Orthogonal polynomials and Möbius transformations

R. S. Vieira, V. Botta
Universidade Estadual Paulista – UNESP, Faculdade de Ciências e Tecnologia, Departamento de Matemática e Computação,
CEP. 19060-900, Presidente Prudente, SP, Brasil.
arXiv:1904.10766v1 [math.CV] 15 Apr 2019

Abstract
Given an orthogonal polynomial sequence on the real line, another sequence of polynomials can be found by
composing these polynomials with a general Möbius transformation. In this work, we study the properties
of such Möbius-transformed polynomials. We show that they satisfy an orthogonality relation in given curve
of the complex plane with respect to a varying weight function and that they also enjoy several properties
common to the orthogonal polynomial sequences on the real line — e.g. a three-term recurrence relation,
Christoffel-Darboux type identities, their zeros are simple, lie on the support of orthogonality and have
the interlacing property, etc. Moreover, we also show that the Möbius-transformed polynomials obtained
from classical orthogonal polynomials also satisfy a second-order differential equation, a Rodrigues’ type
formula and generating functions. As an application, we show that Hermite, Laguerre, Jacobi, Bessel
and Romanovski polynomials are all related to each other by a suitable Möbius transformation. New
orthogonality relations for Bessel and Romanovski polynomials are also presented.
Keywords: Orthogonal polynomials, Möbius transformations, varying weight functions, classical
orthogonal polynomials, Bessel polynomials, Romanovski polynomials.
2010 MSC: 42C05, 33C47, 30C35

Contents

1 Introduction 2

2 Möbius-transformed polynomials 3

3 Möbius-transformed orthogonal polynomials 4

3.1 Orthogonality relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.2 Three-term recurrence relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.3 Christoffel-Darboux identities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.4 Properties of the zeros. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

4 Classical Möbius-transformed orthogonal polynomials 10

4.1 Second-order differential equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2 Weight function differential equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.3 Rodrigues’ type formulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.4 Generating functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

5 Applications 16
5.1 From Laguerre to Bessel polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.2 From Jacobi to Romanovski polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.3 From Jacobi to Laguerre polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.4 From Jacobi to Hermite polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1. Introduction

Since its introduction in mathematical sciences, the concept of orthogonal polynomials has been extended
and generalized in many ways. The first concept of polynomial orthogonality referred to sequences of
polynomials orthogonal with respect to a real, non-negative and continuous weight function (the classical
orthogonal polynomials of Hermite, Laguerre and Jacobi are the most known examples). This concept was
soon after generalized in order to cover the cases of discrete measures (the Charlier polynomials being the
prototype in this class). The range of known families of orthogonal polynomials increased exponentially after
Askey showed that hypergeometric functions lead to several families of orthogonal polynomials; together with
their q-analogues, a concept that was introduced by Hahn, this resulted in Askey scheme for the classification
of hypergeometric orthogonal polynomials [1]. All those polynomials are orthogonal on the real line; of course,
an obvious generalization is to consider polynomials that are orthogonal in other curves of the complex plane.
This was already taken into account, the most known example being the orthogonal polynomial sequences
on the unit circle [2, 3]. Since then, many other concepts of polynomial orthogonality were defined: we
can cite, for instance, quasi-orthogonal polynomials [4], polynomials that are orthogonal with respect to
a varying measure [5], exceptional orthogonal polynomials [6, 7], para-orthogonal polynomials [8], among
others.
In this work, we present another generalized class of orthogonal polynomials which we call Möbius-
transformed orthogonal polynomials. As the name says, these sequences of polynomials are generated by the
action of a general Möbius transformation over an orthogonal polynomial sequence on the real line. In what
follows, we shall show that such sequences of polynomials are orthogonal in a given curve of the complex
plane, with respect to a simple varying weight function. Besides, these Möbius-transformed orthogonal
polynomials enjoy many properties of the usual orthogonal polynomials on the real line as, for example, their
zeros are all simple, lie on the support of orthogonality and have the interlacing property; they also satisfy
a three-term recurrence relation and identities similar to those of Christoffel-Darboux can be derived. In
particular, the Möbius-transformed polynomials obtained from a classical orthogonal polynomial sequence
also satisfy second-order differential equations, their weight function also satisfy a first-order differential
equation of Pearson’s type, there also exist Rodrigues’ type formula for them and they can also be found
through simple generating functions.
We highlight that these Möbius-transformed orthogonal polynomials can be of importance in Applied
Mathematics and Physics. In fact, the composition of polynomials with a Möbius transformation is already
considered in many applications: we can cite, for example, its role in some of the fastest polynomial isolating
and root-finding methods available to date (e.g., Akritas method [9]). We can also find the number of
zeros that a complex polynomial has on the unit circle through Sturm theorem if we previously consider
the composition of the testing polynomial with the following pair of Möbius transformations (often called
Cayley transformations),
x+1 x−i
M (x) = −i , W (x) = ,
x−1 x+i
that map, respectively, the real line onto the unit circle and vice versa [10]. Moreover, if we act with the
first of these special Möbius transformations over an orthogonal polynomial sequence on the real line, then
we can verify (from the results that will be reported in this work) that a new sequence of polynomials whose
zeros lie all on the unit circle is obtained. Consequently, we shall get an infinite sequence of self-inversive or
self-reciprocal polynomials (which are complex or real polynomials whose zeros are all symmetric with respect
to the unit circle). It is worth to mention that self-inversive and self-reciprocal polynomials are important in
many fields of Physics and Mathematics — see [11] and references therein. Finally, such Möbius-transformed
polynomials are also orthogonal on the unit circle with respect to a varying measure; because of this, we
can conclude that they are intimately related with the usual orthogonal polynomials on the unit circle and
also with para-orthogonal polynomials. The study of this special case will be reported a separated paper
due to particular importance (in this work we shall consider a general Möbius transformation).

Email addresses: rs.vieira@unesp.br (R. S. Vieira), vanessa.botta@unesp.br (V. Botta)

2
 Another field where Möbius-transformed polynomials can be of relevance is related to boundary value
problems and second-order differential equations, which are usually found in physical problems. It is well
known the role of classical orthogonal polynomials in Sturm-Liouville problems on the real line; because
Möbius transformations preserve the main characteristics of the polynomials, we can conclude from this
that the corresponding Möbius-transformed polynomials will be associated with boundary values problems
defined in some curves (arc circles or line segments) of the complex plane and that they will also obey
second-order differential equations with complex coefficients. We can cite the so-called relativistic classical
orthogonal polynomials, which are orthogonal polynomials on the real line with respect to a varying measure
that appear in some problems of relativistic quantum mechanics [12, 13, 14]. These relativistic classical
orthogonal polynomials are actually special cases of Möbius-transformed orthogonal polynomials — in fact,
it was showed by Ismail that they can be obtained from Jacobi polynomials through the action of a specific
Möbius transformation with real parameters [15].
This work is organized as follows: the concept of Möbius-transformed polynomials is introduced in
Section 2. In Section 3, we begin our study on Möbius-transformed orthogonal polynomials, where their
orthogonality, the three-term recurrence relations and the corresponding Christoffel-Darboux type identities
are discussed. The classical Möbius-transformed polynomials are studied in Section 4, where we show that
they also satisfy second-order differential equations, that their weight functions satisfy a first-order Pearson’s
type differential equation, we derive Rodrigues’ like formulas for them and the respective generating functions.
Finally, in Section 5 we show, as an application, that all the sequences of Hermite, Laguerre, Jacobi, Bessel
and Romanovski polynomials are related to each other essentially by a Möbius transformation; this also
provided new orthogonality relations for Bessel and Romanovski polynomials.

