2.1. A linear operator A is called relatively closed in a Banach space X if
there exists an auxiliary Banach space X
A
such that D(A) X
A
X and A is closed in
X
A
X. A relatively closed operator will also be called (X
A
X)-closed.
Examples 2.2 (Sums and Products). An important class of examples of relatively
closed operators are sums and products of closed operators A, B. In general, the sum
S := A + B with domain D(S) = D(A) D(B) and the composition C := BA with
domain D(C) = x D(A) : Ax D(B) will not be closed or closable in X. However,
both S and C are relatively closed if we choose as X
S
and X
C
the Banach space [D(A)]
with the graph norm |x|
A
:= |x| +|Ax|.
We remark that S and C can be relatively closed even if the operators A, B are not
closed themselves. As example take a pair of jointly closed operators A, B on a Banach
space X; i.e., D(A) D(B) x
n
x, and Ax
n
y
1
, and Bx
n
y
2
implies that x
D(A)D(B), Ax = y
1
, and Bx = y
2
(see also N. Sauer (1982)). Then X
S
:= [D(A)D(B)]
with norm |x|
X
S
:= |x| + |Ax| + |Bx| is a Banach space and D(S) X
S
X. It is
easy to see that the sum S is (X
S
X)-closed.
22
As an example of an operator which is not closable in X, but relatively closed, we
consider the following composition of closed operators. Let A be the rst derivative on
X = C[0, 1] with maximal domain and let B be the bounded operator Bf := f(0)g,
where 0 ,= g X. As seen above, the composition Cf := BAf = f

(0)g with domain

D(C) = D(A) = C
1
[0, 1] is relatively closed. However, since there is a sequence f
n
D(C)
with f
n
0 and f

n
(0) = 1, it follows from Cf
n
= f

n
(0)g = g ,= 0 that C is not closable;
i.e., the closure of the graph of C in X X which is given by X Cg is not the graph
of a single-valued operator. Because the multivalued closure of C does not contain any
information about the original operator, and because closedness is absolutely necessary for
most operations, it is necessary to consider the graph as a subset of X
A
X.
In this example one might consider the domain D(C
max
) := f X : f

(0) exists
instead of C
1
[0, 1]. Dene bounded operators on X by A
t
f :=
f(t)f(0)
t
g. Then, for each
f D(C
max
) one has that A
t
f C
max
f as t 0. Because C
max
is the pointwise limit
of closed operators, it follows from the next theorem that it is relatively closed (for proofs
of the following theorems see B. Baumer and F. Neubrander (1994)).
Theorem 2.3. For all t I := (0, 1] let A
t
be an (X
A
t
X)-closed operator. Suppose
there exists a Banach space Y such that Y X
A
t
for all t I. Then Ax := lim
t0
A
t
x
with D(A) = x

tI
D(A
t
) Y : lim
t0
A
t
x exists is (X
A
X)-closed, where
X
A
= x

tI
D(A
t
) Y : |x|
X
A
:= |x|
Y
+ sup
tI
|A
t
x| < .
Theorem 2.4. For all n IN let A
n
be a (X
n
X)-closed operator. Suppose there
exists a Banach space Y such that Y X
n
for all n IN. Then Ax :=

n=0
A
n
x
with D(A) = x

nIN
D(A
n
) Y :

n=0
A
n
x exists is (X
A
X)-closed, where
X
A
= x

nIN
D(A
n
) Y : |x|
X
A
= |x|
Y
+ sup
nIN
|

n
i=1
A
i
x| < .
Theorem 2.5. For all n IN let A
n
be a (X
n
X)-closed operator. Then for all n IN,
C
n
x := A
n
A
0
x with D(C
n
) = x D(A
0
) : A
k
A
0
x D(A
k+1
) for all 0 k
n 1 is (X
C
N
X)-closed, where X
C
n
:= x D(A
0
) : A
k
A
0
x D(A
k+1
) for all
0 k n 2, A
n1
A
0
x X
n
and |x|
C
n
:= |x|
X
0
+ sup
0kn1
|A
k
A
0
x|
X
k+1
<
. If there exists a Banach space Y such that Y X
C
n
for all n IN, then Cx :=
lim
n
C
n
x with D(C) := x

