2

The polynomial $x^3+y^4$ is irreducible in $\mathbb C[x, y]$, thus $\mathbb C[x, y]/(x^3+y^4)$ is an integral domain. My question is: Is $\mathbb C[[x, y]]/(x^3+y^4)$ also an integral domain?

Further thoughts: $\mathbb C[[x, y]]/(x^3+y^4)$ is the completion of the excellent reduced local ring $\mathbb C[x, y]_{(x,y)}/(x^3+y^4)$, hence $\mathbb C[[x, y]]/(x^3+y^4)$ is reduced ... being an integral domain, however, is a much stronger property.

uno
  • 1,832
  • You probably know the site guidelines by now after 5 years of membership. Please add what you have tried, and give some context. Also add what you already know about the ring of formal power series. Pure problem statement questions are likely to get closed. – Martin Brandenburg Nov 15 '24 at 09:36
  • @MartinBrandenburg: haven't I already added enough context? Completion of an integral domain need not be an integral domain, so I think it is pretty nontrivial – uno Nov 15 '24 at 09:39
  • Now you have added some context. I mean did you just went ahead and tried to follow the definition of a prime element / irreducible element? (The ring is UFD!) I am not saying that this will definitely work. I am just saying that these kind of remarks belong to a good question. – Martin Brandenburg Nov 15 '24 at 09:44
  • @MartinBrandenburg: I am pretty sure the ring is not UFD ... one dimensional local UFD (hence normal) must be regular, and my ring in question is not regular – uno Nov 15 '24 at 09:48
  • 2
    I am talking about $\mathbb{C}[[x,y]]$, which is an UFD. Your question is if a specific element is irreducible in that ring. So of course one should try checking that. – Martin Brandenburg Nov 15 '24 at 09:49
  • 1
    Does this answer your question? https://math.stackexchange.com/questions/509121/when-does-an-irreducible-polynomial-stay-irreducible-as-a-power-series – Martin Brandenburg Nov 15 '24 at 10:32

1 Answers1

2

Consider $\Bbb C[x,y]/(x^3+y^4)$. Let's first compute the normalization of this. We have $(\frac{x}{y})^{4}+x=0$, so the normalization contains $z:=\frac{x}{y}$. As $z^4=-x$ and $z^3=-y$, we get $\Bbb C[z] \supseteq R$. As a polynomial ring over a field, $\Bbb C[z]$ is integrally closed. Thus $\Bbb C[z]$ is the normalization of $\Bbb C[x,y]/(x^3+y^4)$ and expressing $x$ and $y$ in terms of $z$ as above, we get that $\Bbb C[x,y]/(x^3+y^4) \cong \Bbb C[z^3,z^4]$. Clearly the only prime ideal in $\Bbb C[z]$ lying over $(z^3,z^4)$ with respect to the normalization map $\Bbb C[z^3,z^4] \hookrightarrow \Bbb C[z]$ is $(z)$. Thus we get an induced map of local rings $\Bbb C[z^3,z^4]_{(z^3,z^4)} \hookrightarrow \Bbb C[z]_{(z)}$. By applying [stacks 0394], we get that $\Bbb C[[z]]$ is the $(z^3,z^4)$-adic completion of $\Bbb C[z]_{(z)}$. (It's easy to see that the required condition is satisfied.) But because $\Bbb C[z]_{(z)}$ is a finite $\Bbb C[z^3,z^4]_{(z^3,z^4)}$-module, this completion may be expressed as a tensor product $\Bbb C[z]_{(z)} \otimes_{A} \widehat{A}$, where $A=\Bbb C[z^3,z^4]_{(z^3,z^4)}$. To finish the argument, we use that $A \to \widehat{A}$ is flat, so the injection $A \to \Bbb C[z]_{(z)}$ induces an injection $A \otimes_A \widehat{A} \hookrightarrow \Bbb C[z]_{(z)} \otimes_{A} \widehat{A} \cong \Bbb C[[z]]$. Being isomorphic to a subring of an integral domain, $\widehat{A}$ is also an integral domain. In fact, if we know that $\widehat{A}$ is isomorphic to a subring of $\Bbb C[[z]]$, it's not hard to figure out that $\widehat{A}=\Bbb C[[z^3,z^4]]$.

Lukas Heger
  • 25,763