ON COEFFICIENT PROBLEMS FOR Sρ∗

S. SIVAPRASAD KUMAR∗ , ARYA TRIPATHI, AND SNEHAL PANNU

Abstract. Logarithmic and inverse logarithmic coefficients play a crucial role in the theory of
arXiv:2412.17403v1 [math.CV] 23 Dec 2024

univalent functions. In this study, we focus on the class of starlike functions Sρ∗ , defined as
 
∗ zf ′ (z)
Sρ = f ∈ A : ≺ ρ(z), z ∈ D ,
f (z)
where ρ(z) := 1 + sinh−1 (z), which maps the unit disk D onto a petal-shaped domain. This
investigation aims to establish bounds for the second Hankel and Toeplitz determinants, with
their entries determined by the logarithmic coefficients of f and its inverse f −1 , for functions
f ∈ Sρ∗ .

1 Introduction

Let A denote the class of all normalized analytic functions f defined on the open unit disk
D := {z ∈ C : |z| < 1}, such that f (0) = 0 and f ′ (0) = 1. Each function f ∈ A admits a Taylor
series expansion of the form:
X∞
f (z) = z + an z n . (1.1)
n=2
Let S ⊂ A, where S represents the collection of analytic univalent functions. In particular, the
class S ∗ , comprising starlike functions, maps the unit disk D onto a starlike domain, represented
as:
zf ′ (z)
 
∗
S = f ∈ A : Re > 0, z ∈ D .
f (z)
Let f and g be two analytic functions, we say that f is subordinate to g, denoted by f ≺ g, if
there exists a Schwarz function w with w(0) = 0 and |w(z)| < 1 such that f (z) = g(w(z)). In
1992, Ma and Minda [8] introduced a broader subclass of S ∗ , denoted by S ∗ (ϕ), and defined as:
zf ′ (z)
 
S ∗ (ϕ) = f ∈ A : ≺ ϕ(z) ,
f (z)
where ϕ is an analytic univalent function satisfying the conditions Re ϕ(z) > 0, with ϕ(D) being
starlike with respect to ϕ(0) = 1, ϕ′ (0) > 0, and the domain ϕ(D) being symmetric about the
real axis. Various subclasses of S ∗ arise when the choice of ϕ is varied in the class S ∗ (ϕ).
Arora and Kumar [1] introduced a class of starlike functions associated with the petal-shaped
domain ρ(D), by choosing ϕ(z) := ρ(z), where ρ(z) = 1 + sinh−1 (z), defined by:

*Corresponding author
2020 Mathematics Subject Classification. 30C45 · 30C50 .
Key words and phrases. Univalent functions · Starlike functions · Petal-shaped domain · Hankel determinant
· Toeplitz determinant · Logarithmic coefficient · Inverse Logarithmic coefficient.
1
2 S. S. KUMAR, A. TRIPATHI, AND S. PANNU

zf ′ (z)
 
Sρ∗ = f ∈A: ≺ 1 + sinh−1 (z), z∈D .
f (z)
Clearly, the function ρ(z) is defined over the branch cuts along the segments (−∞, −i) ∪ (i, ∞)
on the imaginary axis. Consequently, the function is analytic in the unit disk D, geometrically,
ρ(D) = Ωρ := {Ω ∈ C : | sinh(Ω − 1)| < 1}. Certain coefficient problems of the analytic functions
associated with the class Sρ∗ were discussed in [6]. The function f0 (z) serves as an extremal
function for the family Sρ∗ which helps in obtaining the sharp results, given by
sinh−1 (t)
Z z 
1 1 1 1 6
f0 (z) = z exp dt = z + z 2 + z 3 + z 4 − z 5 − z + ... . (1.2)
0 t 2 9 72 225
The logarithmic coefficients γn associated with f ∈ S are defined as follows:
  ∞
f (z) X
Ff (z) = log =2 γn (f )z n ; z ∈ D \ {0}, log 1 = 0.
z
n=1
We denote γn (f ) by γn . For functions in class S, the sharp bounds for the logarithmic coefficients
when n = 1 and n = 2 are:
1 1
|γ1 | ≤ 1, |γ2 | ≤ + 2 .
2 e
A significant motivation behind studying logarithmic coefficients lies in estimating the sharp
bounds for the class S. Currently, these sharp bounds have only been established for γ1 and
γ2 . The problem of determining the sharp bounds for γn when n ≥ 3 is still open. The Koebe
1/4-theorem, enables us to define the inverse function Ff ∈ A in a neighborhood of the origin
with the Taylor series expansion:
X∞
−1
F (w) := f (w) = w + An wn , |w| < 1. (1.3)
n=2

