Let $A$ and $B$ be $C^*$-algebras. Suppose that there exists a projection $p$ in $\mathcal{M}(B)$, the multiplier algebra of $B$, such that $A=pBp$. That is, $A$ is a corner of $B$.

Question: Is it true that the corner $p(\mathcal{M}(B))p$ contains the multiplier algebra of $A$ (which we view as a subalgebra of $B$)? In other words, does every multiplier of the corner come from the corner of the multiplier algebra?


1 Answer 1


Yes. We have $$ M(A) = M(pBp) = pBp \subseteq pM(B)p. $$ The second equality holds since $pBp$ is a unital C*-algebra.

Corrected version: There is an obvious map $\iota \colon pBp \hookrightarrow p M(B)p$. This inclusion extends to the multiplieralgebra $\bar \iota \colon M(pBp) \to pM(B)p$. The extension exists since $\iota$ is non-degenerate. Indeed, if $(e_n)$ is an approximate unit for $B$, then $pe_np \to p$ strictly, which is the unit of $pM(B)p$. Furthermore, it is clear that $\bar \iota$ is still injective.

  • $\begingroup$ Why must $pBp$ be unital if we start off with $B$ non-unital? $\endgroup$
    – geometricK
    Jun 10, 2019 at 14:22
  • $\begingroup$ Ah okay, I made a mistake. It is not the case that that $pBp$ is strictly closed in $M(B)$, e.g. $ \mathbb K = 1_{B(H)} \mathbb K 1_{B(H)}$. $\endgroup$
    – Epsilon
    Jun 10, 2019 at 14:49
  • $\begingroup$ I have corrected the proof. Thanks for pointing out. $\endgroup$
    – Epsilon
    Jun 10, 2019 at 14:56

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.