nIN
D(C
n
) Y : lim
n
C
n
x exists is (X
C
X)-
closed, where X
C
:= x

nIN
D(C
n
) Y : |x|
X
C
:= |x|
Y
+ sup
nIN
|C
n
x| < .
23
Clearly, most linear operators that appear in applications can be decomposed into
sums, products and/or limits of relatively closed operators. In order to nd linear operators
which are not relatively closed one may consider small domains. For example, consider
the identity operator I on X = C[0, 1] with the polynomials T as its domain. Clearly, I is
closable. Assume it were relatively closed. Then there exists a Banach space X
I
such that
the graph ( = (p, p) : p T of I is closed in X
I
X. Thus ( is a complete metric space
of rst category, which is a contradiction to Baires theorem. It follows that the operator
I with domain T is not relatively closed.
The following is a classical result of functional analysis due to E. Hille. For a proof,
see E. Hille and R.S. Phillips (1957), Theorem 3.7.12.
Proposition 2.6. Let A be a (X
A
X)-closed operator. Assume that u() maps an
interval I into D(A). If u() : I X
A
and Au() : I X are (improperly) Bochner
integrable, then
_
I
u(t) dt D(A) and
_
I
Au(t) dt = A
_
I
u(t)dt .
3. THE LAPLACE TRANSFORM METHOD
Let A be a relatively closed operator on a Banach space X, x X and 0 < T .
In this section we will study the abstract Cauchy problem
(ACP) u

(t) = Au(t); u(0) = x; 0 t < T

in terms of the asymptotic characteristic equation
(CE) (kI A)y
k
= x a
k
(k
0
k IN),
where a
k
is a sequence with exponential decay T in X; i.e., limsup
k
1
k
ln|a
k
| T.
If T = , we set a
k
= 0 for all k IN. To obtain necessary and sucient conditions for
the existence of solutions, we include the integral equations
(ACP
n
) v(t) = A
_
t
0
v(s) ds +
t
n
n!
x ; 0 t < T
in our considerations (n IN
0
). Formally, the (n + 1)-times integrated Cauchy problem
(ACP
n
) can be obtained by integrating (ACP) (n+1)-times from 0 to t. Then a solution
v() of (ACP
n
) can be thought of as the n-th normalized antiderivative
u
[n]
(t) :=
_
t
0
(t s)
n1
(n 1)!
u(s) ds
of a solution u() of (ACP).
24
For v C([0, T]); X) let v
T
be the 0-continuation of v to IR
+
; i.e., v
T
:= v on [0, T]
and v
T
:= 0 on (T, ). For v L
1
loc
([0, ); X) we denote by abs
X
(v) the inmum of all
a IR for which the Laplace transform
v() :=
_

0
e
t
v(t) dt = lim
r
_
r
0
e
t
v(t) dt
of v exists for Re > a. If 0, then abs
X
(v) if and only if
X
(v
[1]
) , where the
exponential growth bound
X
(v
[1]
) is dened to be the inmum of the numbers

IR for
which there exists a constant M such that |v
[1]
(t)|
X
Me

t
for all t 0 (see G. Doetsch
(1950), Satz 7 [2.2]). Notice that abs
X
(v
T
) =
X
(v
T
) = if T < . We denote by IN

the set of all integers larger than .

The following lemma is the foundation of the Laplace transform method. It shows
that the Laplace transform allows us to switch freely between the Cauchy problem and
the asymptotic characteristic equation for any relatively closed operator A (see B. Baumer
and F. Neubrander (1994)). Notice that the following lemma contains also a uniqueness
result.
Lemma 3.1 (Fundamental Lemma). Let A be (X
A
X)-closed, x X, n IN
0
,
0 < T and 0. Let v C([0, T]; X) with v
[1]
C([0, T]; X
A
), abs
X
(v
T
) and
abs
X
A
(v
[1]
T
) . The following are equivalent.
(i)
_
t
0
v(s) ds D(A) and v(t) = A
_
t
0
v(s) ds +
t
n
n!
x for all t [0, T].
(ii) v
T
() D(A) and (I A)
n
v
T
() = xe
T
_

