Zeta and q-Zeta Functions and Associated Series and Integrals

Zeta and q-Zeta Functions and Associated Series and Integrals

H.M. Srivastava

Department of Mathematics, University of Victoria, Victoria, Canada

Junesang Choi

Department of Mathematics, Dongguk University, Gyeongju, Republic of Korea

Table of Contents

Cover image

Title page

Copyright

Preface

Acknowledgements

1. Introduction and Preliminaries

1.1 Gamma and Beta Functions

1.2 The Euler-Mascheroni Constant γ

1.3 Polygamma Functions

1.4 The Multiple Gamma Functions

1.5 The Gaussian Hypergeometric Function and its Generalization

1.6 Stirling Numbers of the First and Second Kind

1.7 Bernoulli, Euler and Genocchi Polynomials and Numbers

1.8 Apostol-Bernoulli, Apostol-Euler and Apostol-Genocchi Polynomials and Numbers

1.9 Inequalities for the Gamma Function and the Double Gamma Function

Problems

2. The Zeta and Related Functions

2.1 Multiple Hurwitz Zeta Functions

2.2 The Hurwitz (or Generalized) Zeta Function

2.3 The Riemann Zeta Function

2.4 Polylogarithm Functions

2.5 Hurwitz–Lerch Zeta Functions

2.6 Generalizations of the Hurwitz–Lerch Zeta Function

2.7 Analytic Continuations of Multiple Zeta Functions

Problems

3. Series Involving Zeta Functions

3.1 Historical Introduction

3.2 Use of the Binomial Theorem

3.3 Use of Generating Functions

3.4 Use of Multiple Gamma Functions

Problems

4. Evaluations and Series Representations

4.1 Evaluation of

4.2 Rapidly Convergent Series for

4.3 Further Series Representations

4.4 Computational Results

Problems

5. Determinants of the Laplacians

5.1 The n-Dimensional Problem

5.2 Computations Using the Simple and Multiple Gamma Functions

5.3 Computations Using Series of Zeta Functions

5.4 Computations using Zeta Regularized Products

5.5 Remarks and Observations

Problems

6. q-Extensions of Some Special Functions and Polynomials

6.1 q-Shifted Factorials and q-Binomial Coefficients

6.2 q-Derivative, q-Antiderivative and Jackson q-Integral

6.3 q-Binomial Theorem

6.4 q-Gamma Function and q-Beta Function

6.5 A q-Extension of the Multiple Gamma Functions

6.6 q-Bernoulli Numbers and q-Bernoulli Polynomials

6.7 q-Euler Numbers and q-Euler Polynomials

6.8 The q-Apostol-Bernoulli Polynomials of Order

6.9 The q-Apostol-Euler Polynomials of Order

6.10 A Generalized q-Zeta Function

6.11 Multiple q-Zeta Functions

Problems

7. Miscellaneous Results

7.1 A Set of Useful Mathematical Constants

7.2 Log-Sine Integrals Involving Series Associated with the Zeta Function and Polylogarithms

7.3 Applications of the Gamma and Polygamma Functions Involving Convolutions of the Rayleigh Functions

7.4 Bernoulli and Euler Polynomials at Rational Arguments

7.5 Closed-Form Summation of Trigonometric Series

Problems

Bibliography

Copyright

Elsevier

32 Jamestown Road, London NW1 7BY

225 Wyman Street, Waltham, MA 02451, USA

First edition 2012

Copyright © 2012 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangement with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

ISBN: 978-0-12-385218-2

For information on all Elsevier publications visit our website at elsevierdirect.com

This book has been manufactured using Print On Demand technology. Each copy is produced to order and is limited to black ink. The online version of this book will show color figures where appropriate.

Preface

This book is essentially a thoroughly revised, enlarged and updated version of the authors’ work: Series Associated with the Zeta and Related Functions (Kluwer Academic Publishers, Dordrecht, Boston and London, 2001). It aims at presenting a state-of-the-art account of the theories and applications of the various methods and techniques which are used in dealing with many different families of series associated with the Riemann Zeta function and its numerous generalizations and basic (q) extensions. Systematic accounts of only some of these methods and techniques, which are widely scattered in journal articles and book chapters, were included in the above-mentioned book.

