J. Austral. Math. Soc.

(Series A) 35 (1983), 1-17

A SUBCLASS OF UNIVALENT FUNCTIONS

R. M. GOEL and BEANT SINGH MEHROK

(Received 23 December 1980, revised 19 June 1981)

Communicated by R. Vyborny

Abstract

Sharp results for the coefficient estimates, distortion theorems, radius of convexity, arc-length and
area of the image curve are obtained for the class R(A, B) of regular functions whose derivative is
subordinate to (1 + Az)/(\ + Bz), - 1 *C B < A < 1, in the unit disc E = {z: | z | < 1}. We also
establish a convolution theorem for this class.

1980 Mathematics subject classification (Amer. Math. Soc): 30 A 32, 30 A 34, 30 A 42.

Keywords and phrases: quasi-subordination, subordination, majorization, subclass of univalent func-

tions, convolution (Hadamard product).

1. Introduction

Let U denote the class of functions

(1.1) w(z)=5c/
k=\

which are regular inE — (z: \z\< 1} and satisfying there the conditions w (0) = 0
and | w(z) \< 1.
Let S denote the class of functions

(1.2) f(z) = z+ i
k= 2

regular and univalent in E.

© Copyright Australian Mathematical Society 1983

1

[Link] Published online by Cambridge University Press

 2 R. M. Goel and Beant Singh Mehrok [21

Let R(A, B) denote the class of functions f(z) — z + 2J? =2 akzk which are
regular in E and satisfying there

(1.3) f f(
( zzL±J£
) < t -KB<A<\,zGE.

Obviously / ? ( + l , - l ) coincides with R, the class of functions f(z) — z +

2J°=2 akzk regular in E and satisfying Re/'(z) > 0, z G E. Thus /?(^, B) is a
subclass of R(\, -1). To avoid repetition we lay down, once for all, that -1 < B <
A^l, and z e E.
Let us set

(1.4) f'(z) = P(z) = I + ^pkzk.

k=\
Then b y definition of subordination, / e R(A, B) if and only if f\z) has the
representation

An easy computation shows that/ G /?(v4, B) if and only if

(1.6) \f'{z)-\\<\A-Bf'{z)\.
Alexander [1] and Wolff [17] made an early study of the class R. It follows
from the Noshiro-Warschawski theorems [12, 16] that functions of the class R are
equivalent in E. Hence R(A, B) is a subclass of S.
MacGregor [9] investigated the properties of the class R, and subsequently, the
same author [10] studied the subclass R(l) of R of regular functions /(z)
satisfying the condition
(1.7) |/'(z)-l|<l.
The first author [5, 6] developed some properties of the subclass S(a) of R of
regular functions/(z) which satisfy the condition
(1.8) |/'(z)-«|<«, («>*).
Padmanabhan [13] investigated the subclass R(a) of R of regular functions/(z)
satisfying
f'(z) - 1
(1.9) <a, 0 <a < 1.
f'(z) + 1
Capling and Causey [4] also studied the class R(a) and improved some of the
results due to Padmanabhan [13].
The following observations are obvious:

[Link] Published online by Cambridge University Press

 [3] A subclass of univalent functions 3

(iii)/t(l,l/a-l) = S(a),(a>i),
(iv) R(a, -a) = R(a), (0 < a < 1).
Thus, R(A, B) contains all the above mentioned classes, and therefore the view
of Brickman [3, page 341], "idea of subordination has unified the geometric
theory of functions" is strengthened.
In this paper, we obtain sharp result for coefficient estimates, distortion
theorems, radius of convexity, arc-length and area of the image curve for the class
R(A, B). We also prove that if
00 00

/(*) = * + 2aHz\ h(z)=z+ 2bnz"

n=2 n=2

belong to R(A, B), then so does F(z) = z + ^2~ = 2 nanKz"-

Results due to MacGregor [9, 10], Padmanabhan [13], Capling and Causey [4]
and the first author [5, 6] follow as special cases from our theorems.

