The ladder operators for number states, $\alpha_{\ell}^{\dagger}$, and $\alpha_{\ell}$ have the following properties when working on mode $\ell$: $$\begin{array}{l} \hat{\alpha}_{\ell}\left|n_{\ell}\right\rangle=\sqrt{n_{\ell}}\left|n_{\ell}-1\right\rangle \quad \\ \hat{\alpha}_{\ell}\left|0_{\ell}\right\rangle=0 \\ \hat{\alpha}_{\ell}^{\dagger}\left|n_{\ell}\right\rangle=\sqrt{n_{\ell}+1}\left|n_{\ell}+1\right\rangle \end{array}$$

And considering a pure wavefunction $\psi_{\ell}$, the expectation value $$\left\langle\alpha_{\ell}^{\dagger} \alpha_{\ell}\right\rangle=\left\langle\psi_{\ell}\left|\alpha_{\ell}^{\dagger} \alpha_{\ell}\right| \psi_{\ell}\right\rangle$$ gives the number of photons in mode $\ell$. However, what would be the effect of $\alpha_{\ell}^{\dagger} \alpha_{\ell}$ on a wavefunction over some different mode $\ell'$? Put differently what would I get if I tried to calculate $\left\langle\psi_{\ell^{\prime}}\left|\alpha_{\ell}^{\dagger} \alpha_{\ell}\right| \psi_{\ell^{\prime}}\right\rangle$?

  • $\begingroup$ What is |psi_l> (with or without ')? $\endgroup$ Mar 30, 2021 at 18:53
  • $\begingroup$ e.g. $|\psi\rangle_{\ell}=\frac{1}{2}\left(\sqrt{3}|n\rangle_{\ell}+|n+1\rangle_{\ell}\right)$ (where I took the $\ell$ outside the bracket with the same meaning) $\endgroup$
    – Logi
    Mar 30, 2021 at 18:59
  • $\begingroup$ But if you have more modes you have to specify what the state is for all modes. And then you know what the operators do. $\endgroup$ Mar 30, 2021 at 19:02
  • $\begingroup$ I consider mixed states using the density matrix $\hat{\rho}=\sum_{i} p_{i}\left|\psi_{i}\right\rangle\left\langle\psi_{i}\right|$ $\endgroup$
    – Logi
    Mar 30, 2021 at 19:03
  • $\begingroup$ Say I want to calculate the expectation value of the operator $\widehat{N}:=\alpha_{\ell}^{\dagger} \alpha_{\ell}+\alpha_{\ell^{\prime}}^{\dagger} \alpha_{\ell^{\prime}}$. The expectation value of a general operator, $\hat{A}$, can be calculated as $\langle\hat{A}\rangle=\sum_{i} p_{i}\left\langle\psi_{i}|\hat{A}| \psi_{i}\right\rangle$. But that means that when calculating $\hat{N}$ I will get "expectation values" of $\alpha_{\ell^{\prime}}^{\dagger} \alpha_{\ell^{\prime}}$ over $\psi_{\ell}$ and vice versa. $\endgroup$
    – Logi
    Mar 30, 2021 at 19:07

1 Answer 1


Multimode states are product states $\vert n_1n_2\ldots,n_{\ell'}\ldots,n_\ell\ldots \rangle =\vert n_1\rangle\otimes\ldots \vert n_{\ell'}\rangle \otimes \ldots \otimes \vert n_\ell\rangle\otimes \ldots $ and the working assumption is probably that if $\langle n_1\ldots,n_\ell=0,\ldots\vert\psi_i\rangle = 0$ then $\hat \alpha_\ell \vert\psi_i\rangle=0$ so $\langle \hat \alpha^\dagger_\ell \hat\alpha_\ell\rangle=0$.

  • $\begingroup$ Now that I think about it, that does make sense, since $\hat{\alpha}_{\ell}\left|0_{\ell}\right\rangle=0$ as written in my question. $\endgroup$
    – Logi
    Mar 30, 2021 at 19:18
  • $\begingroup$ so then it's a matter of figuring out if you multimode state has anything in mode $\ell$. $\endgroup$ Mar 30, 2021 at 19:20
  • $\begingroup$ And $\alpha_{\ell}$ commutes with $|n\rangle_{\ell^{\prime}}$? So that I can write $\left\langle\left. n\right|_{\ell}\left\langle\left. n\right|_{\ell^{\prime}} \alpha_{\ell}^{\dagger} \alpha_{\ell} \mid n\right\rangle_{\ell^{\prime}} \mid n\right\rangle_{\ell}=\left\langle n\left|\alpha_{\ell}^{\dagger} \alpha_{\ell}\right| n\right\rangle_{\ell}\langle n \mid n\rangle_{\ell^{\prime}}$ $\endgroup$
    – Logi
    Mar 30, 2021 at 19:31
  • $\begingroup$ yes. $\hat \alpha_\ell$ and $\hat \alpha^\dagger_\ell$ only "see" the part of the state that refers to mode $\ell$. $\endgroup$ Mar 30, 2021 at 19:35
  • $\begingroup$ Okay, then I think it adds up. Thank you for the help. $\endgroup$
    – Logi
    Mar 30, 2021 at 19:40

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.