This question is a follow-up to both About Goldbach's conjecture and Question in Proof of Hardy Ramanujan theorem about $\omega(n) =\sum_{p|n} 1$.

Can one derive from Robin's criterion for RH an inequality of the form $\sigma(n)\ll n\omega(n)$? Moreover, as Carl Pomerance told me in a private communication that the quantity $r_{0}(n):=\inf\{r>0,(n-r,n+r)\in\mathbb{P}^{2}\}$ is expected to fulfill $r_{0}(n)\ll\log^{2}n\log\log n$, can one get from this criterion an upper bound for $r_{0}(n)$ of the form $r_{0}(n)\ll\frac{\sigma(n)}{N_{2}(n)}$ where $N_{2}(n)=\sharp\{0<r<n-\sqrt{2n-3},(n-r,n+r)\in\mathbb{P}^{2}\}$ under RH? Finally, whenever $n$ is a perfect number, does $\sigma(n)=2n$ imply (always assuming RH) that $r_{0}(n)<C\log^{2}n$ for some positive constant $C$?


1 Answer 1


The asymptotic $\sigma(n)\ll n\omega(n)$ is true and has nothing to do with RH or Robin's criterion.

Let us rewrite it as $\frac{\sigma(n)}{n}\ll\omega(n)$. By writing $n=\prod_{i=1}^k p_i^{e_i}$, where $p_1<\dots<p_k$ are primes and $e_i>0$, we want to show $$\prod_{i=1}^k\frac{\sigma(p_{e_i}^{e_i})}{p_{e_i}^{e_i}}\ll k.$$ But $\frac{\sigma(p^k)}{p^k}=1+\frac{1}{p_i}+\dots+\frac{1}{p_i^{e_i}}<1+\frac{1}{p_i-1}\leq 1+\frac{1}{i}$, so $$\frac{\sigma(n)}{n}=\prod_{i=1}^k\frac{\sigma(p_{e_i}^{e_i})}{p_{e_i}^{e_i}}\leq\prod_{i=1}^k\left(1+\frac{1}{i}\right)=k+1.$$

I have no idea about the quantities related to Goldbach conjecture but I strongly doubt RH can prove anything like what you ask for.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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