ON A CLASS OF UNIVALENT FUNCTIONS DEFINED BY A

DIFFERENTIAL INEQUALITY

MD FIROZ ALI, VASUDEVARAO ALLU, AND HIROSHI YANAGIHARA

P∞
Abstract. For 0 < λ ≤ 1, let U(λ) be the class analytic functions f (z) = z + n=2 an z n in
the unit disk D satisfying |f ′ (z)(z/f (z))2 − 1| < λ and U := U(1). In the present article, we
prove that the class U is contained in the closed convex hull of the class of starlike functions
arXiv:2006.15577v1 [math.CV] 28 Jun 2020

and using this fact, we solve some extremal problems such as integral mean problem and arc
length problem for functions in U. By means of the so called theory of star functions, we also
solve the integral mean problem for functions in U(λ). We also obtain the estimate of the
Fekete-Szegö functional and the pre-Schwarzian norm of certain nonlinear integral transform
of functions in U(λ). Further, for the class of meromorphic functions which are defined in
∆ := {ζ ∈ Cb : |ζ| > 1} and associated with the class U(λ), we obtain a sufficient condition for
a function g to be an extreme point of this class.

1. Introduction
We denote the complex plane by C and the extended complex plane by C b = C ∪ {∞}. Let
H denote the class of analytic functions in the unit disk D := {z ∈ C : |z| < 1}. Then H is a
locally convex topological vector space endowed with the topology of uniform convergence over
compact subsets of D. Denote by A the subclass of H of functions f with Taylor series
∞
X
(1.1) f (z) = z + an z n .
n=2

The subclass S of A, consisting of univalent (one-to-one) functions has attracted much interest
for over a century, and is a central area of research in the theory of complex analysis.
Although the class S is the main attraction, various geometric subclasses (e.g. starlike,
convex and close-to-convex) have been extensively studied, some of which appear naturally in
different areas in the theory of quasiconformal mappings. A function f ∈ A is called starlike
(respectively convex), if f (D) is starlike with respect to the origin (respectively convex). Let
S ∗ and C denote the class of starlike and convex functions in S, respectively. It is well-known
that a function f ∈ A belongs to S ∗ if, and only, if Re (zf ′ (z)/f (z)) > 0 for z ∈ D. Similarly,
a function f ∈ A belongs to C if, and only if, Re (1 + (zf ′′ (z)/f ′ (z))) > 0 for z ∈ D. From the
above, it is easy to see that f ∈ S ∗ if, and only if, J[f ] ∈ C, where J[f ] denotes the Alexander
transform of f ∈ A defined by
Z z Z 1
f (ξ) f (tz)
J[f ](z) := dξ = dt.
0 ξ 0 t
In 1960, Biernacki claimed that f ∈ S implies J[f ] ∈ S, but this turned out to be wrong (see
[6, Theorem 8.11]). Later, Kim and Merkes [9] have considered the nonlinear integral transform
Jα defined by
Z z α
f (ξ)
(1.2) Jα [f ](z) := dξ
0 ξ
File: U_lambda-Final.tex, printed: 2020-6-30, 1.07
2010 Mathematics Subject Classification. Primary 30C45, 30C55.
Key words and phrases. Analytic, univalent, starlike, convex functions, extreme points, closed convex hull.

1
2 Md Firoz Ali, Vasudevarao Allu and Hiroshi Yanagihara

for complex numbers α and for the functions f in the class

ZF = {f ∈ A : f (z) 6= 0 for all 0 < |z| < 1}
and proved that Jα (S) := {Jα [f ] : f ∈ S} ⊂ S for |α| ≤ 1/4. For starlike functions, it is known
that Jα (S ∗ ) ⊂ S when either |α| ≤ 1/2 or α ∈ [1/2, 3/2] (see [3, 19]).
For 0 < λ ≤ 1, let U(λ) be the class of functions f ∈ A satisfying
 2
′ z
f (z) − 1 < λ for z ∈ D.
f (z)

Since f ′ (z)(z/f (z))2 6= 0 in D, it follows that every f ∈ U(λ) is non-vanishing in D\{0}. We set
U := U(1). It is also clear that every f ∈ U(λ) is locally univalent. Moreover, it is well-known
that every f ∈ U is univalent in D, i.e., U ( S (see [2]) and hence, for 0 ≤ λ < 1, one has
∗ ∗
U(λ) ( S. On the other hand, it follows from √ [7, 15] that neither U is included in S nor S
is included in U. However, for 0 < λ < 1/ 2, functions in U(λ) are starlike with the with the
additional assumption that f ′′ (0) = 0, . In 1977, Singh [18] obtained an estimate for the radius
of starlikeness of the class U which is surprisingly close to unity.

