Journal of Inequalities and Applications

, 2013:281

Bounds for the second Hankel determinant of certain univalent functions

  • See Keong Lee
  • V Ravichandran
  • Shamani Supramaniam
Open AccessResearch

DOI: 10.1186/1029-242X-2013-281

Cite this article as:
Lee, S.K., Ravichandran, V. & Supramaniam, S. J Inequal Appl (2013) 2013: 281. doi:10.1186/1029-242X-2013-281
Part of the following topical collections:
  1. Proceedings of the International Congress in Honour of Professor Hari M. Srivastava

Abstract

The estimates for the second Hankel determinant a 2 a 4 a 3 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq1_HTML.gif of the analytic function f ( z ) = z + a 2 z 2 + a 3 z 3 + https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq2_HTML.gif , for which either z f ( z ) / f ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq3_HTML.gif or 1 + z f ( z ) / f ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq4_HTML.gif is subordinate to a certain analytic function, are investigated. The estimates for the Hankel determinant for two other classes are also obtained. In particular, the estimates for the Hankel determinant of strongly starlike, parabolic starlike and lemniscate starlike functions are obtained.

MSC:30C45, 30C80.

Dedication

Dedicated to Professor Hari M Srivastava

1 Introduction

Let A https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq5_HTML.gif denote the class of all analytic functions
f ( z ) = z + a 2 z 2 + a 3 z 3 + https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ1_HTML.gif
(1)
defined on the open unit disk D : = { z C : | z | < 1 } https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq6_HTML.gif. The Hankel determinants H q ( n ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq7_HTML.gif ( n = 1 , 2 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq8_HTML.gif , q = 1 , 2 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq9_HTML.gif) of the function f are defined by
H q ( n ) : = [ a n a n + 1 a n + q 1 a n + 1 a n + 2 a n + q a n + q 1 a n + q a n + 2 q 2 ] ( a 1 = 1 ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equa_HTML.gif
Hankel determinants are useful, for example, in showing that a function of bounded characteristic in D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif, i.e., a function which is a ratio of two bounded analytic functions with its Laurent series around the origin having integral coefficients, is rational [1]. For the use of Hankel determinants in the study of meromorphic functions, see [2], and various properties of these determinants can be found in [[3], Chapter 4]. In 1966, Pommerenke [4] investigated the Hankel determinant of areally mean p-valent functions, univalent functions as well as of starlike functions. In [5], he proved that the Hankel determinants of univalent functions satisfy
| H q ( n ) | < K n ( 1 2 + β ) q + 3 2 ( n = 1 , 2 , , q = 2 , 3 , ) , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equb_HTML.gif

where β > 1 / 4000 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq11_HTML.gif and K depends only on q. Later, Hayman [6] proved that | H 2 ( n ) | < A n 1 / 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq12_HTML.gif ( n = 1 , 2 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq8_HTML.gif ; A an absolute constant) for areally mean univalent functions. In [79], the estimates for the Hankel determinant of areally mean p-valent functions were investigated. ElHosh obtained bounds for Hankel determinants of univalent functions with a positive Hayman index α [10] and of k-fold symmetric and close-to-convex functions [11]. For bounds on the Hankel determinants of close-to-convex functions, see [1214]. Noor studied the Hankel determinant of Bazilevic functions in [15] and of functions with bounded boundary rotation in [1619]. In the recent years, several authors have investigated bounds for the Hankel determinant of functions belonging to various subclasses of univalent and multivalent functions [2027]. The Hankel determinant H 2 ( 1 ) = a 3 a 2 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq13_HTML.gif is the well-known Fekete-Szegö functional. For results related to this functional, see [28, 29]. The second Hankel determinant H 2 ( 2 ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq14_HTML.gif is given by H 2 ( 2 ) = a 2 a 4 a 3 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq15_HTML.gif.

An analytic function f is subordinate to an analytic function g, written f ( z ) g ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq16_HTML.gif, if there is an analytic function w : D D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq17_HTML.gif with w ( 0 ) = 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq18_HTML.gif satisfying f ( z ) = g ( w ( z ) ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq19_HTML.gif. Ma and Minda [30] unified various subclasses of starlike ( S https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq20_HTML.gif) and convex functions ( C https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq21_HTML.gif) by requiring that either of the quantity z f ( z ) / f ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq22_HTML.gif or 1 + z f ( z ) / f ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq23_HTML.gif is subordinate to a function φ with a positive real part in the unit disk D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif, φ ( 0 ) = 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq24_HTML.gif, φ ( 0 ) > 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq25_HTML.gif, φ maps D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif onto a region starlike with respect to 1 and symmetric with respect to the real axis. He obtained distortion, growth and covering estimates as well as bounds for the initial coefficients of the unified classes.

The bounds for the second Hankel determinant H 2 ( 2 ) = a 2 a 4 a 3 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq26_HTML.gif are obtained for functions belonging to these subclasses of Ma-Minda starlike and convex functions in Section 2. In Section 3, the problem is investigated for two other related classes defined by subordination. In proving our results, we do not assume the univalence or starlikeness of φ as they were required only in obtaining the distortion, growth estimates and the convolution theorems. The classes introduced by subordination naturally include several well-known classes of univalent functions and the results for some of these special classes are indicated as corollaries.

Let P https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq27_HTML.gif be the class of functions with positive real part consisting of all analytic functions p : D C https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq28_HTML.gif satisfying p ( 0 ) = 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq29_HTML.gif and Re p ( z ) > 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq30_HTML.gif. We need the following results about the functions belonging to the class P https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq27_HTML.gif.

