GAL Daniel Eceizabarrena - Distribution Theory and Fundamental Solutions of Differential Operators

Final Degree Dissertation
Degree in Mathematics
Distribution Theory and Fundamental Solutions of

Differential Operators
Author:
Daniel Eceizabarrena Pérez
Supervisor:
Javier Duoandikoetxea Zuazo
Leioa, June 24th 2015

Contents
Contents i
Introduction i
1 Distributions 1
1.1 The space D of testing functions . . . . . . . . . . . . . . . . . . . . 1
1.2 The space D0 of distributions . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.3 Convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.4 Operations on distributions . . . . . . . . . . . . . . . . . . . 8
1.2.5 Differentiation of distributions . . . . . . . . . . . . . . . . . 10
2 Tempered Distributions 15
2.1 The Schwartz space S . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2 The space S 0 of tempered distributions . . . . . . . . . . . . . . . . . 20
2.2.1 Definition and some basic properties . . . . . . . . . . . . . . 20
2.2.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.3 Basic operations on tempered distributions . . . . . . . . . . 24
2.2.4 Convolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.5 Direct product . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3 The Fourier Transform 33

3.1 The Fourier transform in L1 . . . . . . . . . . . . . . . . . . . . . . . 33
3.2 The Fourier transform in S . . . . . . . . . . . . . . . . . . . . . . . 34
3.2.1 Some properties of the Fourier transform in S . . . . . . . . . 34
3.2.2 The Fourier transform of the Gaussian function . . . . . . . . 37
3.2.3 The inverse Fourier transform in S . . . . . . . . . . . . . . . 39
3.2.4 More properties . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3 The Fourier transform in S 0 . . . . . . . . . . . . . . . . . . . . . . . 43
3.3.1 Definition and properties . . . . . . . . . . . . . . . . . . . . 43
3.3.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4 Fundamental Solutions of Differential Operators 49

4.1 Generalised and fundamental solutions . . . . . . . . . . . . . . . . . 49
i
ii CONTENTS
4.2 Three fundamental solutions . . . . . . . . . . . . . . . . . . . . . . . 52

4.2.1 Fundamental solution of the heat operator . . . . . . . . . . . 52
4.2.2 Fundamental solution of the Laplace operator . . . . . . . . . 54
4.2.3 Fundamental solution of the Cauchy-Riemann operator . . . 58
4.3 Fundamental solution of the Cauchy problem . . . . . . . . . . . . . 58
4.3.1 Fundamental solution of the Cauchy problem for the heat op-
erator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.3.2 Fundamental solution of the Cauchy problem for the Schrödinger
operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.4 The Malgrange-Ehrenpreis theorem . . . . . . . . . . . . . . . . . . . 61
A Laurent Schwartz: Life and Work 63
Bibliography 67
Introduction
The objective of this dissertation is to study the theory of distributions and some
of its applications. Even if the vast majority of the theoretical results developed
throughout the Degree correspond to the time before 20th century, we have to focus
on the last 100 years to understand the matter of this document.
Certain concepts which we would include in the theory of distributions nowadays
have been widely used in several fields of mathematics and physics. For example,
many approaches to the very famous delta function were known to be used by
mathematicians such as Cauchy and Poisson, and some more recent as Kirchoff and
Hermite. In their works, they make use of some functions which they call impulse
functions. Nevertheless, it was Dirac who first introduced the delta function as we
know it, in an attempt to keep a convenient notation in his works in quantum mech-
anics. Their work contributed to open a new path in mathematics, as new objects,
similar to functions but not of their same nature, were being used systematically.
Distributions are believed to have been first formally introduced by the Soviet
mathematician Sergei Sobolev1 , in an attempt to find weak solutions to partial
differential equations. Nevertheless, it is the name of Laurent Schwartz2 which is
most deeply connected to distribution theory. His book Théorie des distributions,
published in 1950, is the source of the first systematic development of distributions,
and it highlighted their utility.
The aim of this project is to show how distribution theory can be used to obtain
what we call fundamental solutions of partial differential equations. The connection
between these two areas is given by the Fourier transform, probably one of the most
useful inventions of mathematics.
In short, distributions are linear and continuous functionals that assign a complex
number to every function of a certain space. The definition of the space of testing
functions, D, and of the space of distributions, D0 , will be given first. Neverthe-
less, the Fourier transform does not fit with the first and more basic definition of
distributions, so it will be necessary to introduce a more general space of functions,
the Schwartz space, S, in which the Fourier transform behaves properly. This fact
will naturally imply the existence of a new space of distributions, S 0 , which will be
known as the space of tempered distributions.
1
Sergei Lvovich Sobolev (1908, Saint Petersburg - 1989, Moscow) worked in analysis and partial
differential equations. The very famous Sobolev spaces are named after him.
2
See Appendix A.
i
ii INTRODUCTION
Definition and properties of the Fourier transform in S will be described af-

terwards. This will allow to easily define the Fourier transform of tempered dis-
tributions, being the next and last step to work with differential operators and
their fundamental solutions. To illustrate the usefulness of the developed theory,
some examples concerning the heat equation, the Schrödinger equation, the Laplace
equation and the Cauchy-Riemann equations will be covered.
Chapter 1
Distributions
1.1 The space D of testing functions

As it will be developed in this chapter, distributions are operators from a certain
space of functions to the field of real or complex numbers. Several function spaces
can be defined to do so; the space of testing functions will be our first approach.
Definition 1.1. Let Ω ⊂ Rn be an open set and ϕ : Ω → C a function. We say that

ϕ is a testing function if:
1. supp ϕ is compact, and
2. ϕ is C ∞ .
The space of testing functions in Ω is denoted by D(Ω).
Remark 1.2. Recall that the support of a continuous function is the closure of the
set of points in which the value of the function is not zero. In other words, supp ϕ =
Cl{x ∈ Ω | ϕ(x) 6= 0}. Because of this, it is enough to ask the support to be
bounded, as it will always be closed.
The first immediate property that can be obtained from the definition is given in
the following lemma. From now on, most of the times we will consider t = (t1 , · · · , tn )
to be the general variable in Rn , even if sometimes we will use x = (x1 , · · · , xn ) too.
∂ϕ
Lemma 1.3. If ϕ ∈ D(Ω), then ∂ti ∈ D(Ω), ∀i = 1, · · · , n.
∂ϕ
Proof. It is immediate, as ∂ti
is C ∞ because of the definition of ϕ. Moreover, if
t ∈
/ supp ϕ, as the complementary of the support is an open set, there exists a
∂ϕ
neighbourhood of t in which ϕ = 0. Thus, in that neighbourhood, ∂t i
= 0 and
∂ϕ ∂ϕ ∂ϕ
= 0, which implies that t ∈
∂ti (t) / supp ∂t i
. In other words, supp ∂t i
⊆ supp ϕ,
which completes the proof.
We will be able to give the space D an algebraic structure. The following propos-
ition shows why:
1
2 CHAPTER 1. DISTRIBUTIONS
Proposition 1.4. Let ϕ, φ ∈ D(Ω) and a, b ∈ C. Then, aϕ + bφ ∈ D(Ω). Therefore,

D(Ω) is a linear space.
Proof. Let A = supp ϕ and B = supp φ. Then, aϕ(x) + bφ(x) = 0 for x ∈ / A ∪ B.

This, together with the fact of A∪B being closed, implies that supp aϕ+bφ ⊆ A∪B.
What is more, as A and B are bounded, supp aϕ + bφ is bounded too.
Also, the fact that both ϕ and φ are in C ∞ (Ω) implies aϕ + bφ ∈ C ∞ (Ω). Finally,
being working with functions with values in a field, every property for a linear space
holds trivially.
Once we have defined the concept of testing function, let us show an example.
Example 1.5. Let ζ(t) be defined by parts:
( 1
e t2 −1 |t| < 1,
ζ(t) = (1.1)
0 |t| ≥ 1.
Clearly supp ζ = [−1, 1], and ζ is infinitely smooth whenever |t| 6= 1, so we only
have to check the situation at points |t| = 1. Anyway, as the function is even, it is
enough to analyse the case t = 1. We need to work out the derivatives in (−1, 1).
First,
2t 1
ζ 0 (t) = − 2 e t2 −1 , |t| < 1,
(t − 1)2
and it is fairly easy to observe that every derivative at |t| < 1 will be of the form
Pk (t) 1
ζ (k) = e t2 −1 , |t| < 1,
(t2− 1) 2k
where every Pk is a polynomial. Now, after a change of variables,
Pk (t) 1
t2 −1 =
Pk (1) e1/2t
lim e lim = 0.
t→1− (t2 − 1)2k 4k t→0− t2k
This limit shows that every derivative of ζ is continuous, as ζ (k) (t) = 0 in |t| > 1.
As a consequence, ζ(t) ∈ D(R).
Remark 1.6. Most of the times, we will consider Ω = Rn , so from now on, we will use
the notation D(Rn ) (or simply D if it is clear the space we are working in) instead
of the more general D(Ω).
An interesting fact about testing functions is that one can produce new ones fairly
easy. The following proposition advocates so.
Proposition 1.7. Let ϕ ∈ D(Rn ), τ ∈ Rn , a ∈ R − {0} and g ∈ C ∞ (Rn ). Then,
1. ϕ(t + τ ), ϕ(−t), ϕ(at) ∈ D(Rn ).
2. g(t)ϕ(t) ∈ D(Rn ).
1.2. THE SPACE D0 OF DISTRIBUTIONS 3
Proof. Every new function is infinitely smooth, as they are nothing but composi-
tions and products of C ∞ functions. Also, their supports are compact for being
translations, transpositions or dilations of that of ϕ, which is compact.
We will now define the concept of convergence in D, for it is key to understand
distributions, as we will later see.
Definition 1.8. Let {ϕm (t)}m∈N be a sequence of testing functions. We say it
converges to zero in D(Rn ) if
1. there exists K ⊆ Rn a bounded domain such that supp ϕm ⊆ K, ∀m ∈ N, and
2. each sequence {Dk ϕm (t)}m∈N converges uniformly to zero as m → ∞.
More generally, we say that the sequence {ϕm (t)}m∈N converges to ϕ(t) in D(Rn )
if every supp ϕm is contained in a single bounded domain K ⊂ Rn and the sequence
{ϕm (t) − ϕ(t)}m∈N converges to zero in D. In this case, we will write {ϕm } → ϕ, or
simply ϕm → ϕ.
Notation. We will use the notation Dk ϕ(t) with a multi-index k = (k1 , · · · , kn ) of
non-negative integers to represent the partial derivatives of the function ϕ. In other
words,
∂ k1 +···+kn
Dk ϕ(t) = ϕ(t1 , · · · , tn ).
∂tk11 · · · ∂tknn
We will also write |k| = k1 + · · · + kn .
Once we have defined the basics of testing functions, we are ready to introduce
the space of distributions.
1.2 The space D0 of distributions

As we stated at the beginning of Section 1.1, distributions are a special kind of
functionals, which assign a complex number to each function from a particular space
of functions. Let us define formally what a distribution is.
1.2.1 Definition
Definition 1.9. A mapping T : D(Rn ) → C is called a distribution if:
1. it is linear, in the sense that if ϕ, φ ∈ D(Rn ) and a, b ∈ C, then
T (aϕ + bφ) = aT (ϕ) + bT (φ).
2. it is continuous, in the sense that if {ϕm } → ϕ in D(Rn ), then

{T (ϕm )} → T (ϕ)
in C.
The space of distributions is denoted by D0 (Rn ). Some authors also call distri-
butions generalised functions.
Remark 1.10. In the same way we could define the space of testing functions over
an open set Ω ⊂ Rn , we can also define distributions over Ω. Anyway, as stated in
Remark 1.6, we will consider the whole space Rn .
A distribution can be represented by capital letters such as T , but it is usual to
use the same letters as for functions, such as f and g. Also, if ϕ ∈ D(Rn ), we will
write its image with f ∈ D0 (Rn ) as hf, ϕi.
As an immediate consequence of the definition, we can make the condition of
continuity less restrictive if we know a functional is linear.
Proposition 1.11. If a functional f : D(Rn ) → C is known to be linear, it is

enough to see that, if ϕm → 0 in D(Rn ), then hf, ϕm i → 0 in C in order for f to be
continuous, and thus a distribution.
Proof. Let ϕm → ϕ be a convergent sequence in D(Rn ). Then, by definition, {ϕm −

ϕ} → 0 in D. As f holds the property of continuity for sequences which converge
to zero, we can write hf, ϕm − ϕi → 0. Because of linearity, hf, ϕm − ϕi = hf, ϕm i −
hf, ϕi, and we can assert that hf, ϕm i → hf, ϕi.
1.2.2 Examples
After defining what a distribution is, it will be useful to present some examples of
distributions that will be useful later on.
Example 1.12. We say a function on Rn is locally integrable if it is integrable on
every compact subset of Rn . In this situation, every locally integrable function
can be treated as a distribution. In fact, let f be a locally integrable function
on Rn . Then, for any ϕ ∈ D(Rn ), we can define the integral
Z
hf, ϕi = f (x)ϕ(x)dx. (1.2)
Rn
R
Observe that if K = supp ϕ, then hf, ϕi = K f (x)ϕ(x)dx, and as K is compact
and ϕ is infinitely smooth, f ϕ is integrable on K, and therefore on Rn .
So once seen the functional is well-defined, let us prove it is a distribution. The
linearity of the integral makes it to be linear. Thus, we must check it is continuous.
By Proposition 1.11, we only have to consider the situation of sequences which
converge to zero. So let ϕn → 0 in D. We have to see that
Z
lim hf, ϕn i = lim f (x)ϕn (x)dx = 0.
n→∞ n→∞ Rn
The convergence in D implies that the sequence {ϕn (t)} converges to zero uniformly,
and also that there exists a compact K ⊂ Rn in which the support of every ϕn is
contained. In other words,
∀ > 0, ∃N ∈ N | ∀n > N, |ϕn (t)| < , ∀t ∈ Rn ,

and
Z Z
f (x)ϕn (x)dx = f (x)ϕn (x)dx.
Rn K
So ∀n > N we get
Z Z Z

f (x)ϕn (x)dx ≤ |f (x)||ϕn (x)|dx ≤ |f (x)|dx

K K K
R R
and as K |f (x)|dx is finite, limn→∞ Rn f (x)ϕn (x)dx = 0.
Therefore, making the identification (1.2), we can say that f ∈ D0 (Rn ). Some
authors also write Tf to denote this functional. This kind of distributions are called
regular distributions.
The question that arises is whether all distributions are regular. Fortunately, there
are many more. This fact is the reason to call distributions generalised functions,
as they take in more objects than usual functions. The next two examples illustrate
this fact.
Example 1.13. Let ϕ be a testing function in Rn . The functional δ : D(Rn ) → C
given by
hδ, ϕi = ϕ(0) (1.3)

and called Dirac delta function is a distribution, and it plays a very important
role in many fields not only in mathematics, but also in physics.
It is straightforward to see that the delta function is linear, as for every ϕ, φ ∈ D
and a, b ∈ C,
hδ, aϕ + bφi = aϕ(0) + bφ(0) = ahδ, ϕi + bhδ, φi.
To see it is continuous, let ϕn → 0 be a convergent sequence in D. Observe that
hδ, ϕn i = ϕn (0), so we have to prove that ϕn (0) → 0 in C. But convergence in D
implies uniform convergence, what means that limn→∞ ||ϕn ||∞ = 0, where || · ||∞
stands for the supremum norm. Therefore, limn→∞ |ϕn (0)| = 0.
Now that we have seen that the delta function (1.3) defines a distribution, we want
to see that it is not regular. Let us check this fact for R. By way of contradiction,
let f be a locally integrable function such that
Z
f (x)ϕ(x)dx = ϕ(0), ∀ϕ ∈ D(R).
R
Consider the testing function (1.1), and define
ζn (t) = ζ(nt), ∀n ∈ N.
It holds for every n ∈ N that ζn (0) = ζ(0) = e−1 , and supp ζn = [− n1 , n1 ]. So we can
write Z Z 1/n
f (t)ζn (t)dt = f (t)ζn (t)dt = e−1 .
R −1/n
If we take limits on both sides, we get

Z 1/n
lim f (t)ζn (t)dt = e−1 . (1.4)
n→∞ −1/n
By the definition of ζ, it is clear that Im ζ = [0, e−1 ], so every ζn is bounded by e−1 .

