Olympiad Maths Prep

Library / /2 of 6

, 2011

Number theory Difficulty 5.1 AIME, harder Prove it Czech-Polish-Slovak Mathematical Match

Let aa be a given integer. Prove that there exist infinitely many prime numbers pp such that
pn2+3,pm3a p \mid n^2 + 3, \quad p \mid m^3 - a
for some integers nn and mm.

Solution

Let kk be an arbitrary integer. Note that
(9a2k3)2+3=3(27a4k6+1) (9a^2k^3)^2 + 3 = 3(27a^4k^6 + 1)
and
(9a3k4)3a=a(36a8k121)=a(27a4k61)(27a4k6+1) (9a^3k^4)^3 - a = a(3^6a^8k^{12} - 1) = a(27a^4k^6 - 1)(27a^4k^6 + 1)
It follows that for every kZk \in \mathbb{Z} the number 27a4k6+127a^4k^6 + 1 is a common divisor of the numbers n2+3n^2 + 3 and m3am^3 - a with n=9a2k3n = 9a^2k^3 and m=9a3k4m = 9a^3k^4. So it is enough to prove that there are infinitely many primes pp such that p27a4k6+1p \mid 27a^4k^6 + 1 for some integer kk.
Suppose that there are only finitely many such primes and these are p1,p2,,prp_1, p_2, \dots, p_r. If we take k=p1p2pr+1k = p_1p_2\dots p_r + 1, then it is clear that the number 27a4k6+127a^4k^6 + 1 is not divisible by any pip_i for 1ir1 \le i \le r and that it is also greater than 1. It follows that it has a prime divisor pp, which is different from every pip_i for 1ir1 \le i \le r. We have obtained a contradiction, which finishes the proof. \square

Looking for a route rather than an archive? The track puts 2,000 problems in a working order, from AMC 10 level to the IMO shortlist.

Source: MathNet, licensed CC-BY-4.0. Statement and solution reproduced as published; topic and difficulty added by this site.