Sākums

WW.IMOSHL.2022.N4   en

Find all triples \((a,b,p)\) of positive integers with \(p\) prime and \(a^p=b!+p\).

Hide solution

Solution-1

Answer: \((2,2,2)\) and \((3,4,3)\).

Clearly, \(a>1\). We consider three cases.

Case 1: We have \(a<p\). Then we either have \(a \leqslant b\) which implies \(a \mid a^{p}-b!=p\) leading to a contradiction, or \(a>b\) which is also impossible since in this case we have \(b!\leqslant a!<a^{p}-p\), where the last inequality is true for any \(p>a>1\).

Case 2: We have \(a>p\). In this case \(b!=a^{p}-p>p^{p}-p \geqslant p!\) so \(b>p\) which means that \(a^{p}=b!+p\) is divisible by \(p\). Hence, \(a\) is divisible by \(p\) and \(b!=a^{p}-p\) is not divisible by \(p^{2}\). This means that \(b<2 p\). If \(a<p^{2}\) then \(a / p<p\) divides both \(a^{p}\) and \(b!\) and hence it also divides \(p=a^{p}-b!\) which is impossible. On the other hand, the case \(a \geqslant p^{2}\) is also impossible since then \(a^{p} \geqslant\left(p^{2}\right)^{p}>(2 p-1)!+p \geqslant b!+p\).

Comment. The inequality \(p^{2 p}>(2 p-1)!+p\) can be shown e.g. by using

\[(2p-1)!=[1 \cdot(2 p-1)] \cdot[2 \cdot(2 p-2)] \cdots \cdots[(p-1)(p+1)] \cdot p < \left(\left(\frac{2 p}{2}\right)^{2}\right)^{p-1} \cdot p=p^{2 p-1}\]

where the inequality comes from applying AM-GM to each of the terms in square brackets. **Case 3:** We have \(a=p\). In this case \(b!=p^{p}-p\). One can check that the values \(p=2,3\) lead to the claimed solutions and \(p=5\) does not lead to a solution. So we now assume that \(p \geqslant 7\). We have \(b!=p^{p}-p>p!\) and so \(b \geqslant p+1\) which implies that

\[\nu_{2}((p+1)!) \leqslant \nu_{2}(b!) = \nu_{2}\left(p^{p-1}-1\right) \stackrel{LTE}{=} 2\nu_{2}(p-1) + \nu_{2}(p+1)-1 = \nu_{2}\left(\frac{p-1}{2} \cdot(p-1) \cdot(p+1)\right),\]

where in the middle we used lifting-the-exponent lemma. On the RHS we have three factors of \((p+1)!\). But, due to \(p+1 \geqslant 8\), there are at least \(4\) even numbers among \(1,2, \ldots, p+1\), so this case is not possible.

Solution-2

The cases \(a \neq p\) are covered as in solution 1 , as are \(p=2,3\). For \(p \geqslant 5\) we have \(b!=p\left(p^{p-1}-1\right)\). By Zsigmondy's Theorem there exists some prime \(q\) that divides \(p^{p-1}-1\) but does not divide \(p^{k}-1\) for \(k<p-1\). It follows that \(\operatorname{ord}_{q}(p)=p-1\), and hence \(q \equiv 1\) \(\bmod (p-1)\). Note that \(p \neq q\). But then we must have \(q \geqslant 2 p-1\), giving

\[b!\geqslant(2 p-1)!=[1 \cdot(2 p-1)] \cdot[2 \cdot(2 p-2)] \cdots \cdots[(p-1) \cdot(p+1)] \cdot p>(2 p-1)^{p-1} p>p^{p}>p^{p}-p,\]

a contradiction.

Solution-3

The cases \(a \neq p\) are covered as in solution 1, as are \(p=2,3\). Also \(b>p\), as \(p^{p}>p!+p\) for \(p>2\). The cases \(p=5,7,11\) are also checked manually, so assume \(p \geqslant 13\). Let \(q \mid p+1\) be an odd prime. By LTE

\[\nu_{q}\left(p^{p}-p\right) = \nu_{q}\left(\left(p^{2}\right)^{\frac{p-1}{2}}-1\right) = \nu_{q}\left(p^{2}-1\right) + \nu_{q}\left(\frac{p-1}{2}\right) = \nu_{q}(p+1)\]

But \(b \geqslant p+1\), so then \(\nu_{q}(b!) > \nu_{q}(p+1)\), since \(q, a contradiction. This means that \(p+1\) has no odd prime divisor, i.e. \(p+1=2^{k}\) for some \(k\). Now let \(q \mid p-1\) be an odd prime. By LTE

\[\nu_{q}\left(p^{p}-p\right)=2 \nu_{q}(p-1)\]

Let \(d = \nu_{q}(p-1)\). Then \(p \geqslant 1+q^{d}\), so

\[\nu_{q}(b!) \geqslant \nu_{q}(p!) \geqslant \nu_{q}\left(q^{d}!\right)>q^{d-1} \geqslant 2d\]

provided \(d \geqslant 2\) and \(q>3\), or \(d \geqslant 3\). If \(q=3, d=2\) and \(p \geqslant 13\) then \(\nu_{q}(b!) \geqslant \nu_{q}(p!) \geqslant \nu_{q}(13!)=5 > 2d\). Either way, \(d \leqslant 1\). If \(p>2 q+1\) (so \(p>3 q\), as \(q \mid p-1\) ) then

\[\nu_{q}(b!) \geqslant \nu_{q}((3q)!)=3\]

so we must have \(q \geqslant \frac{p}{2}\), in other words, \(p-1 = 2q\). This implies that \(p=2^{k}-1\) and \(q=2^{k-1}-1\) are both prime, but it is not possible to have two consecutive Mersenne primes.

Solution-4

Let \(a=p, b>p\) and \(p \geqslant 5\) (the remaining cases are dealt with as in solution \(3\)). Modulo \((p+1)^{2}\) it holds that

\(p^{p}-p=(p+1-1)^{p}-p \equiv\binom{p}{1}(p+1)(-1)^{p-1}+(-1)^{p}-p=p(p+1)-1-p=p^{2}-1 \not \equiv 0 \bmod \left((p+1)^{2}\right)\).

Since \(p \geqslant 5\), the numbers 2 and \(\frac{p+1}{2}\) are distinct and less than or equal to \(p\). Therefore, \(p+1 \mid p!\), and so \((p+1)^{2} \mid(p+1)!\).

But \(b \geqslant p+1\), so \(b!\equiv 0 \not \equiv p^{p}-p \bmod (p+1)^{2}\), a contradiction.