2. Möbius-transformed polynomials

A Möbius transformation is a conformal mapping defined of the extended complex plane C∞ = C ∪ {∞}
(i.e., the Riemann sphere) by the formula [16, 17]:
ax + b
M (x) = , (1)
cx + d
where x ∈ C∞ and a, b, c and d are complex numbers such that ∆ = ad − bc 6= 0. Its inverse, which is also
a Möbius transformation, is given by the expression,
dx − b
W (x) = − . (2)
cx − a
Together with the relations

M (−d/c) = ∞, M (∞) = a/c, W (a/c) = ∞, W (∞) = −d/c,

transformations (1) and (2) become continuous on C∞ . The Möbius transformations include as special cases
many important geometric transformations as, for instance, translations M (x) = x + b, rotations M (x) = ax
(|a| = 1), inversions M (x) = 1/x and linear transformations M (x) = ax+b. They also have many interesting
properties which can be found in [16, 17].
Given a complex polynomial p(x) = p0 + p1 x + · · · + pn−1 xn−1 + pn xn of degree n, we can compose p(x)
with M (x) so that rational function
   n−1  n
ax + b ax + b ax + b
r(x) = p0 + p1 + · · · + pn−1 + pn ,
cx + d cx + d cx + d
is is obtained as the result. Thus, multiplying r(x) by (cx + d)n , a new polynomial is obtained:
n n
q(x) = (cx + d) r(x) = (cx + d) p(M (x))
n n−1 n−1 n
= p0 (cx + d) + p1 (ax + b) (cx + d) + · · · + pn−1 (ax + b) (cx + d) + pn (ax + b) . (3)

3
We shall refer to r(x) as a Möbius-transformed rational function and to q(z) as a Möbius-transformed
polynomial. Notice that the original polynomial p(x) can be retrieved from r(x) and q(x) by the formula

q(W (x)) (−1)n n

p(x) = r(W (x)) = n = (cx − a) q(W (x)). (4)
(cW (x) + d) ∆n

Not always the Möbius-transformed polynomial q(x), as defined in (3), has the same degree as the original
polynomial p(x). The precise condition for this to be true is the following:
Theorem 1. Let p(x) = pn xn + · · ·+ p0 be a complex polynomial of degree n. Then, the Möbius-transformed
polynomial q(x) will be polynomial of degree n if, and only if, c = 0 or, either, if c 6= 0 and p(a/c) 6= 0.
Proof. If c = 0 the proof is trivial; thereby, assume c 6= 0. Expanding the binomials in (1) we find that the
leading coefficient qn of q(x) is given by the expression,
n
X n
X  a n−k h  a n i a
qn = pn−k an−k ck = cn pn−k = cn p n + · · · + p 0 = cn p ,
c c c
k=0 k=0

from which we conclude that q(x) will be degree n whenever a/c is not a zero of p(x).
We can also find how the zeros of the Möbius-transformed polynomial q(x) are related with the zeros of
the original polynomial p(x) through the following:
Theorem 2. Let ξ1 , . . . , ξn denote the zeros of a polynomial p(x) of degree n and ζ1 , . . . , ζn the respective
zeros of the Möbius-transformed polynomial q(x). Assuming that p(a/c) 6= 0, we have that,

ζ1 = W (ξ1 ) , . . . , ζn = W (ξn ).

Proof. For any zero ξk (1 6 k 6 n) of Pn (x) we have from (4) that:

n
(−1) n
p (ξk ) = (cξk − a) q (W (ξk )) = 0, 1 6 k 6 n.
∆n
This plainly shows us that ζk = W (ξk ) will be a zero of Qn (x) whenever ξk 6= a/c.
From this we can also conclude that the Möbius-transformed polynomial q(x) will be a polynomial of
degree n − m whenever the original polynomial p(x) of degree n has a zero of multiplicity m at the point
x = a/c. This, of course, is due to the cancellation of the m poles generated by the mapping W (a/c) = ∞
with the corresponding m zeros of the factors (cξ − a) generated by the action of the Möbius transformation
over p(z).

3. Möbius-transformed orthogonal polynomials

Now, let us consider an orthogonal polynomial sequence defined in an given interval (l, r) ⊆ R with
∞
respect to a given weight function w(x). That is, let us consider a sequence P = {Pn (x)}n=0 in which each
polynomial Pn (x) ∈ P has degree n and such that the following orthogonality relation is satisfied [18, 19]:
Z r
Pm (x)Pn (x)w(x)dx = Km δm,n , (5)
l

where δm,n is the Kronecker symbol, Km are positive constants and the interval (l, r) ∈ R can be either
finite, semi-infinite or infinite1 .

1 We remark that definition above can also be generalized to the case where the weight function w(x) is discrete by interpreting

the integral in the sense of Stieltjes or Lebesgue, although in this work we shall only consider the case where w(x) is continuous.

4
 Composing each polynomial Pn (x) ∈ P with the Möbius transformation (1), we shall obtain a sequence
∞
R = {Rn (x)}n=0 of rational functions, where,
Rn (x) = Pn (M (x)) , n > 0. (6)
n
Multiplying each rational function Rn (x) by (cx + d) , a sequence of Möbius-transformed polynomials Q =
∞
{Qn (x)}n=0 is obtained:
Qn (x) = (cx + d)n Rn (x) = (cx + d)n Pn (M (x)), n > 0. (7)
As before, the relationship between Pn (x), Qn (x) and Rn (x) is the following:
n
Qn (W (x)) (−1)
Pn (x) = Rn (W (x)) = n = (cx − a)n Qn (W (x)), n > 0. (8)
(cW (x) + d) ∆n
The properties of such sequences of Möbius-transformed rational functions and polynomials will be
analyzed in what follows.

3.1. Orthogonality relations

Let us begin our analysis by presenting the orthogonality relation satisfied by the Möbius-transformed
rational functions Rn (x) and polynomials Qn (x).
∞
Theorem 3. Let P = {Pn (x)}n=0 be an orthogonal polynomial sequence on the real line whose polynomials
Pn (x) satisfy the orthogonality relation (5) for a given weight function w(x) which does not vanish along
∞
the orthogonality interval I = (l, r) ⊆ R. Then, the sequence R = {Rn (x)}n=0 of the Möbius-transformed
rational functions Rn (x) defined in (6) will satisfy the following orthogonality relations:
Z
Rm (x)Rn (x)ω(x)dx = Km δm,n , (9)
Γ

where
w(M (x))
ω(x) = w(M (x))M ′ (x) = ∆ 2. (10)
(cx + d)
Similarly, the sequence Q = {Qn (x)}∞n=0 of Möbius-transformed polynomials Qn (x) defined in (7) will satisfy
orthogonality relations of the form,
ω(x)
Z
Qm (x)Qn (x)ωm,n (x)dx = Km δm,n where ωm,n (x) = m+n . (11)
Γ (cx + d)
In both cases, the path of integration is
Γ = {z ∈ C : z = W (x), λ < z < ρ}, where λ = W (l) and ρ = W (r), (12)
which corresponds to a curve (a straight line or a circle) in the complex plane whose initial and final points
are λ = W (l) and ρ = W (r), respectively. Finally, the new weight functions ω(x) and ωm,n (x) do not vanish
along the curve Γ , except at infinity.
Proof. The orthogonality relation (9) for the Möbius-transformed rational functions Rn (x) follows from
the change of variable x = M (y) in the integral (5), taking into account that the Jacobian is M ′ (x) =
2
∆/ (cx + d) . The corresponding orthogonality condition (11) for the Möbius-transformed polynomials
Qn (x) follows from their definitions given in (7) and from (9), after we absorb the factors (cx + d)m
n
and (cx + d) into the varying weight function ωm,n (x). Besides, notice that the path of integration
I = {x ∈ R : l < x < r} is mapped through the inverse Möbius-transformation, W (x), onto the curve
Γ , as defined in (12), as well as the limits x = l and x = r are respectively mapped to λ = W (l) and
ρ = W (r). Finally, consider a point y on the curve Γ so that we can write y = W (x) where x ∈ I; in
this case, (10) simplifies to ω(W (x)) = −w(x) (cx − a). Thus, provided that w(x) does not vanish on I, we
conclude that ω(W (x)) can vanish only at x = a/c; but this means that y = W (a/c) = ∞ so that ω(y) can
vanish only at y = ∞ along the curve Γ , and the same line of reasoning holds for ωm,n (x).
5
 Notice that integrals (9) and (11) should be regarded as a complex contour integrals. Whenever the
point x = a/c is not on the orthogonality curve I = (l, r) ⊆ R, there is nothing to worry about, as in
this case the orthogonality curve Γ will be finite, open and connected, so that these integrals will depend
only on their extreme points λ = W (l) and ρ = W (r). However, some care should be taken otherwise. In
fact, if x = a/c is on I, then the curve Γ will no longer be connected: instead, it will be constituted by
two disjoint straight lines in the complex plane passing through the point W (a/c) = ∞ — the first one
starts at λ = W (l) and goes towards the complex infinity, while the other starts at infinity and ends at
ρ = W (r). Notice as well that if I comprehends the whole real line (as in the case of Hermite polynomials,
for instance), then Γ become either an infinite straight line or a circle in the complex plane; in fact, we
have that W (−∞) = W (∞) = −d/c. Notice that in the extended complex plane C∞ , Γ will be always a
connected curve.
The weight function ωm,n (x) that appears in the orthogonality relation (11) is a special type of which is
called a varying weight function. Varying weight functions are measures that depend on some parameters of
the polynomials entering in the orthogonality relation. Such kind of weights arises naturally from sequences
∞
of orthogonal polynomials P = {Pnα (x)}n=0 whose members depend on one (or more) parameter α (e.g.,
the associated Laguerre polynomials Lα n (x)). In fact, provided that, for fixed α, these polynomials satisfy
an orthogonality relation of the usual form (5) with respect with a given weight function wα (x), we can
∞
construct from P a modified sequence P {α} = {Pnαn (x)}n=0 in which each polynomial Pnαn (x) has a different
parameter αn ; it follows in this case that the polynomials Pnαn (x) will satisfy an orthogonality relation with
respect to a varying weight function, namely:

Z r wαn (x), m < n,

αm αn
Pm (x)Pn (x)wαm ,αn (x)dx = Km δm,n where, wαm ,αn (x) = wαm (x), m > n,
l 
wα (x), m = n,


for some positive constants Km , m > 0. This works because every polynomial of degree n that belongs to
a sequence of orthogonal polynomials on the real line is orthogonal to any polynomial of degree less than n
— thus, to establish the orthogonality of a sequence of polynomials with different parameters we just need
to pick up the weight function belonging to the polynomial of higher degree appearing on the orthogonality
relation; besides, when the two polynomials have the same degree, they will also have the same value of α so
that the value of the integral is the same as that one for the fixed value αm = αn of α. Finally, we remark
that sequences of polynomials orthogonal with respect to varying measures appear in several applications,
both in mathematics [5] as in physics [20].