Now observe that
Z 1/n Z
f (t)ζn (t)dt = χ[− 1 , 1 ] (t)f (t)ζn (t)dt,
n n
−1/n R
and |χ[− 1 , 1 ] (t)f (t)ζn (t)| ≤ e−1 χ[−1,1] (t)|f (t)|, so we can bound the integrand by an
n n
integrable function. Thus, we can take the limit in (1.4) inside by the dominated
convergence theorem, and
Z 1/n Z
lim f (t)ζn (t)dt = lim χ[− 1 , 1 ] (t)f (t)ζn (t)dt = 0.
n→∞ −1/n R n→∞ n n
This is a contradiction with (1.4), so the delta function is not a regular distribution.
Distributions which are not regular are called singular distributions.
Example 1.14. In general, we cannot define a distribution through the expression
(1.2) if the function we choose is not locally integrable. This is the case of f (t) = 1/t,
as it is not integrable around the origin in R. We can think of a similar definition
anyway. Let T : D(R) → R be defined the following way:
Z
ϕ(x)
hT, ϕi = lim dx. (1.5)
→0 |x|≥ x
This expression defines a distribution, and it is called Cauchy principal value of

1/x. We denote it by T = Pv x1 . In order to prove it is a distribution, we must
ensure the limit exists. For that, a change of variable allows us to write
− ∞ ∞
ϕ(x) − ϕ(−x)
Z Z Z Z
ϕ(x) ϕ(x) ϕ(x)
dx = dx + dx = dx
|x|≥ x −∞ x x x
Now, we know that ϕ has compact support, so we can find a positive constant k
such that supp ϕ ⊆ [−k, k], and thus the limit (1.5) turns into
k
ϕ(x) − ϕ(−x)
Z
lim dx.
→0 x
By setting χ for the characteristic function,

Z ∞
1 ϕ(x) − ϕ(−x)
hPv , ϕi = lim χ[ 1 ,k] (x) dx.
x n→∞ 0 n x
We also know that ϕ ∈ C ∞ (R), so by the mean value theorem we know that for any
x > 0 we can find ξ ∈ (−x, x) such that ϕ(x) − ϕ(−x) = 2xϕ0 (ξ). Moreover, ϕ ∈ D
implies that ϕ0 ∈ D, so it is bounded and thus
|ϕ(x) − ϕ(−x)|
≤ 2||ϕ0 ||∞ .
x
This allows us to write

χ 1 (x) ϕ(x) − ϕ(−x) ≤ 2||ϕ0 ||∞ χ[0,k] (x)

[ n ,k] x
and the integrand remains bounded by an integrable function on R. By the domin-

ated convergence theorem, limit and integral may be interchanged and
Z ∞ Z k
1 ϕ(x) − ϕ(−x) ϕ(x) − ϕ(−x)
hPv , ϕi = lim χ[ 1 ,k] (x) dx = dx. (1.6)
x 0 n→∞ n x 0 x
Finally, the integral is finite because the singularity at x = 0 represents a removable
discontinuity. Indeed, L’Hôpital’s rule leads to
ϕ(x) − ϕ(−x)
lim = 2ϕ0 (0).
x→0 x
After seeing the functional is well defined, linearity is obtained using expression
(1.6) and the linearity of the integral. To check continuity, let {ϕn } converge to
zero in D. We know that their supports are contained in a compact subset K ⊂ R.
Choose k > 0 big enough so that K ⊂ [−k, k]. Therefore,
Z k
1 ϕn (x) − ϕn (−x)
hPv , ϕn i = dx, ∀n ∈ N.
x 0 x
Again, by the same reasoning as before,
ϕn (x) − ϕn (−x)
≤ 2||ϕ0n ||∞ .
x
Observe also that convergence in D implies that the sequence {ϕ0n } converges uni-
formly to zero. As every ϕ0n is bounded, then there exists M > 0 such that
||ϕ0n ||∞ < M, ∀n ∈ N. This way, the dominated convergence theorem allows us
to write Z k
1 |ϕn (x) − ϕn (−x)|
lim hPv , ϕn i ≤ lim dx = 0.
n→∞ x 0 n→∞ x
So eventually, limn→∞ hPv x1 , ϕn i = 0 and the Cauchy principal value is continuous.

As a consequence, the Cauchy principal value is a distribution.

1.2.3 Convergence
The space of distributions can be given a concept of convergence, which in some
cases will be useful.
Definition 1.15. Let {fn } be a sequence of distributions. We say that the sequence
converges to the distribution f if for every testing function ϕ, the sequence hfn , ϕi
converges to hf, ϕi.
We also say that the sequence {fn } simply converges if every sequence hfn , ϕi
converges.
It can be proven that if a sequence of distributions converges, then the functional
f which assigns to every testing function the limit value of the sequence defines a
distribution. A proof can be consulted in [10, p. 37-39].
1.2.4 Operations on distributions

Once we have analysed several examples of distributions, it is time to define some
operations on them. First, we will define the most simple ones, which will allow to
determine an algebraic structure over D0 (Rn ). So let f, g ∈ D0 (Rn ) and α ∈ C. We
define:
1. The sum of distributions, f + g, as

hf + g, ϕi = hf, ϕi + hg, ϕi, ∀ϕ ∈ D.
2. The multiplication by a constant, αf , as

hαf, ϕi = αhf, ϕi, ∀ϕ ∈ D.
It is straightforward to check that these two functionals define distributions. With

them, we have defined an internal operation in D0 and an external product over C.
Again, for C being a field, D 0 (Rn ) is a linear space.
These first two operations are trivial to define and they stand naturally. There
are some others that also arise naturally if we work with regular distributions, and
we will be able to generalise them to even singular distributions. So let τ ∈ Rn and
a ∈ R − {0}. We define:
3. The shifting of a distribution, f (t − τ ), as

hf (t − τ ), ϕi = hf, ϕ(t + τ )i, ∀ϕ ∈ D.
This operator is well-defined, as we saw in Proposition 1.7 that a shifted testing
function is a testing function. It is fairly simple to justify this definition.
Indeed, if f is a regular distribution, then changing variables,
Z Z
hf (t − τ ), ϕ(t)i = f (t − τ )ϕ(t)dt = f (y)ϕ(y + τ )dy
Rn Rn
= hf (t), ϕ(t + τ )i.
4. The transposition of a distribution, f (−t), as
hf (−t), ϕi = hf, ϕ(−t)i, ∀ϕ ∈ D.
This is also a well-defined functional, because ϕ(−t) is a testing function, as

we checked in Proposition 1.7. The justification is again given in terms of
regular distributions; if f is so, then
Z Z
hf (−t), ϕ(t)i = f (−t)ϕ(t)dt = f (y)ϕ(−y)dy = hf (t), ϕ(−t)i.
Rn Rn
5. The dilation of a distribution, f (at), as

1 t
hf (at), ϕi = hf, n ϕ i, ∀ϕ ∈ D.
|a| a
Proposition 1.7 ensures that this functional is well-defined. As in the previous
definitions, if f is a regular distribution,
Z Z y 1
hf (at), ϕ(t)i = f (at)ϕ(t)dt = f (y)ϕ dy
Rn n a |a|n
R
1 t
= hf (t), n ϕ i.
|a| a
Apart from seeing that these new functionals are well-defined, to check they define
distributions is an easy exercise. Linearity is granted by that of f . To see they are
continuous, let {ϕn } be convergent to zero in D. If we prove that the sequences
{ϕn (t + x)}, {ϕn (−t)}, {ϕn at } converge to zero in D, we will get the desired result

automatically by the continuity of f . Now, if K ⊂ Rn is the domain containing

every supp ϕn , then a translation, a transposition or a dilation of K contains every
support of the elements of the new sequences. Furthermore, every derivative of
the new functions is a translation, transposition or a dilation of the derivative of
the original functions, without taking constant factors into account. This implies
that the suprema of the new derivatives are the same as the ones of the original
derivatives (avoiding amplification constants), making them converge uniformly to
zero. Thus, the new sequences converge to zero in D.
There is one more operation that we can define. Let g ∈ C ∞ . We define:
6. The multiplication of a distribution by a smooth function, gf , as
hgf, ϕi = hf, gϕi, ∀ϕ ∈ D.
This is a completely trivial definition. Indeed, if we consider f to be a regular

distribution, then as g is infinitely smooth, the product gf is also locally
integrable, and it defines a regular distribution,
Z Z
hgf, ϕi = g(t)f (t)ϕ(t)dt = f (t)(g(t)ϕ(t))dt = hf, gϕi,
Rn Rn
as Proposition 1.7 asserts that gϕ ∈ D. The general proof that gf is indeed a

distribution is obtained by very similar arguments as in previous cases.
A particularly interesting case is the one concerning the delta function. In
fact,
hgδ, ϕi = hδ, gϕi = g(0)ϕ(0) = g(0)hδ, ϕi,
so eventually,
gδ = g(0)δ, ∀g ∈ C ∞ (Rn ). (1.7)
After being able to define a product of a distribution with a certain kind of func-
tions, we wonder whether a general product between distributions can p be defined.
Unfortunately, this is not the case. Think of the function f (t) = 1/( |t|). This is
a locally integrable function on R, thus defines a distribution. But f 2 (t) = 1/|t|,
which is known not to be integrable on neighbourhoods of the origin. Hence, we
cannot define a distribution through the expression
Z
1 ϕ(t)
h , ϕi = dt,
|t| R |t|
because the integral will not exist for every testing function ϕ. So in general, we
will not be able to define the product of distributions. This is one of the
major drawbacks, if not the principal, of the theory of distributions, although many
attempts to fix it have been carried.
Up to now, we have defined what we can call basic operations in the space D0 . But
it is possible to define much more powerful operations which will play an important
role in reaching our ultimate objective. Differentiation will be discussed next, and
some others as convolution and the direct product will be treated in Chapter 2.
1.2.5 Differentiation of distributions

Distributions have many convenient properties which do not hold for usual functions,
making them very useful, for they can reach results we would not be able to explore
considering only standard functions. Differentiation is one of them. In fact, even if
a locally integrable function may not be differentiable at some points, it is possible
to define derivatives of every order for distributions. So let us define the distribution
differentiation process.
Definition 1.16. Let f be a distribution in D0 (Rn ). The partial derivatives of f

are defined as
∂f ∂ϕ
h , ϕi = hf, − i, ∀ϕ ∈ D(Rn ),
∂ti ∂ti
for every i ∈ {1, · · · , n}.
As most of the properties for distributions, this definition has its origins in the
behaviour of regular distributions. To see this, let us consider a regular distribution
f ∈ D0 (R), for which the function f is smooth. Then, its derivative f 0 is also smooth,
so it defines a distribution and the integration by parts yields
Z Z Z
0 0 0
hf , ϕi = f (t)ϕ(t)dt = (f ϕ) (t)dt − f (t)ϕ0 (t)dt.
R R R
But as ϕ is a testing function, its support is compact, and thus the fundamental
theorem of calculus forces the first integral to be zero. Because of this, we can write
hf 0 , ϕi = −hf, ϕ0 i, ∀ϕ ∈ D.
which is the one dimensional version of Definition 1.16.

We can expect the derivative of a distribution to be a distribution too. This is in
fact true, as it will be shown in the next proposition.
∂f
Proposition 1.17. Let f be a distribution. Then, the expression for ∂ti in Defini-
tion 1.16 defines a distribution.
Proof. The functional is well-defined, Lemma 1.3 showing that every partial deriv-
ative of a testing function is also a testing function. Its linearity is provided by that
of the derivative. Now, take a sequence of testing functions {ϕn } which converges
to zero in D. Then, as seen in the proof of Lemma 1.3, the support of the derivative
of a testing function is contained in the support of the testing function itself, so
as there exists a subset K ⊂ Rn containing every supp ϕn , it also contains every
supp ∂ϕ α ∞
∂ti . Moreover, we know that every {D ϕn }n=1 converges to zero uniformly,
n
α being a multi-index. So we are allowed to say that the sequence { ∂ϕ

∂ti } converges
n
to zero in D. Finally, as f is a distribution,

∂f ∂ϕn
h , ϕn i = −hf, i
∂ti ∂ti
converges to zero in C. This shows that the functional is continuous, and thus a
distribution.
Once we have fixed the definition of the derivatives of distributions, we are ready to
analyse some properties, which in some cases will enhance the performance of usual
functions. For example, recall Schwarz’ theorem, which allows to commute partial
derivatives of a twice differentiable function. Not only is this true for distributions,
but it also holds with no restrictions at all.
Theorem 1.18. Let f be a distribution. Then, for every i, j ∈ {1, · · · , n},
∂2f ∂2f
=
∂ti ∂tj ∂tj ∂ti
Proof. It is an immediate consequence of the fact that testing functions are infinitely
smooth, thus Schwarz’ theorem holds for them.
It is also interesting to note that the product rule also holds for distributions.
Proposition 1.19. Let f ∈ D0 (Rn ) and g ∈ C ∞ (Rn ). Then, for any i ∈ {1, · · · , n},
∂ ∂g ∂f
(gf ) = f +g .
∂ti ∂ti ∂ti
Proof. Let ϕ be a testing function in Rn . Then,
∂(gf ) ∂ϕ ∂ ∂g ∂f ∂g
h , ϕi = −hf, g i = −hf, (gϕ)i + hf, ϕi = hg , ϕi + h f, ϕi.
∂ti ∂ti ∂ti ∂ti ∂ti ∂ti
The derivative of distributions, by the property seen in Proposition 1.17, can be
seen as an operator in the space of distributions. In some moments, it will be useful
to know it is linear and continuous.
Proposition 1.20. Differentiation is a linear operation in the space of distributions,

and it is continuous in the sense that if a sequence of distributions fn converges to
f in D0 , then Dk fn converges to Dk f in D0 .
Proof. Linearity is trivial. For continuity, Proposition 1.17 and convergence of fn

implies that
hDk fn , ϕi = (−1)|k| hfn , Dk ϕi → (−1)|k| hf, Dk ϕi = hDk f, ϕi.
An interesting question is what will the derivatives of some particular distributions

look like. We have seen that if we are dealing with regular distributions which come
from a differentiable function, the usual derivative coincides with the distributional
derivative, but we know nothing about the remaining ones, such as the delta function.
Example 1.21 (Delta function). We know that for every ϕ ∈ D(Rn ),
hδ, ϕi = ϕ(0),
so if we write the derivative with respect to variable i ∈ {1, · · · , n}, we get
∂δ ∂ϕ ∂ϕ
h , ϕi = −hδ, i = − (0).
∂ti ∂ti ∂ti
In general, from the above we immediately get that if k = (k1 , · · · , kn ) ∈ (Z+ )n ,

then
hDk δ, ϕi = (−1)|k| Dk ϕ(0).
Another interesting example is the Heaviside function, whose derivative will be
the delta function.
Example 1.22 (Heaviside function). We define the Heaviside function as the char-
acteristic function of the interval [0, +∞] ⊂ R. It is usually denoted by H(t), but
also as 1+ (t) or θ(t). If we compute its derivative, for ϕ ∈ D(R),
Z Z ∞
hH 0 (t), ϕi = −hH(t), ϕ0 i = − H(t)ϕ0 (t)dt = − ϕ0 (t)dt.
R 0
Now, as ϕ has compact support, we can find k > 0 such that the support is contained
in [−k, k], and
Z ∞ Z k
0
− ϕ (t)dt = − ϕ0 (t)dt = −ϕ(k) + ϕ(0) = ϕ(0) = hδ, ϕi.
0 0
Therefore, in terms of distributions, we get
H 0 (t) = δ(t). (1.8)
Notice that if we consider H as a function, H 0 (t) = 0 holds almost everywhere.

This example shows that, in general, the distributional derivative of a function does
not coincide with the distribution associated to its derivative if the latter exists only
almost everywhere. Therefore, the derivative of a function must exist in every point
of its domain in order to ensure the distributional derivative coincides with the usual
derivative.
Chapter 2
Tempered Distributions
In this chapter we will introduce a new space of distributions, the space of tempered
distributions. Its importance will be covered in the next chapter, as the main mo-
tivation to define it is to be able to define the Fourier transform for distributions.
The space of tempered distributions, which we call S 0 , is a proper subspace of the
space D0 we already know. It can be defined, as we are to show, through a wider
class of testing fuctions.
2.1 The Schwartz space S

As we have proposed, we want to introduce a space of functions containing the space
of testing functions described in Section 1.1.
Definition 2.1. Let φ : Rn → C be a function. We say that φ is a Schwartz
function if
1. φ is C∞ , and
2. for every m ∈ Z+ , k ∈ (Z+ )n , there exists a constant Cm,k > 0 such that
|t|m |Dk φ(t)| ≤ Cm,k , ∀t ∈ Rn .
The space of all Schwartz functions is called Schwartz space, and it is denoted by
S(Rn ).
Remark 2.2. The Schwartz functions are also called functions of rapid descent.
The reason for this is that Condition 2 in Definition 2.1 is equivalent to
Cm,k
|Dk φ(t)| ≤ .
|t|m
This way, every Schwartz function and its derivatives must decrease to zero as |t|
tends to infinity, not slower than any power of 1/|t|.
Once we have defined the Schwartz space, it is time to present some very basic
properties.
15
16 CHAPTER 2. TEMPERED DISTRIBUTIONS
Proposition 2.3. Let φ ∈ S(Rn ) and k ∈ (Z+ )n a multi-index. Then,
1. S(Rn ) is a linear space.
2. Dk φ ∈ S(Rn ).
Proof. By the triangle inequality, it is trivial to see that the sum of two Schwartz
functions and the product of a Schwartz function with a complex constant are again
Schwartz functions. This way, as every operation is performed in a field, S(Rn ) is a
linear space. Finally, to see 2 holds it is enough to check that for every i ∈ {1, · · · , n},
∂φ n ∞ implies that the partial derivative is so.
∂ti ∈ S(R ). Again, the fact that φ is C
For the second condition, we must see that for every m ∈ Z+ and k 0 ∈ (Z+ )n , there
0
exists Cm,k > 0 such that

k0 ∂φ
m
|t| D ≤ C0 0, ∀t ∈ Rn .
m,k
∂ti
Now, if k 0 = (k1 , · · · , kn ), and we call k = (k1 , · · · , ki + 1, · · · , kn ), we can write

m k0 ∂φ
= |t|m |Dk φ| ≤ Cm,k , ∀t ∈ Rn ,

|t| D
∂ti
0
so if Cm,k is the constant corresponding to φ and m, k, then Cm,k 0 = Cm,k , which
completes the proof.
In the beginning of the section, we have claimed our intention to define a space
wider than the space of testing functions. Let us state this fact in the following
proposition.
Theorem 2.4. If D(Rn ) is the space of testing functions and S(Rn ) is the Schwartz
space, then
D(Rn ) ⊂ S(Rn ),
the inclusion being strict.
Proof. Let φ ∈ D be a usual testing function. Then, as it is C ∞ , we only have to

check the second condition in Definition 2.1. So let m ∈ Z + and k ∈ (Z+ )n . The
objective is to see that |t|m |Dk φ(t)| is bounded. We know that supp φ is bounded,
so there exists r > 0 such that supp φ ⊂ B(r), where B(r) denotes the closed
ball centered in the origin and of radius r. We also proved in Lemma 1.3 that
supp Dk φ ⊂ supp φ. So we analyse two different situations:
/ B(r), then Dk φ(t) = 0, and thus |t|m |Dk φ(t)| = 0.