2. Some preliminary lemmas

LEMMA \. If g(z) and G(z) are regular in | z |< 1 and g(z) is subordinate to G(z)
(g(z) < G(z)) with g(0) = G(0), then for X > 0, 0 < r < 1,
(2.1) f \J ( \ [
J
o o
This lemma is due to Littlewood and its proof can be found in [7, page 484,
Theorem 2; 8(1944), Theorem 210].
Robertson [14] introduced the concept of quasi-subordination. Let g(z) and
G(z) be analytic in E. Let <f>(z) be analytic and |<|>(z)|< 1 in E, such that
g(z)/<j>(z) is regular and subordinate to G(z), for z e E. Then g(z) is said to be
quasi-subordinate to G(z), written as g(z) <q G(z), z BE.
An equivalent condition for this is
g(z)=*(z)G(w(z)), \<t>(z)\< l,wGU,zEE.
If <j>(z) = 1, then g(z) = G(w(z)) so that g(z) < G(z) in E. If w(z) - z, then
g(z) = <f>(z)G(z), we say that g(z) is majorized by G(z) and we write it as
g(z) « G(z), z e E.

LEMMA 2. Ifg(z) = l^=odkzk <qG{z) = l?=0Dkzk, then

(2.2) 2 K I 2 < 2 I'M 2 (« = 0,l,2,...).

k=0 k=0

This lemma is due to Robertson [14].

[Link] Published online by Cambridge University Press

 4 R. M. Goel and Beant Singh Mehrok [41

In particular (2.2) holds also when

(i) g(z) < G(z),
(n)g(z) «G(z).
By the application of Lemma 2, we establish

LEMMA 3. Forf e R(A, B), iff'(z) = P(z) = 1 + lf=lpkzk, then

(2.3) \pn\*(A-B), n>\.
The bounds are sharp.

PROOF. From (1.5) we have

,k

k=\ L k=\
By the application of Lemma 2(2.2), we get
n n 1

k=\ k=\

or

\Pn |2 < {A - Bf - (1 - 2? 2 )V \pk |2 < {A - B)\

k=\

This yields (2.3).

Equality signs in (2.3) are attained for the functions Pn(z) defined by

_, , 1 +ASzn , „, ,

In order to determine the radius of convexity, we need the following two

lemmas.

LEMMA 4. For w G Uand \z\= r,we have

2
-'w(z)|2
(2.4) \zw'(z)-w(z)\<^—

This result is due to Singh and Goel proved in [15].

LEMMA 5. Let
_,_,_! +Bw(z)

[Link] Published online by Cambridge University Press

 [s] A subclass of univalent functions

Then for \z\- r < 1,

(\~r2)\p(z)\

AB(A + g)r 2
-Br)
2 2
[(1 -ABr )-({\ -A){\ - B)(\ +Ar2)(\ + Br2))l/2],
R} >R0,A

where
Oft R -LzlL ni - (1 ~ *)(l + Br2)
(2.6) * . - T ^ . *°-(l
The bounds are sharp.

PROOF. It is easy to see that the transformation

1 + Bw{z)

maps | w{z) |< r onto the circle \p(z) — a |< </, where

fl^
(l-A r ) 2 2
and d =l
and
(l-A2r2)
d

Putting ^(z) = i?e'* (- f < 0 < f) and denoting the left hand side of (2.5) by

B/R)cos6+ ^—

(1 -r2) (1 -r2)R "

For extreme values of T(R, 6), dT/dR - 0 = 97/30 which yield respectively

~ B/R2)

and
(2.8)

[Link] Published online by Cambridge University Press

 6 R. M. Goel and Beant Singh Mehrok [6]

where

1{
(2.9) L(R) = (AR + B/R) + )~ ^ .
(1 - r1)
Now we prove that L(R) remains positive. If B > 0, A > 0, then L{R) > 0.
We now consider the case when B < 0. The following cases arise.

Case I. B < 0, A > 0. Using the fact that 0 < cos 0 < 1, it follows from (2.7)
that

L'(R) = (A — B/R2) > 0 and hence L(R) attains its minimum value at
\ >/2
( 1 _
1
~
1—A r I
D2_2
= *2, say.

