Maths Olympiad Prep

Library / Mongolian Mathematical Olympiad /54 of 97

Mongolian Mathematical Olympiad

Number theory Difficulty 6.1 National Olympiad Prove it Mongolia

Given a∈Na \in \mathbb{N}. How many positive integer solutions following equation an≡−1(modn2)a^n \equiv -1 \pmod{n^2}?

Solution

Lemma: Let p>2p > 2 be a prime number and a∈Na \in \mathbb{N}.
(i) pα∣ap+1p^{\alpha}|a^p + 1, α≥1⇔pα+1∣ap+1\alpha \ge 1 \Leftrightarrow p^{\alpha+1}|a^p + 1
(ii) If a>2a > 2 then there exist q>2q > 2 prime number such that q∣ap+1q|a^p + 1, q∤a+1q \nmid a + 1.
Proof: See 11.6.

If a=1a = 1 then above equation has only one solution.
If a=2a = 2 then n>1n > 1 and n=pα⋅n′n = p^{\alpha} \cdot n', here pp is the smallest prime divisor of nn, (n′,p)=1(n', p) = 1.
Using the lemma, we get p2α∣2pα⋅n′+1⇒pα∣2n′+1p^{2\alpha}|2^{p^{\alpha} \cdot n'} + 1 \Rightarrow p^{\alpha}|2^{n'} + 1.
By Euler's theorem
pα∣4pα−1(p−1)−1pα∣4n′−1}⇒thus pα∣4(n′,pα(p−1))−1=3, hence α=3,p=3. \left. \begin{array}{l} p^{\alpha}|4^{p^{\alpha-1}(p-1)} - 1 \\ p^{\alpha}|4^{n'} - 1 \end{array} \right\} \Rightarrow \text{thus } p^{\alpha}|4^{(n',p^{\alpha}(p-1))} - 1 = 3, \text{ hence } \alpha = 3, p = 3.

By the Fermat's theorem
q∣64q−1−1q∣64n′−1}⇒we get that q∣64(q−1,n′)−1⇒q∣63⇒q=7. \left. \begin{array}{l} q|64^{q-1} - 1 \\ q|64^{n'} - 1 \end{array} \right\} \Rightarrow \text{we get that } q|64^{(q-1,n')} - 1 \Rightarrow q|63 \Rightarrow q = 7.
But 8n′+1≡9≢0(mod7)8^{n'} + 1 \equiv 9 \not\equiv 0 \pmod{7} this is contradiction.

The above cases, we get 2 solutions. Let a+1=2α>3a + 1 = 2^{\alpha} > 3 and p∣np|n, pp is the smallest prime divisor of nn.
If p=2p = 2 then an+1≡2≢0(mod4)a^n + 1 \equiv 2 \not\equiv 0 \pmod{4}. This leads to contradiction.

Hence, we know that p>2p > 2 and pp is an odd prime number.

p∣a2n−1p∣ap−1−1}⇒p∣a2−1=a(a−1)(a+1)⇒p∣a−1 and thus foran+1≡2≢0(modp) this leads to contradiction. \left. \begin{array}{l} p|a^{2n} - 1 \\ p|a^{p-1} - 1 \end{array} \right\} \Rightarrow p|a^2 - 1 = a(a - 1)(a + 1) \Rightarrow p|a - 1 \text{ and thus for} \\ a^n + 1 \equiv 2 \not\equiv 0 \pmod{p} \text{ this leads to contradiction.}

If there exist p1>2p_1 > 2 then by Lemma there exist p2p_2 such that p1∣a+1p_1|a+1, p2∣ap1+1p_2|a^{p_1+1}, p2∣a+1p_2|a+1.
Using Lemma, we have p12∣ap1+1+1p_1^2|a^{p_1+1} + 1 and p2∣ap1+1+1p_2|a^{p_1+1} + 1 from here p22∣ap1p2+1p_2^2|a^{p_1p_2+1} so on ... p12...pk2∣(ap1...pk−1)pk+1p_1^2...p_k^2|(a^{p_1...p_{k-1}})^{p_k} + 1, k≥1k \ge 1.
By the Lemma there exist pk+1p_{k+1} prime number such that
pk+1∣(ap1⋯pk−1)pk+1,pk+1∤a1p1⋯pk−1−1+1,pk+1>2. p_{k+1}|(a^{p_1 \cdots p_{k-1}})^{p_k} + 1, \quad p_{k+1} \nmid a_1^{p_1 \cdots p_{k-1}-1} + 1, \quad p_{k+1} > 2.
Also, (pk+1,p1...pk−1)=1(p_{k+1}, p_1...p_{k-1}) = 1 and if pk+1=pkp_{k+1} = p_k. Therefore, pk2∣(ap1...pk−1)pk+1p_k^2|(a^{p_1...p_{k-1}})^{p_k} + 1, thus for if we put there pk+1=pkp_{k+1} = p_k then pk2∣(ap1...pk−1)pk+1⇒pk∣ap1...pk−1+1p_k^2|(a^{p_1...p_{k-1}})^{p_k} + 1 \Rightarrow p_k|a^{p_1...p_{k-1}} + 1. This leads to contradiction. Thus we can construct infinitely many solutions of mentioned equation.

Want a route through all this instead of an archive? The track puts 2,604 problems in a working order, from Junior Challenge level to the IMO shortlist.

Source: MathNet, licensed CC-BY-4.0. Statement reproduced verbatim; metadata (topic, difficulty) added by this project.