• If t ∈
• If t ∈ B(r), then we are considering a C ∞ function in a compact domain. This

implies that Dk φ(t) is bounded in B(r), so there exists K > 0 such that
|t|m |Dk φ(t)| ≤ rm K < ∞.

2.1. THE SCHWARTZ SPACE S 17
So we see that |t|m |Dk φ(t)| is bounded, what shows that φ ∈ S, and thus D ⊆ S.
We want to prove that indeed D ⊂ S. So let φ be
2
φ(t) = e−|t| . (2.1)
We know that φ ∈ C ∞ . Besides, it is clear that its support is not bounded, as

φ(t) 6= 0, ∀t ∈ Rn . Thus, φ ∈/ D. The question is whether φ ∈ S. It is easy to see
that for every k ∈ (Z+ )n , Dk φ(t) = Pk (t)φ(t), where Pk (t) represents a polynomial.
So, we can write
|t|m |Pk (t)|
|t|m |Dk φ(t)| = , (2.2)
e|t|2
expression that tends to zero as |t| tends to infinity. This means that for every > 0,
we can find a constant N > 0 such that if |t| > N ,
|t|m |Pk (t)|

< .
e|t|2
Thus, if we consider |t| ≤ N , the expression (2.2) is bounded for being a C ∞ function
on a compact domain, say by a constant M > 0. Therefore, consider Cm,k =
max{M, }, and so,
|t|m |Dk φ(t)| ≤ Cm,k ,
which shows that φ ∈ S.
Knowing that continuity plays an important role in distributions, next step is to

describe convergence in the space S.
Definition 2.5. Let {φn (t)}∞ n=1 be a sequence in the Schwartz space S. We say that
the sequence converges to zero in S if for every m ∈ Z+ , k ∈ (Z+ )n , the sequence
{|t|m Dk φn (t)}∞
n=1 converges to zero uniformly.
Following this definition, we say that the sequence {φn (t)}∞n=1 converges to φ if
the sequence {φn (t) − φ(t)}∞n=1 converges to zero.
Remark 2.6. As in the case of D, for simplicity, the sequence {φn (t)}∞
n=1 will also
be denoted φn , and convergence to the function φ will be written as φn → φ.
We have to take special care with the relationship between convergences in D and
in S. Suppose we are working with a sequence in S. Then, this sequence may not
be in D, so it makes no sense to ask about convergence in D. On the other hand,
every sequence in D is a sequence in S, so in order to keep a coherent structure, we
will have to ensure that convergence in D implies convergence in S. That is what
we present in the following proposition.
Proposition 2.7. Let {φn } be a convergent sequence in D. Then, it is a convergent

sequence in S.
Proof. Remember that as φn ∈ D ⊂ S, we only have to prove that the sequences

{|t|m Dk φn (t)} converge uniformly, for every choice of m ∈ Z+ and k ∈ (Z+ )n . Also
observe that as the general concept of convergence depends on convergence to zero,
it is enough to check this case.
So let {φn } be a sequence convergent to zero in D. We know that there exists a
bounded set K ⊂ Rn which contains the support of every φn , and thus, we can find
T > 0 such that K ⊂ B(T ). Therefore, we can write
|t|m |Dk φn (t)| ≤ T m sup |Dk φn (t)|, ∀t ∈ Rn , (2.3)
|t|≤T
as the derivatives are zero outside K. Now, the convergence in D implies that
||Dk φn ||∞ → 0,
so the right hand side of (2.3) decays to zero as n → ∞. As a consequence,
sup |t|m |Dk φn (t)| → 0, as n → ∞.
Rn
This means that the sequence {|t|m Dk φn (t)} converges uniformly to zero. The proof
is now complete.
Granted that the definition of the Schwartz space is consistent with everything we
have defined before, we will introduce some operations on S, in a similar way as in
Proposition 1.7. But before that, we will present a little lemma, which will be useful
both to prove the following proposition and to get some results in next sections.
Lemma 2.8. Let a, b > 0 and m > 0. Then,
(a + b)m ≤ 2m (am + bm ).
Proof. We can prove it directly if we observe that
(a + b)m ≤ (2 max{a, b})m = 2m (max{a, b})m ≤ 2m (am + bm ).
Proposition 2.9. Let φ ∈ S(Rn ), τ ∈ Rn , a ∈ R − {0} and p(t) a polynomial in
Rn . Then,
1. φ(t + τ ), φ(−t), φ(at) ∈ S(Rn ).
2. p(t)φ(t) ∈ S(Rn ).
Proof. All of the newly defined functions are infinitely smooth for φ and p being so.
Now, for each of the cases, and for every m ∈ Z+ , k ∈ (Z+ )n , if we consider the
constants Cm,k for φ,
• Using the triangle inequality and Lemma 2.8, for φ(t + τ ),
|t|m |Dk φ(t + τ )| ≤ (|t + τ | + |τ |)m |Dk φ(t + τ )|
≤ 2m (|t + τ |m |Dk φ(t + τ )| + |τ |m |Dk φ(t + τ )|)
≤ 2m (Cm,k + |τ |m C0,k )
2.1. THE SCHWARTZ SPACE S 19
• For ϕ(t) = φ(−t),
|t|m |Dk ϕ(t)| = |t|m |(−1)|k| Dk φ(−t)| = | − t|m |Dk φ(−t)|

≤ Cm,k .
• For ϕ(t) = φ(at),
|t|m |Dk ϕ(t)| = |t|m |a|k| Dk φ(at)| = |a||k|−m |at|m |Dk φ(at)|
≤ |a||k|−m Cm,k .
• For p(t)φ(t), because of Proposition 2.3, it is enough to show ti φ(t) ∈ S, ∀i ∈

{1, · · · , n}. So, as the variable ti only has influence in derivatives with respect
to variable ti , it is easy to see that
∂k ∂ k−1 ∂k
(ti φ(t)) = k φ(t) + ti φ(t),
∂tki ∂tk−1
i
∂tki
and therefore, if k = (k1 , · · · , ki , · · · , kn ), and k 0 = (k1 , · · · , ki − 1, · · · , kn ),

then
0
Dk (ti φ(t)) = ki Dk φ(t) + ti Dk φ(t),
and it is straightforward that the constant we need to bound |t|m |Dk (ti φ(t))|
is ki Cm,k0 + Cm,k .
There is one more property that plays a very important role when working with
Schwartz functions.
Theorem 2.10. Let p ∈ [1, ∞]. Then, S(Rn ) ⊆ Lp (Rn ).
Proof. Let φ be a Schwartz function. We know that it is infinitely smooth, and for
every m ∈ Z+ and k ∈ (Z+ )n , there exists a constant Cm,k > 0 such that
|t|m |Dk φ(t)| ≤ Cm,k .
In particular, for m = 0, k = (0, · · · , 0), |φ(t)| ≤ C0,0 , so φ ∈ L∞ (Rn ).

Now let 1 ≤ p < ∞. Write
|φ(t)|p (1 + |t|N )p
Z Z
p
|φ(t)| dt = dt,
Rn Rn (1 + |t|N )p
for some N ∈ N. Notice that the numerator can be bounded as follows:
|φ(t)|p (1 + |t|N )p = [|φ(t)| + |t|N |φ(t)|]p ≤ (C0,0 + CN,0 )p < ∞.
Call C = (C0,0 + CN,0 )p . Then,

Z Z
p 1
|φ(t)| dt ≤ C dt. (2.4)
Rn Rn (1 + |t|N )p
So we need to check that the integral in (2.4) is finite. By a change to polar co-
ordinates, we can write
∞
rn−1
Z Z
1
dt = σ(Sn−1 ) dr
Rn (1 + |t|N )p 0 (1 + rN )p
Z 1 Z ∞ (2.5)
n−1 n−1 dr
≤ σ(S ) r dr + ,
0 1 rN p−n+1
where σ is a measure over the sphere Sn−1 . The first integral in (2.5) is finite, and
the second will be so if N p − n + 1 > 1, or equivalently, N > n/p. Thus, choosing
N as asserted, φ ∈ Lp (Rn ).
We have already seen enough about Schwartz functions to be able to introduce

the space of tempered distributions.
2.2 The space S 0 of tempered distributions

Remember that we defined the space D0 of distributions as a dual space of the space
of testing functions D. In Section 2.1, we have extended the latter into a larger one,
S, in which we encountered many more testing functions. So it is natural to try to
define distributions in this new space.
2.2.1 Definition and some basic properties

Definition 2.11. A mapping T : S(Rn ) → C is called a tempered distribution
if
1. it is linear, in the sense that if φ, ϕ ∈ S(Rn ) and a, b ∈ C, then
T (aφ + bϕ) = aT (φ) + bT (ϕ).
2. it is continuous, in the sense that if {φn } → φ in S(Rn ), then
{T (φn )} → T (φ)
in C.
The space of tempered distributions is denoted by S 0 (Rn ). Some authors also call
tempered distributions distributions of slow growth. In general, tempered dis-
tributions will be represented using letters such as f and g as usual functions, and
the image f (φ) will be written as hf, φi.
As we see, the concept of tempered distribution is completely analogous to that of

usual distribution. In fact, corresponding to Proposition 1.11, we also have a similar
result.
2.2. THE SPACE S 0 OF TEMPERED DISTRIBUTIONS 21
Proposition 2.12. If a functional f : S(Rn ) → C is known to be linear, it is

enough to see that, if {φm } → 0 in S, then {hf, φm i} → 0 in C in order for f to be
continuous, and thus a tempered distribution.
Proof. It is completely analogous to the one corresponding to Proposition 1.11.
In Theorem 2.4 we asserted that every testing function is a Schwartz function. It

is natural to wonder whether a similar relation holds for spaces D0 and S 0 . Observe
that if we consider a tempered distribution f ∈ S 0 , it assigns a complex number to
every Schwartz function, and thus to every testing function. So we can expect that
f ∈ D0 . Indeed, the following proposition shows so.
Theorem 2.13. Every tempered distribution is a distribution. In short,
S 0 (Rn ) ⊂ D0 (Rn ),
where the inclusion is proper.
Proof. Let f ∈ S 0 be a tempered distribution. We know by Theorem 2.4 that for

every ϕ ∈ D, the image hf, ϕi is well-defined, and because of the definition of being a
tempered distribution, f is linear. So we only have to check whether it is continuous.
Consider ϕn to be a sequence which converges to zero in D. By Proposition 2.7, ϕn
also converges in S, so for the continuity of f in S 0 , the sequence hf, ϕn i converges
to zero in C. This shows that S 0 ⊆ D0 .
2
Now consider the function f (t) = e|t| in Rn . This function is locally integrable
for being C ∞ , so according to (1.2), it defines a distribution in D0 (Rn ), given by
Z
hf, ϕi = f (t)ϕ(t)dt.
Rn
But the function φ defined by (2.1) in Theorem 2.4 is a Schwartz function, for which
Z Z Z
|t|2 −|t|2
hf, φi = f (t)φ(t)dt = e e dt = dt = ∞.
Rn Rn Rn
This means that f cannot be a tempered distribution, thus S 0 ( D0 . This completes

the proof.
Very briefly, we will define convergence in the space of tempered distribu-

tions, very similarly as we did in section 1.2.3.
Definition 2.14. Let fn be a sequence of tempered distributions. We say that the

sequence converges to the tempered distribution f if for every Schwartz function φ,
the sequence hfn , φi converges to hf, φi.
2.2.2 Examples
In the same way we did in Section 1.2.2, it is interesting to analyse some examples
of tempered distributions, for they will be important in the upcoming sections.
Example 2.15 (Delta function). Let us start with the most representative distri-
bution, the delta function. We saw in Section 1.2.2 that it defines a distribution.
We want to check it also defines a tempered distribution. We know that for every
φ ∈ S, hδ, φi = φ(0), so clearly, it is well-defined. We also checked that it is linear.
Now, consider a convergent sequence φn → 0 in S. Then, we know that choosing
constants m = 0 and k = (0, · · · , 0), the sequence {φn (t)} converges uniformly to
zero. Because of this,
|hδ, ϕn i| = |φn (0)| ≤ sup φn → 0,

Rn
while n → ∞. As a conclusion, the delta function defines a tempered distri-

bution.
Example 2.16. The proof of Theorem 2.13 clearly shows that, in contrast to usual
distributions, not every locally integrable function defines a tempered distribution.
Anyway, we are interested in knowing for which kind of functions f an expression
as Z
f (t)φ(t)dt (2.6)
Rn
defines a tempered distribution. It seems clear that the problem resides in the
behaviour of the function f (t)φ(t) when |t| tends to infinity. We also know that
φ ∈ S represent a Schwartz function, or as we called it alternatively, a function of
rapid descent. This means that the values φ(t) tend to zero as |t| increases. So it
seems that the only property we need to ask f is that it does not grow too much; it
should grow not faster than the pace in which φ decreases.
With this idea in mind, we call f a function of slow growth if there exists a
natural number N such that
|f (t)|
lim |t|−N |f (t)| = lim = 0. (2.7)
|t|→∞ |t|→∞ |t|N
We are interested in proving that every locally integrable function of slow

growth defines a tempered distribution through the expression (2.6). So con-
sidering f to be so, the first we have to check is whether the functional is well-defined.
In other words, we need to prove that the integral exists for every φ ∈ S(Rn ). From
(2.7) we can assert that for any > 0 we can find M > 0 such that
∀|t| > M, |f (t)| < |t|N . (2.8)
This property can be exploited if we split the integral (2.6) into two terms:
Z Z Z
f (t)φ(t)dt = f (t)φ(t)dt + f (t)φ(t)dt. (2.9)
Rn |t|≤M |t|>M
Observe that if |t| ≤ M , then as φ ∈ C∞ (B(M )), it is bounded, and thus, the local
integrability of f implies that the first term of the right-hand side of (2.9) is finite.
It remains to deal with the second term. We can work with 1/|t|, because we are
far away from the origin. Thus, using (2.8),
Z Z Z

f (t)φ(t)dt ≤ |f (t)||φ(t)|dt ≤ |t|N |φ(t)|dt

|t|>M |t|>M |t|>M
(2.10)
Z
|t|N +n+1 |φ(t)|
= dt.
|t|>M |t|n+1
Now, as φ ∈ S(Rn ), there exists a constant CN +n+1,0 > 0 such that |t|N +n+1 |φ(t)| <
CN +n+1,0 , and thus, it follows from (2.10) that
Z Z
dt
f (t)φ(t)dt ≤ CN +n+1,0 .

