5
$\begingroup$

If $\rho$ is the density matrix of a system then $Tr(\rho ^2) \leq 1$. If the equality holds the system is in a pure state and it is in a mixed state otherwise. But, what is the physical meaning of $Tr(\rho^2)$ ?

$Tr(\rho) = 1$ for all valid density matrices. This stems from the normalization constraint that total probability must be one. Is there any such interpretation for $Tr(\rho ^2)$ ?

$\endgroup$
5
  • 1
    $\begingroup$ This quantity can be measured experimentally, you might be interested by this paper arxiv.org/pdf/1509.01160v1.pdf $\endgroup$
    – Adam
    Commented Mar 30, 2016 at 8:41
  • $\begingroup$ What do you mean by "physical meaning"? -- Note that one could well argue that $\mathrm{tr}(\rho)$ has no physical meaning, since states must always be normalized. $\endgroup$ Commented Mar 30, 2016 at 11:34
  • $\begingroup$ @Adam Note that this quantity can only be measured given at least two copies of $\rho$. $\endgroup$ Commented Mar 30, 2016 at 11:34
  • $\begingroup$ One possible interpretation is the overlap of $\rho\otimes \rho$ with the symmetric subspace. Would this be a satisfactory answer?? (After all, $\rho$ has to be pure in order for $\rho\otimes \rho$ to be fully contained in the symmetric subspace.) $\endgroup$ Commented Mar 30, 2016 at 11:34
  • $\begingroup$ It's not clear what you're asking. The inequality is saturated if and only if $\rho$ is pure, what more of a "physical meaning" do you seek? $\endgroup$
    – ACuriousMind
    Commented Mar 30, 2016 at 11:50

2 Answers 2

3
$\begingroup$

You can interpret $1-Tr[\rho^2]$ as a kind of entropy. If you think of $\rho$ as a classical mixture of quantum states, then $Tr[\rho^2] = \sum_i p_i^2$, where $p_i$ is probability of being found in eigenstate $i$. The more dispersed the states, the smaller the quantity. So for a completely mixed state you have $1/n$, whereas for a pure state you have 1.

$\endgroup$
1
$\begingroup$

If the system is in a pure state, the density operator is just the projector onto that state, and so $\rho^2 = \rho$. Since, $\mathrm{tr}\,\rho=1$, in a pure state clearly $\mathrm{tr}\,\rho^2=1$.

Since probabilities must be non-negative, $\rho$ has only non-negative eigenvalues ($\rho$ is positive semidefinite). With this and the trace/total probability condition, each eigenvalue $\lambda_i$ for a mixed state satisfies $\lambda_i < 1$. Then $\mathrm{tr}\,\rho^2<1$ for a mixed state.

$\endgroup$
4
  • 1
    $\begingroup$ It's not clear to me how this constitutes an interpretation of the quantity $\mathrm{tr}(\rho^2)$. $\endgroup$ Commented Mar 30, 2016 at 7:35
  • 1
    $\begingroup$ It does: Pure states are 1, therefore it measures some degree of "purity". $\endgroup$
    – Martin
    Commented Mar 30, 2016 at 8:52
  • 1
    $\begingroup$ @Martin But I understand the OP understands this. He wants some interpretation "beyond that" (whatever that might be -- I'm not saying that the question makes sense, I just don't feel that this provides an answer). $\endgroup$ Commented Mar 30, 2016 at 11:33
  • $\begingroup$ @NorbertSchuch: You are right. I overread that in the original question. $\endgroup$
    – Martin
    Commented Mar 30, 2016 at 11:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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