Letβs do this proof in two cases: first the easy case when
\(n\) is a prime number itself. Then
\(n\) is its own prime factorization, and we are done.
Now let
\(n\) be a composite number. There are two things to show here: first, that we can factor
\(n\) into a product of primes (the
existence of a prime factorization), and secondly, that there is only one way to do so (the
uniqueness of the prime factorization). We will do these two parts separately.
Existence: By
TheoremΒ 7.3.10 we know that
\(n\) has at least one prime factor, letβs call one of them
\(p\text{.}\) So we can factor this out so
\(n=pc\) for some number
\(c\text{.}\) If
\(c\) is itself prime, we are done. If it isnβt, then
\(c\) is composite and we can use
TheoremΒ 7.3.10 again. And we can repeat this process, which must end since
\(n\) is a finite number and
\(c < n\text{,}\) until we have completely factored
\(n\) into a product of primes.
ExampleΒ 7.4.5 is an example of this part of the theorem.
Uniqueness: we will show this by contradiction, like in some previous proofs in this section. If there are numbers with two distinct prime factorizations, there must be a smallest one. Weβll show that if we pick the smallest one, we can find an even smaller number with two different prime factorizations, which is our contradiction.
Let
\(n\) be the smallest number to have have two different prime factorizations, say
\(n=p_1p_2\ldots p_s\) and
\(n=q_1q_2\ldots q_t\) where the
\(p\)βs and
\(q\)βs are (not necessarily distinct) primes, and letβs say that
\(s \leq t\text{.}\) We know that since
\(p_1 \mid n\text{,}\) \(p_1 \mid q_1q_2\ldots q_t\) since
\(n=q_1q_2\ldots q_t\)By Euclidβs Lemma,
LemmaΒ 7.4.6, we know that
\(p_1 \mid q_i\) for one of the primes
\(q_i\text{.}\) But since
\(q_i\) is prime, its only factors are
\(1\) and
\(q_i\text{.}\) Since
\(p_1 \neq 1\text{,}\) since it is a prime as well, it must be that
\(p_1=q_i\text{.}\) Now letβs divide both sides by this common prime, remembering that
\(p_1=q_i\text{:}\)
\begin{equation*}
(n)\div p_1 = (p_1p_2p_3 \ldots p_s) \div p_1 = (q_1 q_2 \ldots q_{i-1}q_{i}q_{i+1} \ldots q_t) \div p_1
\end{equation*}
\begin{equation*}
\rightarrow n\div p_1 = p_2p_3 \ldots p_s = q_1 q_2 \ldots q_{i-1}q_{i+1} \ldots q_t.
\end{equation*}
Since \(p_1 > 1\) itβs clear that \(n > (n \div p_1)\text{.}\) But the equation above is two different prime factorizations for \(n \div p_1\text{,}\) so \(n\) was not the smallest number with two different prime factorizations. So this is a contradiction, which means that our original statement is true.