n
v(T) +

n1
i=0
(T)
i
i!
x
_
if Re > .
(iii) v
T
(k) D(A) for all k IN

and there exists a sequence a

k
of exponential decay T
such that (kI A)k
n
v
T
(k) = x a
k
for all k IN

.
PROOF. We outline the proof given in B. B aumer and F. Neubrander (1994). Assume
(i) holds. If Re > , then (ii) follows from
v
T
() =
_
T
0
e
t
v(t) dt =
_
T
0
e
t
Av
[1]
(t) dt +
_
T
0
e
t
t
n
n!
xdt
=
1

e
T
v(T) +
1

A v
T
() +
1

n+1
x
e
T

n+1
n1

i=0
(T)
i
i!
x.
Clearly, (ii) implies (iii).
25
Assume (iii) holds. Let

> . Then v
[2]
T
Lip

([0, ); X), v
[3]
T
Lip

([0, ); X
A
)
and
y() :=
n
v
T
() =
n+1
_

0
e
t
dv
[2]
T
(t) =
n+2
_

0
e
t
dv
[3]
T
(t).
Since lim
k

j=1
(1)
j+1
j!
e
tkj 1
(kj)
n+2
a
kj
= 0 and Ay(k) = ky(k)x+a
k
for all suciently
large k IN it follows from the (X
A
X)-closedness of A that v
[3]
(t) D(A) and
Av
[3]
(t) = lim
k

j=1
(1)
j+1
j!
e
tkj
Ay(kj)
(kj)
n+2
= lim
k

j=1
(1)
j+1
j!
e
tkj
_
y(kj)
(kj)
n+1

x
(kj)
n+2
+
a
kj
(kj)
n+2
_
= v
[2]
T
(t)
t
n+2
(n + 2)!
x
Now (i) follows from the (X
A
X)-closedness of A.
As the Fundamental Lemma 3.1 shows, a solution of (ACP) or (ACP
n
) can only exist
on [0, T] if the initial value x is in the T-approximate range of the operators I A ;
i.e., there exists a sequence a
k
X of exponential decay T and a sequence y
k
such that
y
k
D(A) and
(3.1) (kI A)y
k
= x a
k
for all suciently large k IN. If there exists a function y which is analytic and polyno-
mially bounded in a right half-plane such that y(k) = y
k
for all suciently large k IN,
then (3.1) is also sucient for the existence of a solution of (ACP
n
) for some n IN
0
(see
B. Baumer and F. Neubrander (1994)).
Theorem 3.2 (Existence and Uniqueness). Let A be a (X
A
X)-closed operator,
x X, and 0 < T . The following are equivalent.
(i) There exists n IN
0
and v C([0, T]; X) with v
[1]
C([0, T]; X
A
), abs
X
(v
T
) < ,
abs
X
A
(v
[1]
T
) < ,
_
t
0
v(s) ds D(A) and
v(t) = A
_
t
0
v(s) ds +
t
n
n!
x for all t [0, T].
(ii) There exist a sequence a
k
in X of exponential decay T and, for some IR, an
analytic, polynomially bounded y : Re > X
A
such that y(k) D(A) and
(kI A)y(k) = x a
k
for all suciently large k IN.
26
Now let T = . It follows from (3.1) that
(3.2) x

Re>
Im(I A)
is a necessary range condition for the existence of a global, Laplace transformable solution
of (ACP
n
) for some n IN
0
.
Assume now that (3.2) holds; i.e., there exists y() D(A) such that (I A)y() = x
for Re > . Then, by the Fundamental Lemma 3.1, (ACP
n
) has a solution for some
n IN
0
if and only if
1

n
y() = v() for some v C([0, ); X) with v
[1]
C([0, ); X
A
),
abs
X
(v) < and abs
X
A
(v
[1]
) < . Given a local resolvent y() solving the char-
acteristic equation (I A)y() = x for some x X, any of the Laplace Representation
Theorems 1.4., 1.8 and 1.9 will lead to a Hille-Yosida type existence result. For example,
Widders Theorem 1.4 leads to the following local version of the Hille Yosida theorem.
Corollary 3.3 (Local Hille-Yosida Theorem). Let A be (X
A
X)-closed, x X,
n IN and 0. The following are equivalent.
(i) There exists a solution v Lip