One of the important notions in geometric function theory is subordination. Let f and g be
two analytic functions in D. We say that f is subordinate to g, written as f ≺ g or f (z) ≺ g(z),
if there exists an analytic function ω : D → D with ω(0) = 0 such that f (z) = g(ω(z)) for z ∈ D.
In the case, g is univalent in D, subordination is equivalent to f (0) = g(0) and f (D) ⊆ g(D).
In 1995, Obradović [14] proved that if f ∈ U then f (z)/z ≺ k(z)/z, where k(z) = z/(1−z)2 is
the Koebe function. Recently, this result has been generalized for the class U(λ) by Obradović
et al. [16], where it has been proved that if f ∈ U(λ) then
f (z) kλ (z) z
(1.3) ≺ with kλ (z) = .
z z (1 − z)(1 − λz)
Clearly, k1 (z) = k(z) is the well-known Keobe function.
Suppose X is a linear topological space and U ⊆ X. The closed convex hull of U, denoted
by co U, is defined as the intersection of all closed convex sets containing U. For U ⊆ V ⊆ X,
we say that U is an extremal subset of V if u = tx + (1 − t)y, where u ∈ U, x, y ∈ V and
0 < t < 1 then x and y both belong to U. An extremal subset of U consisting of just one point
is called an extreme point of U. We denote the set of all extreme points of U by E(U). For a
general reference and for many important results on this topic, we refer to [8].

In the present article, among other results, we prove that the class U is contained in the
closed convex hull of S ∗ , i.e., U ( co S ∗ and using this fact, we solve some extremal problems
such as integral mean problem and arc length problem for functions in U. We also consider
the nonlinear integral transform Jα defined by (1.2) for functions in U(λ) and obtain its pre-
Schwarzian norm estimates. We also discuss the Fekete-Szegö problem for functions in U(λ).
Finally, we consider the class M0 (λ) of meromorphic functions defined in ∆ := {ζ ∈ C b : |ζ| > 1}
and associated with the class U(λ). For the class M0 (λ), we obtain a sufficient condition for a
function g to be an extreme point of the class M0 (λ).

2. Properties of functions in U(λ)

We need the following results to prove our main results.
Proposition 2.1. [1] Let F be an analytic and univalent function in the unit disk D. If F (z) 6= 0
in D and C \ F (D) is convex domain, then any analytic function f in s(F n ) := {g : g ≺ F n },
 On a class of univalent functions defined by a differential inequality 3

n = 1, 2, . . ., can be expressed as
Z
f (z) = F n (xz) dµ(x),
|x|=1

for some probability measure µ on the unit circle |x| = 1.

Proposition 2.2. [8, Theorem 5.8] The set co S ∗ consists of all functions represented by
Z
z
f (z) = 2
dµ(x),
|x|=1 (1 − xz)

where µ varies over the set of all probability measures on the unit circle |x| = 1. Also,
 
∗ z
E(co S ) = : |x| = 1 .
(1 − xz)2
Using Propositions 2.1 and 2.2, we prove the following integral representation for functions
in U.
Theorem 2.1. Each function f ∈ U has an integral representation of the form
Z
z
(2.1) f (z) = 2
dµ(x)
|x|=1 (1 − xz)

for some probability measure µ on the unit circle |x| = 1. Moreover, U ( co S ∗ .