The logarithmic inverse coefficients Γn , for n ∈ N, associated with the function F , are defined
by the relation
  ∞
F (w) X 1
Ff −1 (w) := log =2 Γn wn , |w| < .
w n=1
4
Ponnusamy et al. [13] established sharp bounds for the initial logarithmic coefficients of the
inverse function for certain subclasses of S ∗ .
Hankel matrices (and determinants) have emerged as fundamental elements in different areas
of mathematics, finding a wide array of applications [15]. For a function f ∈ A, the q th Hankel
determinant Hq,n (Ff ), where q, n ∈ N, with entries as logarithmic coefficients, is given by:

γn γn+1 . . . γn+q−1
γn+1 γn+2 . . . γn+q
Hq,n (Ff /2) = .. .. .. .. . (1.4)
. . . .
γn+q−1 γn+q . . . γn+2(q−1)
Kowalczyk et al. [5] investigated the Hankel determinants formed using logarithmic coefficients.
Specifically, the expression H2,1 (Ff /2) shares a notable resemblance to H2,1 (f ) = a2 a4 − a23 ,
where f ∈ A. Hankel determinants are beneficial, for example, in determining whether certain
 ON COEFFICIENT PROBLEMS FOR Sρ∗ 3

coefficient functionals related to functions are bounded in D and if they achieve sharp bounds,
see [14]. Furthermore, Mundalia and Kumar [9] explored the problem of logarithmic coefficients
for certain subclasses of close-to-convex functions. The logarithmic coefficients for a function
f ∈ S are represented as follows:
   
1 1 1 2 1 1 3
γ1 = a2 , γ2 = a3 − a2 , γ3 = a4 − a2 a3 + a2 . (1.5)
2 2 2 2 3
In 2016, Ye and Lim [15] established that any n×n matrix over C can be represented, in general,
as a product of certain Toeplitz or Hankel matrices. Toeplitz matrices, along with their deter-
minants, play a significant role in both applied and theoretical mathematics. Their applications
extend to various fields such as analysis, quantum physics, image processing, integral equations,
and signal processing. A defining feature of Toeplitz matrices is that their elements remain
constant along each diagonal. Giri and Kumar [4] explored bounds on Toeplitz determinants for
different subclasses of normalized univalent functions in higher-dimensional spaces. Inspired by
these contributions, we define the Toeplitz determinant of the logarithmic coefficients for f ∈ S
as follows:

γn γn+1 . . . γn+q−1
γn+1 γn . . . γn+q−2
Tq,n (Ff /2) = .. .. .. .. . (1.6)
. . . .
γn+q−1 γn+q−2 ... γn
Motivated further by the above concepts, we begin the examination of the Hankel determinant
Hq,n (Ff −1 /2) and the Toeplitz determinant Tq,n (Ff −1 /2), where the elements are the logarithmic
coefficients of the inverse functions [12] of f −1 ∈ S. The Hankel determinant is given by

Γn Γn+1 . . . Γn+q−1
Γn+1 Γn+2 . . . Γn+q
Hq,n (Ff −1 /2) = .. .. .. .. , (1.7)
. . . .
Γn+q−1 Γn+q . . . Γn+2(q−1)
and the Toeplitz determinant Tq,n (Ff −1 /2) is given by
Γn Γn+1 . . . Γn+q−1
Γn+1 Γn . . . Γn+q−2
Tq,n (Ff −1 /2) = .. .. .. .. . (1.8)
. . . .
Γn+q−1 Γn+q−2 ... Γn
This work focuses on deriving bounds for the second Hankel and Toeplitz determinants, where
the entries are the logarithmic coefficients associated with the functions f and f −1 . Here, f
belongs to the class Sρ∗ .
4 S. S. KUMAR, A. TRIPATHI, AND S. PANNU

2 Preliminary results
The Carathéodory class P and its coefficient bounds are crucial in determining Hankel and
Toeplitz determinant bounds. Let P denote the class of functions consisting of p, such that
∞
X
p(z) = 1 + pn z n , (2.1)
n=1
which are analytic in the open unit disk D and satisfy Re p(z) > 0 for any z ∈ D. Here p(z) is
called the Carathéodory function, or functions with positive real part [2, 11]. It is known that
pn ≤ 2, n ≥ 1, for a function p ∈ P. In this section, we provide crucial lemmas that will be
utilized to establish the main results of this paper.

