Equation 7.71 of Nielsen and Chuang's Quantum Computation and Quantum Information gives the Hamiltonian for a two level atom and single mode photons in a cavity as:

$H = \hbarωN + δZ + g(a^†σ_− + aσ_+)$

where $ω$ is the photon frequency, $δ = (ω_0 - ω)/2$ with $ω_0$ the atomic transition frequency. $Z$ is the Pauli Z matrix; g is a coupling constant; $a^†$ and $a$ are creation and annihilation operators for photons, and $σ_−$ and $σ_+$ are lowering and raising operators for the atomic energy levels.

Neglecting the $N$ term (since it only contributes a fixed phase) and considering a single excitation in the (photon) field mode, this Hamiltonian is re-written as

$$H = -\begin{bmatrix}\delta&0&0\\0&\delta&g\\0&g&-\delta\end{bmatrix}$$

I've tried but been unable to derive this Hamiltonian from equation 7.71 above. $H$ in 7.71 should be a 2 x 2 matrix as Pauli matrices are 2 x 2. But here it is written as a 3 x 3 matrix. Can you please explain how this can be obtained? Thanks.


While the Pauli-$Z$ matrix is a 2 x 2 matrix, there are more basis states that need to be considered, namely the vacuum. The atomic basis states are $\left|0\right>$ and $\left|1\right>$, representing the number of excitations in the atom (as it's only a two-level atom, there can't be more than 1 excitation) and this is what $Z$ acts on. Similarly, the number of excitations (i.e. photons) in the field forms another basis, denoted by $\left|N\right>$.

However, we're only considering a cavity with a total of number of photons $\leq 1$, so the possible bases are, ordering the Hilbert space by $\mathcal H_{\text{photons}}\otimes\mathcal H_{\text{atom}}$, $\left|00\right>, \,\left|10\right>$ and $\left|01\right>$, as all the other possible states have $> 1$ photon.

The easiest way to go about figuring out what the Hamiltonian on this system looks like is to look at what each operator does to each state, where $N=aa^\dagger$ is ignored.

That is, $Z$ acts on the 'atomic' Hilbert space as $\delta Z\left|0\right> = \delta\left|0\right>$ and $\delta Z\left|1\right> = -\delta\left|1\right>$, $\sigma_+$ acts on the same space as $\sigma_+\left|0\right> = \left|1\right>$ and $\sigma_+\left|1\right> = 0$ and $\sigma_-$ acts as $\sigma_-\left|1\right> = \left|0\right>$ and $\sigma_-\left|0\right> = 0$. Similarly, on the photonic Hilbert space, $a\left|0\right> = 0$, $a\left|1\right> = \left|0\right>$, $a^\dagger\left|0\right> = \left|1\right>$ and $a^\dagger\left|1\right> = \sqrt{2}\left|2\right>$.

As a result, the term involving $g$ acts on the combined system as: $$\begin{align*}g\left(a^\dagger\sigma_- + a\sigma_+\right)\left|00\right> &= 0\\ g\left(a^\dagger\sigma_- + a\sigma_+\right)\left|10\right> &= g\left|01\right>\\ g\left(a^\dagger\sigma_- + a\sigma_+\right)\left|01\right> &= g\left|10\right>\end{align*}$$ and the $\delta Z$ term similarly gives $$\begin{align*}\delta Z\left|00\right>&=\delta\left|00\right>\\\delta Z\left|10\right>&=\delta\left|10\right>\\\delta Z\left|01\right>&=-\delta\left|01\right>\end{align*}.$$

Writing this in matrix form with the basis $\left|00\right>, \,\left|10\right>$ and $\left|01\right>$ as above gives equation 7.76 as expected: $$H = \begin{bmatrix}\delta&0&0\\0&\delta&g\\0&g&-\delta\end{bmatrix}.$$

It does appear that I'm out in what I get for $H$ by a minus sign, but looking at equation 7.77, which has $U=e^{-iHt} = e^{-i\delta t}\left|00\rangle\langle00\right|+...$ suggests that this is presumably a typo (in my version, anyway), unless there's something I'm missing.

  • $\begingroup$ Thanks a lot for the detailed answer. In my copy too the first delta has a negative sign. $\endgroup$
    – Bashir
    Jul 3 '19 at 23:35

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.