3.2. Three-term recurrence relations

∞
It is well-known that any sequence P = {Pn (x)}n=0 of orthogonal polynomials on the real line satisfies
a three-term recurrence relation of the form [18, 19]:

Pn (x) = (An x − Bn ) Pn−1 (x) − Cn Pn−2 (x), n > 1, (13)

with the initial conditions P−1 (x) = 0 and P0 (x) = 1, where An , Bn and Cn are real constants with
∞
An > 0 and Cn > 0 for each n > 1. In the following, we shall show that the sequences R = {Rn (x)}n=0
∞
and Q = {Qn (x)}n=0 of Möbius-transformed rational functions and polynomials also satisfy three-term
recurrence relation.
Theorem 4. Let P = {Pn (x)}∞ n=0 be orthogonal polynomial sequence on the real line whose polynomials
∞
Pn (x) satisfy the three-term recurrence condition (13). Then, the sequence R = {Rn (x)}n=0 of rational
functions Rn (x) defined in (6) will satisfy a three term relation of the following form:
   
ax + b
Rn (x) = [An M (x) − Bn ] Rn−1 (x) − Cn Rn−2 (x) = An − Bn Rn−1 (x) − Cn Rn−2 (x), n > 1,
cx + d

6
 ∞
with the initial conditions R−1 (x) = 0 and R1 (x) = 1. Similarly, the sequence Q = {Qn (x)}n=0 of Möbius-
transformed polynomials Qn (x) defined in (7), will satisfy the three-term recurrence relation:
2
Qn (x) = (An x − Bn ) Qn−1 (x) − Cn (cx + d) Qn−2 (x) n > 1, (14)

with the initial conditions Q−1 (x) = 0 and Q0 (x) = 1, where,

An = (aAn − cBn ) , Bn = (dBn − bAn ) , Cn = Cn .

Proof. Inserting relation (8) into (13), we get that,

Rn (W (x)) = (An x − Bn ) Rn−1 (W (x)) − Cn Rn−2 (W (x)).

Now, making the change of variable x = M (y), we obtain, at once,

   
ay + b
Rn (y) = (An M (y) − Bn ) Rn−1 (y) − Cn Rn−2 (y) = An − Bn Rn−1 (y) − Cn Rn−2 (y).
cy + d

It is clear that the relation above holds for any n > 1 provided we define R−1 (x) = 0 and R1 (x) = 1. Finally,
(14) follows after we insert (7) into the above expression. In fact, we get in this way,
   
Qn (y) ay + b Qn−1 (y) Qn−2 (y)
n = An − B n n−1 − Cn ,
(cy + d) cy + d (cy + d) (cy + d)n−2
so that,
   
ay + b 2
Qn (y) = An − Bn (cy + d) Qn−1 (y) − Cn (cy + d) Qn−2 (y),
cy + d
= [An (ay + b) − Bn (cy + d)] Qn−1 (y) − Cn (cy + d)2 Qn−2 (y),
2
= [(aAn − cBn ) y − (dBn − bAn )] Qn−1 (y) − Cn (cy + d) Qn−2 (y),

which holds for any n > 1 provided we define Q−1 (x) = 0 and Q0 (x) = 1.

3.3. Christoffel-Darboux identities

Another important relation satisfied by any orthogonal polynomial sequence on the real line is the so-
called Christoffel-Darboux identity [18, 19]:
n
X Ak+1 1 Pn+1 (x)Pn (y) − Pn (x)Pn+1 (y)
Pk (x)Pk (y) = , (15)
C0 · · · Ck+1 C0 · · · Cn+1 x−y
k=0

and its confluent form:

n
Ak+1 P ′ (x)Pn (x) − Pn′ (x)Pn+1 (x)
Pk2 (x) = n+1
X
. (16)
C0 · · · Ck+1 C0 · · · Cn+1
k=0

Similar identities also exist for the sequences of Möbius-transformed rational functions and polynomials, as
we shall describe in the following:
∞
Theorem 5. Let P = {Pn (x)}n=0 be an orthogonal polynomial sequence on the real line that satisfies the
∞
Christoffel-Darboux identities (15) and (16). Then the sequence R = {Rn (x)}n=0 of Möbius-transformed
rational functions Rn (x) defined in (6) will satisfy the following identities:
n  
X Ak+1 Rn+1 (x)Rn (y) − Rn (x)Rn+1 (y)
Rk (x)Rk (y) = (cx + d) (cy + d) , (17)
C0 · · · Ck+1 ∆ (C0 · · · Cn+1 ) (x − y)
k=0
7
and
n  ′
Rn+1 (y)Rn (y) − Rn′ (y)Rn+1 (y)

Ak+1
Rk2 (y) = (cy + d)2
X
. (18)
C0 · · · Ck+1 ∆ (C0 · · · Cn+1 )
k=0
∞
Similarly, the sequence Q = {Qn (x)}n=0 of Möbius-transformed polynomials Qn (x) defined in (7) will satisfy
the identities:
n
X Ak+1 Qk (x) Qk (y) Qn+1 (x)Qn (y) (cy + d) − (cx + d) Qn (x)Qn+1 (y)
= n n , (19)
C0 · · · Ck+1 (cx + d)k (cy + d)k ∆ (C0 · · · Cn+1 ) (cx + d) (cy + d) (x − y)
k=0

and
n
Ak+1 Q2k (x) Q′ (x)Qn (x) − Q′n (x)Qn+1 (x) − cQn+1 (x)Qn (x)
= n+1
X
2k 2n−1 . (20)
C0 · · · Ck+1 (cx + d) ∆ (C0 · · · Cn+1 ) (cx + d)
k=0

Proof. From (8) we get that (15) becomes,

n
X Ak+1 1 Rn+1 (W (x))Rn (W (y)) − Rn (W (x))Rn+1 (W (y))
Rk (W (x))Rk (W (y)) = .
C0 · · · Ck+1 C0 · · · Cn+1 x−y
k=0

Now, making the change of variables x = M (z) and t = M (y), we obtain,

n
X Ak+1 1 Rn+1 (z)Rn (t) − Rn (z)Rn+1 (t)
Rk (z)Rk (t) = .
C0 · · · Ck+1 C0 · · · Cn+1 M (z) − M (t)
k=0

But M (z) − M (t) = ∆(z − t)/ [(ct + d) (cz + d)], so that we get,
n  
X Ak+1 1 (cz + d) (ct + d) Rn+1 (z)Rn (t) − Rn (z)Rn+1 (t)
Rk (z)Rk (t) = ,
C0 · · · Ck+1 ∆ C0 · · · Cn+1 z−t
k=0

which is of the same form as (17). Now, to prove (18), we proceed in a similar fashion: using (8) into (16),
we get,
n  
X Ak+1 2 1 d d
R (W (x)) = [Rn+1 (W (x))] Rn (W (x)) − [Rn (W (x))] Rn+1 (W (x)) ,
C0 · · · Ck+1 k C0 · · · Cn+1 dx dx
k=0

that is,
n
X Ak+1 W ′ (x)  ′
Rk2 (W (x)) = Rn+1 (W (x))Rn (W (x)) − Rn′ (W (x))Rn+1 (W (x)) .

C0 · · · Ck+1 C0 · · · Cn+1
k=0

2
Thus, making the change of variables x = M (y) and using the identity W ′ (M (y)) = (cy + d) /∆, we obtain,
n 2
X Ak+1 1 (cy + d)  ′
Rk2 (y) = Rn+1 (y)Rn (y) − Rn′ (y)Rn+1 (y) ,

C0 · · · Ck+1 ∆ C0 · · · Cn+1
k=0

which is identical to (18). Finally, to prove (19) and (20), we just need to use (7). With that, (19) becomes:
n
X Ak+1 Qk (z) Qk (z)
=
C0 · · · Ck+1 (cz + d)k (ct + d)k
k=0
" #
1 (cz + d) (ct + d) 1 Qn+1 (z) Qn (z) Qn (z) Qn+1 (z)
− ,
∆ C0 · · · Cn+1 z − t (cz + d)n+1 (ct + d)n n
(cz + d) (ct + d)n+1

8
so that (18) follows after a simplification. Now, using (7) into (20), we get,

n
X Ak+1 Q2k (y)
=
C0 · · · Ck+1 (cy + d)2k
k=0
( " # )
1 (cy + d)2
 
d Qn+1 (y) Qn (y) d Qn (y) Qn+1 (y)
− ,
∆ C0 · · · Cn+1 dx (cy + d)n+1 (cy + d)n dx (cy + d)n (cy + d)n+1

and (20) follows after we expand the derivatives and simplify the resulting expression.