Lemma 1 [31]

If the function p P https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq31_HTML.gifis given by the series
p ( z ) = 1 + c 1 z + c 2 z 2 + c 3 z 3 + , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ2_HTML.gif
(2)
then the following sharp estimate holds:
| c n | 2 ( n = 1 , 2 , ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ3_HTML.gif
(3)

Lemma 2 [32]

If the function p P https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq31_HTML.gifis given by the series (2), then
2 c 2 = c 1 2 + x ( 4 c 1 2 ) , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ4_HTML.gif
(4)
4 c 3 = c 1 3 + 2 ( 4 c 1 2 ) c 1 x c 1 ( 4 c 1 2 ) x 2 + 2 ( 4 c 1 2 ) ( 1 | x | 2 ) z , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ5_HTML.gif
(5)

for somex, zwith | x | 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq32_HTML.gifand | z | 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq33_HTML.gif.

2 Second Hankel determinant of Ma-Minda starlike/convex functions

Subclasses of starlike functions are characterized by the quantity z f ( z ) / f ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq3_HTML.gif lying in some domain in the right half-plane. For example, f is strongly starlike of order β if z f ( z ) / f ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq3_HTML.gif lies in a sector | arg w | < β π / 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq34_HTML.gif, while it is starlike of order α if z f ( z ) / f ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq3_HTML.gif lies in the half-plane Re w > α https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq35_HTML.gif. The various subclasses of starlike functions were unified by subordination in [30]. The following definition of the class of Ma-Minda starlike functions is the same as the one in [30] except for the omission of starlikeness assumption of φ.

Definition 1 Let φ : D C https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq36_HTML.gif be analytic, and let the Maclaurin series of φ be given by
φ ( z ) = 1 + B 1 z + B 2 z 2 + B 3 z 3 + ( B 1 , B 2 R , B 1 > 0 ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ6_HTML.gif
(6)
The class S ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq37_HTML.gif of Ma-Minda starlike functions with respect toφ consists of functions f A https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq38_HTML.gif satisfying the subordination
z f ( z ) f ( z ) φ ( z ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equc_HTML.gif
For the function φ given by φ α ( z ) : = ( 1 + ( 1 2 α ) z ) / ( 1 z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq39_HTML.gif , 0 < α 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq40_HTML.gif, the class S ( α ) : = S ( φ α ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq41_HTML.gif is the well-known class of starlike functions of order α. Let
φ PAR ( z ) : = 1 + 2 π 2 ( log 1 + z 1 z ) 2 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equd_HTML.gif
Then the class
S P : = S ( φ PAR ) = { f A : Re ( z f ( z ) f ( z ) ) > | z f ( z ) f ( z ) 1 | } https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Eque_HTML.gif
is the parabolic starlike functions introduced by Rønning [33]. For a survey of parabolic starlike functions and the related class of uniformly convex functions, see [34]. For 0 < β 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq42_HTML.gif, the class
S β : = S ( ( 1 + z 1 z ) β ) = { f A : | arg ( z f ( z ) f ( z ) ) | < β π 2 } https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equf_HTML.gif
is the familiar class of strongly starlike functions of orderβ. The class
S L : = S ( 1 + z ) = { f A : | ( z f ( z ) f ( z ) ) 2 1 | < 1 } https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equg_HTML.gif

is the class of lemniscate starlike functions studied in [35].

Theorem 1Let the function f S ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq43_HTML.gifbe given by (1).
  1. 1.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    | B 2 | B 1 , | 4 B 1 B 3 B 1 4 3 B 2 2 | 3 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equh_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | B 1 2 4 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equi_HTML.gif
  1. 2.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    | B 2 | B 1 , | 4 B 1 B 3 B 1 4 3 B 2 2 | B 1 | B 2 | 2 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equj_HTML.gif
     