Proof. If f ∈ U then (1.3) holds, i.e., f (z)/z ≺ k(z)/z = (1 − z)−2 . We note that the function
k(z)/z = (1 − z)−2 is analytic and univalent in D and maps the unit disk D onto a domain
Ω whose complement is convex. Indeed, the function k(z)/z maps the unit circle ∂D onto a
hyperbola given by the equation v 2 = −(u − 1/4). In view of Proposition 2.1, f (z)/z can be
expressed as Z
f (z) k(xz)
= dµ(x),
z |x|=1 xz
which is equivalent to (2.1). From the representation (2.1) and Proposition 2.2 it follows that
U ( co S ∗ . The proper inclusion follows from the fact that U is not included in S ∗ and
S ∗ ( co S ∗ . 
Remark 2.1. It follows from Theorem 2.1 that every function f ∈ U has the integral represen-
tation of the form (2.1). But unfortunately, we are unable to conclude similar integral repre-
sentation for functions in U(λ), 0 < λ < 1, using the subordination relation f (z)/z ≺ kλ (z)/z,
because the function φλ (z) := kλ (z)/z, 0 < λ < 1, maps D onto a domain φλ (D) whose comple-
ment is not convex. Indeed, if we let φλ (eiθ ) = u(θ) + iv(θ), then a simple computation shows
that
1 − λ − 2λ cos θ (1 + λ − 2λ cos θ) cot(θ/2)
(2.2) u(θ) = 2
and v(θ) = ,
2(1 + λ − 2λ cos θ) 2(1 + λ2 − 2λ cos θ)
and so, the equation of the curve φλ (eiθ ), 0 ≤ θ < 2π, is given by
2 (1 + (λ − 1)u)2 (2(λ + 1)u − 1)
(2.3) v =− .
(λ + 1)(2(λ − 1)2 u + 3λ − 1)
From (2.2), we note that for 0 < λ < 1, |v| → +∞ and u → (1 − 3λ)/(2(1 − λ)2 ) as θ → 0,
and so u = (1 − 3λ)/(2(1 − λ)2 ) is an vertical asymptote of the curve (2.3). Hence, C \ φλ (D)
is not convex.
It is important to note that by using the integral representation (2.1) for functions in U, one
can easily obtain the following coefficient estimates for functions in U.
4 Md Firoz Ali, Vasudevarao Allu and Hiroshi Yanagihara

Corollary 2.1. If f ∈ U is of the form (1.1) then |an | ≤ n. Equality holds if, and only if, f is
a rotation of the Koebe function.
If G is a convex subset of H and J : H → R then J is called convex on G if J(tf + (1 − t)g) ≤
tJ(f ) + (1 − t)J(g) whenever f, g ∈ G and 0 ≤ t ≤ 1. We note that co S ∗ is a compact subset
of H and E(co S ∗ ) ( U. Hence, for any real-valued, continuous and convex functional J on
co S ∗ , by [8, Theorems 4.5 and 4.6], we have
max J(f ) ≤ max∗ J(f ) = max J(f ) ≤ max J(f ),
f ∈U f ∈co S f ∈E(co S ∗ ) f ∈U

and therefore,
(2.4) max J(f ) = max J(f ).
f ∈U f ∈E(co S ∗ )