Lemma 2.1. [7, 10, 11]: If p ∈ P is the form (2.1), then

p1 = 2ζ1 , (2.2)
p2 = 2ζ12 + 2(1 − ζ12 )ζ2 , (2.3)
p3 = 2ζ13 + 4(1 − ζ12 )ζ1 ζ2 − 2(1 − ζ12 )ζ1 ζ22 + 2(1 − ζ12 )(1 2
− |ζ2 | )ζ3 ,
where ζ1 ∈ [0, 1] and ζ2 , ζ3 ∈ D := {z ∈ C : |z| ≤ 1}.

Lemma 2.2. [3]: If A, B, C ∈ R, let us consider

Y (A, B, C) := max{|A + Bz + Cz 2 | + 1 − |z|2 , z ∈ D}
1. If AC ≥ 0, then

 |A| + |B| + |C|,
 |B| ≥ 2(1 − |C|),
Y (A, B, C) =
 1 + |A| + B2
4(1−|C|) , |B| < 2(1 − |C|).


2. If AC < 0, then
B2
, −4AC(C −2 − 1) ≤ B 2 ∧ |B| < 2(1 − |C|),


 1 − |A| + 4(1−|C|)




B2
Y (A, B, C) = 1 + |A| + 4(1+|C|) , B 2 < min{4(1 + |C|)2 , −4AC(C −2 − 1)},






R(A, B, C), Otherwise,
where 

 |A| + |B| − |C|, |C|(|B| + 4|A|) ≤ |AB|,





R(A, B, C) = −|A| + |B| + |C|, |AB| ≤ |C|(|B| − 4|A|),



 q
 (|C| + |A|) 1 − B 2 , Otherwise.


4AC

3 Second Hankel Determinant

By applying the definition of the Hankel determinant, we obtain the expression for second Hankel
determinant given by,
H2,1 (f ) = a1 a3 − a22 .
 ON COEFFICIENT PROBLEMS FOR Sρ∗ 5

Since a1 := 1, it follows that H2,1 (f ) = a3 − a22 . By varying the coefficients, we can compute the
second Hankel determinant for different values of the coefficients. For instance, using equation
(1.4), we can derive the result for logarithmic coefficients, given by
 
γ γ 1 1 4
H2,1 (Ff /2) = 1 2 = γ1 γ3 − γ22 = a2 a4 − a23 + a2 . (3.1)
γ2 γ3 4 12
It is important to note that H2,1 (Ff /2) remains invariant under rotation. Since, for the function
fθ (z) := e−iθ f (eiθ z), where f ∈ S and θ ∈ R, we have the following relation:

e4iθ
 
1 4
H2,1 (Ffθ /2) = a2 a4 − a23 + a = e4iθ H2,1 (Ff /2) . (3.2)
4 12 2
Using (1.7), we can derive the logarithmic coefficients for the inverse functions, given by
Γ1 Γ2 1
= Γ1 Γ3 − Γ22 = 13a42 − 12a22 a3 − 12a23 + 12a2 a4 .


H2,1 (Ff −1 /2) = (3.3)
Γ2 Γ3 48
Similarly, it can be observed that H2,1 (Ff −1 /2) is also invariant under rotation.

3.1 Sharp Bound Estimation of |H2,1 (Ff /2)| for Sρ∗

In this section, we will primarily focus on estimating the sharp bound for the Hankel determinant
associated with the logarithmic coefficients for starlike functions (|H2,1 (Ff /2)|) defined in the
petal-shaped domain ρ(D).