or the conditions
| B 2 | B 1 , | 4 B 1 B 3 B 1 4 3 B 2 2 | 3 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equk_HTML.gif
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | 1 12 | 4 B 1 B 3 B 1 4 3 B 2 2 | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equl_HTML.gif
  1. 3.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    | B 2 | > B 1 , | 4 B 1 B 3 B 1 4 3 B 2 2 | B 1 | B 2 | 2 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equm_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | B 1 2 12 ( 3 | 4 B 1 B 3 B 1 4 3 B 2 2 | 4 B 1 | B 2 | + 4 B 1 2 B 2 2 | 4 B 1 B 3 B 1 4 3 B 2 2 | 2 B 1 | B 2 | B 1 2 ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equn_HTML.gif
Proof Since f S ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq47_HTML.gif, there exists an analytic function w with w ( 0 ) = 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq48_HTML.gif and | w ( z ) | < 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq49_HTML.gif in  D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif such that
z f ( z ) f ( z ) = φ ( w ( z ) ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ7_HTML.gif
(7)
Define the functions p 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq50_HTML.gif by
p 1 ( z ) : = 1 + w ( z ) 1 w ( z ) = 1 + c 1 z + c 2 z 2 + , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equo_HTML.gif
or, equivalently,
w ( z ) = p 1 ( z ) 1 p 1 ( z ) + 1 = 1 2 ( c 1 z + ( c 2 c 1 2 2 ) z 2 + ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ8_HTML.gif
(8)
Then p 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq50_HTML.gif is analytic in D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif with p 1 ( 0 ) = 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq51_HTML.gif and has a positive real part in D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif. By using (8) together with (6), it is evident that
φ ( p 1 ( z ) 1 p 1 ( z ) + 1 ) = 1 + 1 2 B 1 c 1 z + ( 1 2 B 1 ( c 2 c 1 2 2 ) + 1 4 B 2 c 1 2 ) z 2 + . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ9_HTML.gif
(9)
Since
z f ( z ) f ( z ) = 1 + a 2 z + ( a 2 2 + 2 a 3 ) z 2 + ( 3 a 4 3 a 2 a 3 + a 2 3 ) z 3 + , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ10_HTML.gif
(10)
it follows by (7), (9) and (10) that
a 2 = B 1 c 1 2 , a 3 = 1 8 [ ( B 1 2 B 1 + B 2 ) c 1 2 + 2 B 1 c 2 ] , a 4 = 1 48 [ ( 4 B 2 + 2 B 1 + B 1 3 3 B 1 2 + 3 B 1 B 2 + 2 B 3 ) c 1 3 a 4 = + 2 ( 3 B 1 2 4 B 1 + 4 B 2 ) c 1 c 2 + 8 B 1 c 3 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equp_HTML.gif
Therefore
a 2 a 4 a 3 2 = B 1 96 [ c 1 4 ( B 1 3 2 + B 1 2 B 2 + 2 B 3 3 B 2 2 2 B 1 ) + 2 c 2 c 1 2 ( B 2 B 1 ) + 8 B 1 c 1 c 3 6 B 1 c 2 2 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equq_HTML.gif
Let
d 1 = 8 B 1 , d 2 = 2 ( B 2 B 1 ) , d 3 = 6 B 1 , d 4 = B 1 3 2 + B 1 2 B 2 + 2 B 3 3 B 2 2 2 B 1 , T = B 1 96 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ11_HTML.gif
(11)
Then
| a 2 a 4 a 3 2 | = T | d 1 c 1 c 3 + d 2 c 1 2 c 2 + d 3 c 2 2 + d 4 c 1 4 | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ12_HTML.gif
(12)
Since the function p ( e i θ z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq52_HTML.gif ( θ R https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq53_HTML.gif) is in the class P https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq27_HTML.gif for any p P https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq31_HTML.gif, there is no loss of generality in assuming c 1 > 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq54_HTML.gif. Write c 1 = c https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq55_HTML.gif, c [ 0 , 2 ] https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq56_HTML.gif. Substituting the values of c 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq57_HTML.gif and c 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq58_HTML.gif respectively from (4) and (5) in (12), we obtain
| a 2 a 4 a 3 2 | = T 4 | c 4 ( d 1 + 2 d 2 + d 3 + 4 d 4 ) + 2 x c 2 ( 4 c 2 ) ( d 1 + d 2 + d 3 ) + ( 4 c 2 ) x 2 ( d 1 c 2 + d 3 ( 4 c 2 ) ) + 2 d 1 c ( 4 c 2 ) ( 1 | x | 2 ) z | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equr_HTML.gif
Replacing | x | https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq59_HTML.gif by μ and substituting the values of d 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq60_HTML.gif, d 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq61_HTML.gif, d 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq62_HTML.gif and d 4 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq63_HTML.gif from (11) yield
| a 2 a 4 a 3 2 | T 4 [ c 4 | 2 B 1 3 + 8 B 3 6 B 2 2 B 1 | + 4 | B 2 | μ c 2 ( 4 c 2 ) + μ 2 ( 4 c 2 ) ( 2 B 1 c 2 + 24 B 1 ) + 16 B 1 c ( 4 c 2 ) ( 1 μ 2 ) ] = T [ c 4 4 | 2 B 1 3 + 8 B 3 6 B 2 2 B 1 | + 4 B 1 c ( 4 c 2 ) + | B 2 | ( 4 c 2 ) μ c 2 + B 1 2 μ 2 ( 4 c 2 ) ( c 6 ) ( c 2 ) ] F ( c , μ ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ13_HTML.gif
(13)
Note that for ( c , μ ) [ 0 , 2 ] × [ 0 , 1 ] https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq64_HTML.gif, differentiating F ( c , μ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq65_HTML.gif in (13) partially with respect to μ yields
F μ = T [ | B 2 | ( 4 c 2 ) + B 1 μ ( 4 c 2 ) ( c 2 ) ( c 6 ) ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ14_HTML.gif
(14)
Then, for 0 < μ < 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq66_HTML.gif and for any fixed c with 0 < c < 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq67_HTML.gif, it is clear from (14) that F μ > 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq68_HTML.gif, that is, F ( c , μ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq69_HTML.gif is an increasing function of μ. Hence, for fixed c [ 0 , 2 ] https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq70_HTML.gif, the maximum of F ( c , μ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq69_HTML.gif occurs at μ = 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq71_HTML.gif, and
max F ( c , μ ) = F ( c , 1 ) G ( c ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equs_HTML.gif
Also note that
G ( c ) = B 1 96 [ c 4 4 ( | 2 B 1 3 + 8 B 3 6 B 2 2 B 1 | 4 | B 2 | 2 B 1 ) + 4 c 2 ( | B 2 | B 1 ) + 24 B 1 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equt_HTML.gif
Let
P = 1 4 ( | 2 B 1 3 + 8 B 3 6 B 2 2 B 1 | 4 | B 2 | 2 B 1 ) , Q = 4 ( | B 2 | B 1 ) , R = 24 B 1 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ15_HTML.gif
(15)
Since
max 0 t 4 ( P t 2 + Q t + R ) = { R , Q 0 , P Q 4 ; 16 P + 4 Q + R , Q 0 , P Q 8  or  Q 0 , P Q 4 ; 4 P R Q 2 4 P , Q > 0 , P Q 8 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ16_HTML.gif
(16)
we have
| a 2 a 4 a 3 2 | B 1 96 { R , Q 0 , P Q 4 ; 16 P + 4 Q + R , Q 0 , P Q 8  or  Q 0 , P Q 4 ; 4 P R Q 2 4 P , Q > 0 , P Q 8 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equu_HTML.gif

where P, Q, R are given by (15). □

Remark 1 When B 1 = B 2 = B 3 = 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq72_HTML.gif, Theorem 1 reduces to [[24], Theorem 3.1].