n+1
|t|>M |t|

|t|>M
Now, integrating in polar coordinates, we get that

Z Z ∞
dt n−1 dr
n+1
= σ(S ) 2
< ∞, (2.11)
|t|>M |t| M r
where σ represents a measure over the (n − 1)-dimensional sphere Sn−1 . This shows
that the operator is well defined.
Apart from that, it is clear that the functional (2.6) is linear. To see that it is
continuous, we must prove that if φm converges to zero in S, then hf, φm i does so
in C. The idea is the same as in proving that the functional is well defined. Indeed,
we split the integral into two parts as is (2.9) (considering φm instead of φ). The
first part is bounded by
Z
sup |φm (t)| |f (t)|dt,
|t|<M |t|<M
in which the integral is finite and the supremum tends to zero, because the conver-
gence of φm is uniform. On the other hand, for the second part we need (2.8) and
a procedure similar to (2.10) to get a bound given by
Z
dt
sup{|t|N +n+1 |φm (t)|} n+1
,
Rn |t|>M |t|
where the integral, the same as (2.11), is also finite and the supremum tends to zero
by the convergence of φm in S. So eventually we get the result and f defines a
tempered distribution.
Example 2.17. As a remark, it turns out that every bounded function is of slow
growth, as it can be easily checked from the definition given in (2.7). In particular,
every Schwartz function is of slow growth, as they are bounded (Theorem
2.10). Moreover, as Schwartz functions are infinitely smooth, they are locally in-
tegrable, so each of them defines a tempered distribution. As a consequence, and
considering Schwartz functions and their corresponding tempered distributions to
be the same, this example and Theorems 2.4 and 2.13 show that we can display a
chain of relationships among every space discussed so far,
D ⊂ S ⊂ S 0 ⊂ D0 .
Example 2.18. Let f ∈ Lp (Rn ) for some p ≥ 1. Then, f defines a tempered distri-
bution. Indeed, for every φ ∈ S(Rn ), from the definition of being Schwartz we can
find a constant K such that
1 + |t|N
Z Z
|f (t)|
Z

f (t)φ(t)dt ≤ |f (t)||φ(t)| dt ≤ K dt. (2.12)
1 + |t|N N
Rn 1 + |t|
n
R Rn
Now, recall the computations in Theorem 2.10. We saw in the proof, in (2.4) and in
(2.5) that the function 1/(1 + |t|N ) is in Lp whenever N > n/p. Therefore, taking
any N > n, it is in every Lp space, for p ≥ 1. Therefore, considering q to be the
Hölder conjugate of p, by Hölder’s inequality we can write

|f (t)|
Z
1
N
dt ≤ ||f ||p

N
< ∞.
R n 1 + |t| 1 + |t|
q
Therefore, the functional is well-defined. Linearity is clear, and continuity is given

by the same procedure as in (2.12) and the dominated convergence theorem. Thus,
we can write
Lp (Rn ) ⊆ S 0 (Rn ).
2.2.3 Basic operations on tempered distributions

We know that we are allowed to make certain operations with distributions in order
to get new distributions. As S 0 ⊂ D0 , we can, for instance, sum f, g ∈ S 0 and obtain
f + g ∈ D0 . A well developed and consistent theory would make f + g ∈ S 0 . Indeed,
ours does so, and other operations defined in subsection 1.2.4 also perform well in
S 0.
In this subsection, we will present some of the most important operations on S 0 ,
which coincide with those on D0 . Notice that we only need to check continuity to
ensure they produce tempered distributions, as the definitions are given in Subsec-
tion 1.2.4, they are well-defined because of the results presented in Propositions 2.3
and 2.9, and the property of linearity does not depend on the nature of the space
of functions. So let f, g ∈ S 0 (Rn ), α ∈ C, τ ∈ Rn and a ∈ R, a 6= 0. We analyse
the continuity of the sum, the multiplication by a constant, the shifting, the trans-
position and the dilation of tempered distributions. For that, consider φn to be a
convergent sequence in S whose limit is zero.
1. The sum of tempered distributions, f + g.

|hf + g, φn i| ≤ |hf, φn i| + |hg, φn i| → 0.
2. The multiplication by a constant, αf .
|hαf, φn i| ≤ |α||hf, φn i| → 0.
3. The shifting of a tempered distribution, f (t − τ ). In this case, as hf (t −

τ ), φi = hf, φ(t + τ )i, it is enough to check that the sequence {φn (t + τ )} → 0
in S. So for every m ∈ Z+ and k ∈ (Z+ )n , we get, using Lemma 2.8,
|t|m |Dk φn (t + τ )| ≤ 2m (|t + τ |m |Dk φn (t + τ )| + |τ |m |Dk φn (t + τ )|),
where both summands tend uniformly to zero because the convergence of φn

in S. This shows that the sequence {φn (t + τ )} converges to zero in S, and
thus |hf, φn (t + τ i| → 0 in C.
4. The transposition of a tempered distribution, f (−t). As in the previous

case, it is enough to see that the sequence {φn (−t)} converges to zero in S.
But this is trivial, because |t| = | − t|.
5. The dilation of a tempered distribution, f (at). Again we only have to

check that the sequence {φn (t/a)} converges to zero in C. It is trivial too,
because m
m k
t mt k
t
|t| D φn = |a| D φ n ,
a a a
and the right-hand side term converges uniformly to zero.
Remark 2.19. Notice that in this case, we cannot define the product of a tempered
distribution with a smooth function. In fact, not every product of a smooth function
and a Schwartz function is a Schwartz function. It is enough to consider φ(t) =
2 2
e−|t| ∈ S as in 2.1 and g(t) = e|t| .
However, it can be proved that a tempered distribution can be multiplied by any
C ∞ function for which itself and all its derivatives are of slow growth. If ψ is of that
kind, then φψ is Schwartz for every Schwartz function φ, and the sequence ψφn will
converge to zero if φn does so. Thus, for every f ∈ S 0 , ψ · f will define a tempered
distribution by the usual definition
hψ · f, φi = hf, ψφi, ∀φ ∈ S 0 .
In particular, as every Schwartz function is of slow growth and as all derivatives are
Schwartz functions and thus of slow growth, multiplication between tempered
distributions and Schwartz functions is permitted.
We cannot forget about one of the main feature of distributions, the derivative.
∂f
6. The derivative of a tempered distribution, ∂ti . It is immediate from the
definition of convergence in S that the sequence { ∂φ
∂ti } converges to zero in S,
n
so the derivative of a tempered distribution is continuous, and thus a tempered

distribution.
So as we have proved that derivatives remain in the space of tempered distribu-

tions, they will hold every property we analysed in section 1.2.5. It is also interesting
to remark that differentiation is, in the same way as seen in Proposition 1.20, a linear
and continuous operator in S.
Proposition 2.20. Differentiation is a linear operation in the space of tempered

distributions, and it is continuous in the sense that if a sequence of tempered distri-
butions fn converges to f in S 0 , then Dk fn converges to Dk f in S 0 .
2.2.4 Convolution
The ones presented in 2.2.3 are not the only operations we can define in the space
of tempered distributions. Indeed, we will need to work with not so basic ones when
we deal with fundamental solutions of differential operators in Chapter 4. The first
new operation we will work with is convolution. It is possible, and useful, to define
the convolution between tempered distributions under certain conditions, and even
on the space of usual distributions. However, we will focus on the convolution
between a tempered distribution and a Schwartz function, as it will be
enough to achieve our goal.
Recall the definition of the convolution between two integrable functions
f and g, Z
(f ∗ g)(x) = f (y)g(x − y)dy.
Rn
It is straightforward to see that convolution is a commutative operation, so f ∗ g =

g ∗ f . What we want to do first is to show that the convolution between two
Schwartz functions is still a Schwartz function. We know that the definition of
a Schwartz function depends in a big way of differentiation. On the other hand,
convolution implies dealing with an integral, so we can expect having troubles with
the commutability of both. The dominated convergence theorem is of great help in
this situation, and it will allow us to present a lemma determining conditions under
which commutation of integration and derivation will be permitted.
Lemma 2.21. Let g be a function in Rn × Rn , and consider, being x, t ∈ Rn ,

Z
f (t) = g(t, x)dx.
Rn
∂g
Let k ∈ {1, · · · , n}, and suppose both g and ∂tk are continuous. Moreover, suppose
∂g
∂tk is bounded by an integrable function in the variable x. In this situation,
Z
∂f ∂g
(t) = (t, x)dx.
∂tk Rn ∂tk
Proof. It is consequence of using the definition of the partial derivative as a limit,

the mean value theorem and the dominated convergence theorem.
Once we know how to manage this situation, we present the following auxiliar
proposition.
Proposition 2.22. Let f, g ∈ S(Rn ). Then, f ∗ g ∈ C∞ (Rn ).
Proof. We know that both f and g are smooth. As we want to compute, for every
k ∈ {1, · · · , n}, the derivative
Z
∂
f (x − y)g(y)dy,
∂xk Rn
we can use Lemma 2.21 to take the derivative inside, and thus
Z Z
∂ ∂
f (x − y)g(y)dy = (f (x − y)g(y))dy
∂xk Rn Rn ∂xk
Z
∂f ∂f
= (x − y)g(y)dy = ∗ g (x),
Rn ∂xk ∂xk
from where the result follows immediately.
Remark 2.23. If we read the previous proof carefully, we will realize that g need not
be so regular. Indeed, it would be enough for it to be continuous and integrable. In
any case, if g is also Schwartz, by commuting the definition of convolution, the same
proof shows that

∂f ∂ ∂g
∗ g (x) = (f ∗ g)(x) = f ∗ (x).
∂xk ∂xk ∂xk
Let us prove that convolution is an internal operation in S.
Theorem 2.24. Let f, g ∈ S(Rn ). Then f ∗ g ∈ S(Rn ).
Proof. Proposition 2.22 shows the convolution is smooth. Thus, take m ∈ Z+ , k ∈

(Z+ )n , and observe that
Z
|x|m |Dk (f ∗ g)(x)| ≤ |x|m |g(x − y)||Dk f (y)|dy.
Rn
Now, writing |x| = |x − y + y|, we can use Lemma 2.8 to assert that
Z
|x|m |Dk (f ∗ g)(x)| ≤ 2m |x − y|m |g(x − y)||Dk f (y)|dy
R n
Z (2.13)
+ |y|m |g(x − y)||Dk f (y)|dy .
Rn
Now, |x − y|m |g(x − y)| and |y|m |Dk f (y)| can be bounded for the definition of f
and g being Schwartz, and as every Schwartz function is integrable (Theorem 2.10),
both Dk f (y) and g(x − y) are so, thus the sum of integrals (2.13) is bounded by a
constant depending on m and k, but not on x.
After seeing some properties of convolution in S, we need to define the convolution

between a tempered distribution and a Schwartz function. Again, it is a good way to
check the situation with regular distributions, so let φ, ϕ ∈ S(Rn ). By Theorem 2.24,
φ∗ϕ ∈ S(Rn ), so it can be seen as a regular distribution, and thus for all ψ ∈ S(Rn ),
Z Z Z
hφ ∗ ϕ, ψi = (φ ∗ ϕ)(x)ψ(x)dx = φ(y)ϕ(x − y)dy ψ(x)dx.
Rn Rn Rn
Observe that φ, ϕ, ψ are integrable and bounded, so Fubini’s theorem can be applied
to revert the order inside the integral, and
Z Z
hφ ∗ ϕ, ψi = φ(y) ϕ(x − y)ψ(x)dx dy.
Rn Rn
Denote as ϕ̃ the transposition of the function ϕ. In other words, ϕ̃(t) = ϕ(−t).

Then, we can write
Z Z Z Z
φ(y) ϕ(x − y)ψ(x)dx dy = φ(y) ϕ̃(y − x)ψ(x)dx dy,
Rn Rn Rn Rn
which clearly shows a convolution between ϕ̃ and ψ, to finally get

hφ ∗ ϕ, ψi = hφ, ϕ̃ ∗ ψi. (2.14)
Notice that ϕ̃ is a Schwartz function, and so, we will be able to extend (2.14) to any
kind of distributions.
Definition 2.25. Let f ∈ S 0 (Rn ) and ψ ∈ S(Rn ). We define the convolution
between a tempered distribution and a Schwartz function, f and ψ, as
follows:
hf ∗ ψ, φi = hf, ψ̃ ∗ φi, ∀φ ∈ S(Rn ),
where ψ̃ represents the transposition of the function ψ.
We can expect, by the reasoning done for Schwartz functions, this definition to
lead to a tempered distribution. Fortunately, that property holds, as we can see in
the following theorem.
Theorem 2.26. For every f ∈ S 0 (Rn ) and ψ ∈ S(Rn ), the convolution f ∗ ψ defines
a tempered distribution.
Proof. The linearity of the functional defined in Definition 2.25 is a result of the
linearity of both the convolution of functions and the distribution f . So let φl be
a sequence convergent to zero in S(Rn ). We know that every ψ̃ ∗ φl is a Schwartz
function, so we only have to check that every sequence |x|m Dk (ψ̃ ∗ φl )(x) converges
uniformly to zero, for usual indices m, k. We will reason as in equation (2.13).
Indeed, substituting f for ψ̃ and g for φl , and taking suprema,
Z
|x|m |Dk (ψ̃ ∗ φl )(x)| ≤ 2m sup |x|m |φl (x)| |Dk ψ̃(y)|dy
x∈R n R n
Z (2.15)
m k
+ sup |φl (x)| |y| |D ψ̃(y)|dy .
x∈Rn Rn
The two integrals on the right-hand side of (2.15) are finite. Indeed, the first one
is concerning a Schwartz, thus integrable function. The second integrand can be
bounded by a factor C/|y|n+1 , where C is a constant, for ψ̃ being Schwartz. There-
fore, the convergence of φl in S makes both suprema tend to zero, what implies the
convergence of ψ̃ ∗ φn . Finally, the continuity of f implies that of f ∗ ψ.
We have seen that the derivatives are fairly easy to handle when working with
Schwartz functions. The next result shows that the situation does not change when
considering tempered distributions.
Proposition 2.27. Let f ∈ S 0 (Rn ) and ψ ∈ S(Rn ). Then,
∂ ∂f ∂ψ
(f ∗ ψ) = ∗ψ =f ∗ , k = 1, · · · , n.
∂xk ∂xk ∂xk
Proof. It is an immediate consequence of the definition of the convolution and remark

2.23.
The most interesting example of convoluting a distribution is that concerning

the delta function. Indeed, let δ be the delta function, and consider any Schwartz
function ψ. We want to work out the value of δ ∗ ψ. For that, if φ represents a
Schwartz function,
Z
hδ ∗ ψ, φi = hδ, ψ̃ ∗ φi = ψ̃ ∗ φ(0) = ψ̃(−y)φ(y)dy
Rn
Z
= ψ(y)φ(y)dy = hψ, φi.
Rn
This means that

δ ∗ ψ = ψ, ∀ψ ∈ S(Rn ). (2.16)
In other words, the delta function works as the identity element with re-
spect to convolution.
Remark 2.28. Even if we will not be able to define the convolution between two
arbitrary distributions, it is possible to do so under certain restrictions, and in a
more general way than we did in this section. We refer the reader to [10, p. 122-137]
or to [8, p. 102-109].
2.2.5 Direct product

Another important operation on the space of tempered distributions is the direct
product. Indeed, the direct product can be defined on the space D0 of usual dis-
tributions. In this subsection, we will briefly discuss the direct product on D0 . The
reason for this is that this new operation will be required for some minor, though
necessary computations in Chapter 4. We will also mention the analogous results
for S 0 , and references for more details will be given.
First of all, let us present what intuitively the direct product could be. Let f and
g be locally integrable functions on Rn . Then, the function given by f (x)g(y) on
Rm+n is also locally integrable, thus defines a distribution,
Z
hf (x)g(y), ϕ(x, y)i = f (x)g(y)ϕ(x, y)dxdy, ∀ϕ ∈ D(Rn+m ).
Rn+m
For ϕ being bounded, Fubini’s theorem can be applied to consider the most con-
venient order. In this case, let us write
Z Z Z
f (x)g(y)ϕ(x, y)dxdy = f (x) g(y)ϕ(x, y) dx,
Rn+m Rn Rm
so considering g as a distribution, we could formally write
hf (x)g(y), ϕ(x, y)i = hf (x), hg(y), ϕ(x, y)ii,
what suggests a definition for the direct product of distributions. In fact, we can
assert that the definition is correct.
Definition 2.29. Let f ∈ D0 (Rn ) and g ∈ D0 (Rm ) be any two distributions. Their
direct product is defined as
hf (x) · g(y), ϕ(x, y)i = hf (x), hg(y), ϕ(x, y)ii, ∀ϕ ∈ D(Rn+m ).
It is also usual to denote the direct product of f and g as f (x) × g(y).
We have seen that the definition works fine for locally integrable functions, but
when it comes to distributions, we have to ensure that this definition makes sense.
For instance, we should check that the expression hg(y), ϕ(x, y)i defines a testing
function on Rn whenever g is a distribution on Rm and ϕ is a testing function on
Rn+m . The following lemma asserts so, although we will not prove it. A proof can
be found in [8, p.96-98].
Lemma 2.30. For every distribution g ∈ D0 (Rm ) and every testing function ϕ ∈
D(Rn+m ), the function
ψ(x) = hg(y), ϕ(x, y)i
is a testing function on Rn . Moreover, its derivatives are given by
Dα ψ(x) = hg(y), Dxα ψ(x, y)i,
for every multi-index α, where the notation Dxα stands for the partial derivatives
with respect to x and with indices α. Also, if ϕn → ϕ in D(Rn+m ), then
ψn = hg(y), ϕn (x, y)i → ψ
in D(Rn ).
This lemma allows us to prove that the direct product is indeed a distribution.
Proposition 2.31. The direct product of two distributions is a distribution.
Proof. Let f and g be distributions on Rn and Rm respectively. Lemma 2.30 ensures

that the functional f (x) · g(y) is well-defined, and it is linear because both f and g
are so. Now, according again to Lemma 2.30, if ϕn → ϕ in D(Rn+m ), we know that
hg(y), ϕn (x, y)i → hg(y), ϕ(x, y)i
in D(Rn ), and as f is continuous,
hf (x), hg(y), ϕn (x, y)ii → hf (x), hg(y), ϕ(x, y)ii,
showing that f (x) · g(y) is continuous.
It can also be proved that the direct product is a commutative operation. This
means that for every distributions f and g, it holds that
f (x) · g(y) = g(y) · f (x). (2.17)
A proof can be found in [8, p.99-100].

