Particular values of the Riemann zeta function explained

In mathematics, the Riemann zeta function is a function in complex analysis, which is also important in number theory. It is often denoted

\zeta(s)

and is named after the mathematician Bernhard Riemann. When the argument

s

is a real number greater than one, the zeta function satisfies the equation\zeta(s) = \sum_^\infty\frac \, .It can therefore provide the sum of various convergent infinite series, such as \zeta(2) = \frac + \frac + \frac + \ldots \, . Explicit or numerically efficient formulae exist for

\zeta(s)

at integer arguments, all of which have real values, including this example. This article lists these formulae, together with tables of values. It also includes derivatives and some series composed of the zeta function at integer arguments.

The same equation in

s

above also holds when

s

is a complex number whose real part is greater than one, ensuring that the infinite sum still converges. The zeta function can then be extended to the whole of the complex plane by analytic continuation, except for a simple pole at

s=1

. The complex derivative exists in this more general region, making the zeta function a meromorphic function. The above equation no longer applies for these extended values of

s

, for which the corresponding summation would diverge. For example, the full zeta function exists at

s=-1

(and is therefore finite there), but the corresponding series would be 1 + 2 + 3 + \ldots \,, whose partial sums would grow indefinitely large.

The zeta function values listed below include function values at the negative even numbers, for which and which make up the so-called trivial zeros. The Riemann zeta function article includes a colour plot illustrating how the function varies over a continuous rectangular region of the complex plane. The successful characterisation of its non-trivial zeros in the wider plane is important in number theory, because of the Riemann hypothesis.

The Riemann zeta function at 0 and 1

At zero, one has\zeta(0)= =-=-\tfrac\!

At 1 there is a pole, so ζ(1) is not finite but the left and right limits are:\lim_\zeta(1+\varepsilon) = \pm\inftySince it is a pole of first order, it has a complex residue\lim_ \varepsilon \zeta(1+\varepsilon) = 1\,.

Positive integers

Even positive integers

For the even positive integers

n

, one has the relationship to the Bernoulli numbers

Bn

:

\zeta(n) = (-1)^\frac \,.

The computation of

\zeta(2)

is known as the Basel problem. The value of

\zeta(4)

is related to the Stefan–Boltzmann law and Wien approximation in physics. The first few values are given by:\begin\zeta(2) & = 1 + \frac + \frac + \cdots = \frac \\[4pt]\zeta(4) & = 1 + \frac + \frac + \cdots = \frac \\[4pt]\zeta(6) & = 1 + \frac + \frac + \cdots = \frac \\[4pt]\zeta(8) & = 1 + \frac + \frac + \cdots = \frac \\[4pt]\zeta(10) & = 1 + \frac + \frac + \cdots = \frac \\[4pt]\zeta(12) & = 1 + \frac + \frac + \cdots = \frac \\[4pt]\zeta(14) & = 1 + \frac + \frac + \cdots = \frac \\[4pt]\zeta(16) & = 1 + \frac + \frac + \cdots = \frac\,.\end

Taking the limit

ninfty

, one obtains

\zeta(infty)=1

.
Selected values for even integers
ValueDecimal expansionSource

\zeta(2)

...

\zeta(4)

...

\zeta(6)

...

\zeta(8)

...

\zeta(10)

...

\zeta(12)

...

\zeta(14)

...

\zeta(16)

...

The relationship between zeta at the positive even integers and the Bernoulli numbers may be written as

A_n \zeta(2n) = \pi^ B_n

where

An

and

Bn

are integers for all even

n

. These are given by the integer sequences and, respectively, in OEIS. Some of these values are reproduced below:
coefficients
nAB
161
2901
39451
494501
5935551
6638512875691
7182432252
83256415662503617
93897929548012543867
101531329465290625174611
1113447856940643125155366
12201919571963756521875236364091
13110944819760305781251315862
145646536601700762736718756785560294
1556608788046690826740700156256892673020804
16624902205710223412072664062507709321041217
1712130454581433748587292890625151628697551

If we let

ηn=Bn/An

be the coefficient of

\pi2n

as above,\zeta(2n) = \sum_^\frac=\eta_n\pi^then we find recursively,

\begin\eta_1 &= 1/6 \\\eta_n &= \sum_^(-1)^\frac+(-1)^\frac\end

This recurrence relation may be derived from that for the Bernoulli numbers.