Theorem 3.1. Let f ∈ Sρ∗ , then

1 1
≤ |H2,1 (Ff /2)| ≤ . (3.4)
72 16
The above inequality is sharp.

Proof. Since f ∈ Sρ∗ , we can find a Swcharz function w(z) such that
zf ′ (z)
= 1 + sinh−1 (w(z)). (3.5)
f (z)
Assume that w(z) = (p(z) − 1)/(p(z) + 1) where, p ∈ P is given by (2.1). Using the expansion
of f from (1.1) and p from (2.1), we get
1 1 1
−p31 − 6p1 p2 + 24p3 .


a2 = p 1 , a3 = p 2 , a4 = (3.6)
2 4 144
Using (1.5), (3.1), (3.6), we obtain the following expression
L := γ1 γ3 − γ22
      2
1 1 1 3 1 1 2
= a2 a4 − a2 a3 + a2 − a3 − a2
2 2 3 2 2
 
1 1
= a2 a4 − a23 + a42 (3.7)
4 12
1
p41 − 12p21 p2 − 36p22 + 48p1 p3


= (3.8)
2304
6 S. S. KUMAR, A. TRIPATHI, AND S. PANNU

′
By the Lemma (2.1) and (3.8), we get an equation in terms of the ζi s, where ζi ∈ D for i = 1, 2, 3:
1
−9ζ22 + 6ζ12 ζ22 + ζ14 (−2 + 3ζ22 ) + 12ζ1 ζ3 − 12ζ13 ζ3 + 12ζ1 (−1 + ζ12 )ζ3 |ζ2 |2


L = (3.9)
.
144
Since ζ1 ∈ [0, 1] from Lemma (2.1), therefore the above expression in (3.9) leads to,

|ζ2 |2 1

 16 ≤ 16 , ζ1 = 0,
|L| =
 1,

ζ1 = 1.
72

By utilising thr triangle inequality in (3.9), for ζ1 ∈ (0, 1) and |ζ3 | ≤ 1, we obtain the following
inequality:
1
|L| ≤ | − 9ζ22 + 6ζ12 ζ22 + ζ14 (−2 + 3ζ22 ) + 12ζ1 ζ3 − 12ζ13 ζ3 | + 12ζ1 (−1 + ζ12 )ζ3 |ζ2 |2
144
1
= ζ1 (1 − ζ12 ) Ψ(A, B, C), (3.10)
12
where
Ψ(A, B, C) := |A + Bζ2 + Cζ22 | + 1 − |ζ22 |,
with
−ζ13 ζ12 + 3
 
A= , B = 0, C=− . (3.11)
6(1 − ζ12 ) 4ζ1

We now examine the following cases in the context of Lemma 2.2, based on the various cases
involving A, B, and C, which are given in (3.11).
I. Suppose ζ1 ∈ X = (0, 1). Clearly, we can observe that AC ≥ 0, also
3 ζ1
|B| − 2(1 − |C|) = −2 + + > 0, ζ1 ∈ X.
2ζ1 2
Therefore, by using Lemma 2.2, we get,
Ψ(A, B, C) = |A| + |B| + |C|.
Now, by utilising above equation, (3.10) can be further reduced in the following manner:
1
|L| ≤ ζ1 (1 − ζ12 ) (|A| + |B| + |C|)
12
−ζ13 −ζ12 − 3
 
1 2
= ζ1 (1 − ζ1 ) +
12 6(1 − ζ12 ) 4ζ1
1
= (9 − 6ζ12 − ζ14 )
144
1
≤ .
16

Thus, from the above result we can see that inequality (3.4) holds. To demonstrate the sharpness
of (3.4), we consider the function f1 ∈ Sρ∗ for the upper bound, given by:
sinh−1 (t2 )
Z z
z3 z5 z7

f1 (z) = z exp dt = z + + − + ... .
0 t 2 8 144
 ON COEFFICIENT PROBLEMS FOR Sρ∗ 7

By comparing the coefficients, we can easily find the values for a2 , a3 , and a4 , which upon
further substitution into (3.7) yields the equality:
1
|H2,1 (Ff /2)| = .
16
For the lower bound case, we consider the extremal function f0 (z), defined in (1.2), which
completes the proof.