Results for the space S 0 of tempered distributions completely analogous to Lemma
2.30 and Proposition 2.31 can be proved. Proofs can be found in [8, 119-121].
To end with this section, we present a very simple example concerning the Delta
function. Let x ∈ Rn and y ∈ Rm . Then,
hδ(x) · δ(y), ϕ(x, y)i = hδ(x), ϕ(x, 0)i = ϕ(0, 0) = hδ(x, y), ϕ(x, y)i,
for every ϕ ∈ D(Rn+m ). Thus, it is clear that
δ(x) · δ(y) = δ(x, y). (2.18)

Chapter 3
The Fourier Transform
In this chapter, we will study what the effect of applying the Fourier transform to
functions in S is, in order to define the Fourier transform of distributions. Several
problems appear when the Fourier transform is defined for the space D, but it works
properly when we extend the space to S. We will discuss this along next pages.
3.1 The Fourier transform in L1

First of all, we will introduce the Fourier transform for integrable functions.
Definition 3.1. Let f : Rn → C be an integrable function. We define the Fourier

transform of f as the function fˆ : Rn → C, given by
Z
ˆ
f (ξ) = f (x)e−i(ξ·x) dx.
Rn
It is also common to denote the Fourier transform of f as F(f ) or simply Ff .

We also define the inverse Fourier transform of f as the function fˇ : Rn → C,
given by Z
ˇ 1
f (ξ) = f (x)ei(ξ·x) dx,
(2π)n Rn
and it is usually denoted by F −1 (f ) or simply F −1 f .
It is important to notice that the Fourier transform of a function in L1 is bounded

and continuous, and thus well-defined. We present this fact in the following propos-
ition.
Proposition 3.2. Let f ∈ L1 (Rn ). Then, Ff ∈ L∞ (Rn ), and it is continuous.
Proof. The boundedness requires a really short proof. Indeed,

Z Z
|fˆ(ξ)| ≤ |f (x)||e−i(ξ·x) |dx = |f (x)|dx < ∞.
Rn Rn
33
34 CHAPTER 3. THE FOURIER TRANSFORM
Continuity follows from the dominated convergence theorem, as we can write

Z
ˆ ˆ
|f (ξ + h) − f (ξ)| ≤ |f (x)||e−i(ξ·x) ||e−i(h·x) − 1|dx.
Rn
The integrand is bounded by 2|f (x)| and the limit can be taken inside. It is clear
that it tends to zero.
It is interesting to notice that, in general, the Fourier transform of a L1 function

need not be integrable, thus we will not be able to define its inverse transform. This
will not be a problem in S nor in S 0 . Moreover, based on this definition, the Fourier
transform can also be defined for L2 functions, through a process of density and
convergence. It can be proven that S is dense in L2 , so it will be enough to work
in the former. There is even another way to define it in L2 . Recall the result in
Example 2.18. Thus, as L2 ⊆ S 0 , it will be enough to define it on S 0 .
3.2 The Fourier transform in S

Defining the Fourier transform in the space S is key to be able to define it for
tempered distributions. We can think of using Definition 3.1 to do so. Remember
that we saw in Theorem 2.10 that S ⊆ L1 , so there is no trouble.
Definition 3.3. The Fourier transform on the Schwartz space S(Rn ) is an operator
F : S(Rn ) → L∞ (Rn ) which asigns to every φ ∈ S(Rn ) the function
Z
Fφ(ξ) = φ̂(ξ) = φ(x)e−i(ξ·x) dx.
Rn
Our next objective is to see that the Fourier transform is a bijective endomorphism
(thus an automorphism). In other words, we want to prove:
1. that the Fourier transform of every Schwartz function is also a Schwartz func-
tion, and
2. that FF −1 = 1S = F −1 F, where 1S represents the identity operator.
To get these two results, it is vital to know how to manage the derivatives of
the Fourier transform of a function, as the definition of S is heavily dependent on
differentiation. For that reason, we will fist develop some properties concerning F.
3.2.1 Some properties of the Fourier transform in S

As we have recently advocated, we have to find a way to manage the derivatives of
the Fourier transform of Schwartz functions. Its definition involves an integral, so we
can expect an environment in which derivatives and integrals will appear at the same
time. The very famous dominated convergence theorem will be extremely useful in
this situation, by means of Lemma 2.21. This lemma will allow us to prove the
3.2. THE FOURIER TRANSFORM IN S 35
properties which are probably the major advantage of working in the Fourier space.
These properties are the ones which make a direct relation between differentiation
and multiplication. Let us analyse them in detail.
Proposition 3.4. Let φ ∈ S(Rn ) be a Schwartz function. Then, for every k ∈

(Z+ )n , the following properties hold:
1. Dk F(φ)(ξ) = (−i)|k| F(xk φ)(ξ).
2. ξ k F(φ)(ξ) = (−i)|k| F(Dk φ)(ξ).
where if k = (k1 , · · · , kn ), we denote xk = xk11 · · · xknn .
Proof. 1. We want to make the derivative of the Fourier transform φ̂(ξ). Let
i ∈ {1, · · · , n}, and
Z
∂ φ̂ ∂
(ξ) = φ(x)e−i(ξ·x) dx.
∂ξi ∂ξi Rn
Observe that

∂ −i(ξ·x)

φ(x)e = φ(x)(−ixi )e−i(ξ·x) = |xi φ(x)|.
∂ξi
By Proposition 2.9, we know that xi φ(x) ∈ S(Rn ), and by Theorem 2.10, it is

integrable. Therefore, Lemma 2.21 asserts that
Z
∂ φ̂ ∂
(ξ) = φ(x)e−i(ξ·x) dx
∂ξi RnZ ∂ξi
(3.1)
−i(ξ·x)
= −i xi φ(x)e dx = −iF(xi φ(x))(ξ).
Rn
So clearly, applying (3.1) |k| times, we get
Dk φ̂(ξ) = (−i)|k| F(xk φ(x))(ξ).
2. Choose again i ∈ {1, · · · , n}. We will compute the Fourier transform of the
derivative. By definition,
Z
∂φ ∂φ
F( )(ξ) = (x)e−i(ξ·x) dx.
∂xi Rn ∂x i

∂φ −i(ξ·x) =
Observe that the function we are integrating is in L1 , because ∂x i
(x)e

∂φ ∂φ
∂xi (x), and ∂xk ∈ S ⊆ L1 , as we saw in Proposition 2.3 and in Theorem 2.10.

So the order of the integral can be changed by Fubini’s theorem, and
Z Z Z
∂φ ∂φ
(x)e−i(ξ·x) dx = (x)e−i(ξ·x) dx. (3.2)
Rn ∂xi R n−1 R ∂x i
Now, integrating by parts, the boundary term vanishes for φ being Schwartz,
and Z Z
∂φ −i(ξ·x)
(x)e dxi = iξk φ(x)e−i(ξ·x) dxi .
R ∂xi R
So rewriting (3.2),
Z Z
∂φ
(x)e−i(ξ·x) dx = iξk φ(x)e−i(ξ·x) dx.
Rn ∂xi Rn
∂φ
In other words, F( ∂xi
)(ξ) = iξk F(φ). Applying this last expression |k| times,
F(Dk φ)(ξ) = i|k| ξ k F(φ)(ξ).
Proposition 3.4 conveys how derivatives work in interaction with the Fourier trans-
form. This is key to prove our first objective.
Theorem 3.5. Let φ ∈ S(Rn ). Then, F(φ) ∈ S(Rn ).
Proof. We must check two facts:
1. φ̂ ∈ C ∞ , and
2. for every m ∈ Z+ , k ∈ (Z+ )n , ∃Cm,k > 0 | |ξ|m |Dk φ̂(ξ)| < Cm,k .
So let us prove each fact separately.
1. By Proposition 3.4,
Dk φ̂(ξ) = (−i)|k| F(xk φ(x))(ξ),
so any derivative can be computed in terms of Fourier transforms of xk φ(x),

which we know it is a Schwartz function by 2.9. The Fourier transform of any
Schwartz function is continuous, so φ̂ is infinitely smooth.
2. We need to use both properties of Proposition 3.4. Consider i ∈ {1, · · · , n}.

Then,
ξim Dk φ̂(ξ) = (−i)|k|+m F(Dim (xk φ(x))).
Then, by the definition of the Fourier transform,
Z
|ξim Dk φ̂(ξ)| ≤ |Dim (xk φ(x))|dx. (3.3)
Rn
Observe now that Dim (xk φ) is a sum of several products between polynomials
and derivatives of φ. For that, Dim (xk φ) ∈ S holds. This means that it is
integrable, and thus, Mi = Rn |Dim (xk φ(x))|dx is finite.
R
√ m
Also observe that |ξ|m ≤ n maxi=1,··· ,n |ξi |m , so we can write
√ m √ m
|ξ|m |Dk φ̂(ξ)| ≤ n max |ξim Dk φ̂(ξ)| ≤ n max Mi < ∞. (3.4)
i=1,··· ,n i=1,··· ,n
Some ideas of this proof yield to the fact that the Fourier transform F is a linear
continuous operator from S to S.
Theorem 3.6. The operator F : S → S is a linear operator, and it is continuous

in the sense that if φn → φ in S, then F(φn ) → F(φ) in S.
Proof. It is trivial to see that the operator F is linear. To see that it is continuous,
for the same reason as in Proposition 1.11, it is enough to check the case {φl }∞
l=1 → 0.
m k
Supposing so, we have to see that every sequence |ξ| D φ̂l (ξ) → 0 uniformly, being
m ∈ Z+ , k ∈ (Z+ )n . But if we consider (3.4) from the proof of Theorem 3.5, we
know that
√ m
|ξ|m |Dk φ̂l (ξ)| ≤ n max Mi,l ,
i=1,··· ,n
where Z
Mi,l = |Dim (xk φl (x))|dx. (3.5)
Rn
Therefore, it is enough to check that maxi=1,··· ,n Mi,l tends to zero as l → ∞.
As the result of expanding the derivative of (3.5) will be a finite sum of products of
polynomials and derivatives of φl , the convergence of φl implies that the derivatives
will converge uniformly to zero.
Moreover, if we choose q ∈ N, the expression |x|q |Dim (xk φl )(x)| will be bounded by
the finite sum mentioned above times |x|q , which, by the convergence of φl in S, will
be bounded by a constant Cq,m (it will be the sum of the constants corresponding
to each summand). This way,
|x|q |Dim (xk φl )(x)| ≤ Cq,m .
Now, as we can have a bound with a integrable function, we can apply the dominated
convergence theorem to assert that
lim Mi,l = 0, ∀i = 1, · · · , n.
l→∞
The proof is now complete.
3.2.2 The Fourier transform of the Gaussian function

Before going on, it is convenient to present an example of the Fourier transform
of a particular function. This example has great importance in several proofs and
solutions of differential equations. The function we are talking about is the Gaussian
function.
Definition 3.7. Let x ∈ Rn . We call Gaussian function the following function:

2
g(x) = e−k|x| , for some k > 0. (3.6)
Instead of using the definition, we will think in an alternative way. First of all,
observe that, if we allow a little abuse of notation when considering g in one variable,
2 2 2 2 2
g(x) = e−k|x| = e−k(x1 +...+xn ) = e−kx1 · . . . · e−kxn = g(x1 ) · . . . · g(xn ). (3.7)
Also observe that as g ∈ S (recall Theorem 2.4, equation (2.1)), it is integrable, and
Fubini’s theorem allows us to split the integral using (3.7), and get
ĝ(ξ) = ĝ(ξ1 ) · . . . · ĝ(ξn ). (3.8)
Therefore, it is enough to work out the transform in one variable. So consider g in

one variable. Then, it is easy to check that the following equation holds:
g 0 (x) + 2kxg(x) = 0. (3.9)
Now, let us apply the Fourier transform to (3.9). We get
F(g 0 (x)) + 2kF(xg(x)) = 0. (3.10)
Proposition 3.4 is offering a way to proceed:
F(g)0 (ξ) = −iF(xg)(ξ) and ξFg(ξ) = −iF(g 0 )(ξ),
so (3.9) is equivalent to
iξFg(ξ) + 2kiF(g)0 (ξ) = 0,
which in turn is
ξ
F(g)0 (ξ) +
Fg(ξ) = 0.
2k
A solution can be easily found; it must be of the form
ξ2
Fg(ξ) = Ce− 4k , C ∈ R.
We need to determine that constant C. Indeed, by the definition,

Z r
π
ĝ(0) = g(x)dx = .
R k
The last integral value can be found by first squaring it and then using polar co-
ordinates. Thus, considering (3.8),
r n
π 2
ĝ(x) = e−|x| /4k , in Rn , (3.11)
k
the Fourier transform of the Gaussian.

It is also interesting to analyse the case in which the constant k is not real.
Consider k = a + bi = z with a > 0, and the Gaussian
2 2
g(x) = e−z|x| = e−(a+bi)|x| . (3.12)
By repeating every calculation for the case where k > 0, we can assert that
2 /4z
ĝ(ξ) = Ke−ξ ,
where K is a constant to determine. In fact,

Z Z
2
K = ĝ(0) = g(x)dx = e−zx dx.
R R
This integral cannot be computed as we did in the real case. It needs an special
treatment. So, for a > 0, consider
Z
2
Ia (t) = e−(a+ti)x dx.
R
Lemma 2.21 and integration by parts allow to write

−i
Ia0 (t) = Ia (t),
2(a + ti)
and therefore we get, for a function k(a),
k(a)
Ia (t) = √ ,
a + ti
where we work with the principal branch of the square root in C. Finally, the value
√
Ia (0) gives the value of k(a). In fact, k(a) = π, so
√
π
Ia (t) = √ , ∀a > 0.
a + ti
√ √
This little calculation conveys that K = π/ z,
r
π −ξ2 /4z
ĝ(ξ) = e ,
z
for every z with Re(z) > 0, and in Rn ,
r n
π 2
ĝ(ξ) = e−|ξ| /4z . (3.13)
z
3.2.3 The inverse Fourier transform in S

Remember that we gave a definition for the inverse Fourier transform in Definition
3.1. In this subsection, our objective will be to prove that it is indeed the inverse
operator corresponding to the Fourier transform F. Our first approach could be to
try to get the result directly. This way, if φ ∈ S,
Z Z
−1 1 −i(ξ·x)
F (F(φ))(y) = φ(x)e dx e−i(z·ξ) dξ
(2π)n Rn Rn
Z Z (3.14)
1 −iξ·(x−z)
= φ(x)e dxdξ.
(2π)n Rn Rn
Anyway, the integral in 3.14 is not absolutely convergent, and thus, Fubini’s theorem
cannot be applied. To solve this problem, we will need some properties of the Fourier
transform, concerning its behaviour with respect to certain operations.
Proposition 3.8. Let φ ∈ S(Rn ). Then,
1. If τ ∈ Rn , then
F(φ(x − τ ))(ξ) = e−iξ·τ Fφ(ξ). (3.15)
2. If λ ∈ R, then
1 ξ
F(φ(λx))(ξ) = Fφ . (3.16)
|λ|n λ
3. If φ, ϕ ∈ S(Rn ), then
Z Z
Fφ(x)ϕ(x)dx = φ(x)Fϕ(x)dx. (3.17)
Rn Rn
Proof. Properties 1 and 2 are a direct result of simply applying a change of variables
in the integral of the definition of the Fourier transform. Finally, property 3 is a
direct consequence of Fubini’s theorem. Observe that Fubini’s results can be applied
as every Schwartz function is integrable (see Theorem 2.10).
Having presented these properties, we are ready to face our main objective.
Theorem 3.9. Let φ ∈ S(Rn ). Then, the inversion formula given by

Z
1
F −1 (φ)(ξ) = φ(x)ei(ξ·x) dx
(2π)n Rn
is the inverse operator of F, in the sense that
F −1 (F(φ)) = φ = F(F −1 (φ)).
Proof. We already noticed that a direct proof cannot be achieved. So consider

property (3.17) in Proposition 3.8, for f, g ∈ S(Rn ):
Z Z
Ff (x)g(x)dx = f (x)Fg(x)dx.
Rn Rn
Consider λ 6= 0 and g(x/λ) instead of g(x). Then, by property (3.16),

x
F g (ξ) = |λ|n Fg(λξ),
λ
and coming back to (3.17),
Z x Z
fˆ(x)g dx = |λ|n f (x)ĝ(λx)dx. (3.18)
Rn λ Rn
Now, changing variables y = λx on the right hand side of (3.18), we get