3.4. Properties of the zeros.

∞
It is well-known that the zeros of polynomials belonging to a sequence P = {Pn (x)}n=0 of orthogonal
polynomials on the real line are all simple and lie in the orthogonality interval I = (l, r) ⊆ R; moreover, for
n > 1, the zeros of Pn (x) interlace with the zeros of Pn+1 (x) [18, 19]. In this section, we shall show that the
Möbius-transformed polynomials defined in (7) also present similar properties.
∞
Theorem 6. Let Pn (x) be a polynomial of degree n > 1 that belongs to a sequence P = {Pn (x)}n=0 of
orthogonal polynomials defined in a given interval I = (l, r) ⊆ R of the real line and such that Pn (a/c) 6= 0
for all n. Then, the zeros of the corresponding Möbius-transformed polynomials Qn (x) defined in (7), for
n > 1, will all lie on the curve of orthogonality Γ , as defined in (12). Moreover, these zeros are all simple.
Proof. From Theorems 1 and 2, it follows that the zeros ζ1n , . . . , ζnn of the Möbius-transformed polynomials
Qn (x) are related to the zeros ξ1n , . . . , ξnn of the original polynomials Pn (x) by the relations,

ζ1n = W (ξ1n ), ζ2n = W (ξ2n ), ··· ζnn = W (ξnn ),

provided that a/c is not a zero of Pn (x). This shows us that zeros of Qn (x) will be all finite and will lie on
the curve Γ given in (12), which corresponds to the image of the interval I = (l, r) ⊆ R under the action
of the inverse Möbius-transformation W (x). Now, these zeros must be also simple because the Möbius
transformation is one-to-one and onto on the extended complex plane.

Theorem 7. The zeros of any two consecutive polynomials Qn (x) and Qn+1 (x), for n > 1, that belong to
∞
the sequence Q = {Qn (x)}n=0 of Möbius-transformed orthogonal polynomials Qn (x) defined in (7) interlace
on the new curve of orthogonality Γ given by (12).
Proof. The Möbius transformations are one-to-one and onto on the extended complex plane C∞ . In par-
ticular, this implies that the image of the interval I = (l, r) ⊆ R under the mapping y = W (x), which is
the curve Γ , is a not-self-intersecting curve. Thus, the ordination of any set of points lying in the interval
I is either unaltered or only inverted by a Möbius transformation. Because the zeros of Pn+1 (x) interlaces
with the zeros of Pn (x) on I, we conclude from what was said that the zeros of the Möbius-transformed
polynomial Qn+1 (x) will also interlace with the zeros of the polynomial Qn (x) over Γ .
We remark that the ordination above should be understood as performed in the extended complex plane
C∞ . In the usual complex plane, the ordination should be interpreted with care because the curve Γ can be
disconnected, as already commented. In fact, this will occur whenever the point x = a/c lies in the interval
I = (l, r) ⊆ R as, in this case, the curve Γ will cross the infinite. Thus, in this case the curve Γ will be
constituted by to disconnected branches in the complex plane and the ordination described above should be
understood as beginning in the point λ = W (l), passing through the infinity and ending in point ρ = W (r)
along the curve Γ .

9
 Chebyshev polynomials Inversion-transformed Chebyshev polynomials
T0 (x) = 1 T0 (x) = 1
T1 (x) = x T1 (x) = 1
T2 (x) = −1 + 2x2 T2 (x) = 2 − x2
T3 (x) = −3x + 4x3 T3 (x) = 4 − 3x2
T4 (x) = 1 − 8x2 + 8x4 T4 (x) = 8 − 8x2 + x4
T5 (x) = 5x − 20x3 + 16x5 T5 (x) = 16 − 20x2 + 5x4
T6 (x) = −1 + 18x2 − 48x4 + 32x6 T6 (x) = 32 − 48x2 + 18x4 − x6
T7 (x) = −7x + 56x3 − 112x5 + 64x7 T7 (x) = 64 − 112x2 + 56x4 − 7x6
T8 (x) = 1 − 32x2 + 160x4 − 256x6 + 128x8 T8 (x) = 128 − 256x2 + 160x4 − 32x6 + x8

Table 1: The first Möbius-transformed orthogonal polynomials Tn (x) = xn Tn (M (x)), where Tn (x) are the Chebyshev polyno-
mials and M (x) = 1/x is the inversion transformation. Notice that the polynomials Tn (x) for odd n have a zero at x = 0,
which implies that the Möbius-transformed polynomials Tn (x) for odd n will have even degree n′ = n − 1. Notwithstanding,
these polynomials are orthogonal
√ on the interval J = (−∞, −1) ∪ (1, ∞) of the real line with respect to the varying weight
function ωm,n (x) = x−m−n−2 / 1 − x−2 .

∞
A very interesting case occurs when some of the polynomials belonging to the sequence P = {Pn (x)}n=0
of orthogonal polynomials on the real line have a zero at the point x = a/c. Indeed, according to Theorem
1, if Pn (a/c) = 0 for some polynomials Pn (x) ∈ P of degree n, then the corresponding Möbius-transformed
polynomials Qn (x) ∈ Q will have degree n′ = n − 1. Thus, Q will be in this case a defective sequence of
orthogonal Möbius-transformed polynomials, that is, a sequence containing polynomials of repeated degrees
whereas other polynomials of specific degrees are absent. This particularity resembles what is called a
sequence of exceptional orthogonal polynomials [6, 7], which is a sequence of orthogonal polynomials on the
real line that does not contain some polynomials of specific degrees. A sequence Q of defective orthogonal
polynomials, however, differ from the sequences of exceptional orthogonal polynomials in two points: firstly
they usually are orthogonal in curves of the complex plane (not necessarily on the real line); secondly, even
though the sequence Q contain polynomials of repeated degrees, the orthogonality condition (11) is always
satisfied, even for the defective polynomials2 .
As an example of defective orthogonal polynomials arising from a Möbius transformation, we can consider
any sequence S of orthogonal polynomials defined in a symmetrical interval of the real line under the action of
the inversion map M (x) = 1/x, so that we have a/c = 0 in this case. In fact, we know that every polynomial
Pn (x) ∈ S of odd degree (say, n = 2m + 1) will have a zero at x = 0, thence the corresponding Möbius-
transformed polynomials Q2m+1 (x) ∈ Q will have actually an even degree (say, n′ = 2m), i.e. they will
be defective. The sequence Q of Möbius-transformed orthogonal polynomials will be, therefore, constituted
by polynomials of even degrees only, and in such a way that two any consecutive polynomials belonging to
this sequence will have the same degree. In Table 1 we present the first orthogonal Möbius-transformed
polynomials for the Chebyshev polynomials under the inversion map.

4. Classical Möbius-transformed orthogonal polynomials

Among all the sequences of orthogonal polynomials on the real line, the so-called classical orthogonal
polynomial sequences — which include the sequences of Jacobi, Laguerre and Hermite polynomials — are
of particular interest. In fact, the classical families of orthogonal polynomials were the first ones to be
systematically studied, which is justified by their great importance in applied mathematics, physics and
engineering. Besides, the classical orthogonal polynomials have many properties that are not satisfied by
any other sequences of orthogonal polynomials: for example, they satisfy second-order differential equations,
their weight function also satisfy a linear differential equation, there exist Rodrigues’ type relations and

2 Here we remark that the parameters m and n on the varying weight function ω
m,n (x) do not rely on the degrees of the
polynomials Qm (x) and Qn (x), but only to the order in which they appear in the sequence Q = {Qn (x)}∞ n=0 of Möbius-
transformed polynomials.

10
generating functions from which the classical orthogonal polynomials can be derived and so on [18, 19]. In
this section, we shall show that the Möbius-transformed rational functions and polynomials that arise from
a classical orthogonal polynomial sequence on the real line enjoy similar properties.