3.2 Sharp bound of |H2,1 (Ff −1 /2)| for Sρ∗

In this section, we will focus on estimating the sharp bound of |H2,1 (Ff −1 /2)| within ρ(D).

Theorem 3.2. Let f ∈ Sρ∗ , then

1 1
≤ |H2,1 (Ff −1 /2)| ≤ . (3.12)
16 9
The above inequality is sharp.

Proof. Let us suppose that f ∈ Sρ∗ satisfy (3.5). Ponnusamy et al. [3] examined the logarithmic
coefficients of the inverses of univalent functions, defined as
−1 −1 3 −1 5
Γ1 = a2 , Γ2 = a3 + a22 , Γ3 = a4 + 2a2 a3 − a32 . (3.13)
2 2 4 2 3
′
By utilising (3.3), (3.13), and Lemma 2.1, we express Γ1 Γ3 − Γ22 in terms of ζi s by substituting
′
pi s with ζi′ s (i = 1, 2, 3):
L := Γ1 Γ3 − Γ22
1
13a42 − 12a22 a3 − 12a23 + 12a2 a4


=
48
1
37p41 − 48p21 p2 − 36p22 + 48p1 p3


=
2304
1
= (6ζ 2 (−3 + ζ2 )ζ2 − 9ζ22 + ζ14 (16 + 18ζ2 + 3ζ22 ) + 12ζ1 ζ3 − 12ζ13 ζ3 (3.14)
144 1
+12ζ1 (−1 + ζ12 )ζ3 |ζ2 |2 ).
As |ζ3 | ≤ 1, the above inequality provides

|ζ2 |2 1

 16 ≤ 16 , ζ1 = 0,
|L| =
 1,

ζ1 = 1.
9

Let ζ1 ∈ (0, 1). By applying the triangle inequalitiy in (3.14) and the fact that |ζ3 | ≤ 1, we
obtain
1
|L| = ζ1 (1 − ζ12 ) Ψ(A, B, C), (3.15)
12
where
Ψ(A, B, C) := |A + Bζ2 + Cζ22 | + 1 − |ζ22 |,
with
4ζ13 3 + ζ12
 
3
A= , B = − ζ1 , C = − . (3.16)
3(1 − ζ12 ) 2 4ζ1
8 S. S. KUMAR, A. TRIPATHI, AND S. PANNU

In light of Lemma 2.2 and the expressions of A, B and C obtained in (3.16), we study the
following cases. Let us assume that ζ1 ∈ X where X = (0, 1). Clearly, we can observe that
AC < 0 for ζ1 ∈ X. This case demands the following sub-cases:

(a) Simply for each ζ1 ∈ X

3
T1 (ζ1 ) := |B| − 2(1 − |C|) = 2ζ1 − 2 + > 0,
2ζ1
implies
|B| > 2(1 − |C|).

−ζ12 (225 + 11ζ12 )

 
1 2
T2 (ζ1 ) := −4AC − 1 − B = ≤ 0,
C2 12(3 + ζ12 )
implies  
1
−4AC − 1 ≤ B 2.
C2
We can infer from the above computations that T1 (X ∗ )∩T2 (X ∗ ) is empty. Consequently,
for every ζ1 ∈ X, this case does not occur, as stated by Lemma 2.2.

1
(b) In view of ζ1 ∈ X, the relation 4(1 + |C|)2 and −4AC


C2
− 1) become
2
(3 + 4ζ1 + ζ1 )2
T3 (ζ1 ) := 4(1 + |C|)2 = > 0,
4ζ12
4ζ12 (−9 + ζ12 )
 
1
T4 (ζ1 ) := −4AC − 1 = < 0.
C2 3(3 + ζ12 )
Therefore, min{T3 (ζ1 ), T4 (ζ1 )} = T4 (ζ1 ).
From case (a), we know that:
 
1
−4AC − 1 ≤ B 2.
C2
Thus this case fails for ζ1 ∈ X, as stated by Lemma 2.2.

(c) For ζ1 ∈ X we assume that T5 (ζ1 ) := |AB| − |C|(|B| + 4|A|). Therefore,

27 + 78ζ12 − 25ζ14
T5 (ζ1 ) = <0
24(−1 + ζ12 )
implies
|AB| < |C|(|B| + 4|A|).
Thus the case fails to exist for ζ1 ∈ X, as stated by Lemma 2.2.