Corollary 1
  1. 1.

    If f S ( α ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq73_HTML.gif, then | a 2 a 4 a 3 2 | ( 1 α ) 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq74_HTML.gif.

     
  2. 2.

    If f S L https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq75_HTML.gif, then | a 2 a 4 a 3 2 | 1 / 16 = 0.0625 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq76_HTML.gif.

     
  3. 3.

    If f S P https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq77_HTML.gif, then | a 2 a 4 a 3 2 | 16 / π 4 0.164255 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq78_HTML.gif.

     
  4. 4.

    If f S β https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq79_HTML.gif, then | a 2 a 4 a 3 2 | β 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq80_HTML.gif.

     
Definition 2 Let φ : D C https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq36_HTML.gif be analytic, and let φ ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq81_HTML.gif be given as in (6). The class C ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq82_HTML.gif of Ma-Minda convex functions with respect toφ consists of functions f satisfying the subordination
1 + z f ( z ) f ( z ) φ ( z ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equv_HTML.gif
Theorem 2Let the function f C ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq83_HTML.gifbe given by (1).
  1. 1.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    B 1 2 + 4 | B 2 | 2 B 1 0 , | 6 B 1 B 3 + B 1 2 B 2 B 1 4 4 B 2 2 | 4 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equw_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | B 1 2 36 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equx_HTML.gif
  1. 2.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    B 1 2 + 4 | B 2 | 2 B 1 0 , 2 | 6 B 1 B 3 + B 1 2 B 2 B 1 4 4 B 2 2 | B 1 3 4 B 1 | B 2 | 6 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equy_HTML.gif
     
or the conditions
B 1 2 + 4 | B 2 | 2 B 1 0 , | 6 B 1 B 3 + B 1 2 B 2 B 1 4 4 B 2 2 | 4 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equz_HTML.gif
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | 1 144 | 6 B 1 B 3 + B 1 2 B 2 B 1 4 4 B 2 2 | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equaa_HTML.gif
  1. 3.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    B 1 2 + 4 | B 2 | 2 B 1 > 0 , 2 | 6 B 1 B 3 + B 1 2 B 2 B 1 4 4 B 2 2 | B 1 3 4 B 1 | B 2 | 6 B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equab_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | B 1 2 576 ( 16 | 6 B 1 B 3 + B 1 2 B 2 B 1 4 4 B 2 2 | 12 B 1 3 48 B 1 | B 2 | 36 B 1 2 B 1 4 8 B 1 2 | B 2 | 16 B 2 2 | 6 B 1 B 3 + B 1 2 B 2 B 1 4 4 B 2 2 | B 1 3 4 B 1 | B 2 | 2 B 1 2 ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equac_HTML.gif
Proof Since f C ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq84_HTML.gif, there exists an analytic function w with w ( 0 ) = 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq48_HTML.gif and | w ( z ) | < 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq49_HTML.gif in D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif such that
1 + z f ( z ) f ( z ) = φ ( w ( z ) ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ17_HTML.gif
(17)
Since
1 + z f ( z ) f ( z ) = 1 + 2 a 2 z + ( 4 a 2 2 + 6 a 3 ) z 2 + ( 8 a 2 3 18 a 2 a 3 + 12 a 4 ) z 3 + , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ18_HTML.gif
(18)
equations (9), (17) and (18) yield
a 2 = B 1 c 1 4 , a 3 = 1 24 [ ( B 1 2 B 1 + B 2 ) c 1 2 + 2 B 1 c 2 ] , a 4 = 1 192 [ ( 4 B 2 + 2 B 1 + B 1 3 3 B 1 2 + 3 B 1 B 2 + 2 B 3 ) c 1 3 a 4 = + 2 ( 3 B 1 2 4 B 1 + 4 B 2 ) c 1 c 2 + 8 B 1 c 3 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equad_HTML.gif
Therefore
a 2 a 4 a 3 2 = B 1 768 [ c 1 4 ( 4 3 B 2 + 2 3 B 1 1 3 B 1 3 1 3 B 1 2 + 1 3 B 1 B 2 + 2 B 3 4 3 B 2 2 B 1 ) + 2 3 c 2 c 1 2 ( B 1 2 4 B 1 + 4 B 2 ) + 8 B 1 c 1 c 3 16 3 B 1 c 2 2 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equae_HTML.gif
By writing
d 1 = 8 B 1 , d 2 = 2 3 ( B 1 2 4 B 1 + 4 B 2 ) , d 3 = 16 3 B 1 , d 4 = 4 3 B 2 + 2 3 B 1 1 3 B 1 3 1 3 B 1 2 + 1 3 B 1 B 2 + 2 B 3 4 3 B 2 2 B 1 , T = B 1 768 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ19_HTML.gif
(19)
we have
| a 2 a 4 a 3 2 | = T | d 1 c 1 c 3 + d 2 c 1 2 c 2 + d 3 c 2 2 + d 4 c 1 4 | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ20_HTML.gif
(20)
Similar as in Theorems 1, it follows from (4) and (5) that
| a 2 a 4 a 3 2 | = T 4 | c 4 ( d 1 + 2 d 2 + d 3 + 4 d 4 ) + 2 x c 2 ( 4 c 2 ) ( d 1 + d 2 + d 3 ) + ( 4 c 2 ) x 2 ( d 1 c 2 + d 3 ( 4 c 2 ) ) + 2 d 1 c ( 4 c 2 ) ( 1 | x | 2 ) z | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equaf_HTML.gif
Replacing | x | https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq59_HTML.gif by μ and then substituting the values of d 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq60_HTML.gif, d 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq61_HTML.gif, d 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq62_HTML.gif and d 4 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq63_HTML.gif from (19) yield
| a 2 a 4 a 3 2 | T 4 [ c 4 | 4 3 B 1 3 + 4 3 B 1 B 2 + 8 B 3 16 3 B 2 2 B 1 | + 2 μ c 2 ( 4 c 2 ) ( 2 3 B 1 2 + 8 3 | B 2 | ) + μ 2 ( 4 c 2 ) ( 8 3 B 1 c 2 + 64 3 B 1 ) + 16 B 1 c ( 4 c 2 ) ( 1 μ 2 ) ] = T [ c 4 3 | B 1 3 + B 1 B 2 + 6 B 3 4 B 2 2 B 1 | + 4 B 1 c ( 4 c 2 ) + 1 3 μ c 2 ( 4 c 2 ) ( B 1 2 + 4 | B 2 | ) + 2 B 1 3 μ 2 ( 4 c 2 ) ( c 4 ) ( c 2 ) ] F ( c , μ ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ21_HTML.gif
(21)
Again, differentiating F ( c , μ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq69_HTML.gif in (21) partially with respect to μ yields
F μ = T [ c 2 3 ( 4 c 2 ) ( B 1 2 + 4 | B 2 | ) + 4 B 1 3 μ ( 4 c 2 ) ( c 4 ) ( c 2 ) ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ22_HTML.gif
(22)
It is clear from (22) that F μ > 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq85_HTML.gif. Thus F ( c , μ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq69_HTML.gif is an increasing function of μ for 0 < μ < 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq66_HTML.gif and for any fixed c with 0 < c < 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq67_HTML.gif. So, the maximum of F ( c , μ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq69_HTML.gif occurs at μ = 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq71_HTML.gif and
max F ( c , μ ) = F ( c , 1 ) G ( c ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equag_HTML.gif
Note that
G ( c ) = T [ c 4 3 ( | B 1 3 + B 1 B 2 + 6 B 3 4 B 2 2 B 1 | B 1 2 4 | B 2 | 2 B 1 ) + 4 3 c 2 ( B 1 2 + 4 | B 2 | 2 B 1 ) + 64 3 B 1 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equah_HTML.gif
Let
P = 1 3 ( | B 1 3 + B 1 B 2 + 6 B 3 4 B 2 2 B 1 | B 1 2 4 | B 2 | 2 B 1 ) , Q = 4 3 ( B 1 2 + 4 | B 2 | 2 B 1 ) , R = 64 3 B 1 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ23_HTML.gif
(23)
By using (16), we have
| a 2 a 4 a 3 2 | B 1 768 { R , Q 0 , P Q 4 ; 16 P + 4 Q + R , Q 0 , P Q 8  or  Q 0 , P Q 4 ; 4 P R Q 2 4 P , Q > 0 , P Q 8 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equai_HTML.gif