([0, ); X) of (ACP
n
) with v
[1]
Lip

([0, ); X
A
).
(ii) There exists y C

((, ); X
A
) and M > 0 such that
(a) (I A)y() = x for all IN , > ,
(b) sup
>
|
1
k!
( )
(k+1)
_
1

n1
y()
_
(k)
|
X
M for all k IN
0
,
(c) sup
>
|
1
k!
( )
(k+1)
_
1

n
y()
_
(k)
|
X
A
M for all k IN
0
.
Moreover, if X
A
= X, then condition (c) can be dropped.
PROOF. Assume (i) holds. For Re > dene y() :=
n
v(). Then
1

n1
y() = v()
and
1

n
y() = v
[1]
() for Re > . The statement (a) in (ii) follows from the Fundamental
Lemma 3.1 and the statements (b) and (c) in (ii) follow from Widders Theorem 1.4.
Assume (ii) holds. It follows from Widders Theorem that there exist v Lip

([0, ); X)
and w Lip

([0, ); X
A
) such that
1

n1
y() = v() and
1

n
y() = w() for all > .
Because
1

n
y() = v() = v
[1]
() for all > , it follows from the Uniqueness Theorem
of Laplace transform theory that v
[1]
= w. Since
n
(I A) v() = (I A)y() = x for
all IN, > , the statement (i) follows from the Fundamental Lemma 3.1.
27
If a closed operator A is dissipative, i.e.,
|(I A)x| ( )|x|
for all > and for all x D(A), the growth conditions on y in the corollary above
are automatically satised. In fact, more can be said. It follows from the closedness and
dissipativity of A that for > and x
n
D(A)
(3.3) (I A)x
n
z implies x
n
x D(A) and (I A)x = z.
As mentioned above, it is necessary for the existence of global, Laplace transformable
solutions for the abstract Cauchy problem that the initial data x is an element of

X :=

>
Im(I A).
Let

X x
n
x and > . Then there exists y
n
D(A) such that (I A)y
n
= x
n
x.
By (3.3), there exists y D(A) with (I A)y = x. Thus

X is closed in X whenever (3.3)
holds.
Let Abe closed and dissipative and let x

X. Then there exists y : (, ) D(A)
such that
(I A)y() = x and |y()|
1

| x|
for all > . It follows from
y(
0
) = (I A)
_
y() y(
0
)

_
( > )
that y() D(A)

X for all > . Let

A be the part of A in

X; i.e., D(

A) =
x D(A)

X : Ax

X and

Ax = Ax. Since (I A)(y() x) = Ax for all
x D(A)

X, we obtain that D(

A) = D(A)

X. Furthermore, it follows that y() D(

A)
and (I

A)y() = x for all x

X. Thus I

A is one-to-one (by dissipativity), onto

X, and |R(,

A) x| = |y(|
1

| x| for all x

X. This, the Fundamental Lemma 3.1
and Widders Theorem 1.4 imply the following.
Theorem 3.4 (Dissipative Operators). Let A be closed and dissipative. Then
(a)

X :=

>
Im(I A) is a closed subspace of X.
(b) A leaves

X invariant; i.e., if x D(A)

X then Ax

X. Let

A denote the part of A
in

X. Then D(

A) = D(A)

X, (, ) (

A) and |R(,

A)|
1

( > ).
In particular, if A is densely dened, then

A generates a semigroup on

X.
28
A similar result can be proved for closed, inverse positive operators on Banach lattices
with order continuous norm (see B. B aumer and F. Neubrander (1994)).
We conclude this section with the following local analytic generation theorem.
Corollary 3.5 (Analytic Solutions). Let A be (X
A
X)-closed, 0 <

2
and
IR. Assume that y() D(A) and (I A)y() = x for all > . If y : +
+

2
X
A
is analytic and sup
+
+

2
|( )y()|
X
A
< for all 0 < < , then there exists
an analytic function u :