Also, there is another recurrence:

\zeta(2n)=\frac \sum_^ \zeta(2k)\zeta(2n-2k) \quad \text \quad n>1 which can be proved, using that

d
dx

\cot(x)=-1-\cot2(x)

The values of the zeta function at non-negative even integers have the generating function:\sum_^\infty \zeta(2n) x^ = -\frac \cot(\pi x) = -\frac + \frac x^2 + \frac x^4+\fracx^6 + \cdotsSince \lim_ \zeta(2n)=1 The formula also shows that for

n\inN,n → infty

,\left|B_\right| \sim \frac

Odd positive integers

The sum of the harmonic series is infinite.\zeta(1) = 1 + \frac + \frac + \cdots = \infty\!

The value is also known as Apéry's constant and has a role in the electron's gyromagnetic ratio.The value also appears in Planck's law.These and additional values are:

Selected values for odd integers
ValueDecimal expansionSource

\zeta(3)

...

\zeta(5)

...

\zeta(7)

...

\zeta(9)

...

\zeta(11)

...

\zeta(13)

...

\zeta(15)

...

It is known that is irrational (Apéry's theorem) and that infinitely many of the numbers, are irrational.[1] There are also results on the irrationality of values of the Riemann zeta function at the elements of certain subsets of the positive odd integers; for example, at least one of is irrational.[2]

The positive odd integers of the zeta function appear in physics, specifically correlation functions of antiferromagnetic XXX spin chain.[3]

Most of the identities following below are provided by Simon Plouffe. They are notable in that they converge quite rapidly, giving almost three digits of precision per iteration, and are thus useful for high-precision calculations.

Plouffe stated the following identities without proof.[4] Proofs were later given by other authors.[5]

ζ(5)

\begin\zeta(5)&=\frac\pi^5 -\frac \sum_^\infty \frac-\frac \sum_^\infty \frac\\\zeta(5)&=12 \sum_^\infty \frac -\frac \sum_^\infty \frac+\frac \sum_^\infty \frac\end

ζ(7)

\zeta(7)=\frac\pi^7 - 2 \sum_^\infty \frac\!

Note that the sum is in the form of a Lambert series.

ζ(2n + 1)

By defining the quantities

S_\pm(s) = \sum_^\infty \frac

a series of relationships can be given in the form

0=A_n \zeta(n) - B_n \pi^ + C_n S_-(n) + D_n S_+(n)

where An, Bn, Cn and Dn are positive integers. Plouffe gives a table of values:

coefficients
nABC D
318073600
514705302484
756700191134000
9185238906253712262474844
1142567525014538513505000
132574321758951492672062370
15390769879500136877815397590000
1719044170077432506758333380886313167360029116187100
19214386125140687507708537428772250281375000
2118810638157622592531256852964037337621294245721105920001793047592085750

These integer constants may be expressed as sums over Bernoulli numbers, as given in (Vepstas, 2006) below. A fast algorithm for the calculation of Riemann's zeta function for any integer argument is given by E. A. Karatsuba.[6] [7] [8]

Negative integers

In general, for negative integers (and also zero), one has

\zeta(-n)=(-1)^\frac

The so-called "trivial zeros" occur at the negative even integers:

\zeta(-2n)=0 (Ramanujan summation)

The first few values for negative odd integers are

\begin\zeta(-1) &=-\frac \\[4pt]\zeta(-3) &=\frac \\[4pt]\zeta(-5) &=-\frac \\[4pt]\zeta(-7) &=\frac \\[4pt]\zeta(-9) &= -\frac \\[4pt]\zeta(-11)&= \frac \\[4pt]\zeta(-13)&= -\frac \end

However, just like the Bernoulli numbers, these do not stay small for increasingly negative odd values. For details on the first value, see 1 + 2 + 3 + 4 + · · ·.