Z x Z x
ˆ
f (x)g dx = f ĝ(x)dx.
Rn λ Rn λ
We want to take limits when λ tends to infinity in both sides. Observe that λ 6= 0
ensures we are working with Schwartz functions all the time, so we are allowed to
write x
|fˆ(x)g | ≤ |fˆ(x)|||g||∞ ,
λ
where of course, by Theorem 2.10, fˆ ∈ L1 . With a similar procedure for the right
hand side integral, we are allowed to apply the dominated convergence theorem and
Z Z
fˆ(x)g(0)dx = f (0)ĝ(x)dx. (3.19)
Rn Rn
This is the time when the Gaussian comes into play. If we consider g to be the
Gaussian function (3.6) with k = 1/2,
2 /2
g(x) = e−|x| ,
then (3.11) and (3.19) convey that
Z √ nZ
ˆ
f (x)dx = f (0)( 2π) g(x)dx = (2π)n f (0).
Rn Rn
Reordering, Z
1
f (0) = fˆ(x)dx, (3.20)
(2π)n Rn
which is the result we seek at ξ = 0, valid for every Schwartz function. The last step
is to use (3.15). Let ξ ∈ Rn . Considering f (x + ξ) ∈ S, then by (3.20) and (3.15),
Z Z
1 1
f (ξ) = F(f (x + ξ))(y)dy = e−i(y·(−ξ)) fˆ(y)dy
(2π)n Rn (2π)n Rn
Z
1
= fˆ(x)ei(ξ·x) dx,
(2π)n Rn
which is the definitive result. This shows that F −1 Ff = f for every Schwartz
function f . The symmetrical result is achieved in a similar way.
With this result, we are allowed to say that the inverse Fourier transform of every
Schwartz function is again a Schwartz function, and that the operator F −1 is also
linear and continuous, by the same arguments used for F.
Theorem 3.10. The inverse Fourier transform of every Schwartz function is a
Schwartz function, and the operator F −1 : S → S is a linear and continuous oper-
ator.
Proof. That F −1 is an endomorphism in S is a direct consequence of Theorem 3.9.
The linearity and continuity of the inverse operator can be obtained by a similar
procedure as in Theorem 3.6.
3.2.4 More properties

Once we have discovered the inverse operator, we can deduce some interesting prop-
erties, which may be useful in some calculations.
Proposition 3.11. Let φ, ϕ ∈ S(Rn ). Then,
1. F −1 φ(ξ) = 1
(2π)n F φ̃(ξ).
2. F φ̃(ξ) = Fφ(−ξ).
3. F 2 φ(ξ) = (2π)n φ̃(ξ).
4. F 4 φ(ξ) = (4π 2 )n φ(ξ).
Proof. To prove properties 1 and 2, it is enough to transpose the variable in the

integral. Property 3 uses the same change of variables, Theorem 3.9 and property
2. Using property 3 twice yields to property 4 directly.
As we said in Section 2.2.4, convolution will play an important role. For this
reason, we need to know how to manage its Fourier transform.
Proposition 3.12. Let φ, ϕ ∈ S(Rn ). Then,
1.
F(φ ∗ ϕ)(ξ) = Fφ(ξ)Fϕ(ξ). (3.21)
2.
1
F(φϕ)(ξ) = (Fφ ∗ Fϕ)(ξ). (3.22)
(2π)n
Proof. Formula (3.21) has to be justified by Fubini’s theorem, to eventually use a

change of variables. For property 2, we know by (3.21) that
F(φ ∗ ϕ) = Fφ · Fϕ.
Now, by properties 1 and 2 in Proposition 3.11,

1
F −1 (φ ∗ ϕ) = F −1 φ · F −1 ϕ,
(2π)n
and if we consider Fourier transforms φ̂, ϕ̂ instead, we get

1
φ · ϕ = F −1 φ̂ · F −1 ϕ̂ = F −1 (φ̂ ∗ ϕ̂).
(2π)n
So finally, by Theorem 3.9,

1
F(φ · ϕ) = (φ̂ ∗ ϕ̂).
(2π)n
3.3. THE FOURIER TRANSFORM IN S 0 43
Another famous property concerning the Fourier transform is the Plancherel

equality. It is an interesting feature, as it can be used to extend the transform
from S to L2 .
Proposition 3.13 (Plancherel equality). Let φ, ϕ ∈ S. Then,

Z Z
1
φ(t)ϕ(t)dt = Fφ(t)Fϕ(t)dt.
Rn (2π)n Rn
In particular, Z Z
1
|φ(t)|2 dt = |Fφ(t)|2 dt.
Rn (2π)n Rn
Proof. Substitute ϕ for Fϕ in (3.17), and observe that

Z
Fϕ(ξ) = ϕ(x)eiξ·x dx = (2π)n F −1 ϕ(ξ).
Rn
The equality ||Fφ||2 = (2π)n/2 ||φ||2 for φ ∈ S and the fact that S is dense in L2
can be used to extend the Fourier transform to L2 . Nevertheless, for our purposes,
we need to extend it to the space of tempered distributions, where we know L2 is
contained.
3.3 The Fourier transform in S 0

Once we have defined the Fourier transform in the space S and studied some of its
properties, it is time to capitalise on them to define the transform in the space S 0 of
tempered distributions. In fact, Theorem 3.6 is a great boon to do so. Thus, to the
operations described in Section 2.2.3 we will be able to add one more. Throughout
this section, we will present many properties concerning the Fourier transform of
tempered distributions. Because of the definition to be given, they will be a direct
consequence of those concerning Schwartz functions, so for the proofs we will offer
references to Section 3.2, without giving many details.
3.3.1 Definition and properties

Definition 3.14. Let f ∈ S 0 (Rn ) be a tempered distribution. We define the Fourier
transform of f as a functional over S(Rn ), denoted by Ff or fˆ, given by
hFf, φi = hf, Fφi, ∀φ ∈ S(Rn ).
In the same way every operation analysed before has a justification in terms of
regular distributions, Definition 3.14 can also be justified by considering Schwartz
functions as tempered distributions. In fact, if φ is a Schwartz function and we treat
it as a distribution, then for every ϕ ∈ S,
Z
hFφ, ϕi = Fφ(x)ϕ(x)dx,
Rn
and by property 3 in Proposition 3.8, we can write

Z Z
Fφ(x)ϕ(x)dx = φ(x)Fϕ(x)dx = hφ, Fϕi,
Rn Rn
thus the definition. It is important to check that we are defining a tempered distri-
bution.
Theorem 3.15. The Fourier transform of every tempered distribution is a tempered

distribution.
Proof. The hardest part of this proof is already done in previous sections. Definition
3.14 and Proposition 3.5 show the new functional Ff is well-defined. The linearity
of both the Fourier transform in S and the tempered distribution f show that Ff is
also linear. Last, we saw in Theorem 3.6 that the Fourier transform is a continuous
operator in S, showing that if a sequence φn converges to zero in S, then so does the
sequence φ̂n . Thus, because of the continuity of f , the sequence hFf, φn i = hf, φ̂n i
converges to zero in C, showing that Ff is continuous.
We can also define the inverse Fourier transform in S 0 in the same way we did it
for the direct one.
Definition 3.16. Let f ∈ S 0 (Rn ) be a tempered distribution. Its inverse Fourier

transform, denoted by F −1 f or fˇ, is defined as follows:
hF −1 f, φi = hf, F −1 φi, ∀φ ∈ S(Rn ).
Theorem 3.17. The inverse Fourier transform of a tempered distribution is a

tempered distribution.
Proof. It is a direct consequence of Theorem 3.10 and an identical reasoning as in

Theorem 3.15.
Theorem 3.18. The Fourier inversion formula still works in S 0 . More precisely,
for every tempered distribution f ,
FF −1 f = f and F −1 Ff = f.
Proof. It is trivial from Definition 3.14 and Theorem 3.9.
Some more properties are given in the following proposition.
Proposition 3.19. Let f be a tempered distribution and denote by f˜ its transposi-

tion. Let also k ∈ (Z+ )n be a multi-index. Then
1. Dk (Ff ) = (−i)|k| F(xk f ).
2. xk Ff = (−i)|k| F(Dk f ).
3. F 2 f = (2π)n f˜.
4. F 4 f = (4π 2 )n f.
Proof. The properties are direct consequence of those which hold for Schwartz func-
tions, given in Propositions 3.4 and 3.11.
It is also easy to show that the Fourier transform is a linear and continuous
operator in the space of tempered distributions. For this, we must recall the concept
of convergence of tempered distributions, which is absolutely identical to that of
usual distributions defined in Section 1.2.3.
Theorem 3.20. The Fourier transform operator F : S 0 (Rn ) → S 0 (Rn ) and the in-
verse Fourier transform operator F −1 : S 0 (Rn ) → S 0 (Rn ) are linear and continuous
operators.
Proof. The linearity being clear, let fn be a sequence of tempered distributions

convergent to f . Then, for F, the continuity is obtained via
hFfn , φi = hfn , Fφi → hf, Fφi = hFf, φi.
The same reasoning is valid for F −1 .
The Fourier transform interacts with the convolution of distributions in a very

similar way it does with that of Schwartz functions.
Proposition 3.21. Let f ∈ S 0 (Rn ) and ψ ∈ S(Rn ). Then,
F(f ∗ ψ) = Ff · Fψ.
Proof. Let φ ∈ S(Rn ). Then,
hF(f ∗ ψ), φi = hf, ψ̃ ∗ Fφi.
Now, as F −1 Ff = f , we can write hf, ψ̃∗Fφi = hFf, F −1 (ψ̃∗Fφ)i . Recall equation

(3.22) from Proposition 3.12. Applying F −1 to that equation, we can write
hFf, F −1 (ψ̃ ∗ Fφ)i = hFf, (2π)n F −1 (ψ̃) · φi.
It enough to observe that, by property 1 of Proposition 3.11,
(2π)n F −1 (ψ̃)(ξ) = Fψ(ξ)
to assert that
hF(f ∗ ψ), φi = hFf, Fψ · φi = hFf · Fψ, φi.
Finally, we want to see what is the effect of applying the Fourier transform to a
direct product of tempered distributions. The result is surprisingly simple.
Proposition 3.22. Let f ∈ S 0 (Rn ) and g ∈ S 0 (Rm ). Then,
F(f (x) · g(y))(ξ, η) = Ff (ξ) · Fg(η).

Proof. Consider φ ∈ S(Rn+m ). By definition, we have

hF(f (x) · g(y))(ξ, η), φ(ξ, η)i = hf (x), hg(y), Fφ(x, y)ii.
In this point, we need to remark the fact that the Fourier transform can be
computed in different stages concerning different variables. This can be
easily checked using the definition. This means that, for example,
Fφ(x, y) = Fη (Fξ [φ])(x, y), (3.23)
where Z
Fξ φ(x, η) = φ(ξ, η)e−i(x·ξ) dξ.
Rn
This method will also be used several times in Chapter 4. This way, by (3.23),
hf (x), hg(y), Fφ(x, y)ii = hf (x), hg(y), Fη (Fξ [φ])(x, y)ii
= hf (x), hFg(η), Fξ (φ)(x, η)ii
= hf (x) · Fg(η), Fξ (φ)(x, η)i.
Remember that we saw in (2.17) that the direct product is commutative. Thus,
hf (x) · Fg(η), Fξ (φ)(x, η)i = hFg(η) · f (x), Fξ (φ)(x, η)i
= hFg(η), hf (x), Fξ (φ)(x, η)ii
= hFg(η), hFf (ξ), φ(ξ, η)ii
= hFf (ξ) · Fg(η), φ(ξ, η)i.
3.3.2 Examples
The Fourier transform plays an important role in the pursuit of fundamental solu-
tions of differential equation, as we will analyse in the following chapters. For this
reason, it is interesting to work out the Fourier transform of some distributions. In
this section, we will work with the delta function and polynomials.
Example 3.23. We want to check that for every multi-index k ∈ (Z+ )n and τ ∈ Rn ,
F(Dk [δ(x − τ )])(ξ) = i|k| ξ k e−iξ·τ . (3.24)
For that, we only need to use properties we presented in Section 3.3.1. Indeed,
choosing any Schwartz function φ,
hF(Dk [δ(x − τ )]), φi = (−1)|k| hδ(x − τ ), Dk φ̂(x)i
= (−1)|k| (−i)|k| hδ(x − τ ), F(ξ k φ)i
(3.25)
Z
|k| |k|
k
= i F(ξ φ)(τ ) = i ξ k φ(ξ)e−i(τ ·ξ) dξ
Rn
|k| k −i(ξ·τ )
= hi ξ e , φi.
In particular, for k = 0 and τ = 0,
Fδ(ξ) = 1(ξ). (3.26)
Example 3.24. Polynomials are of great importance in many areas of analysis, so

it is interesting to be able to work with their Fourier transform. Let us consider a
very general form of a polynomial, with k ∈ (Z+ )n and τ ∈ Rn ,
i|k| xk e−ix·τ .
We can expect to obtain a distribution related to the Delta function, by a similarity

to equation (3.24). In fact,
F(i|k| xk e−ix·τ ) = (−1)|k| (2π)n Dk δ(ξ + τ ), (3.27)
and we will use the mentioned (3.24) to prove so. Indeed, applying F and recovering
results from Proposition 3.19, we get
F(i|k| xk e−ix·τ ) = F 2 (Dk [δ(ξ − τ )])(ξ) = (2π)n (Dk [δ(ξ − τ )])(−ξ).
Now, for any Schwartz function φ,
h(Dk [δ(ξ − τ )])(−ξ), φ(ξ)i = (−1)2|k| Dk φ(−τ ) = (−1)|k| hDk δ(ξ + τ ), φi,
and we get the result. As a direct consequence, we obtain
F(xk ) = (2π)n i|k| Dk δ(ξ), (3.28)
and more basically,

F1 = (2π)n δ(ξ). (3.29)
In the same way, we obtain the Fourier transform of the complex exponential,
F(e−ix·τ )(ξ) = (2π)n δ(ξ + τ ). (3.30)
Finally, if we have a polynomial in R given by P (x) = nk=0 ak xk , where ak ∈ C

P
for every {1, · · · , n}, by linearity, its Fourier transform is given by the following
formula:
Xn
F(P (x)) = 2π ak ik δ (k) (x). (3.31)
k=0
Example 3.25. Recall the Gaussian function from Section 3.2.2. We analysed the
case in which the exponent is a non-real number with positive real part. A particular
case is when the exponent is purely imaginary, say k = bi, b ∈ R. We cannot argue
the same way, but we can use continuity features. Recall that the Fourier transform
F is a continuous operator in S 0 . It is immediate, by the dominated convergence
theorem, that
2 2
g (x) = e−(+bi)|x| → e−bi|x| in S 0 (Rn ) as → 0+ .
For that reason, the convergence does not change even if we apply the Fourier
transform, so by (3.13),
r n
π 2 2
F(g )(ξ) = e−|ξ| /4(+bi) → F(e−bi|x| )
+ bi
as → 0. On the other hand, taking the limit of the sequence, we get

r n
π 2 2
e−|ξ| /4bi = F(e−bi|x| ), (3.32)
bi
what shows that the formula shown in (3.13) is also valid for an imaginary number.
Chapter 4
Fundamental Solutions of
Differential Operators
In this last chapter, we will develop an efficient and general theory to reach what
we will call fundamental solutions of some differential operators. Eventually, we will
apply this theory to analyse some representative cases such as the heat equation,
the Schrödinger equation and the Laplace equation.
4.1 Generalised solutions and fundamental solutions

The first we need is to fix some notation. Let the following be a linear differential
equation of order m:
m
X
aα (x)Dα u = f, (4.1)
|α|=0
in which f ∈ D0 (Rn ) is a distribution and the coefficients aα are C ∞ functions. To

shorten the expressions to be used, we will denote the differential operator
m
X
L(x, D) = aα (x)Dα , (4.2)
|α|=0
thus the equation (4.1) turning into
L(x, D)u = f. (4.3)
In general, we are looking for a distribution u ∈ D0 (Rn ) which satisfies (4.3). Notice
that this solution may not be a usual solution we are used to, for it need not be a
function, and even if it is a function, its derivatives need not be functions. This is
where the concept of generalised solution comes from (also remember that another
name for distributions is generalised functions).
49
CHAPTER 4. FUNDAMENTAL SOLUTIONS OF DIFFERENTIAL
50 OPERATORS
Definition 4.1. Let L(x, D) be a differential operator, f ∈ D0 (Rn ) and an open

subset A ⊆ Rn . We say that a distribution u ∈ D0 (Rn ) is a generalised solution
of the equation L(x, D)u = f in the region A if
hL(x, D)u, ϕi = hf, ϕi
for every testing function ϕ ∈ D(Rn ) whose support is contained in A.
Consider now a differential operator with constant coefficients. If we keep notation

as in (4.2), in this case we will write
m
X m
X
L(x, D) = aα (x)Dα = aα Dα = L(D). (4.4)
|α|=0 |α|=0
In this situation, we will be able to obtain some particular solutions which will be
of extreme importance.
Definition 4.2. Let L(D) be a differential operator with constant coefficients. We

say that a distribution E ∈ D0 (Rn ) is a fundamental solution of the differential
operator L(D) if E satisfies
L(D)E = δ
in D0 (Rn ).
We will work with these fundamental solutions from now on, but of course, it is
important to know the reason for this. Notice that the connection between recently
defined generalised solutions and fundamental solutions remains unknown. The
following theorem will help to enlighten the situation.
Theorem 4.3. Let L(D) be a differential operator with constant coefficients and E
a tempered fundamental solution of it. Let also f be a Schwartz function on Rn .
Then, a solution to the equation L(D)u = f is given by
u = E ∗ f.
Proof. We know that L(D)E = δ, as E is a fundamental solution of the operator