(d) For ζ1 ∈ X we assume that

27 − 114ζ12 − 89ζ14
T6 (ζ1 ) := |AB| − |C|(|B| − 4|A|) = ≤ 0.
24(−1 + ζ12 )
 ON COEFFICIENT PROBLEMS FOR Sρ∗ 9

√
q 
−57+6 157
We observe that the above equation holds true for 0 ≤ ζ1 ≤ ζ′ = 89 ≈
0.451959 ∈ X. Therefore, by using Lemma 2.2, we get,
Ψ(A, B, C) = R(A, B, C) = −|A| + |B| + |C|.
Using (3.15) to proceed towards our desired result,
1
|L| ≤ ζ1 (1 − ζ12 )(−|A| + |B| + |C|)
12
1
9 + 12ζ 2 − 37ζ14 = φ(ζ1 ),


= (3.17)
144
where
1
9 + 12t2 − 37t4


φ(t) :=
144
Since φ′ (t) = 0 for t ∈ (0, 1) holds true for the only possible value of t0 = 6/37 < ζ1′ , we
p

deduce that φ is increasing in [0, t0 ] and decreasing in [t0 , ζ1′ ]. Therefore, for 0 < t < ζ1′ ,
we obtain
φ(ζ1 ) ≤ φ(t0 ) ≈ 0.0692568.
Hence, from (3.17) we see that
|L| ≤ φ(ζ1 ) ≤ φ(t0 ) ≈ 0.0692568 < 1/9.

(e) By applying Lemma 2.2 for ζ1′ < ζ1 < 1, we get

r
1 2 B2
L ≤ ζ1 (1 − ζ1 )(|C| + |A|) 1 −
12 4AC
s
1 75 − 11ζ12 2 4


= 9 − 6ζ 1 + 13ζ 1 = φ2 (ζ1 )
576 3 + ζ12
where
s
1 75 − 11t2
9 − 6t2 + 13t4


φ2 (t) := 2
576 3 + t1
It is observed that φ2 is an increasing function such that φ′2 (t) ≥ 0 for ζ1′ < t < 1,
therefore after a simple computation, we get
|L| ≤ φ2 (ζ2 ) ≤ φ2 (ζ1′ ) ≈ 0.451959 < 1/9.
Based on the results obtained, we can see that the inequality (3.12) holds true.
To demonstrate the sharpness of the bound, we can utilize the functions f0 (z) and f1 (z) for the
upper and lower bounds respectively, similar to those employed in Theorem 3.1. This concludes
the proof.
10 S. S. KUMAR, A. TRIPATHI, AND S. PANNU

4 Second Toeplitz Determinant

By applying the definition of the Toeplitz determinant, we obtain the following expression for
the second Toeplitz determinant:
T2,1 (f ) = a21 − a22 .
By varying the coefficients, the Toeplitz determinant can be computed for different sets of coeffi-
cients. For instance, by using equation (1.6), we can determine the determinant for logarithmic
coefficients:
γ γ 1
T2,1 (Ff /2) = 1 2 = γ12 − γ22 = 4a22 − a42 − 4a23 + 4a22 a3 .


(4.1)
γ2 γ1 16
Similarly, by applying equation (1.8), we can compute the Toeplitz determinant for logarithmic
coefficients of inverse functions:
Γ Γ2 1
T2,1 (Ff −1 /2) = 1 = Γ21 − Γ22 = −9a42 + 4a22 − 4a23 + 12a22 a3 .


(4.2)
Γ2 Γ1 16
Let T2,1 be a functional and Ffθ , Ffθ−1 be functions. Then, we can observe the rotational
invariant characteristic of the functional T2,1 with respect to the functions Ffθ and Ffθ−1 will
be similar to H2,1 (Ff /2) and H2,1 (Ff −1 /2), as found in section 3.

4.1 Sharp bound of |T2,1 (Ff /2)| for Sρ∗

This section explores the sharp bound estimation of the Toeplitz determinant linked to the
logarithmic coefficients of starlike functions (|T2,1 (Ff /2)|) defined on ρ(D).