In recent years, there has been an increasing interest in problems involving closed-form evaluations of (and representations of the Riemann Zeta function at positive integer arguments as) various families of series associated with the Riemann Zeta function ζ(s), the Hurwitz Zeta function ζ(s,a), and their such extensions and generalizations as (for example) Lerch’s transcendent (or the Hurwitz-Lerch Zeta function) Φ(z,s,a). Some of these developments have apparently stemmed from an over two-century-old theorem of Christian Goldbach (1690–1764), which was stated in a letter dated 1729 from Goldbach to Daniel Bernoulli (1700–1782), from recent rediscoveries of a fairly rapidly convergent series representation for ζ(3), which is actually contained in a 1772 paper by Leonhard Euler (1707–1783), and from another known series representation for ζ(3), which was used by Roger Apéry (1916–1994) in 1978 in his celebrated proof of the irrationality of ζ(3).

This revised, enlarged and updated version of our 2001 book is motivated essentially by the fact that the theories and applications of the various methods and techniques used in dealing with many different families of series associated with the Riemann Zeta function, its aforementioned relatives and its many different basic (q) extensions are to be found so far only in widely scattered journal articles published during the last decade or so. Thus, our systematic (and unified) presentation of these results on the evaluation and representation of the various families of Zeta and q-Zeta functions is expected to fill a conspicuous gap in the existing books dealing exclusively with these Zeta and q-Zeta functions.

The main objective of this revised, enlarged and updated version is to provide a systematic collection of various families of series associated with the Riemann and Hurwitz Zeta functions, as well as with many other higher transcendental functions, which are closely related to these functions (including especially the q-Zeta and related functions). It, therefore, aims at presenting a state-of-the-art account of the theory and applications of many different methods (which are available in the rather scattered literature on this subject, especially since the publication of our aforementioned 2001 book) for the derivation of the types of results considered here.

In our attempt to make this book as self-contained as possible within the obvious constraints, we include in Chapter 1 (Introduction and Preliminaries) a reasonably detailed account of such useful functions as the Gamma and Beta functions, the Polygamma and related functions, multiple Gamma functions, the Gauss hypergeometric function and its familiar generalization, the Stirling numbers of the first and second kind, the Bernoulli, Euler and Genocchi polynomials and numbers, the Apostol-Bernoulli, the Apostol-Euler and the Apostol-Genocchi polynomials and numbers, as well as some interesting inequalities for the Gamma function and the double Gamma function. In Chapter 2 (The Zeta and Related Functions), we present the definitions and various potentially useful properties (and characteristics) of the Riemann, Hurwitz and Hurwitz-Lerch Zeta functions and their generalizations, the Polylogarithm and related functions and the multiple Zeta functions, together with their analytic continuations.

In Chapter 3 (Series Involving Zeta Functions), we begin by providing a brief historical introduction to the main subject of this book. We then describe and illustrate some of the most effective methods of evaluating series associated with the Zeta and related functions. Further developments on the evaluations and (rapidly convergent) series representations of ζ(sare presented in Chapter 4 (Evaluations and Series Representations), which also deals with various computational results on this subject.

Chapter 5 (Determinants of the Laplacians) considers the problem involving computations of the determinants of the Laplacians for the n. It is here in this chapter that we show how fruitfully some of the series evaluations (which are presented in the earlier chapters) can be applied in the solution of the aforementioned problem.

In a brand new Chapter 6 (q-Extensions of Some Special Functions and Polynomials), we first introduce the concepts of the basic (q) numbers, the basic (q) series and the basic (q) polynomials. We then proceed to apply these concepts and present a reasonably detailed theory of the various basic (q) extensions of the Gamma and Beta functions, the derivatives, antiderivatives and integrals, the binomial theorem, the multiple Gamma functions, the Bernoulli numbers and polynomials, the Euler numbers and polynomials, the Apostol-Bernoulli polynomials, the Apostol-Euler polynomials and so on.

The last chapter (Chapter 7) contains a wide variety of miscellaneous results dealing with (for example) the analysis of several useful mathematical constants, a variety of Log-Sine integrals involving series associated with the Zeta function and Polylogarithms, applications of the Gamma and Polygamma functions involving convolutions of the Rayleigh functions, evaluations of the Bernoulli and Euler polynomials at rational arguments, and the closed-form summation of several classes of trigonometric series.

Each chapter in this book begins with a brief outline summarizing the material presented in the chapter and is then divided into a number of sections. Equations in every section are numbered separately. While referring to an equation in another section of the book, we use numbers like 3.2(18) to represent Equation (18) in Section 3.2 (that is, the second section of Chapter 3).

At the close of each chapter, we have provided a set of carefully-selected problems, which are based essentially upon the material presented in the chapter. Many of these problems are taken from recent research publications, and (in all such instances) we have chosen to include the precise references for further investigation (if necessary). Another valuable feature of this book is the extensive and up-to-date bibliography on the subject dealt with in the book.

