7
$\begingroup$

If $\mathcal H$ is a Hilbert space and $U_n \in B(\mathcal H)$ is a strong-operator convergent sequence of unitary operators, say $U_n\rightarrow U$, is it true that $U$ is unitary? More explicitly, assume $\|(U_n-U)x\|\rightarrow 0$ as $n \rightarrow \infty$ for all $x\in \mathcal H$ and $U_n U_n^*=1= U_n^*U_n$, is it then true that $UU^*=1= U^*U$?

It is clear that $U$ is an isometry, i.e. $\|Ux\|=\|x\|$ for all $x\in\mathcal H$, so it suffices to show that $U$ is onto, or that $U^*$ is an isometry too. Clearly, the assertion then holds true in the finite dimensional setting.

It is clear that $U$ is unitary if the convergence happens in norm, since this implies that $U_n\rightarrow U$ and $U_n^*\rightarrow U^*$ in the strong operator topology, ensuring that $U$ and $U^*$ are isometries (and therefore that $U$ is unitary).

Also, it is clear that $\langle U_n^* x,y\rangle\rightarrow\langle U^* x,y\rangle$. It follows that if $U_n^*$ converges in the strong operator topology, then $U_n^*\rightarrow U^*$, again finishing the argument.

In order to find a counter example one needs to find a unitary sequence $U_n$ which converges in the strong operator topology but does not converge in norm, while $U_n^*$ converges weakly but not strongly. $U=\lim_{n\rightarrow\infty} U_n$ cannot be onto (but must be an isometry) and $U^*=\lim_{n\rightarrow\infty} U_n^*$ cannot be an isometry.

Any help or hint towards a proof or counter example will be much appreciated!

$\endgroup$
7
$\begingroup$

$U$ need not be unitary. Let $U_n$ be the map on $\ell^2$ that sends $e_j\mapsto e_{j+1}$ for $j=1,\ldots, n$, $e_{n+1}\mapsto e_1$, and $U_n$ is the identity on $L(\{e_j: j>n+1\})$.

Then $U_n$ is unitary (it sends an ONB to an ONB again), but $U=S$ (the shift) is only an isometry because $e_1$ is not in the range.

$\endgroup$
  • $\begingroup$ The corresponding statement for self-adjoint operators is true though. $\endgroup$ – user138530 Sep 1 '14 at 21:18
  • $\begingroup$ Of course! Earlier, I considered the sequence $U_n^*$, but discarded it because it does not converge strongly. In hindsight, that was a poor decision... $\endgroup$ – Jonas Dahlbæk Sep 1 '14 at 21:53

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.