3
$\begingroup$

I'm studying homotopical algebra and I'm trying to prove the following fact:

If $R$ is a quasi-Frobenius ring (for $R$-modules one has "projective module$\iff$ injective module") then the stable category of $R$-modules is (equivalent to) the homotopy category of $R$-mod, once the latter has been equipped with the model structure in which cofibrations are monomorphisms, fibrations are epimorphisms and weak equivalences are stable-isomorphisms. In order to do this, since in this model structure each module is both fibrant and cofibrant, it's enough to show that the homotopy relation is the same as the stable-equivalence relation. More precisely one needs to show that $f$ is left (or right, it's the same) homotopic to $g$ $\iff$ $f$ is stable equivalent to $g$.

Now my problem is in the "$\Leftarrow$" part: how can I construct a left, say, homotopy between $f$ and $g$ once I know that $f-g$ factorizes through a projective (hence also injective) $R$-module? Thanks in advance.

$\endgroup$
0
$\begingroup$

I had this idea:

Denote with $\pi:Mod_R\twoheadrightarrow Ho(Mod_R)$ the projection-localization functor to the homotopy category. Since $0\rightarrowtail M\twoheadrightarrow 0$ every module is both fibrant and cofibrant (or bifibrant), and thus $Ho(Mod_R)$ is isomorphic (not just equivalent) to the classical homotopy category of $Mod_R$ (I guess you read Hovey's book "Model Categories", as it's the only reference I know on this subject, so you're probably familiar with this terminology). In particular two maps $f,g:M\rightarrow N$ are (left-right) homotopic if and only if they are identified by $\pi$.

Moreover, notice that a module $B$ is projective-injective (or bijective) if and only if $0\rightarrow B$ is a trivial cofibration, and if and only if $B\rightarrow 0$ is a trivial fibration: in particular it's a zero object in the Homotopy category.

Then if the morphism $f-g:M\rightarrow N$ factors through a bijective object, necessarily $\pi(f-g)=0$, so all you have to show is that $\pi$ is an additive functor (you can find the proof here).

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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