X
A
such that sup
z

|e
z
u(z)| < for all 0 < < ,
_
t
0
u(s) ds D(A) and
u(t) = A
_
t
0
u(s) ds +x for all t > 0.
PROOF. The statement follows from the Fundamental Lemma 3.1 and the Analytic
Representation Theorem 1.16.
In this section we have indicated how the Laplace transform method can be used to
nd mild or integrated solutions of u

(t) = Au(t), u(0) = x for a particular x X. The

results mentioned are only a small sample of those possible. The Laplace transform method
can also be used to study approximation problems (with approximation theorems similar
to the one of Section 1) as well as the asymptotics of the solutions (with the Abelian and
Tauberian theorems of Laplace transform theory, see for example W. Arendt (1991) or
W. Arendt and C.J.K. Batty (1988)). We have demonstrated how the method works for
the abstract Cauchy problem. However, Laplace transform techniques can also be applied
directly to evolution equations without reducing them to abstract Cauchy problems. Al-
though the direct approach to linear evolution equations has some advantages, the methods
one uses and the problems one encounters remain, in principle, unchanged.
Clearly, the Laplace transform method also allows us to characterize those operators
A for which (ACP) has mild or integrated solutions for all x X (i.e., generators of C
0
-
semigroups or integrated semigroups). A more detailed introduction to Laplace transform
theory, a Laplace transform approach to strongly continuous and integrated semigroups,
and applications of these classes of semigroups to partial dierential equations can be
found in a forthcoming monograph on Laplace Transforms and Evolution Equations by
W. Arendt, M. Hieber and F. Neubrander (1995).
29
REFERENCES
1. S. Agmon and L. Nirenberg (1963). Properties of solutions of ordinary dierential
equations in Banach space. Comm. Pure Appl. Math. 16, 121-239.
2. W. Arendt (1987). Vector valued Laplace transforms and Cauchy problems. Israel J.
Math. 59, 327-352.
3. W. Arendt (1991). Linear Evolution Equations. Lecture Notes, Universit at Z urich.
4. W. Arendt (1993). Vector valued versions of some representation theorems in real
analysis. Preprint.
5. W. Arendt, C.J.K. Batty (1988). Tauberian theorems and stability of one-parameter
semigroups. Trans. Amer. Math. Soc. 306, 837-852.
6. W. Arendt, M. Hieber and F. Neubrander (1995). The Laplace Transform in Banach
Spaces and Evolution Equations. Monograph in preparation.
7. B. Baumer and F. Neubrander (1994). Relatively closed operators and abstract dif-
ferential equations. Preprint. A preliminary version can be found in the LSU Seminar
Notes in Functional Analysis and PDEs. Baton Rouge, 1993.
8. S.R. Caradus (1973). Semiclosed operators. Pacic J. Math. 44, 75-79.
9. I. Cioranescu and G. Lumer (1994). Regularization of evolution equations via ker-
nels K(t), K-evolution operators and convoluted semigroups, generation theorems,
Preprint.
10. M. Deakin (1981). The development of the Laplace transform 1737-1937, I. Euler to
Spitzer, 1737-1880. Arch. Hist. Math., 346-390.
11. M. Deakin (1982). The development of the Laplace transform 1737-1937, II. Poincare
to Doetsch, 1780-1937. Arch. Hist. Math., 351-381.
12. R. deLaubenfels (1994). Existence Families, Functional Calculus and Evolution Equa-
tions. Lect. Notes Math. 1570, Springer Verlag.
13. J. Diestel, J.J. Uhl (1977). Vector Measures. Amer. Math Soc., Providence, Rhode
Island.
14. G. Doetsch (1937). Theorie und Anwendung der Laplace-Transformation. Verlag
Julius Springer, Berlin.
15. G. Doetsch (1950). Handbuch der Laplace-Transformation I. Verlag Birkhauser, Basel.
16. G. Doetsch (1955). Handbuch der Laplace-Transformation II. Verlag Birkhauser,
Basel.
17. G. Doetsch (1956). Handbuch der Laplace-Transformation III. Verlag Birkhauser,
Basel.
30
18. B. Hennig and F. Neubrander (1993). On representations, inversions and approxima-
