The Riemann zeta in terms of the dilogarithm (Q1762309): Difference between revisions

The main result of this paper is a new integral representation of the Riemann zeta function \(\zeta\) for \(\text{Re}\,s > 0\), namely \[ \zeta(s) = \frac{2s(1+s)}{(2\pi)^{1-s}}\left\{\frac{\pi^2}{6} \frac{1}{1+s} - \frac{\pi}{2}\frac{1}{s} - \frac{1}{4}\frac{1}{1-s} - D(-2-s) \right\} \] The function \(D(\cdot)\) is defined as a Mellin transform \[ D(\alpha) = \int_1^{\infty}y^{\alpha}p(y)\,dy \] where \[ p(\theta) = Re \;Li_2(e^{i\theta}) \] and \[ Li_2(z) = \sum_{n=1}^{\infty}\frac{z^n}{n^2} \] is the dilogarithm function. The above representation is complemented by three other similar representations of \(\zeta\) in terms of periodic functions. A symmetrized version of the generalized Müntz formula is derived, as well. For a particular choice of some test function this formula is then related to the integral representation of this paper. For \(0 < \text{Re}(s) <1\), one has \[ M(\mathcal F(\phi)(s))\zeta(s) = \frac{1}{(1-s)(2-s)}\zeta(1-s) \] where \(\phi(s)= 1-| x|\), for \(x \in [-1, 1]\), \(\phi(x) = 0\), for \(x \in (-\infty, -1) \cup (1, \infty)\) and \(\mathcal F(\phi)\) is the Fourier transform of \(\phi\). This also leads to an explicit computation of the Mellin transform of \(\mathcal F(\phi)\). The paper concludes with an original summation formula, involving the gamma function.