Just as its predecessor (that is, the 2001 edition), this book is written primarily as a reference work for various seemingly diverse groups of research workers and other users of series associated with the Zeta and related functions. In particular, teachers, researchers and postgraduate students in the fields of mathematical and applied sciences will find this book especially useful, not only for its detailed and systematic presentations of the theory and applications of the various methods and techniques used in dealing with many different classes of series associated with the Zeta and related functions, or for its stimulating historical accounts of a large number of problems considered here, but also for its well-classified tables of series (and integrals) and its well-motivated presentation of many sets of closely related problems with their precise bibliographical references (if any).

Acknowledgements

Many persons have contributed rather significantly to this thoroughly revised, enlarged and updated version, just as to its predecessor (that is, the 2001 edition), both directly and indirectly. Contribution of subject matter is duly acknowledged throughout the text and in the bibliography. Indeed, we are greatly indebted to the various authors whose works we have freely consulted and who occasionally provided invaluable references and advice serving for the enrichment of the matter presented in this book. The first-named author wishes to express his deep sense of gratitude to his wife and colleague, Professor Rekha Srivastava, for her cooperation and support throughout the preparation of this thoroughly revised, enlarged and updated version of the 2001 book.

The collaboration of the authors on the 2001 book project was conceptualized as long ago as August 1995, and the preparation of a preliminary outline was initiated in December 1997, during the first-named author’s visits to Dongguk University at Gyeongju. The first drafts of some of the chapters in this book were written during several subsequent visits of the first-named author to Dongguk University at Gyeongju. The final drafts of most of the chapters in the 2001 book were prepared during the second-named author’s visit to the University of Victoria from August 1999 to August 2000, while he was on Study Leave from Dongguk University at Gyeongju. The preparation of this thoroughly revised, enlarged and updated version was carried out, in most part, during the period from January 2008 to January 2009, during the second-named author’s visit to the University of Victoria, while he was on Study Leave from Dongguk University at Gyeongju for the second time. Our sincere thanks are due to the appropriate authorities of each of these universities, to the Korea Research Foundation (Support for Faculty Research Abroad under its Research Fund Program) and to the Natural Sciences and Engineering Research Council of Canada, for providing financial support and other facilities for the completion of each of the projects leading eventually to the 2001 edition and this thoroughly revised, enlarged and updated version. We especially acknowledge and appreciate the financial support that was received under the Basic Science Research Program through the National Research Foundation of the Republic of Korea.

We take this opportunity to express our thanks to the editorial (and technical) staff of the Elsevier Science Publishers B.V. (especially the Publisher, Ms. Lisa Tickner, for Serials and Elsevier Insights) for their continued interest in this book and for their proficient (and impeccable) handling of its publication. Springer’s permission to publish this thoroughly revised, enlarged and updated edition of the 2001 book is also greatly appreciated.

Finally, we should like to record our indebtedness to the members of our respective families for their understanding, cooperation and support throughout this project. The second-named author and his family would, especially, like to express their appreciation for the first-named author and his family’s hospitality and every prudent consideration during their stay in Victoria for over one year, first from August 1999 to August 2000 and then again from January 2008 to January 2009, while the second-named author was on Study Leave from Dongguk University at Gyeongju.

H.M. Srivastava

University of Victoria Canada

Junesang Choi

Dongguk University Republic of Korea

February 2011

1

Introduction and Preliminaries

In this introductory chapter, we present the definitions and notations (and some of the important properties and characteristics) of the various special functions, polynomials and numbers, which are potentially useful in the remainder of the book. The special functions considered here include (for example) the Gamma, Beta and related functions, the Polygamma functions, the multiple Gamma functions, the Gaussian hypergeometric function and the generalized hypergeometric function. We also consider the Stirling numbers of the first and second kind, the Bernoulli, Euler and Genocchi polynomials and numbers and the various families of the generalized Bernoulli, Euler and Genocchi polynomials and numbers. Relevant connections of some of these functions with other special functions and polynomials, which are not listed above, are also presented here.

1.1 Gamma and Beta Functions

The Gamma Function

The origin of the Gamma function can be traced back to two letters from Leonhard Euler (1707–1783) to Christian Goldbach (1690–1764), just as a simple desire to extend factorials to values between the integers. The first letter (dated October 13, 1729) dealt with the interpolation problem, whereas the second letter (dated January 8, 1730) dealt with integration and tied the two together.

The Gamma function Γ(z) developed by Euler is usually defined by

(1)

one by Weierstrass:

(2)

where γ denotes the Euler-Mascheroni constant defined by

(3)

and the other by Gauss:

(4)

since

) by

(5)

the definition (4) can easily be written in an equivalent form:

(6)

By taking the reciprocal of (2) and applying the definition (3), we have

which yields Euler’s product form of the Gamma function:

(7)

When t in (1) is also written in an equivalent form:

(8)

This representation of the Gamma function as well as the symbol Γ are attributed to Legendre.

