3
$\begingroup$

I am trying to construct a controlled Z gate using elementary gates. This is what I have so far: \begin{pmatrix} -i & 0 & 0 & 0\\ 0 & -1 & 0 & 0\\ 0 & 0 & 1 & 0\\ 0 & 0 & 0 & -i \end{pmatrix} can someone help me figure out how to get rid of the unwanted phase shift?

I can use combinations of a single-qubit Hamiltonian $\hat{H}=\frac{1}{2}\Omega_x\hat{X}+\frac{1}{2}\Omega_y\hat{Y}+\frac{1}{2}\Omega_z\hat{Z}$ and $\hat{H}=\frac{1}{2}\Omega_{zz}\hat{Z}\otimes\hat{Z}$. I can construct the $\hat{I}\otimes\hat{H}$ matrix above but I cannot construct the CNOT gate so far.

$\endgroup$
3
  • $\begingroup$ Are you looking for ways to make a $CZ$ gate as the evolution under these Hamiltonians? (i.e. $CZ = exp(H_{cons})$ with $H_{cons}$ some constructed Hamiltonian from these two base Hamiltonians? In that case check what is $exp(\frac{1}{2}\Omega_{zz}Z \otimes Z)$ and equate this to the $CZ$ gate; you will find that they are equal for some $\Omega_{zz}$. $\endgroup$
    – JSdJ
    Sep 20 at 9:36
  • $\begingroup$ If you in fact want an identity like $CZ = H_{cons}$ with $H_{cons}$ some constructed Hamiltonian from these two base Hamiltonians, it will be harder to make this, as $CZ$ is a unitary operation, and these wo Hamiltonians are Hermitian, which might give you some problems to construct them. $\endgroup$
    – JSdJ
    Sep 20 at 9:38
  • $\begingroup$ What I'm saying is, you ask how to make the $CZ$ using gates from two base elements, but you provide two Hamiltonians, which are generally speaking not (unitary) gates - they invoke unitary operations through the Schrodinger equation, where $U = exp(H)$. $\endgroup$
    – JSdJ
    Sep 20 at 9:41
4
$\begingroup$

It depends on what you mean by "elementary gates". Using only Pauli gates (that is, using only $X$, $Y$ and $Z$ gates), you cannot build a controlled $Z$ gate. One way to prove it is simply to show that the controlled $Z$ gate cannot be written as $A\otimes B$, with $A$ and $B$ being unitary matrices.

However, if we allow ourselves to use gates from the Clifford group, it is easy to build a controlled $Z$ gate using $I\otimes H$ gates and a CNOT gate: $$\begin{align}(I\otimes H)\mathsf{C}X(I\otimes H) &= \frac12\begin{pmatrix}1&1&0&0\\1&-1&0&0\\0&0&1&1\\0&0&1&-1\end{pmatrix}\begin{pmatrix}1&0&0&0\\0&1&0&0\\0&0&0&1\\0&0&1&0\end{pmatrix}\begin{pmatrix}1&1&0&0\\1&-1&0&0\\0&0&1&1\\0&0&1&-1\end{pmatrix}\\&=\frac12\begin{pmatrix}1&1&0&0\\1&-1&0&0\\0&0&1&1\\0&0&-1&1\end{pmatrix}\begin{pmatrix}1&1&0&0\\1&-1&0&0\\0&0&1&1\\0&0&1&-1\end{pmatrix}\\&=\begin{pmatrix}1&0&0&0\\0&1&0&0\\0&0&1&0\\0&0&0&-1\end{pmatrix}\end{align}$$ Note that the circuit we are building is a CNOT applied on the second qubit, controlled by the first one, and surrounded by two $H$ gates on the second qubit.

It is however way easier to reason on the basis states. A common relation that may be useful to know is $HXH=Z$. Thus, if the first qubit is in state $|0\rangle$, we apply on the second qubit $HH=I$, but if the first qubit is in state $|1\rangle$, we apply $HXH=Z$, which is the behavior of a controlled-$Z$ gate.

$\endgroup$
3
  • $\begingroup$ Thank you for the response Tristan. I am able to use the Pauli gates and a Hamiltonian of the form $ \hat{H}=\frac{1}{2}\theta\hat{Z}\otimes\hat{Z} $. I am not sure if I can construct the same matrices in the Clifford group from these. I will give it a try. $\endgroup$
    – Anne
    Sep 19 at 22:37
  • $\begingroup$ Based on what Tristan has said, I wonder if the problem is solvable since all of the matrices are unitary. $\endgroup$
    – Anne
    Sep 19 at 23:26
  • 1
    $\begingroup$ @Anne can you develop about how you are allowed to use this Hamiltonian directly in your question? The beginning of my answer supposed that you could only use one-qubit gates (since I thought you were only allowed to use Pauli gates). $\endgroup$ Sep 19 at 23:43

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.