4.1. Second-order differential equation

It is a well-known result that any sequence P = {Pn (x)}∞
n=0 of classical orthogonal polynomial satisfy a
second-order differential equation of the form [18, 19]:
f (x)Pn′′ (x) + g(x)Pn′ (x) + hn (x)Pn (x) = 0. (21)
Routh [21], Bochner [22] and Brenke [23], have shown, independently, that polynomial solutions of (21)
exist only when f (x), g(x) and hn (x) are polynomials of degree 2, 1 and 0, respectively, where hn (x) can
be different for each n. Besides, they showed that the only polynomial solutions of (21), up to a linear
transformation, are the classical orthogonal polynomial sequences of Jacobi, Laguerre and Hermite, besides
other few sequences of polynomials which nevertheless are not orthogonal on the real line in the usual sense3 .
In the following, we shall derive the differential equations that are satisfied by the sequences of the Möbius-
transformed rational functions and polynomials when the original sequence of orthogonal polynomials is of
the classical type.
∞
Theorem 8. Let P = {Pn (x)}n=0 be orthogonal polynomial sequence on the real line in which each poly-
nomial Pn (x) for n > 0 satisfy a second order differential equation of the form (21). Then, each Möbius-
∞
transformed rational function Rn (x), as defined by (6), belonging to the sequence R = {Rn (x)}n=0 will
satisfy a second-order differential equation of the form,
F (x)Rn′′ (x) + G(x)Rn′ (x) + Hn Rn (x) = 0, (22)
where, F (y) and G(y) are polynomials independent of n whose degree are at most 4 and 3, respectively, and
Hn is a constant which may depend on n. Similarly, each Möbius-transformed polynomial Qn (x), as defined
by (7), belonging to the sequence Q = {Qn (x)}∞ n=0 will satisfy a second-order differential equation of the
form,
F (x)Q′′n (x) + Gn (x)Q′n (x) + Hn (x)Qn (x) = 0, (23)
where, now, F (x) has degree at most 4 and does not depend on n, while Gn (x) and Hn (x) are polynomials
with degree at most 3 and 2, respectively, that in general depend on n.
Proof. Inserting relation (8) on the differential equation (21), we get,
 2   
d d
f (x) R n (W (x)) + g(x) Rn (W (x)) + hn Rn (W (x)) = 0,
dx2 dx
Computing the derivatives with the help of chain’s rule, we get,
" 2 2 #
d2 W (x) dRn (W (x))
  
dW (x) d Rn (W (x)) dW (x) dRn (W (x))
f (x) + + g(x) + hn Rn (W (x)) = 0,
dx2 dW (x) dx dW (x)2 dx dW (x)

Now, the change of variables x = M (y) provides,

" 2 2 #
d2 W (M (y)) dRn (y)

dW (M (y)) d Rn (y)
f (M (y)) +
dM (y)2 dy dM (y) dy 2
 
dW (M (y)) dRn (y)
+ g(M (y)) + hn Rn (y) = 0.
dM (y) dy

3 The first of such non-orthogonal sequences is constituted by polynomials of the form π (x) = xn + αxm , where α is a real
n
parameter and m any integer satisfying 0 6 m 6 n − 1. The second sequence is related to the so-called Bessel polynomials,
which were consistently studied for the first time in [24]. Finally, the third sequence consists of the Romanovski polynomials
[25].

11
But,
dW (M (y)) (cy + d)2 d2 W (M (y)) 2c(cy + d)3
= , 2
= ,
dM (y) ∆ dM (y) ∆2
so that we get, after gathering the derivatives of Rn (y),
(cy + d)4 d2 Rn (y) 2c(cy + d)3 (cy + d)2
 
dRn (y)
f (M (y)) + f (M (y)) + g(M (y)) + hn Rn (y) = 0,
∆2 dy 2 ∆2 ∆ dy
which is of the same form of (22), provided we define,
(cy + d)4 2c(cy + d)3 (cy + d)2
F (y) = f (M (y)), G(y) = f (M (y)) + g(M (y)), Hn = hn . (24)
∆2 ∆2 ∆
Now, because f (x) and g(x) are polynomials with degree at most 2 and 1 respectively, it is convenient to
introduce the polynomials
Φ(x) = (cx + d)2 f (M (x)), and Γ (x) = (cx + d)g(M (x)), (25)
which are, as a matter of a fact, the Möbius-transformed polynomials of f (x) and g(x), respectively. Notice
that Φ(x) and Γ (x) have the same degree as f (x) and g(x), from which we conclude that F (y) and G(y)
are polynomials with degree at most 4 and 3, respectively.
Let us now prove (23). For this we can proceed as follows: by inserting (7) into (22) we get,
d2
     
Qn (x) d Qn (x) Qn (x)
F (x) 2 + G(x) + H n n = 0.
dx (cx + d)n dx (cx + d)n (cx + d)
Computing the derivatives and simplifying, this becomes,
" #
c2 n(n + 1)
 
2cn ′ ′′ ′ cn
F (x) Qn (x) − Q (x) + Qn (x) + G(x) Qn (x) − Qn (x) + Hn Qn (x) = 0,
(cx + d)2 cx + d n cx + d

that is, gathering the derivatives of Qn (x),

" #
c2 n(n + 1)
 
′′ 2cn ′ cn
F (x)Qn (x) + G(x) − F (x) Qn (x) + 2 F (x) − cx + d G(x) + Hn Qn (x) = 0,
cx + d (cx + d)
which is of the same form than (23), provided we define,
(cx + d)4
F (x) = F (x) = f (M (x)), (26)
∆2
2cn 2c(n − 1)(cx + d)3 (cx + d)2
Gn (x) = G(x) − F (x) = − f (M (x)) + g(M (x)), (27)
cx + d ∆2 ∆
c2 n(n + 1) cn c2 n(n − 1)(cx + d)2 cn(cx + d)
Hn (x) = 2 F (x) − cx + d G(x) + Hn = 2
f (M (x)) − g(M (x)) + hn .
(cx + d) ∆ ∆
(28)
Notice that in terms of the Möbius-transformed polynomials Φ(x) and Γ (x) given by (25), we can rewritten
(26), (27) and (28) as,
(cx + d)2
F (x) = Φ(x),
∆2
2c(n − 1)(cx + d) (cx + d)
Gn (x) = − Φ(x) + Γ (x),
∆2 ∆
c2 n(n − 1) cn
Hn (x) = 2
Φ(x) − Γ (x) + hn ,
∆ ∆
which proves that F (x), Gn (x) and Hn (x) are polynomials with degree at most 4, 3 and 2, respectively. This
concludes the proofs.
12
 At first glance, Theorem 8 seems to contradicts Routh-Bochner-Brenke’s Theorem because we found
polynomial solutions of a second-order differential equation whose coefficients are polynomials of degree
greater than 2. However, this contradiction is only apparent because of the following reason: while Routh-
Bochner-Brenke’s theorem concerns with the solutions of the second-order differential equation (21), where
only constant term hn is supposed to be dependent of n, both the coefficients Gn (x) and Hn (x) in the
differential equation (23) can depend on n, so that (21) and (23) are actually of different forms. Furthermore,
it is clear that no restriction whatsoever can be imposed on the degrees of the polynomial coefficients of a
second-order differential equation when all these coefficients are allowed to depend on n.

4.2. Weight function differential equations

The weight function w(x) of any classical orthogonal polynomial also satisfy a differential equation. In
this case, however, we have the following first-order homogeneous differential equation [18, 19]:

w′ (x) g(x) − f ′ (x)

= , (29)
w(x) f (x)

where f (x) and g(x) are the same polynomials appearing in (21). In fact, (29) is obtained as a condition to
the differential equation (21) be written in the self-adjoint form.
 
d dy(x)
p(x) + qn (x)y(x) = 0. (30)
dx dx

It is well-known that any second-order differential equation can be written in the self-adjoint form by
multiplying it by some integrating factor w(x). In fact, multiplying (21) by w(x) and comparing with
(30) we conclude that p(x) = w(x)f (x), p′ (x) = w(x)g(x) and qn (x) = w(x)h(x), so that the relation
′
w(x)g(x) = [w(x)f (x)] must hold, from which (29) immediately follows. Besides, whenever we have

lim f (x)w(x)xn = 0, and lim f (x)w(x)xn = 0, n ∈ N, (31)

x→a x→b

Sturm-Liouville theory will ensure that the sequence P = {Pn (x)}∞ n=0 of polynomials (which we suppose
to be the solutions of a well-behaved Sturm-Liouville problem), will indeed be an orthogonal polynomial
sequence in the interval I = (a, b) of the real line, with respect to the weight function w(x).
In this section, we shall show that weight functions of the Möbius-transformed rational functions and
polynomials also satisfy similar first-order differential equations.
∞
Theorem 9. Let w(x) be the weight function of a classical orthogonal polynomial sequence P = {Pn (x)}n=0 ,
so that it satisfies the homogeneous first-order differential equation (29). Then, the corresponding weight
∞
function ω(x), given by (10), of the sequence R = {Rn (x)}n=0 of Möbius-transformed rational functions, as
defined in (6), will satisfy the following first-order differential equation:

ω ′ (x) G(x) − F ′ (x)

= , (32)
ω(x) F (x)

where F (x) and G(x) are the same coefficients that appear in the differential equation (22). Similarly, the
degree-dependent weight function ωm,n (x) given in (11), that is associated with the sequence Q = {Qn (x)}∞ n=0
of Möbius-transformed polynomials, as defined in (7), will satisfy the following first-order differential equa-
tion: ′
ωm,n (x)
 
1 Gm (x) + Gn (x) ′
= − F (x) , (33)
ωm,n (x) F (x) 2
where F (x), Gm (x) and Gn (x) are the same polynomials defined in (26) and (27), respectively.