Another problem which has an independent interest in the theory of univalent functions is
the estimation of the Lp mean for certain classes of analytic functions. Corresponding to each
analytic function f in D, we let
Z 2π
1
J(f ) = |f (n) (reiθ )|p dθ,
2π 0
where 0 < r < 1, p > 0 and n = 0, 1, 2, . . .. It is our aim to maximize the functional J(f ) over
the class U. In general, it is more convenient to consider the functional kf k = [J(f )]1/p . In
particular, if p ≥ 1 then ktf +(1−t)gk ≤ tkf k+(1−t)kgk because of Minkowski’s inequality. In
other words, if p ≥ 1 then kf k is a convex functional. The above argument is due to MacGregor
[13].
Theorem 2.2. If f ∈ U and k(z) = z/(1 − z)2 then
Z 2π Z 2π
1 (n) iθ p 1
(2.5) |f (re )| dθ ≤ |k (n) (reiθ )|p dθ,
2π 0 2π 0
whenever 0 < r < 1, p ≥ 1 and n = 0, 1, 2, . . .. Moreover, for n = 0, the inequality (2.5) holds
for any real number p.
Proof. For p ≥ 1, as mentioned above, it is sufficient to consider functions of the form
z
f (z) = , |x| = 1.
(1 − xz)2
For these functions f (z) = x−1 k(xz) and hence f (n) (z) = xn−1 k (n) (xz). Thus
Z 2π Z 2π
1 (n) iθ p 1
|f (re )| dθ = |k (n) (xreiθ )|p dθ
2π 0 2π 0
Z 2π
1
= |k (n) (reiθ )|p dθ.
2π 0
Hence (2.5) holds for any f ∈ U.
Moreover, if f ∈ U then, as mentioned above, f (z)/z ≺ k(z)/z and also z/f (z) ≺ z/k(z).
Therefore, for n = 0, the inequality (2.5) follows from Littlewood’s subordination theorem (see
[6, Theorem 6.1]) for every real p. 
For f ∈ H and 0 < r < 1, let
Z 2π
L(r, f ) = r|f ′(reiθ )| dθ
0
denote the arclength of the image of the circle |z| = r under f . Substituting n = 1 in Theorem
2.2, we obtain the following sharp estimate of L(r, f ) for functions f in U.
 On a class of univalent functions defined by a differential inequality 5

Corollary 2.2. If f ∈ U then L(r, f ) ≤ L(r, k), where k is the Koebe function.
We now review some of the standard facts on the theory of *-functions developed by Baern-
stein [4]. For more on *-functions we refer to Duren [6]. For g ∈ L1 [−π, π], the *-function of g
is defined by Z
g ∗ (θ) = sup g(x) dx, 0 ≤ θ ≤ π,
|E|=2θ E

where |E| denote the Lebesgue measure of the set E. Here the supremum is taken over all
Lebesgue measurable subsets of [−π, π] with |E| = 2θ.
Lemma 2.1. For g, h ∈ L1 [−π, π], the following statements are equivalent.
(a) For every convex nondecreasing function Φ on R,
Z π Z π
Φ(g(θ)) dθ ≤ Φ(h(θ)) dθ.
−π −π

(b) For every t ∈ R,

Z π Z π
+
[g(θ) − t] dθ ≤ [h(θ) − t]+ dθ.
−π −π

(c) g ∗(θ) ≤ h∗ (θ), 0 ≤ θ ≤ π.

Here [g(θ) − t]+ = max{g(θ) − t, 0}.
Lemma 2.2. [11, Lemma 2] Let u and v be two subharmonic functions in D and there exists
an analytic function ω : D → D with ω(0) = 0 such that u(z) = v(ω(z)) for z ∈ D. Then for
each 0 < r < 1,
u∗(reiθ ) ≤ v ∗ (reiθ ).
By using Lemmas 2.1 and 2.2 we prove the following interesting result.
Theorem 2.3. Let Φ be a convex and nondecreasing function in R. Then for f ∈ U(λ) with
0 < λ ≤ 1 and 0 < r < 1, we have
Z π Z π
iθ
Φ(± log |f (re )|) dθ ≤ Φ(± log |kλ(reiθ )|) dθ
−π −π

where kλ is defined by (1.3).

Proof. For f ∈ U(λ), it is known that f (z)/z ≺ kλ (z)/z. Since kλ (z) 6= 0 in D \ {0}, it follows
that
f (z) kλ (z)
(2.6) ± log ≺ ± log .
z z
Since log |f (z)/z| is subharmonic in D, an application of Lemma 2.2 in (2.6) gives
 ∗  ∗
f (reiθ ) kλ (reiθ )
± log ≤ ± log .
reiθ reiθ
For any g ∈ L1 [−π, π] and a constant c ∈ R, we note that (g(reiθ )+c)∗ = g ∗ (reiθ )+c. Therefore,

∗
∗
(2.7) ± log |f (reiθ )| ≤ ± log |kλ (reiθ )| .
The conclusion now follows from (2.7) together with Lemma 2.1. 
Corollary 2.3. If f ∈ U(λ) and kλ is defined by (1.3) then for any 0 < r < 1, p ∈ R, we have
Z 2π Z 2π
1 iθ p 1
|f (re )| dθ ≤ |kλ (reiθ )|p dθ.
2π 0 2π 0
6 Md Firoz Ali, Vasudevarao Allu and Hiroshi Yanagihara