Integration of (1) by parts easily yields the functional relation:

(9)

so that, obviously,

(10)

which enables us to define Γ(zThus, Γ(z) can be continued analytically to the whole complex z

The representation (2) in conjunction with the well-known product formula:

(11)

also yields the following useful relationship between the Gamma and circular functions:

(12)

which incidentally provides an immediate analytic continuation of Γ(z) from right to the left half of the complex z-plane.

when x is real, are worthy of note. Indeed, from (1) we note that

(13)

for all x Thus, immediately yields

(14)

which, in view of (1), implies

(15)

or, equivalently,

(16)

By making use of the relation (10), we obtain

(17)

The last two results in when x is a positive integer and when x is half an odd integer, positive or negative.

From (2) or (4), it also follows that Γ(z) is a meromorphic function on the whole complex zwith their respective residues given by

(18)

which implies that w the Gamma function Γ(z) has an essential singularity at infinity. Furthermore, it follows immediately from has no poles, and, therefore, Γ(z) is never zero.

Pochhammer’s Symbol and the Factorial Function

defined by is also referred to as the shifted factorial.

In terms of the Gamma function, we have (cf. Definition (5))

(19)

which can easily be verified. Furthermore, the binomial coefficient may now be expressed as

(20)

or, equivalently, as

(21)

It follows from (20) and (21) that

yields

(22)

Equations (19) and (22) suggest the definition:

(23)

Equation (19) also yields

(24)

which, in conjunction with (23), gives

(25)

we have

(26)

which may alternatively be written in the form:

(27)

Multiplication Formulas of Legendre and Gauss

In view of the definition (5), it is not difficult to show that

(28)

which follows also from Legendre’s duplication formula for the Gamma function, viz

(29)

For every positive integer m, we have

(30)

which reduces to it can be proved that

(31)

which is known in the literature as Gauss’s multiplication theorem for the Gamma function.

Stirling’s Formula for and its Generalizations

For a large positive integer n, it naturally becomes tedious to compute n!. An easy way of computing an approximate value of n! for large positive integer n was initiated by Stirling in 1730 and modified subsequently by De Moivre, who showed that

(32)

or, more generally, that

(33)

where e is the base of the natural logarithm.

For a complex number z, we have the following asymptotic expansion:

(34)

which, upon taking exponentials, yields an asymptotic formula for the Gamma function:

(35)

The asymptotic formula (35), in conjunction with the recurrence relation (9), is useful in computing the numerical values of Γ(z) for large real values of z.

Some useful consequences of (34) or (35) include the asymptotic expansions:

(36)

and

(37)

where α and β are bounded complex numbers.

Yet another interesting consequence of :

(38)

where x and y take on real values.

The Beta Function

The Beta function is a function of two complex variables α defined by

(39)

or, equivalently, by

(40)

which follows from

The integrals in (39) and (1) are known as the Eulerian integrals of the first and second kind, respectively.

in as an infinite integral:

(41)

The Beta function is closely related to the Gamma function; in fact, we have

(42)

which not only confirms the symmetry property in (39), but also continues the Beta function analytically for all complex values of α . Thus, we may write

(43)

Next we combine the relationship (42) with (40) and (41), and we obtain the following useful integral formulas:

(44)

and

(45)

It should be remarked in passing that the integral , yields the familiar infinite integral:

(46)

which is usually evaluated in the literature by contour integration (see, e.g., Copson [341, p. 139, Example 1]).

In addition to (41) and (45), by means of suitable substitutions, a number of definite integrals are expressible in terms of the Beta function:

(47)

(48)

(49)

(50)

The following functional equations for the Beta function can be deduced easily from (39) and (42):

(51)

(52)

(53)

or, more generally,

(54)

(55)

The Incomplete Gamma Functions

The incomplete Gamma function and its complement (also known as Prym’s function) are defined by

(56)

(57)

so that

(58)

is an entire (integral) function of zis a meromorphic function of z.

The following recursion formulas are worthy of note:

(59)

(60)

The Incomplete Beta Functions

The incomplete Beta function is defined by

(61)

For the associated function:

(62)

we note here the following properties that are easily verifiable:

(63)

(64)

(65)

(66)

(67)

The Error Functions

The error function , also known as the probability integral is defined for any complex z by

(68)

and its complement by

(69)

Clearly, we have

(70)

and, in view of the well-known result (16), we also have

(71)

The following alternative notations:

(72)

respectively, defined by for their complements.