13
Proof. Inverting relation (10), we can express w(x) in terms of ω(W (x)):
ω(W (x))
w(x) =  . (34)
dM(W (x))
dW (x)

Thereby, taking the derivative of w(x), we obtain,

 
d2 M(W (x))
′
dω(W (x)) W (x) dW (x)2
w′ (x) =   − ω(W (x))W ′ (x)  2 . (35)
dW (x) dM(W (x)) dM(W (x))
dW (x) dW (x)

Inserting (34) and (35) into (29) we shall get, after simplification,
  dω(W (x))   d2 M(W (x))  
dW (x) dW (x)2 g(x) − f ′ (x)
w′ (x) =  −    W ′ (x) = .
ω(W (x)) dM(W (x)) f (x)
dW (x)

Now, making the change of variables x = M (y), we shall find,

 
df (M(y))
 ′
ω (y) M (y) ′′

dW (M (y)) g(M (y)) − dM(y)
− = ,
ω(y) M ′ (y) dM (y) f (M (y))
so that,  
df (M(y))

′
ω (y)  g(M (y)) − ′′
=
dM(y)
 M ′ (y) + M (y) .
ω(y) f (M (y)) M ′ (y)

Now, using the relations

∆ 2c∆ df (M (y)) 1 df (M (y))
M ′ (x) = , M ′′ (x) = − , and = ′ ,
(cy + d)2 (cy + d)3 dM (y) M (y) dy
we can verify, taking (24) into account, that
 
g(M (y)) − dfdM(y)
(M(y))

′′ ′
  M ′ (y) + M (y) = G(y) − F (y) ,
f (M (y)) M ′ (y) F (y)

which establishes (32). Finally, to prove (33) we can proceed as follows: from (11), we know that ω(x) =
ωm,n (x)(cy + d)n+n . Thus, inserting this expression into (32) and computing the derivative, we obtain,
′
ωm,n (x) c (m + n) G(x) − F ′ (x)
+ = .
ωm,n (x) (cx + d) F (x)
Now, from (26) we get that F (x) = F (x) and, hence, F ′ (x) = F ′ (x) as well; besides, from (27) we can see
that the term c (m + n) /(cx + d) is exactly canceled by expressing G(x) in terms of 12 [Gm (x) + Gn (x)], from
which (33) is obtained. This conclude the proof.

4.3. Rodrigues’ type formulas

Another interesting property of the classical orthogonal polynomial sequences is that they can be defined
∞
through the so-called Rodrigues’ formulas [18, 19]. In fact, if P = {Pn (x)}n=0 is such a sequence, then the
polynomials Pn (x) ∈ P can be found by taking derivatives:
ǫn dn
Pn (x) = [f (x)n w(x)] , (36)
w(x) dxn
14
where ǫn are some standardizing constants, w(x) the corresponding weight function and f (x) the same
polynomial appearing in the differential equation (21).
In this section, we shall show that the Möbius-transformed rational functions and polynomials also can
be found by taking derivatives through a Rodrigues’-type formula.
∞
Theorem 10. Let P = {Pn (x)}n=0 be a sequence of orthogonal polynomials on the real line satisfying a
Rodrigues’ type formula as given in (36). Then, the rational functions Rn (x), as defined in (6), and the
Möbius-transformed polynomials Qn (x), as defined in (7), will satisfy, respectively, the following Rodrigues’
type formulas: " #
∆2n ǫn n F (x)n
Rn (y) = 2 Dx 4n−2 ω(x) , (37)
(cx + d) ω(x) (cx + d)
and " #
∆2n ǫn n F (x)n
Qn (y) = D ωn,n (x) , (38)
(cx + d)
n+2 ωn,n (x) x (cx + d)2n−2

where F (x) = F (x) is the same function defined in (24), ω(x) and ωn,n (x) are the new weight functions
defined in (10) and (11), respectively, and we introduced the differential operator,
2
1 d (cx + d) d
Dx = ′
= .
M (x) dx ∆ dx

Proof. Formulas (37) and (38) follow in a straightforward way by making the change of variable x = M (y)
into (36) and using (6), (7), (10), (11) and (24).
We remark that formulas (37) and (38) can be rewritten as well in the following equivalent form:

ǫn dn

∆ n

Rn (y) = 2 ω(y) dxn [f (x)w(x)] 
 ,
(cy + d) x=M(y)
dn

∆ ǫn n

Qn (y) = n+2 ω n
[f (x)w(x)]  .
(cy + d) n,n (y) dx

x=M(y)

They can also be rewritten in many other ways — for example, we can expand the derivatives by invoking
Faà-di-Bruno formula for the nth derivative of a composed function, Leibniz formula for the derivative of
a product, etc —, however, we shall not go through these lines because the resulting expressions are very
cumbersome.

4.4. Generating functions

As a last property, let us discuss the existence of generating functions for the Möbius-transformed rational
functions and polynomials. As it is well-known, the sequences of classical orthogonal polynomials can be
generated by generating function of the following forms [18, 19]:
∞ ∞
X X tn
φ(x, t) = Pn (x)tn , or ψ(x, t) = Pn (x) . (39)
n!
k=0 k=0

Similar formulas hold for the Möbius-transformed rational functions and polynomials:
∞
Theorem 11. Let P = {Pk (x)}k=0 be a classical orthogonal polynomial sequence whose generating function
is one of those given in (39). Then, the corresponding generating functions of the sequence R = {Rk (x)}∞
k=0
of Möbius-transformed rational functions Rn (x) defined in (6) are the following:

Φ(y, t) = φ(M (y), t), or Ψ (y, t) = ψ(M (y), t),

15
 ∞
respectively. Similarly, the corresponding generating functions of the sequence Q = {Qk (x)}k=0 of Möbius-
transformed polynomials Qn (x) defined in (7) are:
Φ(y, τ ) = φ(M (y), (cy + d)τ ), or Ψ (y, τ ) = ψ(M (y), (cy + d)τ ),
respectively.
Proof. We have at once that,
∞
X ∞
X
n
Φ(y, t) = φ (M (y), t) = Pn (M (y))t = Rn (y)tn ,
k=0 k=0

and
∞ ∞
X tn X tn
Ψ (y, t) = ψ (M (y), t) = Pn (M (y)) = Rn (y) .
n! n!
k=0 k=0
In the same manner, we have that,
∞
X ∞
X
Φ(y, τ ) = φ (M (y), (cy + d)τ ) = Pn (M (y))(cy + d)n τ n = Qn (y)τ n ,
k=0 k=0

and
∞ ∞
X τn X τn
Ψ (y, τ ) = ψ (M (y), (cy + d)τ ) = Pn (M (y))(cy + d)n = Qn (y) .
n! n!
k=0 k=0

5. Applications
We close this paper with some applications of the theory developed above. More specifically, we show
that the sequences of Hermite, Laguerre, Jacobi, Bessel and Romanovski polynomials are all related with
each other by a Möbius transformation. We also show that generalized Bessel polynomials enjoy a finite
orthogonality on the real line, as well as the Romanovski polynomials satisfy an orthogonality relation on
the imaginary axis.

5.1. From Laguerre to Bessel polynomials

The associated Laguerre polynomials Lα n (x) are polynomials orthogonal on the interval I = (0, ∞) of
the real line with respect to the weight function w(x) = xα e−x , for any α > −1 [18, 19]. Performing the
re-scaling transformation M (x) = βx, we can easily generate the so-called generalized Legendre polynomials
(α,β)
Ln (x) = Lα n (βx), which are orthogonal on the same interval I = (0, ∞) of the real line, for α > −1
and β > 0, but with respect the weight function w(x) = xα e−βx . Now let us consider the generalized
Laguerre polynomials under the inversion map M (x) = 1/x. The resulting transformed polynomials will
have, therefore, the following form:
L(α,β)
n (x) = xn L(α,β)
n (1/x).
From the results of Section 3, it follows that these polynomials are also orthogonal on the interval I = (0, ∞)
but with respect to a varying measure, provided that α > −1 and β > 0. Namely, we have the following:
Z ∞ (α,β) (α,β)  α+2
Lm (x)Ln (x) 1 1 Γ (n + 1 + α)
m+n
e−β/x dx = α+1 δm,n , α > −1, β > 0.
0 x x β Γ (n + 1)
(α,β)
In the following, we shall show that the polynomials Ln (x) are related with the so-called Bessel polyno-
mials, which were studied extensively in [24]. To this end, notice that, according to the results of Section
(α,β)
4.1, the polynomials Ln (x) = y(x) will satisfy the differential equation:
x2 y ′′ (x) + [β + (1 − 2n − α) x] y ′ (x) + n (n + α) y(x) = 0. (40)
16
 (γ,β)
Comparing with the differential equation given in for [24] the generalized Bessel polynomials Bn , namely,

x2 y ′′ (x) + (β + γx) y ′ (x) + n (n + γ − 1) y(x) = 0, (41)

we see that a complete agreement between (40) and (41) is achieved when we set α = 1 − 2n − γ. Thus, the
generalized Bessel polynomials are given by the formula

Bn(γ,β) (x) = L(1−2n−γ,β)

n (x),
(γ,β)
up to a normalization. Thereby, each polynomial Bn (x) is associated with a generalized Laguerre poly-
nomial with a different value of α. From the discussion given in Section 3.1, we can conclude that the
(γ,β)
polynomials Bn (x) will be satisfy the orthogonality relation
Z ∞
(γ,β) Γ (2 − n − γ)
Bm (x)Bn(γ,β) (x)Ωm,n (x)dx = β 2n−2+γ δm,n , Ωm,n (x) = x|m−n|+γ−3 e−β/x , (42)
0 Γ (n + 1)
(γ,β)
provided that γ < 2 − 2 max(m, n) and β > 0. From this we can see that the polynomials Bm (x) can at
most enjoy a finite orthogonality on the interval I = (0, ∞) of the real line4 .

