5
$\begingroup$

I would like to know the behavior of the Riemann zeta function values at even and odd integers for studying irrationality between those values. I have tried using wolfram alpha to check the value of this sum: $$ \sum_{n = 1}^{\infty}\left[\zeta\left(2n\right)-\zeta\left(2n + 1\right)\right]. $$ It tells me it equals $\frac{1}{2}$ .

Note: I don't have any method to show if the titled sum is true . Maybe I find who is help me here for evaluating the titled sum.

Thanks for any help.

$\endgroup$
11
$\begingroup$

Note that , due to the absolute convergence, we have$$\sum_{n\geq1}\left(\zeta\left(2n\right)-\zeta\left(2n+1\right)\right)=\sum_{n\geq1}\left(\sum_{k\geq1}\frac{1}{k^{2n}}-\sum_{k\geq1}\frac{1}{k^{2n+1}}\right) $$ $$=\sum_{n\geq1}\sum_{k\geq1}\frac{k-1}{k^{2n+1}}=\sum_{k\geq2}\sum_{n\geq1}\frac{k-1}{k^{2n+1}} $$ $$=\sum_{k\geq2}\frac{1}{k\left(k+1\right)}=\sum_{k\geq2}\left(\frac{1}{k}-\frac{1}{k+1}\right)=\color{red}{\frac{1}{2}}.$$

$\endgroup$
  • $\begingroup$ @FelixMarin Yes, I'm an idiot :D $\endgroup$ – Marco Cantarini Nov 19 '16 at 8:51
  • 3
    $\begingroup$ It happens sometimes: we says $2 + 2 =5$ several times. After sleeping, we recover the $4$. You're not an idiot. You are very clever. $\endgroup$ – Felix Marin Nov 19 '16 at 8:53
3
$\begingroup$

The computation is not hard:

\begin{align*} \sum_{n=1}^{\infty}[\zeta(2n) - \zeta(2n+1)] &= \sum_{n=1}^{\infty}(-1)^{n+1}[\zeta(n+1) - 1] \\ &= \sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n!}\sum_{k=2}^{\infty} \frac{n!}{k^{n+1}} \\ &= \sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n!}\sum_{k=2}^{\infty} \int_{0}^{\infty} x^n e^{-kx} \, dx \\ &\stackrel{(1)}{=} \sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n!}\int_{0}^{\infty} \frac{x^n e^{-2x}}{1 - e^{-x}} \, dx \\ &\stackrel{(2)}{=} \int_{0}^{\infty} \bigg( \sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n!} x^n \bigg)\frac{e^{-2x}}{1 - e^{-x}} \, dx \\ &= \int_{0}^{\infty} e^{-2x} \, dx = \frac{1}{2}. \end{align*}

For $\text{(1)}$, we utilized Tonelli's theorem to interchange the order of integral and summation. For $\text{(2)}$, we utilized Fubini's theorem along with the estimate:

$$ \int_{0}^{\infty} \bigg( \sum_{n=1}^{\infty}\frac{x^n }{n!} \bigg)\frac{e^{-2x}}{1 - e^{-x}} \, dx = \int_{0}^{\infty} e^{-x} \, dx < \infty. $$

$\endgroup$
  • 2
    $\begingroup$ "The computation is not hard": haha. $\endgroup$ – Yves Daoust Nov 18 '16 at 16:59
2
$\begingroup$

$\newcommand{\bbx}[1]{\,\bbox[15px,border:1px groove navy]{\displaystyle{#1}}\,} \newcommand{\braces}[1]{\left\lbrace\,{#1}\,\right\rbrace} \newcommand{\bracks}[1]{\left\lbrack\,{#1}\,\right\rbrack} \newcommand{\dd}{\mathrm{d}} \newcommand{\ds}[1]{\displaystyle{#1}} \newcommand{\expo}[1]{\,\mathrm{e}^{#1}\,} \newcommand{\ic}{\mathrm{i}} \newcommand{\mc}[1]{\mathcal{#1}} \newcommand{\mrm}[1]{\mathrm{#1}} \newcommand{\pars}[1]{\left(\,{#1}\,\right)} \newcommand{\partiald}[3][]{\frac{\partial^{#1} #2}{\partial #3^{#1}}} \newcommand{\root}[2][]{\,\sqrt[#1]{\,{#2}\,}\,} \newcommand{\totald}[3][]{\frac{\mathrm{d}^{#1} #2}{\mathrm{d} #3^{#1}}} \newcommand{\verts}[1]{\left\vert\,{#1}\,\right\vert}$ \begin{align} &\sum_{n = 1}^{\infty}\bracks{\zeta\pars{2n}-\zeta\pars{2n + 1}} = \overbrace{% \sum_{n = 1}^{\infty}\bracks{\zeta\pars{2n} - 1}}^{\ds{3 \over 4}}\ -\ \overbrace{% \sum_{n = 1}^{\infty}\bracks{\zeta\pars{2n + 1} - 1}}^{\ds{1 \over 4}} = \bbx{1 \over 2} \end{align}

The 'even' and 'odd' above sums are well known identities.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.