where P, Q, R are given in (23). □

Remark 2 For the choice of φ ( z ) = ( 1 + z ) / ( 1 z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq86_HTML.gif, Theorem 2 reduces to [[24], Theorem 3.2].

3 Further results on the second Hankel determinant

Definition 3 Let φ : D C https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq36_HTML.gif be analytic, and let φ ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq81_HTML.gif be as given in (6). Let 0 γ 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq87_HTML.gif and τ C { 0 } https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq88_HTML.gif. A function f A https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq89_HTML.gif is in the class R γ τ ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq90_HTML.gif if it satisfies the following subordination:
1 + 1 τ ( f ( z ) + γ z f ( z ) 1 ) φ ( z ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equaj_HTML.gif
Theorem 3Let 0 γ 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq87_HTML.gif, τ C { 0 } https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq88_HTML.gif, and let the functionfas in (1) be in the class R γ τ ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq91_HTML.gif. Also, let
p = 8 9 ( 1 + γ ) ( 1 + 3 γ ) ( 1 + 2 γ ) 2 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equak_HTML.gif
  1. 1.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    2 | B 2 | ( 1 p ) + B 1 ( 1 2 p ) 0 , | B 1 B 3 p B 2 2 | p B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equal_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | | τ | 2 B 1 2 9 ( 1 + 2 γ ) 2 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equam_HTML.gif
  1. 2.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    2 | B 2 | ( 1 p ) + B 1 ( 1 2 p ) 0 , 2 | B 1 B 3 p B 2 2 | 2 ( 1 p ) B 1 | B 2 | B 1 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equan_HTML.gif
     