Generalized Bessel polynomials Bessel polynomials

(γ,β)
B0 (x) = 1 B0 (x) = 1
(γ,β)
B1 (x) = 1 + βγ x B1 (x) = 1 + x
(γ,β) 2(1+γ) (1+γ)(2+γ) 2
B2 (x) = 1 + β x + β2 x B2 (x) = 1 + 3x + 3x2
(γ,β) 3(2+γ) 3(2+γ)(3+γ) 2 (2+γ)(3+γ)(4+γ) 3
B3 (x) = 1 + β x + β2 x + β3 x B3 (x) = 1 + 6x + 15x2 + 15x3

Table 2: The first generalized Bessel polynomials and Bessel polynomials (normalized with the constant terms equal to unity).

Following [24], the usual Bessel polynomials are obtained by setting γ = 2 and β = 2, from which
(−1−2n,2)
we get that Bn (x) = Ln (x), up to a normalization, while the weight function becomes Ωm,n (x) =
x|m−n|−1 e−2/x . In this case, however, the condition for the validity of the orthogonality relation (42) would
hold only for n < 0, which is not possible. Thus, we conclude that the Bessel polynomials are not orthogonal
on the real line. The first generalized and usual Bessel polynomials are presented in Table 2.

5.2. From Jacobi to Romanovski polynomials

Romanovski polynomials consist of an infinite sequence of real polynomials studied by Romanovski in
[25], although they have appeared before in Routh’s work [21]. Romanovski polynomials also appear in
physical problems, which justified a more detailed study of them — see [26, 27] and references therein. In
(α,β)
what follows we shall show that Romanovski polynomials are also related with Jacobi polynomials Jn (x)
via a Möbius transformation. In this case, however, we should consider the mapping M (x) = ix with the
following values for the Möbius parameters: a = 1, b = 0, c = 0 and d = −i.
(α,β)
Remember that Jacobi polynomials Jn (x) are orthogonal on the interval I = (−1, 1) of the real line
α β
with respect to the weight function w(x) = (1 − x) (1 + x) for α > −1 and β > −1 [18, 19]. Thus,
according with the results of Section 3, the Möbius-transformed polynomials
n
Jn(α,β) (x) = (−i) Jn(α,β) (ix)

will be orthogonal on the interval J = (−i, i) along the imaginary line, provided that α > −1 and β > −1.
(α,β)
The polynomials Jn (x), however, do not have real coefficients for real α and β; nonetheless, a real

4 Notice (γ,β) (γ,β)

as well that, for γ a negative integer, the following identity holds: Bn (x) = B1−γ−n (x). Thus, in these cases the
(γ,β)
sequence B = {Bn (x)}∞
n=0 of generalized Bessel polynomials will be a defective sequence of orthogonal polynomials.

17
polynomial can be obtained if we allow α and β to assume non-real values. The precise condition for this is
that α be the complex conjugate of β, so that we can write α = γ + iδ, β = γ − iδ with γ, δ ∈ R and δ 6= 0.
In fact, up to a normalization, this gives place to the Romanovski polynomials 5 :
n
R(γ,δ)
n (x) = Jn(γ+iδ,γ−iδ) (x) = (−i) Jn(γ+iδ,γ−iδ) (ix),

The first Romanovski polynomials are presented in Table 3.

(γ,δ)
Now, in order to investigate the orthogonality of the polynomials Rn (x) along the imaginary axis
we need to verify under what conditions the Jacobi polynomials with complex conjugate parameters are
(γ+iδ,γ−iδ)
orthogonal on the real line. It is not difficult to verify that Jacobi polynomials Jn (x) are orthogonal
on the interval I = (−1, 1) of the real line provided that γ > −1 and δ ∈ R. In fact, this follows because
the limit relations (31) are satisfied for any non-negative integer n whenever γ > −1 and δ any real number.
R −1 γ+iδ γ−iδ
Besides, we can verify that all the moments µ(n) = −1 xn (1 − x) (1 + x) dx exist under these
conditions as well, which is enough to guarantee the convergence of the integrals (5) for any m, n ∈ N, for
the same range of γ and δ.

Romanovski polynomials
(γ,δ)
R0 (x) = 1
(γ,δ)
R1 (x) = δ + (1 + γ)x
(γ,δ)
R2 (x) = 14 2 + γ + 2δ 2 + 12 δ(3 + 2γ)x + 14 (2 + γ)(3 + 2γ)x2

(γ,δ) (γ+iδ,γ−iδ)
Table 3: The first Romanovski polynomials Rn (x) = (−i)n Jn (ix).

(γ,δ)
Consequently, we conclude that Romanovski polynomials Rn (x) will be orthogonal on the interval
J = (−i, i) of the imaginary axis whenever γ > −1 and δ is a real number6 . The corresponding weight
function can be found either directly from (10) and (11) or by solving the differential equation (33). We
have that, up to a multiplicative constant,
γ+iδ γ−iδ
γ
ω(x) = (1 − ix) (1 + ix) = x2 + 1 e2δ arctan(x) .
Thus, the orthogonality relation reads:
Z i

γ 22γ+1 Γ (n + 1 + γ + iδ)Γ (n + 1 + γ − iδ)
R(γ,δ)
m (x)Rn(γ,δ) (x) x2 + 1 e2δ arctan(x) dx = δm,n .
−i 2n + 2γ + 1 Γ (n + 1 + 2γ)Γ (n + 1)
which holds for γ > −1 and real δ.
Finally, it should be mentioned that the limit relations (31), if applied to this Möbius-transformed weight
function ω(x), provide as well a finite orthogonality for the Romanovski polynomials on the real line. In
(γ,δ) (γ,δ)
fact, in this case we can verify that only those polynomials Rm (x) and Rn (x) satisfying the relation
m + n + 2γ < −1 will form an orthogonal polynomial sequence on the interval J = (−∞, ∞) of the real line.

5.3. From Jacobi to Laguerre polynomials

The Möbius transformation is rich enough to map any finite point to the infinite — in fact, on the
extended complex plane C∞ there is no special distinction between the infinite and any other point. Thereby,
we can transform a given finite interval of the real line to a semi-infinite one. Following this idea, we shall
show in the sequel that Laguerre polynomials can be obtained from Jacobi polynomials through a Möbius
transformation followed by an adequate limit of their parameters.

5 We remark that some authors adopt different values for the parameters γ and δ in the definition of Romanovski polynomials.
6 Notice (γ,δ)
that the zeros of Romanovski polynomials Rn (x) usually do not lie on the imaginary axis, which can be explained
(γ+iδ,γ−iδ)
by the fact that the Jacobi polynomials Jn (x) are not real polynomials for γ and δ real, so that their zeros usually
do not lie on the real line either.

