Section 3.3 Prime and maximal ideals
We prove the following.
Proof.
More generally we have the following.
Proposition 3.3.2. (Existence of a maximal left ideal).
Let \(R\) be a ring and let \(I\) be a proper left ideal in \(R\text{.}\) Then there exists a maximal left ideal \(\mfm\) in \(R\) containing \(I\text{.}\)Lemma 3.3.3.
In a commutative ring every maximal ideal is a prime ideal.Proof.
Proposition 3.3.4.
Let \(A\) be a commutative ring. Then we have the following.- A proper ideal \(\mfp\) in \(A\) is a prime ideal if and only if \(A/\mfp\) is an integral domain
- A proper ideal \(\mfm\) in \(A\) is a maximal ideal if and only if \(A/\mfm\) is a field.
Proof.
- Let \(\mfp\) be a prime ideal, and let \(a,b\in A\) be such that \((a+\mfp)(b+\mfp)=0+\mfp\text{,}\) i.e., \(ab\equiv 0\mod\mfp\text{.}\) Thus, \(ab\in\mfp\) and either \(a\in\mfp\) or \(b\in\mfp\) because \(\mfp\) is a prime ideal. Hence, \(A/\mfp\) is an integral domain. Conversely, assume that \(A/\mfp\) is an integral domain for a proper ideal \(\mfp\text{.}\) Let \(ab\in\mfp\text{.}\) Thus \(ab\equiv 0\mod\mfp\text{.}\) Since \(A/\mfp\) is an integral domain we have either \(a\equiv 0\mod\mfp\) or \(b\equiv 0\mod\mfp\text{.}\) Hence either \(a\in\mfp\) or \(b\in\mfp\text{,}\) i.e., \(\mfp\) is a prime ideal.
- Suppose that \(A/\mfm\) is a field. By Theorem 3.2.2, ideals of \(A\) containing \(\mfm\) are \(\mfm\) and \(A\text{.}\) Thus \(\mfm\) is a maximal ideal. Conversely, assume that \(\mfm\) is a maximal ideal. By Lemma 3.3.3 and the first part of this theorem, \(A/\mfm\) is an integral domain. By Theorem 3.2.2 only ideals of \(A/\mfm\) are the trivial ideal and the whole ring. Thus for any nonzero \(x\in A/\mfm\) the ideal generated by \(x\) is \(A/\mfm\text{.}\) Therefore, there exists \(y\in A/\mfm\) such that \(xy=1\in A/\mfm\text{.}\) Thus, \(A/\mfm\) is a field.
In the following proposition we use criterion for prime/maximal ideal obtained in Proposition 3.3.4 without explicitly mentioning it.
Proposition 3.3.5.
Let \(A\) be a commutative ring and let \(\mfa\) be an ideal in \(A\text{.}\) Then there is a one-one correspondence between prime ideals of \(A\) containing \(\mfa\) and prime ideals of \(A/\mfa\text{.}\)
\begin{equation*}
\left\{\mfp:\mfp\text{ is a prime ideal containing }\mfa\right\}\longleftrightarrow{}\left\{\text{prime ideals in }A/\mfa\right\}\quad\text{It is defined by}\quad\mfp\mapsto\mfp/\mfa.
\end{equation*}
Similarly, there is a one-one correspondence between maximal ideals of \(A\) containing \(\mfa\) and maximal ideals of \(A/\mfa\text{.}\)
\begin{equation*}
\left\{\mfm:\mfm\text{ is a maximal ideal containing }\mfa\right\}\longleftrightarrow{}\left\{\text{maximal ideals in }A/\mfa\right\}\quad\text{It is defined by}\quad\mfm\mapsto\mfm/\mfa.
\end{equation*}
Proof.
Let \(\mfa\subseteq\mfp\) be a prime ideal in \(A\text{.}\) Consider the ideal \(\mfp/\mfa\) in \(A/\mfa\text{.}\) By Theorem 3.2.5, \((A/\mfp)\simeq(A/\mfa)\big/(\mfp/\mfa)\text{.}\) As \(A/\mfp\) is an integral domain, we have \(\mfp/\mfa\) is a prime ideal. Conversely assume that \(I\) is a prime ideal in \(A/\mfa\text{.}\) Thus there exists an ideal \(\mfp\) in \(A\) such that \(I=\mfp/\mfa\) (see Theorem 3.2.2). By Theorem 3.2.5,
\begin{equation*}
A/\mfp\simeq(A/\mfa)\big/(\mfp/\mfa)\simeq(A/\mfa)\big/ I.
\end{equation*}
Since \(A/I\) is an integral domain, \(A/\mfp\) is also an integral domain. Hence, \(\mfp\) is a prime ideal.
The mapping \(\mfp\mapsto\mfp/\mfa\) is one-one follows from Theorem 3.2.2.
The statement about maximal ideals follow by similar arguments as given above.
Let \((S,\leq)\) be a nonempty partially ordered set. A subset \(T\subset S\) is a chain of \(S\) if for any \(x,y\in T\) either \(x\leq y\) or \(y\leq x\text{.}\) The Zorn’s lemma asserts that if every chain \(T\) of \(S\) has an upper bound in \(S\text{,}\) i.e., there exists \(s\in S\) such that for every \(t\in T\) we have \(t\leq s\text{,}\) then \(S\) has at least one maximal element.