or the conditions
2 | B 2 | ( 1 p ) + B 1 ( 1 2 p ) 0 , | B 1 B 3 p B 2 2 | B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equao_HTML.gif
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | | τ | 2 8 ( 1 + γ ) ( 1 + 3 γ ) | B 3 B 1 p B 2 2 | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equap_HTML.gif
  1. 3.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    2 | B 2 | ( 1 p ) + B 1 ( 1 2 p ) > 0 , 2 | B 1 B 3 p B 2 2 | 2 ( 1 p ) B 1 | B 2 | B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equaq_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | | τ | 2 B 1 2 32 ( 1 + γ ) ( 1 + 3 γ ) × ( 4 p | B 3 B 1 p B 2 2 | 4 ( 1 p ) B 1 [ | B 2 | ( 3 2 p ) + B 1 ] 4 B 2 2 ( 1 p ) 2 B 1 2 ( 1 2 p ) 2 | B 3 B 1 p B 2 2 | ( 1 p ) B 1 ( 2 | B 2 | + B 1 ) ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equar_HTML.gif
Proof For f R γ τ ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq92_HTML.gif, there exists an analytic function w with w ( 0 ) = 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq48_HTML.gif and | w ( z ) | < 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq49_HTML.gif in D https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq10_HTML.gif such that
1 + 1 τ ( f ( z ) + γ z f ( z ) 1 ) = φ ( w ( z ) ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ24_HTML.gif
(24)
Since f has the Maclaurin series given by (1), a computation shows that
1 + 1 τ ( f ( z ) + γ z f ( z ) 1 ) = 1 + 2 a 2 ( 1 + γ ) τ z + 3 a 3 ( 1 + 2 γ ) τ z 2 + 4 a 4 ( 1 + 3 γ ) τ z 3 + . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ25_HTML.gif
(25)
It follows from (24), (9) and (25) that
a 2 = τ B 1 c 1 4 ( 1 + γ ) , a 3 = τ B 1 12 ( 1 + 2 γ ) [ 2 c 2 + c 1 2 ( B 2 B 1 1 ) ] , a 4 = τ 32 ( 1 + 3 γ ) [ B 1 ( 4 c 3 4 c 1 c 2 + c 1 3 ) + 2 B 2 c 1 ( 2 c 2 c 1 2 ) + B 3 c 1 3 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equas_HTML.gif
Therefore
a 2 a 4 a 3 2 = τ 2 B 1 c 1 128 ( 1 + γ ) ( 1 + 3 γ ) [ B 1 ( 4 c 3 4 c 1 c 2 + c 1 3 ) + 2 B 2 c 1 ( 2 c 2 c 1 2 ) + B 3 c 1 3 ] τ 2 B 1 2 144 ( 1 + 2 γ ) 2 [ 4 c 2 2 + c 1 4 ( B 2 B 1 1 ) 2 + 4 c 2 c 1 2 ( B 2 B 1 1 ) ] = τ 2 B 1 2 128 ( 1 + γ ) ( 1 + 3 γ ) { [ ( 4 c 1 c 3 4 c 1 2 c 2 + c 1 4 ) + 2 B 2 c 1 2 B 1 ( 2 c 2 c 1 2 ) + B 3 B 1 c 1 4 ] 8 9 ( 1 + γ ) ( 1 + 3 γ ) ( 1 + 2 γ ) 2 [ 4 c 2 2 + c 1 4 ( B 2 B 1 1 ) 2 + 4 c 2 c 1 2 ( B 2 B 1 1 ) ] } , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equat_HTML.gif
which yields
| a 2 a 4 a 3 2 | = T | 4 c 1 c 3 + c 1 4 [ 1 2 B 2 B 1 p ( B 2 B 1 1 ) 2 + B 3 B 1 ] 4 p c 2 2 4 c 1 2 c 2 [ 1 B 2 B 1 + p ( B 2 B 1 1 ) ] | , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ26_HTML.gif
(26)
where
T = | τ | 2 B 1 2 128 ( 1 + γ ) ( 1 + 3 γ ) and p = 8 9 ( 1 + γ ) ( 1 + 3 γ ) ( 1 + 2 γ ) 2 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equau_HTML.gif

It can be easily verified that p [ 64 81 , 8 9 ] https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq93_HTML.gif for 0 γ 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq87_HTML.gif.

Let
d 1 = 4 , d 2 = 4 [ 1 B 2 B 1 + p ( B 2 B 1 1 ) ] , d 3 = 4 p , d 4 = 1 2 B 2 B 1 p ( B 2 B 1 1 ) 2 + B 3 B 1 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ27_HTML.gif
(27)
Then (26) becomes
| a 2 a 4 a 3 2 | = T | d 1 c 1 c 3 + d 2 c 1 2 c 2 + d 3 c 2 2 + d 4 c 1 4 | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ28_HTML.gif
(28)
It follows that
| a 2 a 4 a 3 2 | = T 4 | c 4 ( d 1 + 2 d 2 + d 3 + 4 d 4 ) + 2 x c 2 ( 4 c 2 ) ( d 1 + d 2 + d 3 ) + ( 4 c 2 ) x 2 ( d 1 c 2 + d 3 ( 4 c 2 ) ) + 2 d 1 c ( 4 c 2 ) ( 1 | x | 2 ) z | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equav_HTML.gif
Application of the triangle inequality, replacement of | x | https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq59_HTML.gif by μ and substituting the values of d 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq60_HTML.gif, d 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq61_HTML.gif, d 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq62_HTML.gif and d 4 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq63_HTML.gif from (27) yield
| a 2 a 4 a 3 2 | T 4 [ 4 c 4 | B 3 B 1 p B 2 2 B 1 2 | + 8 | B 2 B 1 | μ c 2 ( 4 c 2 ) ( 1 p ) + ( 4 c 2 ) μ 2 ( 4 c 2 + 4 p ( 4 c 2 ) ) + 8 c ( 4 c 2 ) ( 1 μ 2 ) ] = T [ c 4 | B 3 B 1 p B 2 2 B 1 2 | + 2 c ( 4 c 2 ) + 2 μ | B 2 B 1 | c 2 ( 4 c 2 ) ( 1 p ) + μ 2 ( 4 c 2 ) ( 1 p ) ( c α ) ( c β ) ] F ( c , μ ) , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ29_HTML.gif
(29)

where α = 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq94_HTML.gif, β = 2 p / ( 1 p ) > 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq95_HTML.gif.

