Let $B \supseteq A$ be a finite ring extension where $A$ is a domain with field of fractions $k = \operatorname{Frac}(A)$ and $B$ need not be a domain. Further, let $B$ be torsion-free over $A$.

Now I want to show that the ring of total fractions $\operatorname{Frac}(B)$ of $B$ is equal to $B \otimes_A k$.

By assumption, the RHS is contained in the LHS (the elements of $A$ are regular elements of $B$). But why can we write every fraction with denominator contained in $B$ as a fraction where the denominator is an element of $A$?

The case where $B$ is also integral is not that hard, but what if we cannot use that $\operatorname{Frac}(B)$ is a field?

If this is not true in the generality mentioned above, then is it true if we assume $A$ to be a PID (from which we then deduce that $B$ is free over $A$)?

Thank your for your help!


$B\otimes_Ak$ is equal to its own total ring of fractions, because it is a finite dimensional $k$-vector space, so that, if $x$ is a non-zero divisor in $B\otimes_Ak$, the linear map of multiplication by $x$ is injective by definition, hence it is also bijective, and $x$ is invertible.

  • $\begingroup$ This looks great and easy! Thank you very much! But the rank-nullity theorem also holds for non-finite dimensional vector spaces. So do we really need that $B$ is finite over $A$ here? $\endgroup$ – windsheaf Aug 29 '17 at 9:59
  • $\begingroup$ Injective iff surjective iff bijective is valid only in finite dimension. You hve a counter-example with the ring of polynomials: multiplication by $X$ is injective in $K[X]$, but certainly not bijective. $\endgroup$ – Bernard Aug 29 '17 at 10:02

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.