So ζ(m) can be used as the definition of all (including those for index 0 and 1) Bernoulli numbers.

Derivatives

The derivative of the zeta function at the negative even integers is given by

\zeta^(-2n) = (-1)^n \frac \zeta (2n+1)\,.

The first few values of which are

\begin\zeta^(-2) & = -\frac \\[4pt]\zeta^(-4) & = \frac \zeta(5) \\[4pt]\zeta^(-6) & = -\frac \zeta(7) \\[4pt]\zeta^(-8) & = \frac \zeta(9)\,.\end

One also has

\begin\zeta^(0) & = -\frac\ln(2\pi) \\[4pt]\zeta^(-1) & = \frac-\ln A \\[4pt]\zeta^(2) & = \frac\pi^2(\gamma +\ln 2-12\ln A+\ln \pi)\end

where A is the Glaisher–Kinkelin constant. The first of these identities implies that the regularized product of the reciprocals of the positive integers is

1/\sqrt{2\pi}

, thus the amusing "equation"

infty!=\sqrt{2\pi}

.[9]

From the logarithmic derivative of the functional equation,

2\frac = \log(2\pi)+\frac-\frac=\log(2\pi)+\frac+2\log 2+\gamma\,.

Selected derivatives
ValueDecimal expansionSource

\zeta'(3)

...

\zeta'(2)

...

\zeta'(0)

...

\zeta'(-\tfrac{1}{2})

...

\zeta'(-1)

...

\zeta'(-2)

...

\zeta'(-3)

...

\zeta'(-4)

...

\zeta'(-5)

...

\zeta'(-6)

...

\zeta'(-7)

...

\zeta'(-8)

...

Series involving ζ(n)

The following sums can be derived from the generating function:\sum_^\infty \zeta(k) x^=-\psi_0(1-x)-\gammawhere is the digamma function.

\begin\sum_^\infty (\zeta(k) -1) & = 1 \\[4pt]\sum_^\infty (\zeta(2k) -1) & = \frac \\[4pt]\sum_^\infty (\zeta(2k+1) -1) & = \frac \\[4pt]\sum_^\infty (-1)^k(\zeta(k) -1) & = \frac\end

Series related to the Euler–Mascheroni constant (denoted by) are\begin\sum_^\infty (-1)^k \frac & = \gamma \\[4pt]\sum_^\infty \frac & = 1 - \gamma \\[4pt]\sum_^\infty (-1)^k \frac & = \ln2 + \gamma - 1\end

and using the principal value \zeta(k) = \lim_ \fracwhich of course affects only the value at 1, these formulae can be stated as

\begin\sum_^\infty (-1)^k \frac & = 0 \\[4pt]\sum_^\infty \frac & = 0 \\[4pt]\sum_^\infty (-1)^k \frac & = \ln2\end

and show that they depend on the principal value of

Nontrivial zeros

See main article: Riemann hypothesis.

Zeros of the Riemann zeta except negative even integers are called "nontrivial zeros". The Riemann hypothesis states that the real part of every nontrivial zero must be . In other words, all known nontrivial zeros of the Riemann zeta are of the form where y is a real number. The following table contains the decimal expansion of Im(z) for the first few nontrivial zeros:

Selected nontrivial zeros
Decimal expansion of Im(z)Source
...
...
...
...
...
...
...
...
...
...

