1
$\begingroup$

How can I simulate the following 2×2 Hamiltonian $$ e^{i\begin{bmatrix} 8 & 6+i \\ 6-i & -1\end{bmatrix}}|\Psi\rangle$$

ie. how to rewrite that matrix exponential in terms of other, well-used quantum gates such as X, Y, Z, CNOT, and other standard gates in quantum computing?


Proof of identity due to @DaftWullie's answer:

$$\boxed{(aX+bY+cZ)^2 = (a^2+b^2+c^2)I}$$

$$\left(a\begin{bmatrix} 0 & 1 \\ 1 & 0 \end{bmatrix} + b\begin{bmatrix} 0 & -i \\ i & 0 \end{bmatrix} + c\begin{bmatrix} 1 & 0 \\ 0 & -1 \end{bmatrix}\right)^2 = \begin{bmatrix} c & a-bi \\ a+bi & -c \end{bmatrix}\begin{bmatrix} c & a-bi \\ a+bi & -c \end{bmatrix}$$

$$=\begin{bmatrix} c^2 + (a^2 - (bi)^2) & ca-cbi - ca + cbi \\ ca+cbi-ca-cbi & (a^2-(bi)^2 + c^2 \end{bmatrix}=\begin{bmatrix} a^2+b^2+c^2 & 0 \\ 0 & a^2+b^2+c^2 \end{bmatrix} = (a^2+b^2+c^2)I$$

$\endgroup$

1 Answer 1

3
$\begingroup$

Start by decomposing the Hamiltonian in terms of the Pauli matrices: $$ \begin{bmatrix} 8 & 6+i \\ 6-i & -1\end{bmatrix}=6X-Y+\frac{7}{2}I+\frac{9}{2}Z. $$ Now, you can rewrite this as $$ \frac{7}{2}I+\sqrt{6^2+1^2+\left(\frac{9}{2}\right)^2}\sigma $$ where $$ \sigma=\frac{6X-Y+\frac{9}{2}Z}{\sqrt{6^2+1^2+\left(\frac{9}{2}\right)^2}}. $$ Why have I done it like this? The key property here is that $\sigma^2=I$, which means I can rewrite the whole evolution as $$ e^{7i/2}e^{i\sigma\sqrt{229}/2}=e^{7i/2}\left(\cos(\sqrt{229}/2)I+i\sin(\sqrt{229}/2)\sigma\right). $$ This is a rotation by an angle $\sqrt{229}\text{ mod }2\pi$ about the axis $\sigma$. You can neglect the global phase has have no relevance (unless you intend this to be a controlled-unitary). You can now directly decompose this in terms of Euler anglers, i.e. a $Z$ rotation, followed by and $X$ rotation followed by a $Z$ rotation. You just have to find the correct angles.

$\endgroup$
9
  • $\begingroup$ thank you very much! Just 1 small question about $\sigma^2=I$, are you using here some kind of Pauli matrices identity that $(aX+bY+cZ)^2 = (\sqrt{a^2+b^2+c^2})I$ for all $a,b,c$? $\endgroup$
    – James
    Jun 21 at 8:05
  • 1
    $\begingroup$ $(aX+bY+cZ)^2=(a^2+b^2+c^2)I$ $\endgroup$
    – DaftWullie
    Jun 21 at 8:07
  • 1
    $\begingroup$ I'm not good at names of things! You may be able to find a more general statement more easily: $(\vec{n}\cdot\vec{\sigma})\cdot (\vec{m}\cdot\vec{\sigma})=(\vec{n}\cdot\vec{m})I+(\vec{n}\times\vec{m})\cdot\vec{\sigma}$. $\endgroup$
    – DaftWullie
    Jun 21 at 8:13
  • 1
    $\begingroup$ See en.wikipedia.org/wiki/… $\endgroup$
    – DaftWullie
    Jun 21 at 8:17
  • 1
    $\begingroup$ en.wikipedia.org/wiki/… $\endgroup$
    – DaftWullie
    Jun 21 at 10:21

Your Answer

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

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