Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I will be honest. Some play with a weird integral has gotten me to this formulation: $$\lim\limits_{n\mathop\to\infty}\frac{\zeta(2,n)}{\frac 1n+\frac 1{2n^2}}=1$$ It seems true because of the numerical approximations made and the actual evaluation of this limit at Wolfram Alpha. But i am stuck. Does anyone have an idea of where to start?

share|cite|improve this question
Do you consider accepting answers? – Mhenni Benghorbal Aug 31 '14 at 1:02
Of course i will! – Daniel Charry Aug 31 '14 at 1:03
The idea is correct. I made a correction. The limit should be one. – Mhenni Benghorbal Aug 31 '14 at 2:56
up vote 3 down vote accepted

We have that: $$\zeta(2,n)=\int_{0}^{+\infty}\frac{t\,e^{-nt}}{1-e^{-t}}\,dt$$ but since the integral on $[\pi,+\infty)$ is very small (exponentially small in $n$) we can just consider the Taylor series of $\frac{t}{1-e^{-t}}$ with respect to the point $t=0$ in order to have: $$\begin{eqnarray*}\zeta(2,n)&=&\int_{0}^{\pi}\left(1+\frac{t}{2}+\frac{t^2}{12}+O(t^4)\right)e^{-nt}\,dt + O(e^{-n})\\&=&\int_{0}^{+\infty}\left(1+\frac{t}{2}+\frac{t^2}{12}\right)e^{-nt}\,dt+O\left(\frac{1}{n^5}\right)\\&=&\frac{1}{n}+\frac{1}{2n^2}+\frac{1}{6n^3}+O\left(\frac{1}{n^5}\right),\end{eqnarray*}$$ that is much stronger than the limit it is asked to prove.

share|cite|improve this answer
+1 Cool, now I can upvote this. – achille hui Aug 31 '14 at 3:05

If $\zeta(2,n)$ gives the generalized Riemann zeta function, for large values of $n$, the Taylor expansion is $$\zeta(2,n)=\frac{1}{n}+\frac{1}{2 n^2}+\frac{1}{6 n^3}-\frac{1}{30 n^5}+O\left(\left(\frac{1}{n}\right)^7\right)$$ So,$$\frac{\zeta(2,n)}{\frac 1n+\frac 1{2n^2}}=1+\frac{1}{6 n^2}-\frac{1}{12 n^3}+\frac{1}{120 n^4}-\frac{1}{240 n^5}+\frac{29}{1120 n^6}+O\left(\left(\frac{1}{n}\right)^7\right)$$ and then $$\lim\limits_{n\mathop\to\infty}\frac{\zeta(2,n)}{\frac 1n+\frac 1{2n^2}}=1$$

share|cite|improve this answer
Just to be a major nitpicker: how do you prove that $\zeta(2,n)$ has such Taylor series? – Jack D'Aurizio Aug 31 '14 at 3:07
@JackD'Aurizio apply Watson's Lemma to the integral representation of $\zeta(2,n)$ appeared in your answer? – achille hui Aug 31 '14 at 3:13
@achillehui: yes, for sure ;) I was just debating answers like "in virtue of the following unproven relation, your limit equals..." – Jack D'Aurizio Aug 31 '14 at 3:17

You can use the asymptotic expansion of $\zeta(2,n)$

$$ \frac{1}{n}+ \frac{2}{2n^2}+O(\frac{1}{n^3}) .$$

See related techniques.

Added: I provided my answer based on your request for a hint. However here is one technique you can use.

share|cite|improve this answer
Just to be a major nitpicker: how do you prove this identity? – Jack D'Aurizio Aug 31 '14 at 3:06
@JackD'Aurizio: The OP asked for a hint! However you can find some techniques that I used in different answers if you are interested. So the main idea for solving the problem was given in this answer. – Mhenni Benghorbal Aug 31 '14 at 3:09
you just posted a shortened version of C.Leibovici's answer, what is the point in writing it? The OP already had such a hint in a previously posted answer. – Jack D'Aurizio Aug 31 '14 at 3:11

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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