Similarly as in the previous proofs, it can be shown that F ( c , μ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq69_HTML.gif is an increasing function of μ for 0 < μ < 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq96_HTML.gif. So, for fixed c [ 0 , 2 ] https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq56_HTML.gif, let
max F ( c , μ ) = F ( c , 1 ) G ( c ) , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equaw_HTML.gif
which is
G ( c ) = T { c 4 [ | B 3 B 1 p B 2 2 B 1 2 | ( 1 p ) ( 2 | B 2 B 1 | + 1 ) ] + 4 c 2 [ 2 | B 2 B 1 | ( 1 p ) + 1 2 p ] + 16 p } . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equax_HTML.gif
Let
P = | B 3 B 1 p B 2 2 B 1 2 | ( 1 p ) ( 2 | B 2 B 1 | + 1 ) , Q = 4 [ 2 | B 2 B 1 | ( 1 p ) + 1 2 p ] , R = 16 p . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ30_HTML.gif
(30)
Using (16), we have
| a 2 a 4 a 3 2 | T { R , Q 0 , P Q 4 ; 16 P + 4 Q + R , Q 0 , P Q 8  or  Q 0 , P Q 4 ; 4 P R Q 2 4 P , Q > 0 , P Q 8 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equay_HTML.gif

where P, Q, R are given in (30). □

Remark 3 For the choice φ ( z ) : = ( 1 + A z ) / ( 1 + B z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq97_HTML.gif with 1 B < A 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq98_HTML.gif, Theorem 3 reduces to [[36], Theorem 2.1].

Definition 4 Let φ : D C https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq36_HTML.gif be analytic, and let φ ( z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq81_HTML.gif be as given in (6). For a fixed real number α, the function f A https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq89_HTML.gif is in the class G α ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq99_HTML.gif if it satisfies the following subordination:
( 1 α ) f ( z ) + α ( 1 + z f ( z ) f ( z ) ) φ ( z ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equaz_HTML.gif

Al-Amiri and Reade [37] introduced the class G α : = G α ( ( 1 + z ) / ( 1 z ) ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq100_HTML.gif and they showed that G α S https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq101_HTML.gif for α < 0 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq102_HTML.gif. Univalence of the functions in the class G α https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq103_HTML.gif was also investigated in [38, 39]. Singh et al. also obtained the bound for the second Hankel determinant of functions in G α https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq103_HTML.gif. The following theorem provides a bound for the second Hankel determinant of the functions in the class G α ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq99_HTML.gif.

Theorem 4Let the functionfgiven by (1) be in the class G α ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq104_HTML.gif, 0 α 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq105_HTML.gif. Also, let
p = 8 9 ( 1 + 2 α ) ( 1 + α ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equba_HTML.gif
  1. 1.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) + B 1 ( 1 + α 2 p ) 0 , | B 1 4 α ( 2 α 1 p α ) + α B 1 2 B 2 ( 3 2 p ) + ( α + 1 ) B 1 B 3 p B 2 2 | p B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbb_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | B 1 2 9 ( 1 + α ) 2 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbc_HTML.gif
  1. 2.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) + B 1 ( 1 + α 2 p ) 0 , 2 | B 1 4 α ( 2 α 1 p α ) + α B 1 2 B 2 ( 3 2 p ) + ( α + 1 ) B 1 B 3 p B 2 2 | B 1 3 α ( 3 2 p ) 2 ( 1 + α p ) B 1 | B 2 | ( α + 1 ) B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbd_HTML.gif
     
or
B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) + B 1 ( 1 + α 2 p ) 0 , | B 1 4 α ( 2 α 1 p α ) + α B 1 2 B 2 ( 3 2 p ) + ( α + 1 ) B 1 B 3 p B 2 2 | p B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Eqube_HTML.gif
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | | B 1 4 α ( 2 α 1 p α ) + α B 1 2 B 2 ( 3 2 p ) + ( α + 1 ) B 1 B 3 p B 2 2 | 8 ( 1 + α ) ( 1 + 2 α ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbf_HTML.gif
  1. 3.
    If B 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq44_HTML.gif, B 2 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq45_HTML.gifand B 3 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq46_HTML.gifsatisfy the conditions
    B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) + B 1 ( 1 + α 2 p ) > 0 , 2 | B 1 4 α ( 2 α 1 p α ) + α B 1 2 B 2 ( 3 2 p ) + ( α + 1 ) B 1 B 3 p B 2 2 | B 1 3 α ( 3 2 p ) 2 ( 1 + α p ) B 1 | B 2 | ( α + 1 ) B 1 2 0 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbg_HTML.gif
     