Proof. For p = 0, there is nothing to prove. For p 6= 0, by choosing Φ(x) = epx , p > 0, in
Theorem 2.3, we obtain the desired result. 
For a locally univalent function f in D, the pre-Schwarzian derivative Tf is defined by Tf =
f ′′ /f ′ and the pre-Schwarzian norm of f is defined by
kf k = sup(1 − |z|2 )|Tf (z)|.
z∈D

The pre-Schwarzian norm has significant meaning in the theory of Teichmüller spaces. It is
interesting to note that the pre-Schwarzian norm of f is nothing but the Bloch seminorm of
the function log f ′ . In 1976, Yamashita [22] observed that kf k is finite if, and only if, f is
uniformly locally univalent in D, i.e. there exists a positive constant ρ for which f is univalent
in every disk of hyperbolic radius ρ in D. Furthermore, kf k ≤ 6 if f is univalent in D and
conversely, f is univalent in D if kf k ≤ 1 and these bounds are sharp (see [5]). Moreover, if f
can be extended to a k-quasiconformal automorphism of the Riemann sphere C b then we have
kf k ≤ 6k. In 2004, Kim et al. [10] obtained estimate of kJα [f ]k for functions in U(λ) with
second coefficient a2 fixed. Our next result gives estimate of kJα [f ]k for functions in U(λ).
Theorem 2.4. For f ∈ U(λ), 0 < λ ≤ 1, let Jα [f ] be defined by (1.2) where α ∈ C. Then

2|α| for 0 < λ ≤ 13
(2.8) kJα [f ]k ≤ kJα [kλ ]k = √
 3+λ−2 2(1−λ2 )
λ
|α| for 13 < λ ≤ 1,
where kλ is defined by (1.3).
Proof. Taking a logarithmic differentiation in (1.2), we obtain Jα [f ] = αJ[f ] and so
kJα [f ]k = |α|kJ[f ]k.
Hence it suffices to show the inequality (2.8) for the case α = 1. For f ∈ U(λ), let F = J[f ]
and Kλ = J[kλ ]. Then by (1.3), we have
f (z) kλ (z)
F ′ (z) = ≺ = Kλ′ (z).
z z
It is well-known that (see [17, formula 4, page 35]) if g ≺ h then
sup(1 − |z|2 )|g ′ (z)| ≤ sup(1 − |z|2 )|h′ (z)|.
z∈D z∈D
′
Using the fact log F ≺ log Kλ′ , we obtain
F ′′ (z) Kλ′′ (z)
kF k = sup(1 − |z|2 ) ≤ sup (1 − |z| 2
) = kKλ k.
z∈D F ′ (z) z∈D Kλ′ (z)
It remains to find the value of kJ[kλ ]k = kKλ k. Since Kλ′ (z) = kλ (z)/z, by taking logarithmic
differentiation, we obtain
Kλ′′ (z) 1 λ
′
= +
Kλ (z) 1 − z 1 − λz
and so
K ′′ (z) 1 λ
sup(1 − |z|2 ) λ′ = sup(1 − |z|2 ) +
z∈D Kλ (z) z∈D 1 − z 1 − λz
 
2 1 λ
≤ sup(1 − |z| ) +
z∈D 1 − |z| 1 − λ|z|
= sup φ(r),
0<r<1
 On a class of univalent functions defined by a differential inequality 7

where
λ(1 − r 2 )
φ(r) = 1 + r + .
1 − λr
It is important to note that for z = t with 0 < t < 1, we have
Kλ′′ (z) 1 λ
(1 − |z|2 ) ′
= (1 − t2 ) + = φ(t).
Kλ (z) 1 − t 1 − λt
Therefore,
kJ[kλ ]k = kKλ k = sup φ(r).
0<r<1
To find the critical points of φ(r), we solve φ (r) = 0. The roots of φ′ (r) = 0 are given by
′
p p
2 − 2(1 − λ2 ) 2 + 2(1 − λ2 )
r0 = and r1 = .
λ λ
But we note that r1 > 1 for all 0 < λ ≤ 1, while r0 < 1 for all 1/3 < λ ≤ 1. Moreover, it is an
easy exercise to verify that
p
3 + λ − 2 2(1 − λ2 ) 1
φ(r0 ) = ≥ φ(1) for < λ ≤ 1.
λ 3
For 0 < λ ≤ 1/3, clearly φ(0) = 1 + λ and φ(1) = 2. Consequently,

