# How many ordered triple $(p,a,b)$ is possible such that $p^a=b^4+4$?

If we have a prime number $p$ and two natural numbers $a$ and $b$ such that $p^a=b^4+4$, then how many such ordered triplets $(p,a,b)$ exist?

What should be the strategy to solve this one? The only I can see is $(5,1,1)$, is this the only one? if yes, how could we prove that?

-
oeis.org/A057781 – msh210 Dec 23 '11 at 19:33
The following perhaps surprising factorization may be useful: $b^4+4=(b^2-2b+2)(b^2+2b+2)$. – André Nicolas Dec 23 '11 at 19:35
@J. M.:Believe it or not this was a quantitative aptitude problem. ;) – Quixotic Dec 24 '11 at 7:05

This is an old contest problem, I wish I could remember where I first saw it. Anyway, André's comment that $$b^4+4=(b^2-2b+2)(b^2+2b+2)$$ is the key to a solution.

Looking modulo $16$, we see that $b^4+4$ cannot be a power of $2$. For $b>1$, both factors will be strictly greater then $1$, so that if $p^k|(b^4+4)$ then $p$ must divide both $b^2-2b+2$, and $b^2+2b+2$. Since $\gcd(b^2-2b+2,b^2+2b+2)$ must divide $4b$, and $p$ divides both terms, we see that $p|b$. This then implies that $p$ divides $4$ which is impossible.

If $b=1$, then we get the one special case.

-
You are saying $b^4 +4 \equiv 4 \text{ or }5 \mod{16}$, but $4$ is a power of $2$. – Henry Dec 23 '11 at 22:15
@Henry I wasn't counting $b=0$ as a natural number. – Eric Naslund Dec 23 '11 at 23:10

It depends on your definition of natural number, but $(2,2,0)$ may be the only other possibility

-
Perhaps this is more suited as a comment? – Quixotic Dec 24 '11 at 7:13
@MaX: Perhaps, but it is an answer – Henry Dec 24 '11 at 11:51