Now

L(R2) =[(1 - r 2 ) ( ^ + B) + 2(1 - /l 2 r 2 ) 1 / 2 (l - B2r2)x/2\

X
(1 r 2 )[(l - A2r2)(l -B2r2)]V2
which is positive provided

(2.12) (1 - r2)(A +B) + 2((1 - A2r2){\ - B2r2))X/2 > 0.

If (A + B) > 0, there is nothing to prove, so we assume (A + B) < 0. (2.12) will

hold if
4(1 - A2r2){\ - B2r2) - (A + B)2(\ - r2)2 > 0
or if
[(1 + B)(l +Ar2){\ +A)(\ +Br2)][2(\ + ABr2) - (A + B)(l + r2)] > 0
which is always true.

Case II. B < 0, A < 0. Consider the case when L\R) = (A - B/R2) < 0.
Since 0 < cos 6 < 1, it follows from (2.7) that
(1 - B){\ + Br2) ^ R 2 < \-B2r2
2
+Ar ) "" 1 - A2r2
An easy computation would show that this does not hold.

[Link] Published online by Cambridge University Press

 [7 ] A subclass of univalent functions 7

Now, consider the case when (A - B/R2) s* 0. If A - B/R2 = 0, then from

(2.7), we have

(\-A2r2) A
2
and it implies (A — B)(l + ABr ) = 0 which is evidently not possible. Thus the
only case needed to be considered is when (A — B/R2) > 0. Therefore, by (2.10),

T=1$<R'<B/A-
The minimum value of L(R) occurs at R = R2 and L(R2) > 0 if
(1 - r2)(A +B) + 2[(1 - A2r2){\ - B2r2)]i/2 >0
which holds as proved in Case I (when A + B is negative). For extreme values,
from (2.8) and (2.7), we get

\- Br2) _ D 2
(1 - ^ ) ( 1 + ^ r 2 )

It can be easily verified that T(R,6) attains its maximum value at (0 = 0,

R = Ro). So
T(R,6)<T(R0,0)

[(1 - ABr2) - ((1 - ,0(1 - 20(1 + Ar2)(\ + Br2))'/2}.

(1-r2)
It is easy to see that Ro > a - d — (1 + Br)/{\ + Ar). But Ro is not always
less than or equal to a + d. In case Ro f£ [a = d, a + d], the maximum of
T(R,Q) is attained at

and equals
B)r2-4ABr+(A+B)

If /?, < Ro, equality sign in (2.5) holds for the function
] + B z
i \

If /?, > Ro, A =£ 1, equality sign in (2.5) holds for the function
_ 1 - (1 + B)zcos 6 + Bz2
1 - (1 +A)zcos6 + Az2

[Link] Published online by Cambridge University Press

 8 R. M. Goel and Beant Singh Mehrok [8]

where
(2 13) R = l " ^ +B)rcos$ + Br2
0
1 - (1 + A)rcos 0 + Ar2'
Hence the lemma is established.

3. Coefficient estimates

THEOREM 3.1. Letf £ R(A, B) then

(3.1) |flJ<(l_L!), n > 2 .

The bounds are sharp for the functions f(n_ X){z) defined by

(3-2) f(H-Jz) ~-

PROOF. (3.1) follows on equating the coefficients of z" in (1.4) and then using
(2.3).

THEOREM 3.2. / / / £ R(A, B) and if \i is a complex number, then

(3.3) \a,

The estimate is sharp.

PROOF. On equating the coefficients of z2 and z3 in (1.5), we get

2a2
(3.4) c, =
(A-B)'-
3 f , AB
(3 5) 2
' - (A-B)P 3{A-B)"2[
Also

Therefore, for every complex number v, we have

<l+(lvl-l)\C]l2
< max{l, | « |},

[Link] Published online by Cambridge University Press

 19] A subclass of univalent functions

since | c, |< 1. The estimate (3.6) is sharp for w(z) = z and w(z) = z2 respectively
for | v | < 1 and | v \ < 1. From (3.4) and (3.5) we have

(3-7) ,fl3-,fli,= ( l ^ l c-, — v>c\

3
where
4 (v-B)
3(A-B)
or
3(A -
(3-8) v=B+
(3.7) in conjunction with (3.6) and (3.7), yields (3.3). (3.3) is sharp, being attained
for the function/,(z) and/ 2 (z) defined, respectively, by

1 + Az 1 +Az2
and

4. Distortion theorems

THEOREM 4.1. LetfG R(A, B), then for \ z |= r < 1,

(4-1) I/'(Z)I<TT^;
1 -^r
(4.2) Re/'(z)
1 - Br'

(4.3)
= 0,

(4-4)
r — —r = 0.
2 '

All the estimates are sharp.