2 for 0 < λ ≤ 31
kJ[kλ ]k = sup φ(r) = √
0<r<1  3+λ−2 2(1−λ2 )
λ
for 13 < λ ≤ 1.
This completes the proof. 
Corollary 2.4. For f ∈ U, let Jα [f ] be defined by (1.2) where α ∈ C. Then
kJα [f ]k ≤ kJα [k]k = 4|α|
where k is the Koebe function.
Our next result deals with the Fekete-Szegö problem for functions in the class U(λ).
Theorem 2.5. Let f ∈ U(λ), 0 < λ ≤ 1 be of the form (1.1) and µ be a complex number.
Then


|(1 + λ + λ2 ) − µ(1 + λ)2 | for µ − 1+λ+λ
2 1
≥ 1+λ
(1+λ)2
2
(2.9) |a3 − µa2 | ≤

1 + λ for µ − 1+λ+λ
2 1
≤ 1+λ .
(1+λ)2

The first inequality is sharp and equality holds for the function kλ defined by (1.3).
Proof. IfPf ∈ U(λ) then from (1.3) there exists a function ω : D → D with ω(0) and of the form
ω(z) = ∞ n
n=1 cn z such that
f (z) 1
(2.10) = .
z (1 − ω(z))(1 − λω(z))
In terms of series formulation, (2.10) can be written as
∞
X ∞
X
n
an+1 z = An (ω(z))n ,
n=1 n=1
Pn j
where An = 1 + j=1 λ . By comparing the coefficients of z n for n = 1, 2, we obtain
a2 = (1 + λ)c1 and a3 = (1 + λ)c2 + (1 + λ + λ2 )c21 .
8 Md Firoz Ali, Vasudevarao Allu and Hiroshi Yanagihara

Therefore,
(2.11) |a3 − µa22 | = (1 + λ)c2 + ((1 + λ + λ2 ) − µ(1 + λ)2 )c21

≤ (1 + λ)|c2 | + (1 + λ + λ2 ) − µ(1 + λ)2 |c1 |2

≤ (1 + λ)(1 − |c1 |2 ) + (1 + λ + λ2 ) − µ(1 + λ)2 |c1 |2


≤ (1 + λ) + (1 + λ + λ2 ) − µ(1 + λ)2 − (1 + λ) |c1 |2 .
Since |c1 | ≤ 1, the conclusion follows from the inequality (2.11). 
Remark 2.2. The first inequality of (2.9) is sharp and equality holds for the function kλ
defined by (1.3), whereas the second inequality of (2.9) may not be sharp. For the function
g(z) = z/((1 − z 2 )(1 − λz 2 )), it is easy to see that a2 (g) = 0 and a3 (g) = 1 + λ and so
|a3 − µa22 | = 1 + λ. But
 2
′ z
g (z) − 1 = z 2 (1 + λ − 3λz 2 ) > 1
g(z)
when z is purely imaginary number. This shows that g 6∈ U(λ).
If f ∈ U(λ) is of the form (1.1) then a simple exercise shows that
 2
′ z
f (z) − 1 = (a3 − a22 )z 2 + · · ·
f (z)
and so by the Schwarz lemma we obtain |a3 − a22 | ≤ λ. Thus, for µ ≥ 1 we immediately have
(2.12) |a3 − µa22 | ≤ |a3 − a22 | + (µ − 1)|a2 |2

≤ λ + (µ − 1)(1 + λ)2

= µ(1 + λ)2 − (1 + λ + λ2 ).
The equality in (2.12) holds for the function kλ . If we consider µ as a real number in Theorem
2.5 then the condition
1 + λ + λ2 1
µ− ≥
(1 + λ)2 1+λ
is equivalent to



µ ∈ −∞, (λ/(1 + λ))2 ∪ 1 + 1/(1 + λ)2 , ∞ .
From the above discussion, we immediately have
Corollary 2.5. Let f ∈ U(λ) be of the form (1.1) and µ be a real number. Then the following
sharp inequality


|a3 − µa22 | ≤ |(1 + λ + λ2 ) − µ(1 + λ)2 | for µ ∈ −∞, (λ/(1 + λ))2 ∪ (1, ∞) .
holds and equality attains for the function kλ .