Theorem 4.1. Let f ∈ Sρ∗ , then

1 1
≤ |T2,1 (Ff /2)| ≤ .
16 2
The above inequality is sharp.

Proof. Let us suppose that f ∈ Sρ∗ be of the form (1.1). Then there exist a Schwarz function,
w(z) = (p(z) − 1)/(p(z) + 1) such that it satisfies (3.5). From (4.1) and (3.6) we obtain:
T2,1 (Ff /2) = γ12 − γ22
1
−a42 + 4a22 − 4a23 + 4a22 a3


= (4.3)
16
1
−p41 − 4p22 + 4p21 (4 + p2 ) .


= (4.4)
256
Using (2.2) and (2.3) from Lemma 2.1, we derive the relation:
2p2 = p21 + (4 − p21 )ζ2 ,
where |ζ2 | ≤ 1. By substituting the value of p2 in (4.4) and assuming ζ2 := ζ, we derive the
following equation:
1
−p41 ζ 2 − 16ζ 2 + 16p21 + 8p21 ζ 2 .


T2,1 (Ff /2) =
256
In view of triangle inequality, the above equation can be expressed as:
1
|p41 ||ζ 2 | + 16|ζ 2 | + 16|p21 | + 8|p21 ||ζ 2 | .


|T2,1 (Ff /2)| ≤
256
 ON COEFFICIENT PROBLEMS FOR Sρ∗ 11

Given the rotational invariance of P, we restrict our analysis to non-negative values of p1 . Since
|p1 | ≤ 2, we have 0 ≤ p1 ≤ 2. Letting p := |p1 |, this implies p ∈ [0, 2]. Thus,
1
p4 + 24p2 + 16 .


|T2,1 (Ff /2)| ≤
256
Hence for p ∈ [0, 2],
1 1
≤ |T2,1 (Ff /2)| ≤ .
16 2
To complete the proof, it remains to verify the sharpness of the bound. Thus, in order to
establish the upper bound, consider the analytic function f2 ∈ Sρ∗ , defined by:
√ ! √ 7
sinh−1 ( 8it2 ) √ 3
Z z
5 2iz
f2 (z) = z exp dt = z + 2iz − z + + ... .
0 t 9
For the lower bound, we examine the analytic function f3 ∈ Sρ∗ , defined by:
sinh−1 (it2 )
Z z
iz 3 z 5 iz 7

f3 (z) = z exp dt = z + − + + ... .
0 t 2 8 144
By comparing the coefficients, we can easily determine the values of a2 and a3 . Substituting
these into (4.3) yields the desired equality, completing the proof.

4.2 Sharp bound of |T2,1 (Ff −1 /2)| for Sρ∗

This section will center on estimating the sharp bound around the petal shaped domain for
|T2,1 (Ff −1 /2)|, where f ∈ Sρ∗ .

Theorem 4.2. Let f ∈ Sρ∗ , then

1 5
≤ |T2,1 (Ff −1 /2)| ≤ .
16 4
The above inequality is sharp.

Proof. Let f ∈ Sρ∗ be of the form (1.1) satisfying the equation (3.5).In view of (4.2) and (3.13),
a simple computation shows that
T2,1 (Ff −1 /2) = Γ21 − Γ22
1
−9a42 + 4a22 − 4a23 + 12a22 a3


=
16
1
−9p41 − 4p22 + 4p21 (4 + 3p2 ) .


=
256
Using Lemma 2.1 in above equation, we obtain
1
16p21 − 4p41 + 16p21 ζ − 4p41 ζ − 16ζ 2 + 8p21 ζ 2 − p41 ζ 2 .