tions of Laplace transforms in Banach spaces. Applicable Analysis 49, 151-170.
19. M. Hieber, A. Holderrieth and F. Neubrander (1992)). Regularized semigroups and
systems of linear partial dierential equations. Ann. Scuola Norm. Sup. Pisa 19,
363-379.
20. E. Hille (1948). Functional Analysis and Semi-Groups. Amer. Math Soc., Providence,
Rhode Island.
21. E. Hille and R.S. Phillips (1957). Functional Analysis and Semi-Groups. Amer. Math.
Soc. Providence, Rhode Island.
22. C. Knuckles and F. Neubrander (1994). Remarks on the Cauchy problem for multi-
valued linear operators. Partial Dierential Equations, Models in Physics and Biology.
G. Lumer, S. Nicaise, B.-W. Schulze (eds.). Mathematical Research 82, Akademie
Verlag.
23. N.N. Lebedev (1972). Special Functions and their Applications. Dover Publications,
New York.
24. G. Lumer (1994). Evolution equations. Solutions for irregular evolution problems via
generalized solutions and generalized initial values. Applications to periodic shock
models. Annales Saraviensis 5, 1-102
25. Yu. I. Lyubich (1966). The classical and local Laplace transformation in an abstract
Cauchy problem. Uspehi Mat. Nauk 21, 3-51.
26. Sylvia Martis in Biddau (1933). Studio della transformazione di Laplace e della sua
inversa dal punto di vista dei funzionali analitici. Rendiconti del circolo matematico
di Palermo LVII, 1-70.
27. F. Neubrander (1989). Abstract elliptic operators, analytic interpolation semigroups,
and Laplace transforms of analytic functions. Semesterbericht Funktionalanalysis,
Universit at T ubingen.
28. F. Neubrander (1994). The Laplace-Stieltjes transform in Banach spaces and abstract
Cauchy problems. In: Proceedings of the 3rd International Workshop Conference
on Evolution Equations, Control Theory, and Biomathematics, Han-sur-Lesse, Ph.
Clement and G. Lumer (eds.), Lecture Notes in Pure and Applied Math. 155 (417-
431), Marcel Dekker.
29. F. Neubrander (1994b), Laplace transforms and evolution equations. Lecture Notes,
Louisiana State University.
30. J. Pr uss (1993). Evolutionary Integral Equations and Applications. Verlag Birk-
h auser, Basel - Boston - Berlin.
31
31. W. Rudin (1987). Real and Complex Analysis (3rd printing). McGraw-Hill Book Co.,
New York.
32. H.H. Schaefer (1980). Topological Vector Spaces (4th printing). Springer, New York
- Heidelberg - Berlin.
33. N. Sauer (1982). Linear evolution equations in two Banach spaces. Proc. Royal Soc.
Edinburgh 91A, 287-303.
34. M. Sova (1979). The Laplace transform of analytic vector-valued functions. Casopis
pro pestovani matematiky 104, 267-280.
35. M. Sova (1980). Relations between real and complex properties of the Laplace trans-
form. Casopis pro pestovani matematiky 105, 111-119.
36. P. Vieten and L. Weis (1993). Functions of bounded semivariation and classical inte-
gral transforms. Preprint.
37. Yu-Cheng Shen (1947). The identical vanishing of the Laplace integral. Duke Math.
J. 14, 967-973.
38. L. Weis (1993). The inversion of the Laplace transform in L
p
(X)-spaces. In: Pro-
ceedings of the Conference on Dierential Equations in Banach Spaces, Bologna 1991;
North Holland.
39. D.V. Widder (1936). A classication of generating functions. Trans. Amer. Math.
Soc. 39, 244-298.
40. D.V. Widder (1941). The Laplace Transform. Princeton University Press, Princeton,
New Jersey.
41. D.V. Widder (1971). An Introduction to Transform Theory. Academic Press, New
York.
42. S. Zaidman (1960). Sur un theor`eme de I. Miyadera concernant la representation des
functions vectorielles par des integrales de Laplace. Tohoku Math. J. 12, 47-51.
Boris Baumer, Frank Neubrander, Dept. of Mathematics, Louisiana State University,
Baton Rouge, LA, 70803, USA ;
E-mail: [email protected] and [email protected] ;
Phone : 504-388-1612 (oce) ; 504-767-3415 (home); Fax: 504-388-4276.
32