3. Properties of meromorphic functions associated with U(λ)

b : |ζ| > 1} of
Let Σ be the class of meromorphic and univalent functions g on ∆ := {ζ ∈ C
the form
∞
X
(3.1) g(ζ) = ζ + bn ζ −n , 1 < |ζ| < ∞.
n=0
 On a class of univalent functions defined by a differential inequality 9

It is well recognized that the set Σ plays an important role in the study of the class S. The
map f 7→ g(ζ) = 1/f (1/ζ) gives a one-to-one corresponds between the class S and the class
{g ∈ Σ : g(z) 6= 0}. Moreover, we have the following formula
 2
′ z 1
g (ζ) = f ′ (z), ζ = ,
f (z) z
and hence, functions f in U(λ) are associated with the functions g in M(λ) := {g ∈ Σ :
|g ′(ζ) − 1| < λ}. Further, let
Σ0 := {g ∈ Σ : b0 (g) = 0} and M0 (λ) := {g ∈ M(λ) : b0 (g) = 0}.
It is important to note that the classes Σ and M(λ) are not compact. However, the classes
Σ0 and M0 (λ) are compact with respect to the topology of uniform convergence. The classes
M(λ) and M0 (λ) have been extensively studied by Vasudevarao and Yanagihara [21]. For
b \ g(∆). In [21], it has been proved that
g ∈ Σ, let E(g) be the omitted set of g, i.e., E(g) = C
for 0 < λ ≤ 1 and g ∈ M(λ),
∞
!
X
π(1 − λ2 ) ≤ area (E(g)) = π 1 − n|bn |2 ≤ π.
n=1

From this it can be seen that

∞
X
(3.2) n|bn |2 ≤ λ2
n=1

and equality holds in (3.2) for a function g ∈ M(λ) if, and only if, area (E(g)) = π(1 − λ2 ).
In 1955, Springer [20] proved that the set of extreme points of Σ0 contains all functions
g ∈ Σ0 with area (E(g)) = 0. We note that the classes Σ and M(λ) have no extreme points.
Our next result gives a sufficient condition for a function g to be an extreme point of M0 (λ).
Theorem 3.1. Let g ∈ M0 (λ) with area (E(g)) = π(1 − λ2 ). Then g is an extreme point of
M0 (λ).
Proof. Suppose that a function g ∈ M0 (λ) with area (E(g)) = π(1 − λ2 ) has a representation
g = tg1 + (1 − t)g2 , 0 < t < 1, as a proper convex combination of two functions g1 , g2 ∈ M0 (λ).
Let P
g(ζ) be of the form (3.1) andPg1 (z) and g2 (z) have the representation of the form g1 (ζ) =
ζ+ ∞ n=0 cn ζ
−n
and g2 (ζ) = ζ + ∞ n=0 dn ζ
−n
. Then the case of equality in (3.2) gives
∞
X ∞
X
2 2
λ = n|bn | = n|tcn + (1 − t)dn |2
n=1 n=1
X∞

= n t|cn |2 + (1 − t)|dn |2 − t(1 − t)|cn − dn |2
n=1
∞
X
≤ tλ2 + (1 − t)λ2 − t(1 − t) n|cn − dn |2 .
n=1

Since 0 < t < 1, it follows that

∞
X
n|cn − dn |2 ≤ 0
n=1
which implies cn = dn for n = 1, 2, . . .. In other words, g1 = g2 and hence, g is an extreme
point of M0 (λ). 
Acknowledgement: The second author thank SERB, Govt of India for the support.
10 Md Firoz Ali, Vasudevarao Allu and Hiroshi Yanagihara