In terms of the incomplete Gamma functions, it is easily verified that

(73)

The Bohr-Mollerup Theorem

We have already observed that Euler’s definition as noted in conjunction with (35). Since the solution to the interpolation problem is not determined uniquely, it makes sense to add more conditions to the problem. After various trials to find those conditions to guarantee the uniqueness of the Gamma function, in 1922, Bohr and Mollerup were able to show the remarkable fact that the Gamma function is the only function that satisfies the recurrence relationship and is logarithmically convex. The original proof was simplified, several years later, by Emil Artin, and the theorem, together with Artin’s method of proof, now constitute the Bohr-Mollerup-Artin theorem:

Theorem 1.1

Let satisfy each of the following properties:

is a convex function;

for all ;

Then for all .

Instead of giving here the proof of Theorem 1.1 (see Conway [339, p. 179] and Artin [72, p. 14]), we simply state the necessary and sufficient condition for the logarithmic convexity of a given function.

Theorem 1.2

Let , and suppose that for all and that f has a continuous second derivative for . Then f is logarithmically convex, if and only if

Remmert [973] admires the following Wielandt’s uniqueness theorem for the Gamma function: It is hardly known that there is also an elegant function theoretic characterization of Γ(z). This uniqueness theorem was discovered by Helmut Wielandt in 1939. A function theorist ought to be as much fascinated by Wielandt’s complex-analytic characterization as by the Bohr-Mollerup theorem. For further comment and applications for Wielandt’s theorem, see [675, pp. 47–49], [973], and [1065]. Here, without proof, we present

Theorem 1.3

Wielandt’s Theorem

Let be an analytic function in the right half plane having the following two properties:

for all ;

is bounded in the strip .

Then in with .

1.2 The Euler-Mascheroni Constant γ

The divergence of the harmonic series:

(1)

was attributed by James Bernoulli to his brother (see was first established in 1735 by Euler [427] (see Walfisz [1203]), who used the notation C for it and stated that it was worthy of serious consideration. We are surprised at Euler’s foresight that there is a huge amount of literature on this famous mathematical constant γ among which we just refer to the book [543] and the references therein.

The Euler (or, more precisely, the Euler-Mascheroni) constant γ is defined as follows (see Eq. 1.1(3)):

(2)

where Hn are called harmonic numbers defined by

(3)

The symbol γ was first used by Mascheroni in 1790 (see [802]) and the notation C has still been used for the notation γ . Thus, the convergence of the sequence in (2) follows by the monotone convergence theorem (see, e.g., [1202, p. 45]).

It is noted that γ is a constant so chosen that in the Weierstrass product form 1.1(2) of the Gamma function Γ(z), and the constant γ is the very Euler constant in (2).

The relatively more familiar constants are π and e(whether an algebraic or a transcendental number) has not yet been known. This was a part of the famous Hilbert’s seventh problem. David Hilbert [557] announced 23 problems for the twentieth century in the Second International Congress of Mathematicians at Paris in 1900. We introduce here only his seventh problem: Irrationality and Transcendence of Certain Numbers. G. H. Hardy was alleged to have offered to give up his Savilian Chair at Oxford to anyone who proved γ to be irrational (see [543, p. 52]). The degree of possible rationality of γ has been tried (see [1222], [543, p. 97]). Appell [69] gave a proof that γ is irrational. Appell himself’s finding an error, quickly he published a retraction, within a couple of weeks, of his original announcement (see Ayoub [82]).

Euler [427] gave the formula

Bn , he calculated

Mascheroni [802] evaluated the value of Euler constant with 32 figures as follows:

Soldner (see [482]) computed the value of γ as

which differs from Mascheroni’s value in the twentieth place. In fact, Mascheroni’s value turned out to be incorrect. Nonetheless, since Mascheroni’s error has led to eight additional calculations of this celebrated mathematical constant, so γ is often called the Euler-Mascheroni constant. Knuth [678] computed the first 1271 decimals. Gourdon and Demichel [503] computed a record 108 million digits of γ. Kondo [693] has computed γ to 2 billion digits, which is apparently the current world record.

The Euler (or, more popularly, the Euler-Mascheroni) constant γ is considered the third important mathematical constant next to π and e. The mathematical constants π, e and γ are often referred to as the holy trinity. The constant γ has been involved in a variety of mathematical formulas and results. For instance, the book [505] contains about 160 formulas involving γ. Conversely, Wilf [1228] posed as a problem the following elegant infinite product formula, which contains all of the three most important mathematical constants π, e and γ (see Eq. 3.6(19)):

(4)

Choi et al. [275] presented several general infinite product formulas, which include, as their special cases, the product formula (4) of