then the second Hankel determinant satisfies
| a 2 a 4 a 3 2 | B 1 2 32 ( 1 + α ) ( 1 + 2 α ) × [ 4 p [ B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) + B 1 ( 1 + α 2 p ) ] 2 | B 1 4 α ( 2 α 1 p α ) + α B 1 2 B 2 ( 3 2 p ) + ( α + 1 ) B 1 B 3 p B 2 2 | B 1 3 α ( 3 2 p ) ( 1 + α p ) B 1 ( 2 | B 2 | + B 1 ) ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbh_HTML.gif
Proof For f G α ( φ ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq106_HTML.gif, a calculation shows that
| a 2 a 4 a 3 2 | = T | 4 ( 1 + α ) B 1 c 1 c 3 + c 1 4 [ 3 α B 1 2 + α ( 2 α 1 ) B 1 3 + B 1 ( 1 + α ) + 3 α B 1 B 2 + ( 1 + α ) ( B 3 2 B 2 ) p ( α B 1 2 B 1 + B 2 ) 2 B 1 ] 4 p B 1 c 2 2 + 2 c 1 2 c 2 [ 2 ( 1 + α ) B 1 + 3 α B 1 2 + 2 ( 1 + α ) B 2 2 p ( α B 1 2 B 1 + B 2 ) ] | , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ31_HTML.gif
(31)
where
T = B 1 128 ( 1 + α ) ( 1 + 2 α ) and p = 8 9 ( 1 + 2 α ) ( 1 + α ) . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbi_HTML.gif
It can be easily verified that for 0 α 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq105_HTML.gif, p [ 8 9 , 4 3 ] https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq107_HTML.gif. Let
d 1 = 4 ( 1 + α ) B 1 , d 2 = 2 [ 2 ( 1 + α ) B 1 + 3 α B 1 2 + 2 ( 1 + α ) B 2 2 p ( α B 1 2 B 1 + B 2 ) ] , d 3 = 4 p B 1 , d 4 = 3 α B 1 2 + α ( 2 α 1 ) B 1 3 + B 1 ( 1 + α ) + 3 α B 1 B 2 d 4 = + ( 1 + α ) ( B 3 2 B 2 ) p ( α B 1 2 B 1 + B 2 ) 2 B 1 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ32_HTML.gif
(32)
Then
| a 2 a 4 a 3 2 | = T | d 1 c 1 c 3 + d 2 c 1 2 c 2 + d 3 c 2 2 + d 4 c 1 4 | . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ33_HTML.gif
(33)
Similarly as in earlier theorems, it follows that
| a 2 a 4 a 3 2 | = T 4 | c 4 ( d 1 + 2 d 2 + d 3 + 4 d 4 ) + 2 x c 2 ( 4 c 2 ) ( d 1 + d 2 + d 3 ) + ( 4 c 2 ) x 2 ( d 1 c 2 + d 3 ( 4 c 2 ) ) + 2 d 1 c ( 4 c 2 ) ( 1 | x | 2 ) z | T [ c 4 | B 1 3 α ( 2 α 1 p α ) + α B 1 B 2 ( 3 2 p ) + ( α + 1 ) B 3 p B 2 2 B 1 | + μ c 2 ( 4 c 2 ) [ B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) ] + 2 c ( 4 c 2 ) B 1 ( 1 + α ) + μ 2 ( 4 c 2 ) B 1 ( 1 + α p ) ( c 2 ) ( c 2 p 1 + α p ) ] F ( c , μ ) , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ34_HTML.gif
(34)
and for fixed c [ 0 , 2 ] https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq56_HTML.gif, max F ( c , μ ) = F ( c , 1 ) G ( c ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq108_HTML.gif with
G ( c ) = T [ c 4 [ | B 1 3 α ( 2 α 1 p α ) + α B 1 B 2 ( 3 2 p ) + ( α + 1 ) B 3 p B 2 2 B 1 | B 1 2 α ( 3 2 p ) ( 1 + α p ) ( 2 | B 2 | + B 1 ) ] + 4 c 2 [ B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) + B 1 ( 1 + α 2 p ) ] + 16 p B 1 ] . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbj_HTML.gif
Let
P = | B 1 3 α ( 2 α 1 p α ) + α B 1 B 2 ( 3 2 p ) + ( α + 1 ) B 3 p B 2 2 B 1 | P = B 1 2 α ( 3 2 p ) ( 1 + α p ) ( 2 | B 2 | + B 1 ) , Q = 4 [ B 1 2 α ( 3 2 p ) + 2 | B 2 | ( 1 + α p ) + B 1 ( 1 + α 2 p ) ] , R = 16 p B 1 . https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equ35_HTML.gif
(35)
By using (16), we have
| a 2 a 4 a 3 2 | T { R , Q 0 , P Q 4 ; 16 P + 4 Q + R , Q 0 , P Q 8  or  Q 0 , P Q 4 ; 4 P R Q 2 4 P , Q > 0 , P Q 8 , https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_Equbk_HTML.gif

where P, Q, R are given in (35). □

Remark 4 For α = 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq109_HTML.gif, Theorem 4 reduces to Theorem 2. For 0 α < 1 https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq110_HTML.gif, let φ ( z ) : = ( 1 + ( 1 2 α ) z ) / ( 1 z ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq111_HTML.gif. For this function φ, B 1 = B 2 = B 3 = 2 ( 1 α ) https://static-content.springer.com/image/art%3A10.1186%2F1029-242X-2013-281/MediaObjects/13660_2012_Article_729_IEq112_HTML.gif. In this case, Theorem 4 reduces to [[40], Theorem 3.1].

Acknowledgements

The work presented here was supported in part by research grants from Universiti Sains Malaysia (FRGS grants) and University of Delhi as well as MyBrain MyPhD programme of the Ministry of Higher Education, Malaysia.

Copyright information

© Lee et al.; licensee Springer 2013

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Authors and Affiliations

  • See Keong Lee
    • 1
  • V Ravichandran
    • 2
  • Shamani Supramaniam
    • 1
  1. 1.School of Mathematical SciencesUniversiti Sains MalaysiaPenangMalaysia
  2. 2.Department of MathematicsUniversity of DelhiDelhiIndia