The Stacks project

Lemma 10.128.4. Suppose that $R \to S$ is a local homomorphism of local rings. Denote $\mathfrak m$ the maximal ideal of $R$. Let $u : M \to N$ be a map of $S$-modules. Assume

  1. $S$ is essentially of finite presentation over $R$,

  2. $M$, $N$ are finitely presented over $S$,

  3. $N$ is flat over $R$, and

  4. $\overline{u} : M/\mathfrak mM \to N/\mathfrak mN$ is injective.

Then $u$ is injective, and $N/u(M)$ is flat over $R$.

Proof. By Lemma 10.127.13 and its proof we can find a system $R_\lambda \to S_\lambda $ of local ring maps together with maps of $S_\lambda $-modules $u_\lambda : M_\lambda \to N_\lambda $ satisfying the conclusions (1) – (6) for both $N$ and $M$ of that lemma and such that the colimit of the maps $u_\lambda $ is $u$. By Lemma 10.128.3 we may assume that $N_\lambda $ is flat over $R_\lambda $ for all sufficiently large $\lambda $. Denote $\mathfrak m_\lambda \subset R_\lambda $ the maximal ideal and $\kappa _\lambda = R_\lambda / \mathfrak m_\lambda $, resp. $\kappa = R/\mathfrak m$ the residue fields.

Consider the map

\[ \Psi _\lambda : M_\lambda /\mathfrak m_\lambda M_\lambda \otimes _{\kappa _\lambda } \kappa \longrightarrow M/\mathfrak m M. \]

Since $S_\lambda /\mathfrak m_\lambda S_\lambda $ is essentially of finite type over the field $\kappa _\lambda $ we see that the tensor product $S_\lambda /\mathfrak m_\lambda S_\lambda \otimes _{\kappa _\lambda } \kappa $ is essentially of finite type over $\kappa $. Hence it is a Noetherian ring and we conclude the kernel of $\Psi _\lambda $ is finitely generated. Since $M/\mathfrak m M$ is the colimit of the system $M_\lambda /\mathfrak m_\lambda M_\lambda $ and $\kappa $ is the colimit of the fields $\kappa _\lambda $ there exists a $\lambda ' > \lambda $ such that the kernel of $\Psi _\lambda $ is generated by the kernel of

\[ \Psi _{\lambda , \lambda '} : M_\lambda /\mathfrak m_\lambda M_\lambda \otimes _{\kappa _\lambda } \kappa _{\lambda '} \longrightarrow M_{\lambda '}/\mathfrak m_{\lambda '} M_{\lambda '}. \]

By construction there exists a multiplicative subset $W \subset S_\lambda \otimes _{R_\lambda } R_{\lambda '}$ such that $S_{\lambda '} = W^{-1}(S_\lambda \otimes _{R_\lambda } R_{\lambda '})$ and

\[ W^{-1}(M_\lambda /\mathfrak m_\lambda M_\lambda \otimes _{\kappa _\lambda } \kappa _{\lambda '}) = M_{\lambda '}/\mathfrak m_{\lambda '} M_{\lambda '}. \]

Now suppose that $x$ is an element of the kernel of

\[ \Psi _{\lambda '} : M_{\lambda '}/\mathfrak m_{\lambda '} M_{\lambda '} \otimes _{\kappa _{\lambda '}} \kappa \longrightarrow M/\mathfrak m M. \]

Then for some $w \in W$ we have $wx \in M_\lambda /\mathfrak m_\lambda M_\lambda \otimes \kappa $. Hence $wx \in \mathop{\mathrm{Ker}}(\Psi _\lambda )$. Hence $wx$ is a linear combination of elements in the kernel of $\Psi _{\lambda , \lambda '}$. Hence $wx = 0$ in $M_{\lambda '}/\mathfrak m_{\lambda '} M_{\lambda '} \otimes _{\kappa _{\lambda '}} \kappa $, hence $x = 0$ because $w$ is invertible in $S_{\lambda '}$. We conclude that the kernel of $\Psi _{\lambda '}$ is zero for all sufficiently large $\lambda '$!

By the result of the preceding paragraph we may assume that the kernel of $\Psi _\lambda $ is zero for all $\lambda $ sufficiently large, which implies that the map $M_\lambda /\mathfrak m_\lambda M_\lambda \to M/\mathfrak m M$ is injective. Combined with $\overline{u}$ being injective this formally implies that also $\overline{u_\lambda } : M_\lambda /\mathfrak m_\lambda M_\lambda \to N_\lambda /\mathfrak m_\lambda N_\lambda $ is injective. By Lemma 10.99.1 we conclude that (for all sufficiently large $\lambda $) the map $u_\lambda $ is injective and that $N_\lambda /u_\lambda (M_\lambda )$ is flat over $R_\lambda $. The lemma follows. $\square$


Comments (0)


Post a comment

Your email address will not be published. Required fields are marked.

In your comment you can use Markdown and LaTeX style mathematics (enclose it like $\pi$). A preview option is available if you wish to see how it works out (just click on the eye in the toolbar).

Unfortunately JavaScript is disabled in your browser, so the comment preview function will not work.

All contributions are licensed under the GNU Free Documentation License.




In order to prevent bots from posting comments, we would like you to prove that you are human. You can do this by filling in the name of the current tag in the following input field. As a reminder, this is tag 046Y. Beware of the difference between the letter 'O' and the digit '0'.