18
 The associated Laguerre polynomials are orthogonal polynomials on the interval I = (0, ∞) of the
real line with respect to the weight function w(x) = xα e−x [18, 19]. Thus, let us define the the Möbius
(α,β) (α,β)
transformed polynomials Jn (x) = (cx + d)n Jn (M (x)) and look for the transformation W (x) that
maps the points l = −1 and r = 1 respectively to λ = 0 and ρ = ∞. From (1), we plainly see that we
should have c = a and d = −b to this end. With this, we have that the first Möbius-transformed polynomial
becomes:
(α,β)
J1 (x) = a(α + 1)x + b(β + 1).
Comparing this expression with the first Laguerre polynomial, Lβ1 (x) = β + 1 − x, we see that we must have
(α,β)
a = −1/ (α + 1) and b = 1. Then we can verify that the other Möbius transformed polynomials Jn (x)
will match with the Laguerre polynomials Lβn (x) after we take the limit α → ∞, that is,
 n    
β x+α+1 n (α,β) x−α−1 n (α,β) x−α−1
Ln (x) = lim (−1) Jn = lim (−1) Jn .
α→∞ α+1 x+α+1 α→∞ x+α+1
(β,α) n (α,β)
Using the identity Jn (x) = (−1) Jn (−x) we can also write:
 n    
x+β+1 β+1−x β+1−x
Lα
n (x) = lim Jn
(α,β)
= lim Jn
(α,β)
.
β→∞ β+1 β+1+x β→∞ β+1+x
(α,β)
These expressions can be compared with Szegő’s identity [18]: Lα 1 − 2β −1 x .


n (x) = limβ→∞ Jn
Finally, let us show that Laguerre’s weight function can also be obtained from Jacobi’s weight function
through the same procedure. Notice that Ωm,n (x) → ω(x) as α → ∞ because c → 0 in this limit. Thus,
α β
is enough to prove that w(x) = (1 − x) (1 + x) reduces to ω(x) = xβ e−x in the limit α → ∞ after we
perform the Möbius transformation above. From (32) we can verify that the transformed weight function
is,
−α−β−2
ω(x) = xβ (x + α + 1) κ(α, β),
where κ(α, β) is the constant of integration. From this we can see that Laguerre’s weight function is obtained
after we set κ(α, β) = αβ+1 (α + 1)α+1 and take the limit α → ∞. All the other properties of Laguerre
polynomials can be obtained from those of Jacobi by similar arguments.

5.4. From Jacobi to Hermite polynomials

Finally, we can show that Hermite polynomials can also be obtained from Jacobi polynomials through
a Möbius transformation followed by a specific limit. At first sight, we would be tempted to map the finite
interval I = (−1, 1) into the infinite interval J = (−∞, ∞). This, however, it is not possible to be done
with a Möbius transformation because the points x = ±∞ are regarded as equals in the extended complex
plane C∞ and the Möbius transformation is one-to-one and onto in C∞ . Thus, to work this around, we shall
proceed in a somewhat different way: the idea is to first find the Möbius transformation W (x) that maps
the interval I = (−1, 1) into the interval J = (−ξ, ξ) for some ξ > 0 and, then, to take the limit ξ → ∞.
Imposing further that the point 0 is mapped to 0, we find that the required Möbius transformation becomes
W (x) = aξ, whose inverse is M (x) = a/ξ (a is still arbitrary).
α β
Now, according (10), (11) or (33), we get that Jacobi’s weight function w(x) = (1 − x) (1 + x) will be
transformed into  α  β
1 x x
ω(x) = 1− 1+ κ(α, β),
ξ ξ ξ
where κ(α, β) is the constant of integration. From this we can plainly see that Hermite’s weight function
2 √
ω(x) = e−x is achieved by making β = α, κ(α, β) = ξ = α and taking limit α → ∞ (so that we
(α)
get ξ → ∞ as well). With these identifications, the Möbius-transformed polynomials become Jn (x) =
√ n (α,α) √ (α)
(an α) Jn (x/ α). Now, the limits of Jn (x) for α → ∞ only exist if a is inversely proportional

19
to α. In fact, if we set a = 2/α and take the limit α → ∞ then we can verify that we get a complete
(α)
correspondence between the Möbius-transformed polynomials Jn (x) and Hermite polynomials Hn (x):
 n  
1 2 x
Hn (x) = lim Jn(α) (x) = lim √ Jn(α,α) √ .
n! α→∞ α→∞ α α

Acknowledgments

We kindly thank Professors C. F. Bracciali and A. Sri Ranga for the discussions we had about these
ideas and also for their valuable comments. The work of RSV was supported by grants from Coordination
for the Improvement of Higher Education Personnel (CAPES). The work of VB was supported by funds
from São Paulo Research Foundation (FAPESP), grant #2016/02700-8.

Bibliography

References
[1] R. Koekoek, P. A. Lesky, R. F. Swarttouw, Hypergeometric orthogonal polynomials and their q-analogues, Springer Science
& Business Media, 2010 (2010).
[2] B. Simon, Orthogonal polynomials on the unit circle, Part I: Classical theory, American Mathematical Society, 2004 (2004).
[3] B. Simon, Orthogonal polynomials on the unit circle, Part II: Spectral Theory, American Mathematical Society, 2004
(2004).
[4] T. Chihara, On quasi-orthogonal polynomials, Proceedings of the American mathematical Society 8 (4) (1957) 765–767
(1957). doi:10.2307/2033295.
[5] V. Totik, Orthogonal polynomials with respect to varying weights, Journal of Computational and Applied Mathematics
99 (1) (1998) 373–385 (1998). doi:10.1016/S0377-0427(98)00171-X.
[6] D. Gómez-Ullate, N. Kamran, R. Milson, An extended class of orthogonal polynomials defined by a Sturm-Liouville prob-
lem, Journal of Mathematical Analysis and Applications 359 (1) (2009) 352–367 (2009). doi:10.1016/j.jmaa.2009.05.052.
[7] D. Gómez-Ullate, N. Kamran, R. Milson, An extension of Bochner’s problem: exceptional invariant subspaces, Journal of
Approximation Theory 162 (5) (2010) 987–1006 (2010). doi:10.1016/j.jat.2009.11.002.
[8] W. B. Jones, O. Njåstad, W. Thron, Moment theory, orthogonal polynomials, quadrature, and continued frac-
tions associated with the unit circle, Bulletin of the London Mathematical Society 21 (2) (1989) 113–152 (1989).
doi:10.1112/blms/21.2.113.
[9] A. G. Akritas, Elements of computer algebra with applications, Vol. 3, Wiley New York, 1989 (1989).
[10] R. Vieira, How to count the number of zeros that a polynomial has on the unit circle?, arXiv preprint arXiv:1902.04231
(2019).
[11] R. Vieira, Polynomials with symmetric zeros, in: Polynomials – Theory and Application, IntechOpen, 2019 (2019).
doi:10.5772/intechopen.82728.
[12] V. Aldaya, J. Bisquert, J. Navarro-Salas, The quantum relativistic harmonic oscillator: generalized Hermite polynomials,
Physics Letters A 156 (7-8) (1991) 381–385 (1991). doi:10.1016/0375-9601(91)90711-G.
[13] P. Natalini, The relativistic Laguerre polynomials, Rend. Matematica, Ser. VII 16 (1996) 299–313 (1996).
[14] M. He, P. Natalini, Relativistic Jacobi polynomials, Integral Transforms and Special Functions 8 (1-2) (1999) 43–56 (1999).
doi:10.1080/10652469908819215.
[15] M. E. Ismail, Relativistic orthogonal polynomials are Jacobi polynomials, Journal of Physics A: Mathematical and General
29 (12) (1996) 3199 (1996). doi:10.1088/0305-4470/29/12/023.
[16] F. J. Flanigan, Complex variables: harmonic and analytic functions, Dover Publications, 1972 (1972).
[17] L. V. Ahlfors, Complex Analysis: An Introduction to the Theory of Analytic Functions of One Complex Variable, 3rd
Edition, McGraw-Hill Education, 1979 (1979).
[18] G. Szegő, Orthogonal polynomials, Vol. 23, American Mathematical Society, 1939 (1939).
[19] T. S. Chihara, An introduction to orthogonal polynomials, Dover Publications, 2011 (2011).
[20] J. S. Dehesa, A. Martínez-Finkelshtein, J. Sánchez-Ruiz, Quantum information entropies and orthogonal polynomials,
Journal of Computational and Applied Mathematics 133 (1-2) (2001) 23–46 (2001). doi:10.1016/S0377-0427(00)00633-6.
[21] E. J. Routh, On some properties of certain solutions of a differential equation of the second order, Proceedings of the
London Mathematical Society s1-16 (1) (1885) 245–262 (1885). doi:10.1112/plms/s1-16.1.245.
[22] S. Bochner, Über Sturm-Liouvillesche Polynomsysteme, Mathematische Zeitschrift 29 (1) (1929) 730–736 (1929).
doi:10.1007/BF01180560.
[23] W. C. Brenke, On polynomial solutions of a class of linear differential equations of the second order, Bulletin of the
American Mathematical Society 36 (1930) 77–84 (1930). doi:https://doi.org/10.1090/S0002-9904-1930-04888-0.
[24] H. L. Krall, O. Frink, A new class of orthogonal polynomials: The Bessel polynomials, Transactions of the American
Mathematical Society 65 (1) (1949) 100–115 (1949). doi:10.2307/1990516.

20
[25] V. Romanovski, Sur quelques classes nouvelles de polynomes orthogonaux, CR Acad. Sci. Paris 188 (1023) (1929) 1023–
1025 (1929).
[26] D. Alvarez-Castillo, M. Kirchbach, Exact spectrum and wave functions of the hyperbolic Scarf potential in terms of finite
Romanovski polynomials, Revista mexicana de física E 53 (2) (2007) 143–154 (2007).
[27] A. Raposo, H. Weber, D. Alvarez-Castillo, M. Kirchbach, Romanovski polynomials in selected physics problems, Open
Physics 5 (3) (2007) 253–284 (2007). doi:10.2478/s11534-007-0018-5.

21