References
[1] Y. Abu-Muhanna and D.J. Hallenbeck, A class of analytic functions with integral representations,
Complex Var. Theory Appl. 19 (1992) 271–278.
[2] L.A. Aksentév, Sufficient conditions for univalence of regular functions (Russian), Izv. Vysš. Učebn.
Zaved. Matematika 1958 (1958), 3–7.
[3] L.A. Aksentév and I. R. Nezhmetdinov, Sufficient conditions for univalence of certain integral repre-
sentations (Russian), Trudy Sem. Kraev. Zadacham 18 (1982), 3–11.
[4] A. Baernstein, Integral means, univalent functions and circular symmetrization, Acta Math. 133 (1974),
139–169.
[5] J. Becker and Ch. Pommerenke, Schlichtheitskriterien und Jordangebiete, J. Reine Angew. Math. 354
(1984), 74–94.
[6] P. L. Duren, Univalent functions (Grundlehren der mathematischen Wissenschaften 259, New York,
Berlin, Heidelberg, Tokyo), Springer-Verlag, 1983.
[7] R. Fournier and S. Ponnusamy, A class of locally univalent functions defined by a differential inequality,
Complex Var. Elliptic Equ. 52 (2007), 1–8.
[8] D. J. Hallenbeck and T. H. MacGregor, Linear problem and convexity techniques in geometric
function theory, Pitman, 1984.
[9] Y. J. Kim and E. P. Merkes, On an integral of powers of a spirallike function, Kyungpook Math. J. 12
(1972), 249–252.
[10] Y. C. Kim, S. Ponnusamy and T. Sugawa, Geometric properties of nonlinear integral transforms of
certain analytic functions, Proc. Japan Acad. Ser. A Math. Sci. 80 (2004), 57–60.
[11] Y.J. Leung, Integral means of the derivatives of some univalent functions, Bull. London Math. Soc. 11(3)
(1979), 289–294.
[12] R.J. Libera and E.J. Złotkiewicz, Early coefficients of the inverse of a regular convex function, Proc.
Amer. Math. Soc. 85 (2) (1982) 225–230.
[13] T. H. MacGregor, Applications of extreme-point theory to univalent functions, Michigan Math. J. 19
(1972), 361–376.
[14] M. Obradović, Starlikeness and certain class of rational functions, Math. Nachr. 175 (1995), 263–268.
[15] M. Obradović and S. Ponnusamy, New criteria and distortion theorems for univalent functions, Complex
Variables Theory Appl. 44(3) (2001), 173–191.
[16] M. Obradović, S. Ponnusamy and K.-J. Wirths, Geometric Studies on the Class U(λ), Bull. Malays.
Math. Sci. Soc. 39(3) (2016), 1259–1284.
[17] Ch. Pommerenke, Univalent functions, Vandenhoeck and Ruprecht, Göttingen, 1975.
[18] V. Singh, Univalent functions with bounded derivative in the unit disc, Indian J. Pure Appl. Math. 8(11)
(1977), 1370–1377.
[19] V. Singh and P. N. Chichra, An extension of Becker’s criterion of univalence, J. Indian Math. Soc. 41
(1977), 353–361.
[20] G. Springer, Extreme Punkte der konvexen Hülle schlichter Funktionen, Math. Ann. 129 (1955), 230–232.
[21] A. Vasudevarao and H. Yanagihara, On the growth of analytic functions in the class U(λ), Comput.
Methods Funct. Theory 13(4) (2013), 613–634.
[22] S. Yamashita, Almost locally univalent functions, Monatsh. Math. 81 (1976), 235–240.

Md Firoz Ali, Department of Mathematics, National Institute of Technology Calicut,

Calicut- 673601, Kerala, India.
E-mail address: [email protected], [email protected]

Vasudevarao Allu, Discipline of Mathematics, School of Basic Sciences, Indian Institute of

Technology Bhubaneswar, Argul, Bhubaneswar, PIN-752050, Odisha (State), India.
E-mail address: [email protected]

Department of Applied Science, Faculty of Engineering, Yamaguchi University, Tokiwadai,

Ube 755, Japan
E-mail address: [email protected]