PROOF. From (1.5), it is easy to establish (4.1) and (4.2). Using (4.1),
l + At ,

[Link] Published online by Cambridge University Press

 10 R. M. Goel and Beant Singh Mehrok

which yields (4.3). Again using (4.2),

|/(z) |> f Re f'(te^) dt > \r]^~dt

which gives (4.4). (4.1) and (4.3) are sharp, being attained for the function

/.(*) =
z+jz2, 5 = 0.

(4.2) and (4.4) are sharp for the function

A 2
5 = 0.

Let W be any complex number such that

W\<
= 0.

By Rouche's Theorem it follows that/(z) a n d / ( z ) — W have the same number

of zeros in | z | < r, that is, precisely one. Hence we have the following:

COROLLARY. Every function f(z) in R(A, B) maps E onto a domain which covers
the disc

W\<
= 0.

5. Argument of / ' ( z )

THEOREM 5.1. If f £ R(A, 5) then

(5.1) |arg/'(z)|<sin- z = r.
1 - ABr2 '
The result is sharp.

[Link] Published online by Cambridge University Press

 [11] A subclass of univalent functions 11

PROOF. It is easy to show that/'(z) = (1 + Aw(z))/(\ + Bw(z)) maps | w(z) |

r onto the circle
(1 -ABr2) (A-B)r
(5.2)
- B2r2 (1 -B2r2)
(5.1) is an immediate consequence of (5.2). The result is sharp, being attained for
the function fo(z) defined by
,„ s 1 + A8z
(5.3)
1 + BSz '
where
- (A + B)r - A2r2)(\ -
8= - 2
r 1 + ABr

6. Convex set of functions

THEOREM 6.1. Iffandh e R(A, B), then

Xf+(\-X)h<=R(A,B), (O

PROOF. By definition,

(6.1) f'(z 1 + Az
1 + Bz'
1 +Az
(6-2) *'<*-,+&•
Since (1 + Az)/{\ + Bz) is convex univalent in E, it follows by a result due to
Bernardi [2, page 57, Example 2] that

Xf'(z) + (I - X)h'(z) < \

Hence
Xf + (1 -X)h' GR(A,B).

7. Radius of convexity

THEOREM 7.1. Letf e R(A, B), then

(i) for Ao < A < l , / ( z ) is convex in \z\< r0,
where r0 is the smallest positive root of
(7.1) ABr2 - 2Ar + 1 = 0;

[Link] Published online by Cambridge University Press

 12 R. M. G o d and Beant Singh Mehrok [ 12 J

(ii) for - 1 < A < A0,f(z) is convex in | z \ < r,, where rx is the smallest positive
root of
(7.2)

A{\ - r2) - [(1 - ^5r 2 ) - ((1 - ,4)(1 - B){\ + Ar2){\ + Br2))V2] = 0;

_ (2 + B - IB2) + (20 - 36B + 21ff2 - 4B3)i/2
°~~ 2(4 - IB - 5 2 )

77ie results are sharp.

PROOF. Differentiating logarithmically, (1.5) yields

(7.3) together with Lemma 4 gives

(7.4)

-W '*-\«')t
Aw(z))(l + Bw{z)) (1 - r2) | (1 + Aw(z))(\ + Bw(z))

Puttingp(z) — (1 + Bw(z))/(1 + Aw(z)), and using Lemma 5,

2A _ AB(A + B)r2 - 4ABr + (A + B) <

^ - F (v4 - 5 ) ( 1 - ^ r ) ( l - 5 r ) ' '^ °'
I - ((1 - A)(l - B)(\ + Ar ){\ + Br2))l/2
2
- 2
.4 - (l-r2)(A-B)
R{>R0,A*
(7.1) and (7.2) follow by equating the right hand sides of (7.5) to zero.
The equation Ro = Rt yields

(7.6) ABr4 - 2ABr3 + [2(A + B) - AB - l] r2 - 2r + 1 = 0.

