By making use of Mathematica, I detected the following integral expression for zeta(3): $$\int_0^1\frac{\log(x)}{1+x}\log\left(\frac{2+x}{1+x}\right)dx=-\frac5{12}\zeta(3).$$ Any proof of it would be gratefully appreciated.
-
Compare with this post, or this one. – Dietrich Burde Jan 15 '21 at 15:22
-
See here https://math.stackexchange.com/questions/1344455/proving-that-int-01-frac-log-left-frac1t-right-log-t2t1-d?rq=1 – user178256 Jan 16 '21 at 15:47
1 Answers
The proof is simple : the antiderivative exists (ask Mathematica).
The result is not so complicated (beside logarithms, there are just a few polylogarithm functions).
Use the bounds and simplify to get $$\text{Li}_3(-3)-2 \text{Li}_3(3)+i \pi \left(-\text{Li}_2(-3)+2 \text{Li}_2(3)+\log ^2(3)\right)-$$ $$\text{Li}_2(-3) \log (3)+\text{Li}_2(3) \log (9)+\frac{7 \zeta (3)}{4}-\frac{2 i \pi ^3}{3}-\pi ^2 \log (9)$$
This simplifies nicely (just a litle work - I did it by hand).
If you do not know what it is, ask for the numerical value $$I=-0.5008570429831642855832242$$ and the inverse symbolic calculator will tell you that $I$ satisfies the following Z-linear combination : $$ 12 I + 5\, \zeta(3)=0$$
You could have fun with $$J(a,b,c,d)=\int \frac{\log (x+a) \log \left(\frac{x+b}{x+c}\right)}{x+d}\,dx$$ which has a closed form antiderivative and get a bunch of identities.

- 260,315