The value for
(1)
|
can be found using a number of different techniques (Apostol 1983, Choe 1987, Giesy 1972, Holme 1970, Kimble 1987, Knopp and Schur 1918, Kortram 1996, Matsuoka 1961, Papadimitriou 1973, Simmons 1992, Stark 1969, 1970, Yaglom and Yaglom 1987).
is therefore the definite sum version of the indefinite sum
(2)
| |||
(3)
|
where is a generalized harmonic number (whose numerator is known as a Wolstenholme number) and is a polygamma function.
The problem of finding this value analytically is sometimes known as the Basel problem (Derbyshire 2004, pp. 63 and 370) or Basler problem (Castellanos 1988). It was first proposed by Pietro Mengoli in 1644 (Derbyshire 2004, p. 370). The solution
(4)
|
was first found by Euler in 1735 (Derbyshire 2004, p. 64) or 1736 (Srivastava 2000).
Yaglom and Yaglom (1987), Holme (1970), and Papadimitriou (1973) all derive the result, from de Moivre's identity or related identities.
is given by the series
(5)
|
(Knopp 1990, pp. 266-267), probably known to Euler and rediscovered by Apéry.
Bailey (2000) and Borwein and Bailey (2003, pp. 128-129) give a collection of BBP-type formulas that include a number for ,
(6)
| |||
(7)
|
is given by the double series
(8)
|
(B. Cloitre, pers. comm., Dec. 9, 2004).
One derivation for considers the Fourier series of
(9)
|
which has coefficients given by
(10)
| |||
(11)
| |||
(12)
|
where is a generalized hypergeometric function and (12) is true since the integrand is odd. Therefore, the Fourier series is given explicitly by
(13)
|
If , then
(14)
|
so the Fourier series is
(15)
|
Letting gives , so
(16)
|
and we have
(17)
|
Higher values of can be obtained by finding and proceeding as above.
The value can also be found simply using the root linear coefficient theorem. Consider the equation and expand sin in a Maclaurin series
(18)
|
(19)
| |||
(20)
|
where . But the zeros of occur at , , , ..., or , , .... Therefore, the sum of the roots equals the coefficient of the leading term
(21)
|
which can be rearranged to yield
(22)
|
Yet another derivation (Simmons 1992) evaluates using Beukers's (1979) integral
(23)
| |||
(24)
| |||
(25)
| |||
(26)
| |||
(27)
| |||
(28)
| |||
(29)
|
To evaluate the integral, rotate the coordinate system by so
(30)
| |||
(31)
|
and
(32)
| |||
(33)
|
Then
(34)
| |||
(35)
|
Now compute the integrals and .
(36)
| |||
(37)
| |||
(38)
|
Make the substitution
(39)
| |||
(40)
| |||
(41)
|
so
(42)
|
and
(43)
|
can also be computed analytically,
(44)
| |||
(45)
| |||
(46)
|
But
(47)
| |||
(48)
| |||
(49)
| |||
(50)
| |||
(51)
|
so
(52)
| |||
(53)
| |||
(54)
|
Combining and gives
(55)
|