Elimination of r between (7.1) and (7.6) leads to

(7.7) (A-IB- B2)A2 - (2 + B - 2B2)A - (l - B)1 = 0.

[Link] Published online by Cambridge University Press

 [131 A subclass of univalent functions 13

(7.7), on verification of the signs, yields

(2 + B - IB2) + (20 - 3 6 5 + 2 1 5 2 - 4 5 3 ) 1 / 2
A = ^—7-^ ; — = AQ, say.
2(4-2B-B2) °
The results are sharp, being attained respectively, for the functions/,(z) and/ 9 (z)
defined by

,,, x 1 + Az ,,, > I — (I + A)zcos0 + Az2

*«.\- \+Bz' fi\z)=-1 - (1 + B)zcosO + Bz2'

where 6 is defined by (2.13).

REMARK 1. Radii of convexity for the classes R, R(l) and S(a), at once, follow
from (7.1).
4 3
REMARK 2. On taking A = a, B = -a (0 < a *£ 1) in (7.7), we get a - 4a -
2
4a + 4a + 1 = 0 which gives

say.
Hence
(i) for a0 < a < l,/(z) maps \z\< (2 1/2 — l ) / a onto a convex domain;
(ii) for 0 < a < ao,f(z) maps

(a2-
z <
2a(l + a )

onto a convex domain. This result was established by Padmanabhan in [13] and
also by Capling and Causey in [4].

8. Arc-length and area of the image curve

THEOREM 8.1. Let / £ R(A, B) and Lr(f) denotes the length of the image of
| z | = r under f(z), 0 < r < 1, then
(8-1)
A+B (A-B)
•nr log
B 151 1 + Br
Lr(f)
i9
+Are \d0, = Q.

The results are sharp.

[Link] Published online by Cambridge University Press

 14 R. M. Goel and Beant Singh Mehrok [14]

PROOF. In Lemma 1, set g(z) = f'(z), G(z) = (1 + Az)/(l + Bz) and X = 1.

Then

(8.2) 1 + Are1"
d6.
•'o 1 + Brew
Now

z\ = r

By (8.2),
1 + Are'1
dO
1 + Bre'»
A+B _ {A - B){\ - B2r2)
2B 2B(l +2Br cos 0 + B2r
dO
,. (A - B)rsin6
1 + 2BrcosO + B2r2
A+B (A - B)r ri-n (l - B2r2)
dO
B 2|5| ^o 1 +2BrcosO + B2r2

(A-B)r
'o 1 +2Brco$6 + B2r2
A + B TT(A - B)r _ (A ~- B)r p_
— irr 2 2 d6
B B Jo i 2Brcos6 + B r
A + B (A-B) AA-B)
— mr
B
For 5 = 0, result is trivial.
The extremal function/0(z) is given by

1 +A8z
(8.3) 181= 1.
1 + B8z '

COROLLARY. For the class R(a), we deduce, from (8.1),

This is a result established by Capling and Causey [4].

[Link] Published online by Cambridge University Press

 [IS] A subclass of univalent functions 15

T H E O R E M 8.2. / / / £ R(A, B), and if Ar(f) denotes the area of image of\z\=r
under f(z), 0 < r < 1, then
(8.4)

irr
A,(f)
A2r2
mr' 1+ = 0.

The inequalities are sharp.

(8.4) are direct consequences of Lemma 1 (X = 2) and interior area theorem.

Equality sign is attained for the function fo(z) defined by (8.3).
COROLLARY. For the class R(a), we have from (8.4),

This is a result due to Capling and Causey [4].