L(D). Now, if we consider E ∗ f as a candidate solution, and considering notation
(4.4),
Xm
L(D)(E ∗ f ) = aα Dα (E ∗ f ).
|α|=0
Recall the rules of differentiation of the convolution given in (2.23); we can write
m
X m
X
aα Dα (E ∗ f ) = aα (Dα E ∗ f ).
|α|=0 |α|=0
4.1. GENERALISED AND FUNDAMENTAL SOLUTIONS 51
Once we know this, by linearity we get
L(D)(E ∗ f ) = L(D)E ∗ f = δ ∗ f = f, (4.5)
as we saw in (2.16) that the delta function is the identity element with respect to
convolution. As a consequence, the distribution E ∗ f is a solution to the equation
in the statement.
The reason for we bring fundamental solutions into the limelight is clear now. If
we manage to obtain one for an operator, much of the work will be done. In any
case, it is important to remark that, in general, they are not unique. Suppose we
have been able to get a solution to the homogeneous equation L(D)u = 0. If we call
it E0 , then
L(D)(E + E0 ) = L(D)E + L(D)E0 = δ + 0 = δ,
showing that E + E0 is also a fundamental solution.
It would be interesting to get, if possible, a procedure to work out fundamental
solutions. Here is where the Fourier transform comes into play.
Theorem 4.4. Let L(D) be a differential operator with constant coefficients and
E ∈ S 0 (Rn ). Then, E is a fundamental solution of L(D) if and only if
L(iξ)F(E)(ξ) = 1,
where, according to notation (4.4), L(x) = m α α |α| α

P
|α|=0 aα x and (iξ) = i ξ .
Proof. Remember that E is a fundamental solution of L(D) if L(D)E = δ. Knowing

this, apply the Fourier transform to the left hand side, considering the formulas in
Proposition 3.19:
m
X m
X
F(L(D)E)(ξ) = aα F(Dα E)(ξ) = aα i|α| ξ α FE(ξ) = FE(ξ)L(iξ). (4.6)
|α|=0 |α|=0
Once we know this, consider L(D)E = δ. Then, applying the Fourier transform and
by (4.6),
F(E)(ξ)L(iξ) = Fδ = 1.
Conversely, suppose F(E)(ξ)L(iξ) = 1. Then, by the inverse Fourier transform and
(4.6),
F(L(D)E) = 1 ⇒ L(D)E = F −1 Fδ = δ.
The question is how we could exploit this interesting result. It conveys that it is
enough to solve
L(iξ)F(E)(ξ) = 1,
or what we could expect to be the same,
1
F(E)(ξ) = . (4.7)
L(iξ)
52 OPERATORS
But we must be careful at this point. To consider (4.7), we need to be sure that the
expression on the right hand side is a distribution. Indeed, it is a usual function.
If it were locally integrable and of slow growth, we could treat it as a tempered
distribution and work with expression (4.7), being our solution

−1 1
E=F .
L(iξ)
But matters are not so simple if the function turns out not to be locally integrable.
In that case, we will need to manage to obtain a tempered distribution F solving
the equation
L(iξ)F = 1,
to finally obtain the result by the inverse Fourier transform, E = F −1 (F ).
4.2 Three fundamental solutions

In this section, we will make use of the method developed in Section 4.1 to obtain the
fundamental solutions to the heat operator, the Laplace operator and the Cauchy-
Riemann operator.
4.2.1 Fundamental solution of the heat operator

The very well-known heat equation models the evolution of the temperature in a
certain space, which could be a stick or a plane, and even objects in spaces of greater
dimension. Consider variable x ∈ Rn to be representative of space and variable t ∈ R
of time. Then, the heat equation is given as
∂u
(x, t) = a2 ∆x u(x, t), (4.8)
∂t
where ∆x is the Laplace operator concerning variable x and a is a positive constant
representing thermal diffusivity.
Basing on the heat equation (4.8), the heat operator LH (D) is
n
X
LH (D) = Dt − a2 ∆x = Dt − a2 Dx2i , (4.9)
i=1
where Dxi represents the partial derivative with respect to the variable xi . We know
that if we want to obtain a fundamental solution of LH (D), we need to solve the
equation
LH (D)E = Dt E(x, t) − a2 ∆x E(x, t) = δ(x, t).
Instead of applying the general Fourier transform as we did in Section 4.1, we will
only consider the Fourier transform on the variable x. For that,

∂E
Fx − a2 Fx (∆x E) = Fx (δ). (4.10)
∂t
4.2. THREE FUNDAMENTAL SOLUTIONS 53
An application of Lemma 2.21 yields to say that the Fourier transform in x and the
partial derivative in t commute, as they are concerning different variables, so

∂E ∂
Fx = Fx (E)(ξ, t). (4.11)
∂t ∂t
Also observe that
n
X
Fx (∆x E) = Fx (Dx2i E),
i=1
and by properties seen in Proposition 3.19,
Fx (Dx2i E) = i2 ξi2 Fx (E) = −ξi2 Fx (E).
This means that

n
X
Fx (∆x E) = − ξi2 Fx (E) = −|ξ|2 Fx (E). (4.12)
i=1
It remains to handle the right hand side of (4.10). In this case, by properties (2.18)
and (3.22),
Fx (δ(x, t)) = Fx (δ(x) · δ(t)) = Fx (δ(x)) · δ(t),
as can be easily checked. We also know that Fx (δ(x)) = 1(ξ), so
Fx (δ(x, t)) = 1(ξ) · δ(t). (4.13)
Eventually, combining (4.10), (4.11), (4.12) and (4.13), we obtain

∂
Fx (E)(ξ, t) + a2 |ξ|2 Fx (E)(ξ, t) = 1(ξ) · δ(t). (4.14)
∂t
Observe that we have a differential equation with respect to variable t,
∂
F (ξ, t) + k(ξ)F (ξ, t) = 1(ξ) · δ(t), (4.15)
∂t
which in turn, if we fix values for ξ, is similar to
d
G(t) + kG(t) = δ(t).
dt
If G were a function, it could be expressed in terms of an exponential. Nevertheless,
this is not a big setback, as we know that the Delta function is the derivative of the
Heaviside function, H, so we will be able to give a solution in terms of distributions
as
G(t) = H(t)e−kt . (4.16)
Indeed, G0 (t) = δ(t)e−kt − kG(t), and observe that by (1.7), δ(t)e−kt = δ(t), from
where we get the result. Following this idea, an identical calculation shows that a
solution to (4.15) is given by
2 |ξ|2 t
F (ξ, t) = (1(ξ) · H(t))e−a ,
54 OPERATORS
where the only variation is that Dt (1(ξ) · H(t)) = 1(ξ) · Dt H(t). But we can simplify
that expression. In fact,
2 |ξ|2 t 2 2
h(1(ξ) · H(t))e−a , ϕ(ξ, t)i = hH(t), h1(ξ), e−a |ξ| t ϕ(ξ, t)ii
Z Z
2 2
= H(t)e−a |ξ| t ϕ(ξ, t)dξdt
R Rn
2
= hH(t)e−a|ξ| t , ϕ(ξ, t)i.
For this,
2 |ξ|2 t
F (ξ, t) = H(t)e−a .
In short, if the fundamental solution we are looking for is E, then by the inverse
Fourier transform,
2 2
E = Fξ−1 (H(t)e−a |ξ| t ).
Clearly, the transform has no effect on H, and as properties 1 and 2 from Proposition
3.11 can be generalised to distributions,
H(t) 2 2
E= n
Fξ (e−a |ξ| t ).
(2π)
But we know how to work out the Fourier transform of the Gaussian; we analysed
it in section 3.2.2. Thus,
r n
H(t) π 2 2 H(t) 2 2
E(x, t) = n 2
e−|x| /4a t = √
n
e−|x| /4a t . (4.17)
(2π) a t (2a πt)
4.2.2 Fundamental solution of the Laplace operator

The Laplace equation can be used to describe the behaviour of several potentials
such as the electric, the gravitational and the fluid potentials. This fact makes it
also be called the equation of the potential. We are looking for distributions E such
that
∆E(x) = δ(x). (4.18)
As suggested in Theorem 4.4, the Fourier transform will be of great help. Indeed,
we do not need to make any new calculations, as in the case of the heat equation,
in (4.12), we got the Fourier transform of the Laplace operator. Hence,
−|ξ|2 F(E)(ξ) = 1(ξ).
As seen in (4.7), we expect to get
1
F(E)(ξ) = − . (4.19)
|ξ|2
It cannot be done always anyway. In fact, we have problems in the plane. This is
because in the unit ball B in Rn ,
Z Z 1
1 n−1
2
dx = σ(S ) rn−3 dr,
B |x| 0
which is finite if n ≥ 3, but not if n = 2. Thus (4.19) is not locally integrable in R2 ,

and therefore, it is not a distribution. On the other hand, for every n ≥ 3, (4.19) is
locally integrable and of slow growth in Rn , so we can work with it. We will work
with these two cases separately. When n ≥ 3, we will be able to get the solutions in
a direct way. On the other hand, we will check that a particular distribution fulfils
(4.18) in R2 .
• Case n ≥ 3. In this case, according to (4.19), we are looking for

−1 1
E = −F .
|ξ|2
We will make use of the Fourier transform of the Gaussian to get the result.
Indeed, recall from Section 3.2.2 that
r n
−k|x|2 π 2
F(e )(ξ) = e−|ξ| /4k .
k
Also recall that hFf, φi = hf, Fφi for every Schwartz function φ, from the
definition of the Fourier transform in S 0 . So taking f to be the Gaussian, we
can write the following equality.
r n Z Z
π −|ξ|2 /4k 2
φ(ξ)e dξ = φ̂(x)e−k|x| dx. (4.20)
k Rn Rn
Next step is to integrate (4.20) from 0 to infinity with respect to k, to obtain

Z ∞ r n Z Z ∞Z
π 2 2
φ(ξ)e−|ξ| /4k dξdk = φ̂(x)e−k|x| dxdk. (4.21)
0 k Rn 0 R n
We will analyse each side of (4.21) separately.
– The right hand side is a trivial calculation, as by Fubini’s theorem,

∞
Z ∞Z Z
e−k|x|2 Z
φ̂(x)
−k|x| 2
φ̂(x)e dxdk = φ̂(x) dx = dx

2
−|x| 2
0 Rn Rn Rn |x|
0

1 1
= h 2 , φ̂i = hF , φi,
|x| |x|2
and the value is finite for 1/|x|2 being locally integrable and of slow
growth.
– The left hand side is a bit more tricky. Again by Fubini’s theorem, we
can write equivalently
Z Z ∞
π n/2 −|ξ|2 /4k
φ(ξ) e dk dξ.
Rn 0 k
56 OPERATORS
Let us compute the inner integral. With a change of variables given by

r = |ξ|2 /4k, after several steps, we obtain
√
(2 π)n ∞ n −2 −r
Z
r 2 e dr.
4|ξ|n−2 0
This expression should remind us of the Gamma function. Indeed, as

Z ∞
Γ(k) = xk−1 e−x dx,
0
we can write
√ √
(2 π)n ∞ n −2 −r (2 π)n n
Z
r 2 e dr = Γ − 1 .
4|ξ|n−2 0 4|ξ|n−2 2
Thus we can write the whole integral as

√
(2 π)n n Z φ(ξ) n−2 n
n 1
Γ −1 n−2
dξ = 2 π 2Γ − 1 h n−2 , φi.
4 2 Rn |ξ| 2 |ξ|
So after working out each side of equality (4.21),

n
n−2 n
1 1
2 π2Γ − 1 h n−2 , φi = hF , φi, ∀φ ∈ S(Rn ).
2 |ξ| |x|2
Because of this, we have the following equality of distributions:

1 n−2 n
n 1
F = 2 π 2Γ − 1 . (4.22)
|x|2 2 |ξ|n−2
Remember that we seek F −1 (1/|ξ|2 ). We can use equation (4.22) for that,
because we know that
Γ n2 − 1

−1 1 1 1 1
F = F = √ .
|ξ|2 (2π)n |ξ|2 4( π)n |x|n−2
So eventually, the solution to (4.18) in Rn for n ≥ 3 is given by
Γ n2 − 1

1
E(x) = − √ n .
4( π) |x|n−2
• Case n = 2. As we stated before, we cannot make the same reasoning for

R2 . Instead, we will directly prove that the regular distribution E(x) = log |x|
satisfies (4.18) excepting for a constant. We will prove it by a process of
convergence. Consider

1 2 1
En (x) = log |x| + 2 , n ∈ N,
2 n
and observe that every En defines a regular distribution for being continuous
and thus locally integrable, being also of slow growth. We want to see that
En → E in S 0 (R2 ) if n → ∞. Let φ ∈ S(R2 ). Then,
Z
1 1
hEn , φi = log |x|2 + 2 φ(x)dx.
2 R2 n
Observe that | log(|x|2 + 1/n2 )φ(x)| ≤ (| log(|x|2 + 1)| + | log(|x|2 )|)|φ(x)|, and
the right-hand side function is integrable, because logarithms are locally integ-
rable functions of slow growth and φ is Schwartz, so the sum of the products is
integrable. For this, we can use the dominated convergence theorem and say
that
Z Z
1 2
lim hEn , φi = log |x| φ(x)dx = log |x|φ(x)dx = hE, φi.
n→∞ 2 R2 R2
This way, En (x) → E(x) in S 0 (R2 ). Now, observe that, as a function,
∂2 |x|2 + n12 − 2x2i

En (x) = , i = 1, 2,
∂ 2 xi (|x|2 + 1/n2 )2
so
2/n2
∆En (x) = .
(|x|2 + 1/n2 )2
Now, as the derivative is a continuous operator, as we saw in (2.20), we can
write
lim ∆En (x) = ∆E(x).
n→∞
Because of this, for every Schwartz function φ,
2/n2
Z
h∆E(x), φi = lim h∆En (x), φi = lim φ(x)dx.
n→∞ n→∞ R2 (|x|2 + 1/n2 )2
The integral of the right hand side can be transformed to the following by a
change of variables x → x/n:
2/n2
Z Z
2 x
2 2 2
φ(x)dx = 2 2
φ dx.
R2 (|x| + 1/n ) R2 (|x| + 1) n
The question now is if we can commute the limit with the integral. The
answer is positive, as for being Schwartz, φ is bounded, and by a change to
polar coordinates,
Z Z ∞
2 rdr
2 2
dx = 2π = 2π.
R2 (|x| + 1) 0 (1 + r2 )2
So taking every step done together,
Z
2
h∆E(x), φi = 2 2
φ(0)dx = 2πφ(0) = 2πhδ, φi.
R2 (|x| + 1)
58 OPERATORS
In short, we have obtained ∆E(x) = 2πδ(x). It is almost what we need. Even-

tually, linearity of distributions makes it possible to assert that a fundamental
solution of the Laplace operator in R2 is
1
log |x|.
2π
4.2.3 Fundamental solution of the Cauchy-Riemann operator

The Cauchy-Riemann equations are a system of two partial differential equations
which play an important role in the characterisation of holomorphic complex func-
tions. Let f (x, y) = u(x, y) + iv(x, y) be a complex function, where u and v are
real-valued. In this case, the Cauchy-Riemann equations concerning f are given by
ux = vy and uy = −vx .
To obtain an expression for the operator, let us define the following operators in R2 :
1 1
Dz = (Dx + iDy ) and Dz = (Dx − iDy ) .
2 2
One can prove that the condition Dz f = 0 is the same as saying that u and v satisfy
the Cauchy-Riemann conditions. It is also easy to check that
1
Dz Dz f = Dz Dz f = ∆f.
4
1
We have seen in Section 4.2.2 that the function 4π log |x2 + y 2 | is the fundamental
solution of the Laplace operator in the plane. This can be used to obtain the
1
fundamental solutions to both Dz and Dz . Indeed, calling E = 4π log |x2 + y 2 |,
1 1
Dz Dz E = Dz Dz E = ∆E = δ(x, y).
4 4
Therefore, it is clear that the fundamental solutions of the operators Dz and Dz are,
respectively, 4Dz E and 4Dz E, or explicitly,
1 1 1 1
Dz log |x2 + y 2 | = and Dz log |x2 + y 2 | = .
π πz π πz
4.3 Fundamental solution of the Cauchy problem

It is also possible to apply the idea of fundamental solutions to Cauchy problems.
Let L(D) be the differential operator in (4.4). The Cauchy problem associated to
it, with f ∈ S(Rn ), is to find u(x, t) such that
(
Dt u(x, t) − L(D)u(x, t) = 0, for t > 0, x ∈ Rn ,
(4.23)
u(x, 0) = f (x), for x ∈ Rn .
4.3. FUNDAMENTAL SOLUTION OF THE CAUCHY PROBLEM 59
We say that the distribution E(x, t) is a fundamental solution to (4.23) if

(
Dt E(x, t) − L(D)E(x, t) = 0, for t > 0, x ∈ Rn ,
E(x, 0) = δ(x), for x ∈ Rn .
Indeed, if E is a fundamental solution, then the convolution u = E ∗x f solves

