Prime ideals remain distinct

27 Views Asked by At

Let $A$ be a domain (commutative ring with unity and no zero divisors) and let $S$ be a multiplicative subset of $A$. Denote by $S^{-1}A$ the ring of quotients $\frac{a}{s}$ (you define the ring structure as you would expect). Let $\mathfrak{p}_1$ and $\mathfrak{p}_2$ be two distinct prime ideals of $A$ such that $\mathfrak{p}_1 \cap S = \emptyset$ and $\mathfrak{p}_2 \cap S = \emptyset$. Now I would like to prove that $S^{-1}\mathfrak{p}_1$ and $S^{-1}\mathfrak{p}_2$ are two distinct prime ideals of $S^{-1}A$.

Clearly $S^{-1}\mathfrak{p}_1$ and $S^{-1}\mathfrak{p}_2$ are two ideals of $S^{-1}A$. Suppose we have two elements $\frac{a_1}{s_1}$ and $\frac{a_2}{s_2}$ in $S^{-1}A$ such that $\frac{a_1a_2}{s_1s_2} \in S^{-1}\mathfrak{p}_1$. There exists an element $\frac{p}{s_3} \in S^{-1}\mathfrak{p}_1$ such that $\frac{a_1a_2}{s_1s_2} = \frac{p}{s_3}$. Can I conclude that $a_1a_2s_3=ps_1s_2$ or do I have to say that there exists an element $t \in S$ such that $ta_1a_2s_3=tps_1s_2$?

Either way we conclude that $(t)a_1a_2s_3 \in \mathfrak{p}_1$. Because $(t)s_3 \in S$ and $S\cap\mathfrak{p}_1 = \emptyset$ and $\mathfrak{p}_1$ is a prime ideal we find $a_1 \in \mathfrak{p}_1$ or $a_2 \in \mathfrak{p}_1$ and therefore $\frac{a_1}{s_1} \in S^{-1}\mathfrak{p}_1$ or $\frac{a_2}{s_2} \in S^{-1}\mathfrak{p}_1$. We conclude that $S^{-1}\mathfrak{p}_1$ is a prime ideal.

Let $S^{-1}\mathfrak{p}_1 = S^{-1}\mathfrak{p}_2$. Let $x \in \mathfrak{p}_1$, then a fortiori $x \in S^{-1}\mathfrak{p}_1=S^{-1}\mathfrak{p}_2$. Now we find $n \in \mathfrak{p}_2$ and $m \in S$ such that $x=\frac{n}{m}$. Therefore we can find a $t \in S$ such that $txm=tn$ or $txm \in \mathfrak{p}_2$. Since $tx \in S$, $S \cap \mathfrak{p}_2 = \emptyset$ and $\mathfrak{p}_2$ prime, $x\in \mathfrak{p}_2$.

I guess this concludes my proof. Is it correct and isn't there is much more elegant way of proving this?

1

There are 1 best solutions below

3
On

You're certainly right up to your last step, and at that critical stage you don't need to use the additional $t\in S$ since you are assuming that $R$ is a domain -- as such, $t(x-y)=0$ if and only if $x-y=0$ for $t\neq 0$, which is your case since $0\notin S$.

It doesn't seem like your last step is entirely right, and I tried to correct the typo you introduced. You're going for the contrapositive, but directly is good too. If $x\in \mathfrak p_1\setminus \mathfrak p_2$, then I claim $\frac x1\in S^{-1}\mathfrak p_1\setminus S^{-1}\mathfrak p_2$, so indeed the ideals are still distinct. Suppose that $\frac x1\in S^{-1}\mathfrak p_2$. Then it comes from some element $\frac ps$ with $s\in S,p\in\mathfrak p_2$. But this implies that $p=sx\in\mathfrak p_2$. Since $s\notin \mathfrak p_2$, $x\in\mathfrak p_2$, a contradiction.