2
$\begingroup$

I know that I can write a single qubit state in terms of polar co-ordinates $(r,\theta,\phi)$ on a Bloch sphere. \begin{equation} \rho = \begin{pmatrix} \frac{1+r \cos\theta}{2} &\frac{r \exp(-i\phi)\sin\theta}{2} \\ \frac{r \exp(i\phi)\sin\theta}{2} &\frac{1-r \cos\theta}{2} \nonumber \end{pmatrix} \end{equation} Are this kind of polar decomposition exist for two qubits system?

$\endgroup$
1
$\begingroup$

A valid density operator is any Hermitian, trace 1, matrix (with complex entries) and all eigenvalues between 0 and 1. Any two qubit system may be represented therefore by a Hermitian, trace 1 4x4 matrix.

Your qubit representation could be rewritten, more suggestively as:

\begin{align} \rho &= \frac{1}{2}\left(\operatorname{I} + a_1 \sigma_x + a_2 \sigma_y + a_3\sigma_z \right)\\ &= \frac{1}{2}\left(\operatorname{I} + \vec{a} \cdot\vec{\sigma}\right) \end{align}

Where $\operatorname{I}$ is the 2x2 identity matrix and the $\sigma_i$ are the Pauli matrices which are a basis for the space of Hermitian, trace 0 2x2 complex matrices.

Similarly to qubits we can decompose any n-dimensional quantum system using the identity and any basis for the space of Hermitian, traceless nxn complex matrices, a good choice of basis is the Gell-Mann matrices, and a construction in dimension n is given here.

Since they are Hilbert-Schmidt orthogonal you can find the coefficients $a_i$ for your state $\rho$ by taking the inner product:

$$a_i = \operatorname{tr}\sigma_i \rho$$

If you like you can then rewrite the vector $\vec{a}$ in any coordinate system you like (including polars).

There is a problem however because although every valid density operator may be rewritten in this way (with $\vec{a}$ being a n-dimensional sub-normalised vector) it is not the case that every operator of that form will be a valid density operator, the problem is that some of your eigenvalues will become negative and others will become greater than 1 to balance this out. This doesn't happen for qubits (any operator of the form $\rho$ I wrote above will be a valid density operator) and it makes working in higher dimensions fairly annoying sometimes.

$\endgroup$
1
$\begingroup$

Yes. We can do this for any number of qubits using N dimensional spherical coordinates. For two qubits we can write a general density matrix as a linear combination of direct products of Pauli matrices and identity , $$ \rho = \sum_{ij=0}^3 a_{ij}~ \sigma_{i} \otimes \sigma_{j}$$.

Here $\sigma_{0} = I$, and the rest are the usual Pauli matrices. For a valid density matrix $a_{00} = 1/4$. It can be shown that all valid density matrices lie inside a higher dimensional sphere centered at the coordinate corresponding to $a_{00}$. And these points can be represented using higher dimensional spherical coordinates. But unlike the Bloch sphere for a single qubit, all points do not represent valid density matrices as there can be matrices with negative eigenvalues. In all higher dimensions it is a convex body inside this sphere. A detailed explanation can be found here and here.

$\endgroup$
-1
$\begingroup$

A one qubit state may be written in general as \begin{equation} |\psi_1\rangle=\alpha|0\rangle+\beta|1\rangle \end{equation} where $\alpha,\beta \in \mathbb C$ and there is the further restriction that $\langle \psi_1|\psi_1\rangle=|\alpha|^2+|\beta|^2=1$. However, only the relative phase between $|0\rangle$ and $|1\rangle$ is physically meaningful (quantum states are rays in Hilbert space), so we may factor out a phase $e^{i\theta}$. Let's choose that phase so that $a=e^{-i\vartheta}\alpha$ is real while $b=e^{-i\vartheta} \beta$ can remain complex. Our state is then \begin{equation} |\psi_1\rangle=e^{i\vartheta}(a|0\rangle+b|1\rangle). \end{equation} The previous condition is now $\langle \psi_1|\psi_1\rangle=|a|^2+|b|^2=1$. We have three real parameters and one constraint and so we may say that our state $|\psi_1\rangle$ lives on the unit 2-sphere.

Analagously, a two qubit state may be written in general as \begin{equation} |\psi_2\rangle=a|00\rangle+b|01\rangle+c|10\rangle+d|11\rangle \end{equation} where we can choose $a\in \mathbb R$ and $b,c,d \in \mathbb C$ with the further restriction that $|a|^2 + |b|^2 + |c|^2 + |d|^2 =1$. In this case there are seven free parameters and one constraint so a two-qubit system would live on the unit 6-sphere and have a corresponding higher dimensional polar decomposition.

$\endgroup$
  • 1
    $\begingroup$ This answer is for pure states only, for general mixed states the problem is more complicated. $\endgroup$ – or1426 Jun 24 '16 at 11:09

protected by Qmechanic Jun 24 '16 at 15:44

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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