(4.23). By ∗x we are denoting the convolution with respect to the variable x. We
will analyse a couple of examples to illustrate these facts.
4.3.1 Fundamental solution of the Cauchy problem for the heat

operator
When considering the heat equation, it is very common to work in a domain with
some physical meaning. In fact, we are trying to know how will the heat of a domain
evolve with time, so it is natural to consider t > 0. But this extra condition requires
that we fix an initial condition at time t = 0. This is what we call a Cauchy problem.
In the case of the heat equation, it looks like this:
(
ut (x, t) − a2 ∆x u(x, t) = 0, x ∈ Rn , t ≥ 0,
(4.24)
u(x, 0) = f (x), x ∈ Rn .
As stated before, we will work with f ∈ S(Rn ). In the following lines, we will try to
obtain a solution through a fundamental solution. In the case of the heat operator,
we need to solve
(
Et (x, t) − a2 ∆x E(x, t) = 0, for t > 0, x ∈ Rn ,
(4.25)
E(x, 0) = δ(x), for x ∈ Rn .
The method for solving this problem is to use the Fourier transform in variable x in
a similar way we did in 4.2.1. So applying F to the first equation in (4.25),
Fx (Et )(ξ, t) − a2 Fx (∆x E)(ξ, t) = 0,
which by the same reasoning as in 4.2.1 leads to

∂
Fx (E)(ξ, t) + a2 |ξ|2 Fx (E)(ξ, t) = 0.
∂t
If we now consider E to be a standard function, we know that
2 |ξ|2 t
Fx (E)(ξ, t) = c(ξ)e−a , (4.26)
for some function c(ξ). This function can be determined by using the initial condi-
tion. In fact,
Fx (E)(ξ, 0) = Fx (E(x, 0))(ξ) = F(δ)(ξ) = 1(ξ),
and we deduce that,
2 |ξ|2 t
Fx (E)(ξ, t) = e−a . (4.27)
60 OPERATORS
So we have to compute its inverse Fourier transform in order to get the fundamental
solution function E. Observe that the function of which inverse Fourier transform
we have to compute is the same we computed to obtain (4.17), so
2 |ξ|2 t 1 2 2
E(x, t) = Fξ−1 (e−a )= √
n
e−|x| /4a t
(2a πt)
is the fundamental solution to the Cauchy problem of the heat operator. Now, as
we said in the beginning of the section, the solution to the Cauchy problem is given
by a convolution in the variable x,
Z
1 −|x|2 /4a2 t 1 2 2
u(x, t) = f ∗x √
n
e = √
n
f (y)e−|x−y| /4a t dy. (4.28)
(2a πt) (2a πt) Rn
4.3.2 Fundamental solution of the Cauchy problem for the

Schrödinger operator
The Schrödinger equation describes the change in the quantum state of a physical
system over time. It is of great importance in quantum mechanics and in the analysis
of the wave function of several systems. One of the particular forms it takes is the
following, involving a wave function u:
ut (x, t) = ik∆x u(x, t), (4.29)
where k is a positive number related to Planck’s constant. Again, variable t repres-

ents time, and it is usual to have an initial condition,
u(x, 0) = f (x).
In this section, we will try to obtain a solution for the Cauchy problem corresponding
to the Schrödinger equation, so we will try to obtain a fundamental solution E which
fulfils (4.29) and also
E(x, 0) = δ(x).
Once we get that, the solution will be of the form E ∗x f .
In the same way we did for the heat equation, we will make use of the Fourier
transform. The similarity of the Schrödinger operator to the heat operator is obvi-
ous; the only important change is the complex number i. Nevertheless, this little
variation generates a need of caution in every computation, although the result will
be analogous to (4.28). The procedure is the same, so as the Fourier transform does
not affect constants, every step repeats until (4.27). Hence,
2
Fx (E)(ξ, t) = e−ik|ξ| t ,
so we need to compute the inverse Fourier transform of a function which looks like
a Gaussian. In this case, the exponent is a complex number whose real part is zero.
4.4. THE MALGRANGE-EHRENPREIS THEOREM 61
In this moment, we must recall Example 3.25, and more precisely, (3.32). With that
result, and also with Property 1 in Proposition 3.11, we can write
r n
1 π 2
E(x, t) = n
e−|x| /4kti ,
(2π) ikt
which is the fundamental solution to the Cauchy problem for the Schrödinger oper-
ator. Finally, the solution to the original problem is given by
Z
1 2
u(x, t) = E ∗x f = √ f (y)ei|x−y| /4kt dy.
n
(2 πkti) Rn
4.4 The Malgrange-Ehrenpreis theorem

The Malgrange-Ehrenpreis theorem was one of the first results in which the potential
of distributions was clearly shown to the mathematical community. Dated in the
mid 1950s, it was proved independently by B. Malgrange1 and L. Ehrenpreis2 . The
same way we have been able to obtain fundamental solutions to the heat operator
and the Laplace operator in Section 4.2, the theorem states that it is possible to do
so for any partial differential operator with constant coefficients.
Theorem 4.5 (Malgrange-Ehrenpreis). Every non-zero partial differential operator

with constant coefficients admits a fundamental solution in the space of distributions
D0 .
Several different proofs have been published so far, also using completely inde-
pendent arguments. The former proofs were non-constructive, and they made use of
the Hahn-Banach theorem. Other proofs were based on the L2 -theory. Some of them
also assert that the fundamental solutions can be taken from the space of tempered
distributions. In the last decades, constructive proofs have been developed, with
explicit expressions for a solution which have been turning more and more compact.
A recent short proof using Fourier transforms is in [9], where references to other
proofs can be found.
1
Bernard Malgrange (born in 1928, Paris) was a student of Laurent Schwartz. He has been
professor at Strasbourg, Orsay and Grenoble, and he works on differential equations.
2
Leon Ehrenpreis (1930-2010, Brooklyn, NY) was professor at Temple University, Philadelphia,
PA.
Appendix A
Laurent Schwartz: Life and

Work
Laurent Schwartz was a brilliant mathematician who not only focused on scientific
activities during his life. He was also very committed to French and international
social and political problems.
He was born in Paris in 1915 to a non-practicing Jewish family. His father was
the first Jew surgeon in a hospital in Paris, in a time when anti-Semitism was on the
rise in France. Mathematics were present in his childhood; his uncle was a teacher,
and the famous Jacques Hadamard was the uncle of his mother. Even if he was
good at them, he showed better abilities in literature and languages, but advised
by some of his teachers, and after a hard work, he enrolled in the École Normale
Supérieure (ENS), where he met the daughter of the also mathematician Paul Lévy,
Marie-Hélène. They married some years later.
The environment at ENS made him get close to communist, indeed Trotskyist
ideas, an inclination he would never leave behind, and which would cause him some
problems. The young Laurent finished his studies in 1937, and after two years of
military service, he moved, together with his wife, to Toulouse, where both of them
started working in a research institute. Nevertheless, their condition of Jews made
their situation extremely delicate after France fell under Nazi control in 1940 and
the pro-axis Vichy government led by Marshal Philippe Pétain imposed restrictive
laws in the area. They decided to move to the city of Clermont-Ferrand for academic
reasons, where the University of Strasbourg had been transferred during the war.
It was there where Schwartz made contact with the Bourbaki group, a collective
which had a decisive influence in his academic development. After finding himself
forced to move to the Italian border for political reasons, he stayed in Grenoble for
some months. In this time, and during the following years, he started to investigate
in several fields, such as Fréchet spaces and generalised derivatives, after reading
articles about harmonic and polyharmonic functions. One night, he came up with
the idea of generalised functions in the most beautiful night of his life.
Once the war was over, he became professor of the University of Grenoble for a
little time before moving to Nancy, city where he worked in his ideas, and published
63
64 APPENDIX A. LAURENT SCHWARTZ: LIFE AND WORK
several articles in which he presented many concepts and results concerning distri-
butions. These innovative contributions made him well-known in the whole country
and even internationally. In fact, he received a Fields medal for his work in 1950
in Harvard. In the following year he published his two-volume treatise Théorie des
distributions, where he put all his discoveries together.
In 1952, he became professor of the University of Paris. Some years later, in 1958,
his father-in-law Paul Lévy retired as professor in the prestigious, though lately old-
fashioned École Polytechnique. Even if he was not intending to make a request for
the position at first, he was persuaded by representatives of the École, in an attempt
to modernise their institution. He accepted, and from that time on, until he died, he
carried an important work in the modernisation not only of the École, but also of the
whole French university system. He rapidly revitalised the mathematical activity in
a centre which had fallen in a situation of academic paralysis.
He stayed at the École Polytechnique until 1980, but it was not his last work at all.
In 1985, he became chairman of the National Evaluation Committee, an organisation
created by the French government after his advise. He also was member of the French
Academy of Science since 1975. He died in Paris in 2002 at the age of 87.
Distributions were not his unique area of interest, and he worked in several fields
during his life. Before developing the theory which made him famous, he worked on
polynomials, semi-periodic functions, harmonic analysis and exponential sums. His
thesis in 1942 was indeed involving this last matter. Once he published his main
work Théorie des distributions, a second edition with some corrections and more
information was reprinted in 1966, and extended the field with vectorial distributions
and the kernel theorem. Afterwards, he succeeded in discovering applications of his
work to theoretical physics. He also made important contributions to the field of
probability and integration, especially in his latter works.
As stated in the beginning, Laurent Schwartz was a politically active man. He
enrolled in the Trotskyst party in France during his stay in ENS, and excepting the
lapse of the war, he was an active militant. He even presented himself for candidate
in the legislative elections in 1945 and in 1946, but he quitted in 1947. He no longer
took part in political parties, but anyway he had serious trouble when entering the
USA in 1950 to receive his Fields medal. His colleague Marshall Stone discovered
he was classified as a dangerous communist for the federal government.
Schwartz also played an active part in the decolonisation processes of the French
Indochina in 1954 and Morocco and Tunisia in 1956, and in 1960, together with
many intellectuals as Jean Paul Sartre and Simone de Beauvoir, he made a public
manifesto for the right of the French youth not to take part in the Algerian War.
This last action had a professional cost for Schwartz; indeed, he was working at
the Ministry of Defence-run École Polytechnique, from which he was fired in 1961.
Nevertheless, he was readmitted in 1963.
In the next decade, he made a strong campaign against the Vietnam War. He
even travelled to that country and met the communist leader Ho Chi Minh, being
part of the Russell Tribunal to find evidence of war crimes. In the context of the
Afghanistan War after the Soviet invasion, he was president of the International
65
Bureau for Afghanistan.

In a more academic context but in the same path of defence of the human rights,
he tried to make public the repression some intellectuals were suffering in countries
such as Bolivia, Chile, Czechoslovakia, Morocco, Uruguay or the USSR.
As an interesting anecdote, it is remarkable his deep interest in butterflies. In his
journeys to tropical countries in which he was to give conferences, he would try to
obtain new species for his large collection. There are even two species he discovered
for the first time, and as the tradition goes, they were named after him (Xylophanes
schwartzi and Clanis schwartzi ). In his last years, his collection had almost 20.000
insects, which he donated to the Museum of Natural History.
References: [1], [4], [5].

Bibliography
[1] Fernando Bombal. Laurent Schwartz, el matemático que querı́a cambiar el

mundo. La Gaceta de la RSME, 6, 177-201, 2003.
[2] Joan Cerdà. Linear Functional Analysis. American Mathematical Society,

Providence, RI, 2010.
[3] Gerald B. Folland. Introduction to Partial Differential Equations, second edi-

tion. Princeton University Press, Princeton, NJ, 1955.
[4] Jesús Hernández. Matemático, entomólogo, persona decente. Laurent Schwartz,

un mathématicien aux prises avec le siècle. La Gaceta de la RSME, 2, 319-326,
1999.
[5] Gilles Pisier. Vida y obra cientı́fica de Laurent Schwartz. La Gaceta de la

RSME, 7, 691-699, 2004.
[6] Karen Saxe. Beginning Functional Analysis. Springer-Verlag New York, Inc.,
2002.
[7] Robert Strichartz. A Guide to Distribution Theory and Fourier Transforms.

CRC Press, Boca Ratón, FL, 1994.
[8] V. S. Vladimirov. Equations of Mathematical Physics. Marcel Dekker, Inc.,

New York, 1971.
[9] Peter Wagner. A New Constructive Proof of the Malgrange-Ehrenpreis The-

orem. Amer. Math. Monthly, 116, 457-462, May 2009.
[10] A. H. Zemanian. Distribution Theory and Transform Analysis: An Introduc-

tion to Generalized Functions, with Applications, reprint edition. Dover Pub-
lications, New York, 1987.
67

GAL Daniel Eceizabarrena - Distribution Theory and Fundamental Solutions of Differential Operators

Uploaded by

Copyright:

Available Formats

GAL Daniel Eceizabarrena - Distribution Theory and Fundamental Solutions of Differential Operators

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GAL Daniel Eceizabarrena - Distribution Theory and Fundamental Solutions of Differential Operators

Uploaded by

Copyright:

Available Formats

Final Degree Dissertation

Distribution Theory and Fundamental Solutions of

Leioa, June 24th 2015

3 The Fourier Transform 33

4 Fundamental Solutions of Differential Operators 49

4.2 Three fundamental solutions . . . . . . . . . . . . . . . . . . . . . . . 52

A Laurent Schwartz: Life and Work 63

Definition and properties of the Fourier transform in S will be described af-

1.1 The space D of testing functions

Definition 1.1. Let Ω ⊂ Rn be an open set and ϕ : Ω → C a function. We say that

1. supp ϕ is compact, and

The space of testing functions in Ω is denoted by D(Ω).

Proposition 1.4. Let ϕ, φ ∈ D(Ω) and a, b ∈ C. Then, aϕ + bφ ∈ D(Ω). Therefore,

Proof. Let A = supp ϕ and B = supp φ. Then, aϕ(x) + bφ(x) = 0 for x ∈ / A ∪ B.

where every Pk is a polynomial. Now, after a change of variables,

Proposition 1.7. Let ϕ ∈ D(Rn ), τ ∈ Rn , a ∈ R − {0} and g ∈ C ∞ (Rn ). Then,

1. ϕ(t + τ ), ϕ(−t), ϕ(at) ∈ D(Rn ).

1.2 The space D0 of distributions

2. it is continuous, in the sense that if {ϕm } → ϕ in D(Rn ), then

Proposition 1.11. If a functional f : D(Rn ) → C is known to be linear, it is

Proof. Let ϕm → ϕ be a convergent sequence in D(Rn ). Then, by definition, {ϕm −

∀ > 0, ∃N ∈ N | ∀n > N, |ϕn (t)| < , ∀t ∈ Rn ,

hδ, ϕi = ϕ(0) (1.3)

Consider the testing function (1.1), and define

If we take limits on both sides, we get

By the definition of ζ, it is clear that Im ζ = [0, e−1 ], so every ζn is bounded by e−1 .

This expression defines a distribution, and it is called Cauchy principal value of

By setting χ for the characteristic function,

and the integrand remains bounded by an integrable function on R. By the domin-

As a consequence, the Cauchy principal value is a distribution.

1.2.4 Operations on distributions

1. The sum of distributions, f + g, as

2. The multiplication by a constant, αf , as

It is straightforward to check that these two functionals define distributions. With

3. The shifting of a distribution, f (t − τ ), as

4. The transposition of a distribution, f (−t), as

hf (−t), ϕi = hf, ϕ(−t)i, ∀ϕ ∈ D.

This is also a well-defined functional, because ϕ(−t) is a testing function, as

5. The dilation of a distribution, f (at), as

automatically by the continuity of f . Now, if K ⊂ Rn is the domain containing

6. The multiplication of a distribution by a smooth function, gf , as

hgf, ϕi = hf, gϕi, ∀ϕ ∈ D.

This is a completely trivial definition. Indeed, if we consider f to be a regular

as Proposition 1.7 asserts that gϕ ∈ D. The general proof that gf is indeed a

1.2.5 Differentiation of distributions

Definition 1.16. Let f be a distribution in D0 (Rn ). The partial derivatives of f

which is the one dimensional version of Definition 1.16.

α being a multi-index. So we are allowed to say that the sequence { ∂ϕ

to zero in D. Finally, as f is a distribution,

Theorem 1.18. Let f be a distribution. Then, for every i, j ∈ {1, · · · , n},

Proposition 1.20. Differentiation is a linear operation in the space of distributions,

Proof. Linearity is trivial. For continuity, Proposition 1.17 and convergence of fn

hDk fn , ϕi = (−1)|k| hfn , Dk ϕi → (−1)|k| hf, Dk ϕi = hDk f, ϕi.

An interesting question is what will the derivatives of some particular distributions

so if we write the derivative with respect to variable i ∈ {1, · · · , n}, we get

In general, from the above we immediately get that if k = (k1 , · · · , kn ) ∈ (Z+ )n ,

Therefore, in terms of distributions, we get

H 0 (t) = δ(t). (1.8)

Notice that if we consider H as a function, H 0 (t) = 0 holds almost everywhere.

∀ > 0, ∃N ∈ N | ∀n > N, |ϕn (t)| < , ∀t ∈ Rn ,

∀|t| > M, |f (t)| < |t|N . (2.8)