A natural number $n$ is said to have the property $P,$ if, for all $a, n^2$ divides $a^n - 1$ whenever $n$ divides $a^n - 1.$ a.) Show that every prime number $n$ has property $P.$ b.) Show that there are infinitely many composite numbers $n$ that possess property $P.$
Problem
Source: IMO Shortlist 1993, India 5
Tags: modular arithmetic, number theory, Divisibility, prime numbers, composite numbers, IMO Shortlist
15.03.2006 23:14
Non-$\LaTeX$ed problems/solutions: http://www.mathlinks.ro/Forum/viewtopic.php?t=266
06.10.2013 09:08
02.07.2019 20:29
Observe that $2$ has the property $P$, since whenever a square is odd, it is also congruent to $1$ modulo $4$. Henceforth assume that $n$ is odd. Let $n$ be any odd positive integer such that $\gcd(n,\varphi(n)) = 1$. (EDIT 1/29/2023: This implies $\color{red}n$ is squarefree.) The condition that $n$ divides $a^n - 1$ implies that $\gcd(a,n) = 1$, so $a^{\varphi(n)} \equiv 1\pmod n$. This yields $a\equiv a^{\gcd(n,\varphi(n))} \equiv 1\pmod n$. Now for any (odd) prime $p$ dividing $n$, we see that \[ \nu_p(a^n - 1) = \nu_p(a-1) + \nu_p(n) \geq 1 + 1 = 2, \]and so $n^2$ divides $a^n - 1$. This makes both parts easy. For part (a), observe that any prime $p$ satisfies $\gcd(p,\varphi(p)) = \gcd(p,p-1) = 1$. For part (b), recall that the are infinitely many primes $p$ which are $2$ modulo $3$. Now let $n = 3p$; then \[ \gcd(n,\varphi(n)) = \gcd(3p,2p-2) = \gcd(p+2,2p-2) = \gcd(p+2,6) = 1, \]and so $n$ has property $P$.
16.06.2021 02:42
a) The problem is trivial if $n=2$. Otherwise, we have $n|a^n-1\implies a^n\equiv 1\text{(mod n)}$. By FLT, this implies $a\equiv 1\text{(mod n)}\implies v_n(a^n-1)=v_n(a-1)+v_n(n)\geq 2$, as desired. b) Let $n=p_1p_2$ for odd primes $p_1,p_2$ such that $(n,\phi (n))=1$. Then $a^{p_1p_2}\equiv 1\text{(mod }p_1p_2\text{)}$. Since $(n,\phi (n))=1$, we must have $a\equiv 1\text{(mod }p_1p_2\text{)}$. Then, $v_{p_1}(a^n-1)=v_{p_1}(a-1)+v_{p_1}(n)\geq 2$, and similarly $v_{p_2}(a^n-1)\geq 2$, so we're done.
15.08.2021 21:04
We know that there are infinitely many twin primes $(p,p+2)$ and then showed that for all such twin primes $n=p(p+2)$ works....It's okay right? This does prove the infinititude....
14.04.2022 07:38
(a) Let $p$ be any prime. By Fermat's Little Theorem, $a \equiv a^p \equiv 1 \pmod p$. In particular, $a = kp + 1$ for some integer $k$. Then, \[a^p - 1 \equiv (kp + 1)^p - 1 \equiv 0 \pmod{p^2}\]by binomial expansion, as required. (b) We claim that all $n$ of the form $2p$ satisfy $P$, where $p$ is any odd prime. By Fermat's Little Theorem, \[a^2 \equiv (a^2)^p \equiv a^{2p} \equiv 1 \pmod{2p},\]so $a^2 = kp + 1$ for some integer $k$. Then, \[a^{2p} - 1 \equiv (kp + 1)^p - 1 \equiv 0 \pmod{p^2}\]by binomial expansion. All perfect squares leave a residue of $1$ modulo $4$, so \[a^{2p} - 1 \equiv (a^2)^p - 1 \equiv 1^p - 1 \equiv 0 \pmod 4.\]Since $p$ is odd, $4p^2 \mid a^{2p} - 1$, as required.
29.01.2023 05:18