4
$\begingroup$

Define two prime powers $p^m<q^n$, ($m,n\in\mathbb N$) to be consecutive if it doesn't exist a prime power $r^k$ such that $p^m\!<r^k\!<q^n$. It seems that every even number is the difference between two consecutive prime powers, I haven't ran any large tests yet, but what about the odd numbers?

Are there two consecutive prime powers $p^m<q^n$ and $i\in\mathbb N^+$ such that $q^n-p^m=11^i$?

I have tested all consecutive prime powers less than $10,000,000$ without finding any solution to $q^n-p^m=11^i$.

(A pattern seems to suggest that no odd numbers $N\ge 31$ have a solution to $q^n-p^m=N$, consecutive prime powers. But this must be wrong since prime powers has similar asymptotic distribution as the primes).

$\endgroup$
  • $\begingroup$ This is a good question. But "power", not "potency", is the usual word in English. I've edited accordingly. $\endgroup$ – Michael Lugo Mar 3 '17 at 18:13
  • $\begingroup$ @MichaelLugo, thanks! $\endgroup$ – Lehs Mar 3 '17 at 18:13
  • $\begingroup$ So one of the primes has to be even, and there are few of those. $\endgroup$ – hardmath Mar 3 '17 at 18:15
  • 1
    $\begingroup$ For powers of $11$ greater than the first we get into the realm of the abc-conjecture. $\endgroup$ – hardmath Mar 3 '17 at 18:21
  • 1
    $\begingroup$ 'No odd numbers $N\geq 31$ have a solution to $q^n-p^m=N$' is almost certainly false heuristically; near $2^k$ the probability that the next $N$ numbers are composite is approximately $(1-\frac{\ln 2}k)^{N/2}$ and for all $N$ this is bounded away from zero for large $k$, so we should in fact expect infinitely many powers of $2$ whose nearest prime is $\geq N$ away from them. $\endgroup$ – Steven Stadnicki Mar 3 '17 at 22:47
7
$\begingroup$

See the OEIS: https://oeis.org/A013603 gives the difference between $2^n$ and the largest prime less than $2^n$. We can read off that:

  • $2^n - 11$ is prime and there are no primes between it and $2^{n}$ for $n = 42, 78, 114, 190, 322, 546, 3894$.
  • $2^n - 11^2$ is prime and there are no primes between it and $2^n$ for $n = 219, 303, 1443, 2333, 2589, 3315, 3693$.
  • $2^n - 11^3$ is prime and there are no primes between it and $2^n$ for $n = 1390, 1670$.

These are the only examples where $n < 5000$ and $q = 2$. The "mirror image" sequence of the gap between $2^m$ and the smallest prime greater than $2^m$, which doesn't exist in OEIS, would be needed for the $p = 2$ case.

$\endgroup$
  • 1
    $\begingroup$ For example, $139-2^7=11$, but $137$ was already prime, and thus these prime powers are not consecutive. $\endgroup$ – hardmath Mar 3 '17 at 18:36
  • $\begingroup$ See OEIS A013597 for something of a "mirror image" sequence of a(n) = nextprime(2^n) - 2^n. $\endgroup$ – hardmath Mar 21 '17 at 21:29

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.