T2,1 (Ff −1 /2) =
256
In view of triangle inequality, we see above equation can be expressed as:
1
16|p21 | + 4|p41 | + 16|p21 ||ζ| + 4|p41 ||ζ| + 16|ζ 2 | + 8|p21 ||ζ 2 | + |p41 ||ζ 2 | .


|T2,1 (Ff −1 /2)| ≤
256
Assume p := |p1 | where p1 ≥ 0 by analogous reasoning in Theorem 4.1 and for the case |ζ| ≤ 1,
we obtain the following result by taking the help of Lemma 2.1.
1
9p4 + 40p2 + 16 .


|T2,1 (Ff −1 /2)| ≤
256
12 S. S. KUMAR, A. TRIPATHI, AND S. PANNU

Since p ∈ [0, 2]. Therefore,

1 5
≤ |T2,1 (Ff −1 /2)| ≤ . (4.5)
16 4
To demonstrate the sharpness of the upper bound, consider the analytic function f4 ∈ Sρ∗ ,
defined by:
√ !
sinh−1 ( 20it2 ) √ 5√ 7
Z z
5z 5
f4 (z) = z exp dt = z + 5iz 3 − + 5iz + . . . .
0 t 2 18
For this function, a straightforward calculation yields
5
|T2,1 (Ff −1 /2)| = .
4
For the sharpness of the lower bound, consider the function f2 , defined in Theorem 4.1. Thus,
we achieve equality in (4.5) which concludes our proof.

Conflict of interest: The authors declare no conflict of interest.

Authors’ Contribution: Each author contributed equally to the research and preparation of
the manuscript.

References

[1] Arora, K., Kumar, S. S.: Starlike functions associated with a petal shaped domain, Bulletin of the Korean
Mathematical Society. Korean Mathematical Society, 59(4), (2022), 993–1010.
[2] Carathéodory, C.: Über den Variabilitätsbereich der Koeffizienten von Potenzreihen, die gegebene Werte
nicht annehmen, Mathematische Annalen, 64(1), (1907), 95–115.
[3] Choi, J. H., Kim, Y. C., Sugawa, T.: A general approach to the Fekete-Szegö problem, Journal of the
Mathematical Society of Japan, 59(3), (2007), 707–727.
[4] Giri, S., Kumar, S. S.: Toeplitz determinants in one and higher dimensions. Acta Mathematica Scientia, 44,
(2024), 1931–1944.
[5] Kowalczyk, B., Lecko, A.: The Fekete-Szegö inequality for close-to-convex functions with respect to a certain
starlike function dependent on a real parameter, Journal of Inequalities and Applications, (2014), 1–16.
[6] Kumar, S. S., Verma, N.: Coefficient problems for starlike functions associated with a petal shaped domain,
arXiv:2208.14644, (2022).
[7] Libera, R. J., Zlotkiewicz, E. J., Early coefficients of the inverse of a regular convex function, Proceeding of
the American Mathematical Society, 85(2), (1982), 225–230.
[8] Ma, W. C., Minda, D.: A unified treatment of some special classes of univalent functions, Proceedings of the
Conference on Complex Analysis, International Press Inc., (1992).
[9] Mundalia, M., Kumar, S. S.: Coefficient problems for certain close-to-convex functions, Bulletin of the Iranian
Mathematical Society, 49(1), p.5, (2023).
[10] Pommerenke, C.: On the Hankel determinants of univalent functions, Mathematika, 14(1), (1967), 108–112.
[11] Pommerenke, C.: Univalent functions, Vandenhoeck & Ruprecht, Göttingen, (1975).
[12] Ponnusamy, S., Sharma, N. L., Wirths, K. J.: Logarithmic coefficients of the inverse of univalent functions,
Results in Mathematics, 73, (2018), 1–20.
[13] Ponnusamy, S., Sharma, N. L., Wirths, K. J.: Logarithmic coefficients problems in families related to starlike
and convex functions, Journal of the Australian Mathematical Society, 109(2), (2020), 230–249.
[14] Verma, N., Kumar, S. S.: A conjecture on H3 (1) for certain starlike functions, Mathematica Slovaca, 73(5),
(2023), 1197–1206.
[15] Ye, K., Lim, L. H.: Every matrix is a product of Toeplitz matrices, Foundations of Computational Mathe-
matics, 16(3), (2016), 577–598.
 ON COEFFICIENT PROBLEMS FOR Sρ∗ 13

Department of Applied Mathematics, Delhi Technological University, Bawana Road, Delhi-110042,

INDIA
Email address: [email protected]

Department of Applied Mathematics, Delhi Technological University, Bawana Road, Delhi-110042,

INDIA
Email address: [email protected]

Department of Applied Mathematics, Delhi Technological University, Bawana Road, Delhi-110042,

INDIA
Email address: [email protected]