1
$\begingroup$

Let $R$ be a commutative local artin $k$-algebra,where $k$ is a field with characteristic $0$.I wonder whether universal coefficient theorem holds in this case.Namely,if $C$ is a chain of flat $R$-modules. Then does $H(C\otimes_R S)\cong H(C)\otimes_R S$ whenever $H(C)$ is flat $R$-module. $H(C)$ is homology of chain complex $C$.$S$ is commutative ring over $R$

I checked wikipedia and nLab,it seems that there is only the case when $R$ is field or Principal ideal domain.I dont know whether one can use so called Kunneth spectral sequence to make arguments.I just started to understand

Thanks

$\endgroup$
2
  • 2
    $\begingroup$ When $C$ is bounded below, this appears to follow from the two hyperhomology spectral sequences, which both collapse in this case. All that is required is that every term of $C$ and of its homology is a flat $R$-module. For a description of the hyperhomology spectral sequences, see the wikipedia page on hyperhomology or the more detailed proposition 5.7.6 in Weibel's book "An introduction to homological algebra". $\endgroup$ Apr 3, 2014 at 3:23
  • 1
    $\begingroup$ More directly, the required Künneth spectral sequence for bounded below chain complexes of flat modules is given in theorem 5.6.4 of Weibel's book. $\endgroup$ Apr 3, 2014 at 4:00

1 Answer 1

4
$\begingroup$

It doesn't hold. Take $k$ any field (of characteristic zero if you wish), $R=k[\epsilon]/(\epsilon^2)$ its ring of dual numbers, $S=R/(\epsilon)=k$ its residue field, and $C$ the periodic complex $$\cdots\rightarrow R\stackrel{\epsilon}\longrightarrow R\rightarrow\cdots.$$ This is a complex of projective (hence flat) $R$-modules with trivial (hence flat) homology $H(C)=0$, so $H(C)\otimes_RS=0$, but $C\otimes_RS$ is $$\cdots\rightarrow S\stackrel{0}\longrightarrow S\rightarrow\cdots,$$ hence $H_n(C\otimes_RS)=S\neq 0$ in each degree $n\in\mathbb Z$.

$\endgroup$
1
  • 3
    $\begingroup$ If $C$ is bounded on the right, it is a particular case of the Künneth formula, see Bourbaki, Algebra X, § 4, no. 7, Corollary 4. $\endgroup$
    – abx
    Apr 3, 2014 at 5:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.