9. Convolution

THEOREM 9.1. If f(z) = z + 2" = 2 anz" and h(z) = z + 2~ =2 bnz" belong to the
class R(A, B), then so does

PROOF. Since/ e R(A, B), it follows by (1.6) that \f'(z) - 1 | < | A - Bf'(z) | .
It is equivalent to
(9-1) \f'{z)-b\<C
where b = (1 - AB)/{\ - B ), C = (A - B)/(\ - B2). It is easy to see that
2

1 - b< C < b. We know that if H(z) = l^=Qhnzn is regular for | z |< 1 and
| H(z)\*z M, then, by [11, page 101],

(9.2) 2 \hn\2<M2.
Applying (9.2) to (9.1), we get (1 - b)2 + 2 ~ = 2 n2 \an\2 < C2 or

(9.3) I •'!«.!'< (i>"J)1.

[Link] Published online by Cambridge University Press

 16 R. M. God and Beant Singh Mehrok [16]

Similarly

(9.4)

Now

\F'(z)-b\2 =
n=2

n-\ !
2 n2anbnz"-

l/2 l/2

(by Cauchy-Schwarz inequality)

1/2 1/2

n= 2
, / oo oo

7 2«2I«.

+ 1.
(1-B2)3 4
(l-52)4
[using (9.3) and (9.4)]
2 2 3
B (A-B) , g(^t-,g) 1 (^-ff)4
+
,3 4

B2(A-B)2 | 1
|
4

This gives on simplification, (A + B) < 2 which is true. Hence F E R(A, B).

REMARK. The first author [6] proved this theorem for the class S{a).
Acknowledgement

The authors would like to thank the referee for a number of suggestions.

[Link] Published online by Cambridge University Press

 117] A subclass of univalent functions 17

References

[1] J. W. Alexander, 'Functions which map the interior of the unit circle upon simple regions',
Ann. of Math. 17(1915), 12-22.
[2] S. D. Bernardi, 'Special classes of subordinate functions', Duke Math.J. 33 (1966), 55-67.
[3] L. Brickman, 'Subordinate families of analytic functions', Illinois J. Math. 15 (1971), 241-248.
[4] T. R. Capling and W. M. Causey, 'A class of univalent functions', Proc. Amer. Math. Soc. 39
(1973), 357-361.
[5] R. M. Goel, 'A class of univalent functions whose derivatives have positive real part in the unit
disc', Nieuw Arch. Wisk. (3), 15 (1967), 55-63.
[6] R. M. Goel, 'A class of analytic functions whose derivatives have positive real part in the unit
disc', Indian J. Math. 3 (1971), 141-145.
[7] J. E. Littlewood, 'On inequalities in the theory of functions', Proc. London Math. Soc. (2), 23
(1925), 481-519.
[8] J. E. Littlewood, Lectures on the theory of functions, pp. 163-185 (Oxford Univ. Press, London,
1944).
[9] T. H. MacGregor, 'Functions whose derivative has a positive real part', Trans. Amer. Math.
Soc. 104 (1962), 532-537.
[10] T. M. MacGregor, 'A class of univalent functions', Proc. Amer. Math. Soc. 15(1964), 311-317.
[11] Z. Nehari, Conformal mapping (McGraw Hill, New York, 1952).
[12] K. Noshiro, 'On the theory of schlicht functions', J. Fac. Sci. Hokkaido Univ. (1), 2 (1934-35),
129-155.
[13] K. S. Padmanabhan, 'On a certain class of functions whose derivatives have a positive real part
in the unit disc', Ann. Polon. Math. 23 (1970), 73-81.
[14] M. S. Robertson, 'Quasi-subordinate functions', Mathematical essays dedicated to A. J.
Maclntyre, pp. 311 -330 (Ohio Univ. Press, Athens, Ohio, 1967).
[15] V. Singh and R. M. Goel, 'On radii of convexity and starlikeness of some classes of functions',
J. Math. Soc. Japan 23 (1971), 323-339.
[16] S. S. Warschawaki, 'On the higher derivatives at the boundary in conformal mappings', Trans.
Amer. Math. Soc. 8(1935), 310-340.
[17] J. Wolff, 'L'integrale d'une fonction holomorphe et a partie reelle positive dans un demiplan est
univalente', C. R. Acad. Sci. Paris Ser A-B 198 (1934), 1209-1210.

Department of Mathematics
Panjabi University
Patiala-147002
(Panjab State)
India

[Link] Published online by Cambridge University Press