MathOverflow is a question and answer site for professional mathematicians. 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

Given any integer $n$ sufficiently large, I want to prove (or disprove) that there exists another integer $m\ge n$ with the form $m=2^a3^b$ ($a,b$ are no negative integers) such that $m-n=o(n)$, i.e., $n$ can be approximated by $m$.

share|cite|improve this question
I added a vote to close as the question is not intrinsically a research level question, and no additional context is given that might link it to one. – user9072 Mar 16 at 15:06
The growth rate of the semigroup <2,3> was explicitly estimated in the paper by Bourgain-Lindenstrauss-Michel-Venkatesh regarding effective versions of Furstenberg's theorem. The estimation uses the Baker-Wustholz theorem. – Asaf Mar 16 at 16:01

Fix $\varepsilon>0$. We have to prove (for large enough $n$) that there exist non-negative integers $a,b$ such that $\log_2 n\leqslant a+b\log_2 3<\log_2 n+\varepsilon$. This follows from the fact that fractional parts of $b\alpha$, $\alpha:=\log_2 3$, are dense in $(0,1)$. Indeed, choose non-negative integers $b_1,\dots,b_k$ so that fractional parts of $b_i\alpha$ form an $\varepsilon/2$-net of $[0,1]$. Then we may choose $b\in\{b_1,\dots,b_k\}$ and appropriate integer $a$. This $a$ is positive provided that $n$ is large enough.

share|cite|improve this answer
If you want an effective result about the size of the error, you can obtain it from a paper of George Rhin, who bounded the irrationality exponent of $\log_2 3$. – Anthony Quas Mar 16 at 13:38
Is there explicit bound? Just to check the quality of the abc triples. – joro Mar 16 at 13:49

I believe replacing $o(n)$ by $O(n^a)$ for $0 < a < 1$ contradicts the $abc$ conjecture. Fix small natural $b$ coprime to $6$ (say $5$) and set $n=b^k$ for natural $k$.

In the abc triple $(m-n,n,m)$ the radical is $O(n^a)$ while $c$ is $b^k$, giving fixed quality of $\frac1a > 1$ infinitely often as $k$ varies (2,3 and $b$ dont contribute essentially anything to the radical).

This doesn't rule bound of the form something like $n/\log{n}$, but I suspect the answer to the question is negative.

share|cite|improve this answer
This contradicts not only abc, but common sense: there are not enough many numbers of the form $2^a3^b$ for such strong approximations. – Fedor Petrov Mar 16 at 13:07

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.