Andrew Odlyzko computed the first 2 million nontrivial zeros accurate to within 4, and the first 100 zeros accurate within 1000 decimal places. See their website for the tables and bibliographies.[10] [11] A table of about 103 billion zeros with high precision (of ±2-102≈±2·10-31) is available for interactive access and download (although in a very inconvenient compressed format) via LMFDB.[12]

Ratios

Although evaluating particular values of the zeta function is difficult, often certain ratios can be found by inserting particular values of the gamma function into the functional equation

\zeta(s)=2s\pis-1\sin\left(

\pis
2

\right)\Gamma(1-s)\zeta(1-s)

We have simple relations for half-integer arguments

\begin{align} \zeta(3/2)
\zeta(-1/2)

&=-4\pi\\

\zeta(5/2)
\zeta(-3/2)

&=-

16\pi2\\
3
\zeta(7/2)
\zeta(-5/2)

&=

64\pi3\\
15
\zeta(9/2)
\zeta(-7/2)

&=

256\pi4
105

\end{align}

Other examples follow for more complicated evaluations and relations of the gamma function. For example a consequence of the relation

\Gamma\left(\tfrac{3}{4}\right)=\left(\tfrac{\pi}{2}\right)\tfrac{1{4}}{\operatorname{AGM}\left(\sqrt2,1\right)}\tfrac{1{2}}

is the zeta ratio relation

\zeta(3/4)
\zeta(1/4)

=2\sqrt{

\pi
(2-\sqrt{2

)\operatorname{AGM}\left(\sqrt2,1\right)}}

where AGM is the arithmetic–geometric mean. In a similar vein, it is possible to form radical relations, such as from

\Gamma\left(1\right)2
5
\Gamma\left(1
\right)\Gamma\left(3
10
\right)
10

=

\sqrt{1+\sqrt{5
}}

the analogous zeta relation is

\zeta(1/5)2\zeta(7/10)\zeta(9/10)
\zeta(1/10)\zeta(3/10)\zeta(4/5)2

=

(5-\sqrt{5
)\left(\sqrt{10}+\sqrt{5+\sqrt{5}}\right)}{10 ⋅ 2

\tfrac{3{10}}}

Further reading

Notes and References

  1. Rivoal . T. . 2000 . La fonction zeta de Riemann prend une infinité de valeurs irrationnelles aux entiers impairs . Comptes Rendus de l'Académie des Sciences, Série I . 331 . 4 . 267–270 . 2000CRASM.331..267R . 10.1016/S0764-4442(00)01624-4 . math/0008051. 119678120 .
  2. W. Zudilin . One of the numbers ζ(5), ζ(7), ζ(9), ζ(11) is irrational . Russ. Math. Surv. . 2001 . 56 . 4 . 774–776 . 10.1070/rm2001v056n04abeh000427. 2001RuMaS..56..774Z . 250734661 .
  3. H.E. . Boos . V.E. . Korepin . Y. . Nishiyama . M. . Shiroishi . 2002 . Quantum correlations and number theory . J. Phys. A . 35 . 20 . 4443–4452 . cond-mat/0202346 . 2002JPhA...35.4443B . 10.1088/0305-4470/35/20/305. 119143600 . .
  4. Web site: Identities for Zeta(2*n+1) .
  5. Web site: Formulas for Odd Zeta Values and Powers of Pi .
  6. E. A.. Karatsuba. Fast calculation of the Riemann zeta function ζ(s) for integer values of the argument s. 1995. Probl. Perdachi Inf.. 31. 4. 69–80. 1367927.
  7. E. A. Karatsuba: Fast computation of the Riemann zeta function for integer argument. Dokl. Math. Vol.54, No.1, p. 626 (1996).
  8. E. A. Karatsuba: Fast evaluation of ζ(3). Probl. Inf. Transm. Vol.29, No.1, pp. 58–62 (1993).
  9. .
  10. Web site: Odlyzko . Andrew . Tables of zeros of the Riemann zeta function . 7 September 2022.
  11. Web site: Odlyzko . Andrew . Papers on Zeros of the Riemann Zeta Function and Related Topics . 7 September 2022.
  12. https://www.lmfdb.org/zeros/zeta/ LMFDB: Zeros of ζ(s)