3
$\begingroup$

Show that there is an unique norm on $M_n(A)$ making it a $C^*$-algebra, where $A$ is a $C^*$-algebra itself.

It is enough to show this when $A$ is a concrete C^*-algebra. Suppose we are able to find a norm $\|.\|_1$ for such a case. Because for any $C^*$-algebra $B$, there is an isometric isomorphism $\pi$ from $B$ to a concrete $C^*$-algebra. Then for $X \in M_n(B)$, we define $\|X\|=\|\pi(X)\|_1$, where $$\pi(X)=\begin{bmatrix} \pi(b_{11})&\pi(b_{12})&\ldots&\pi(b_{1n})\\ \vdots&\vdots& \vdots &\vdots\\ \pi(b_{n1})&\pi(b_{n2})&\ldots&\pi(b_{nn})\\ \end{bmatrix}$$

when $$X=\begin{bmatrix} (b_{11})&(b_{12})&\ldots&(b_{1n})\\ \vdots &\vdots& \vdots &\vdots\\ (b_{n1})&(b_{n2})&\ldots&(b_{nn})\\ \end{bmatrix}$$

Then $$\|X^*X\|=\|\pi(X^*X)\|_1=\|(\pi(X))^*\pi(X)\|_1=\|\pi(X)\|_1^2=\|X\|^2$$

So the problem now boils down to finding a norm on $M_n(A$) when $A$ is a concrete C^*-algebra. Let $X \in M_n(A)$. Suppose that $X=\begin{bmatrix} (b_{11})&(b_{12})&\ldots&(b_{1n})\\ \vdots &\vdots& \vdots &\vdots\\ (b_{n1})&(b_{n2})&\ldots&(b_{nn})\\ \end{bmatrix}$

Then for $$a=(a_1,a_2,\ldots,a_n) \in H^n, Xa=(\sum_{j=1}^n b_{1j}a_j,\ldots,\sum_{j=1}^nb_{nj}a_j)$$

Naturally I think $$\|Xa\|^2=\langle \sum_{j=1}^nb_{1j}a_j, \sum_{j=1}^nb_{1j}a_j \rangle + \ldots + \langle \sum_{j=1}^nb_{nj}a_j, \sum_{j=1}^nb_{nj}a_j \rangle=\sum_{i=1}^n \left\|\sum_{j=1}^nb_{ij}a_j\right\|^2$$ So I should define $$\|X\|=\sup_{\|a\| \le 1}\|Xa\|$$

where $$\|a\|=\left(\sum_{i=1}^n\|a_i\|^2\right)^{1/2}$$

I am not sure. Is there any other norm? Thanks for the help!!

$\endgroup$
2
$\begingroup$

What you did is correct (and canonical), minus a very minor typo in the last line (you are missing either a square on the left-hand-side, or a square root on the right-hand-side).

As for uniqueness: yes, that's the only possible norm. Once you have a C$^*$-algebra (not just a $*$-algebra) the C$^*$-norm is unique. This is due to $$ \|x\|^2=\|x^*x\|=\text{spr}\,(x ^*x), $$ which shows that the algebraic structure determines the norm. So this only depends on showing that $M_n(A)$ is closed when $A$ is a C$^*$-algebra.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Done. Thanks, Jonas. $\endgroup$ – Martin Argerami Apr 11 '17 at 19:38

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.