1

Let $a$, $b$ be positive integers. Clearly if $a\mid b^2$, $ak = b*b$ for some $k$ in $\mathbb{Z}$. Intuitively, $a$ will be the gcd so it must be $1$. But how can I show this? Is there a more appropriate way to phrase this or maybe some mathematical simplification or steps I can see?

Thanks!

Also, when I get to $a = 1$, how can I show that $b$ also must be $1$?

gebruiker
  • 6,272
  • The only option here is that $a=\pm 1$. You cannot conclude that $b=1$. $b$ can be anything. – Michael Burr Apr 21 '16 at 11:20
  • Not following. If $\gcd (a,b)=1$ and $a|b$ then $a=\pm1$. – lulu Apr 21 '16 at 11:20
  • Right but I'm not sure how to show that. Intuitively it makes sense. With a,b as positive integers the only answer is a=1. – Pingoo Bofasot Apr 21 '16 at 11:20
  • @lulu how did you go from a|b^2 to a|b? – Pingoo Bofasot Apr 21 '16 at 11:21
  • @DietrichBurde how can I adapt that question to this situation? I understand that if a|bc and gcd(a,b) = 1 then a|c but in this case I have a|b^2 and am trying to get to a|b? – Pingoo Bofasot Apr 21 '16 at 11:23
  • I am not clear what you are trying to prove. If $p$ is a prime dividing $a$, then $a|b^2\implies p|b^2\implies p|b$. But that contradicts $\gcd (a,b)=1$ so $p$ can't exist, whence $a=\pm 1$. – lulu Apr 21 '16 at 11:23
  • 1
    @PingooBofasot You can adapt it by setting $c=b$ in the original question. – Dietrich Burde Apr 21 '16 at 11:24
  • How do we know that a prime divides a? I mean of course that holds for all prime factorizations I guess but I don't see how that is relevant. I'm basically trying to prove that with a,b relatively prime, that in a|b^2, a=b=1. An intermediate step seems to be a|b but I don't see how to go from a|b^2 to a|b – Pingoo Bofasot Apr 21 '16 at 11:26

3 Answers3

1

Suppose that $p\mid a$. Since $a\mid b^2$, it follows that $p\mid b^2$. Then, it follows that $p\mid b$. This, however, is impossible since $a$ and $b$ are relatively prime. Therefore, there can be no primes dividing $a$. Therefore, $a=\pm 1$. Since $\pm 1$ divides everything, $a\mid b$ (and $b$ can be any integer).

The proof of if $p\mid b^2$ then $p\mid b$ is the same as the proof that if $2\mid b^2$, then $2\mid b$ (this shows up in the contradiction proof of $\sqrt{2}$ being irrational).

In this case, the proof that if $p\mid b^2$ then $p\mid b$ requires unique factorization. If we prime factorize $b^2$, then since $p\mid b^2$, $p$ must be one of the prime factors of $b^2$. However, since the prime factors of $b^2$ are just the prime factors of $b$ with twice their multiplicity, it follows that $p$ must be a prime factor of $b$ and so $p\mid b$. (Or see the alternative proof in the comments below).

Michael Burr
  • 33,866
1

You are saying that all the primes in $a$ are in $b^2$ and hence in $b$. Moreover you are saying that $a$ and $b$ have no common primes. The only option is that $a$ has no primes, i.e. $a=\pm 1$. But then $a|b$ trivially.

0

Hint: Note that $a\not\mid b\implies b=ka+r,\ r\in {1,2,\cdots, a-1}\implies b^2\equiv j \mod a$ with $j\ne 0$.

Edit: Note that as per my answer $(a,b)=1\implies a=\pm 1$.

Noble Mushtak
  • 18,869