Within the standard model, photons are point particles, i.e., with no spatial distribution.

On the other hand, classical electromagnetic modes have field distributions.

Suppose there are two different single mode nanophotonic waveguides. The two waveguides are gradually bent and become adjacent to each other to form a coupler. enter image description here

Two classical electromagnetic waves propagating along the two waveguides can couple to each other. For two photons, can they show the two photon interference effect in this structure?

If they can, how to calculate the probability that exactly one photon is detected at each exit, taking into consideration that the two waveguide eigenmodes have different field distributions?

I notice that, in the second quantization formalism, the creation operator is $$\hat{A}^{\dagger}(\alpha)=\int_{0}^{\infty} \mathrm{d} \omega \alpha(\omega) \hat{a}^{\dagger}(\omega),$$ where $\alpha(\omega)$ is the frequency distribution function of the excitation. Does this mean that the classical electromagnetic field distribution doesn't matter for quantum optics and the only important property is the frequency?

Update: In this paper, the probability that exactly one photon is detected at each exit is calculated as $$P\left(1_{\mathrm{a}}, 1_{\mathrm{b}}\right)=|t|^{4}+|r|^{4}+\left[t^{2} r^{* 2}+r^{2} t^{* 2}\right] I,$$ where $t$ and $r$ are the complex transmission and reflection coefficients. These coefficients can be calculated using the coupled mode theory which certainly takes the (classical) electromagnetic field distribution into consideration. This looks like a feasible formalism.

However, this calculation should be equivalent to some second-quantization formalism that involves the spatial field distribution, because this calculates the two photon interference possibility which should be a purely quantum mechanical quantity.

I wonder if there has been any literature that thoroughly established something like "waveguide quantum optics" in the language of second quantization.

  • 4
    $\begingroup$ A side note about the 1st sentence: The popular idea of a photon as something that propagates through space as a pointlike object is a figment of the imagination. Relativistic QFTs, such as the standard model, don't say that. They do say things about locality of interactions between fields, about structure-functions in scattering experiments, and about the behavior of localized detectors. The words "point particle" have been used for all of those things, even though those things are only loosely related to each other. QFT doesn't imply the popular idea, and neither does experiment. $\endgroup$ Apr 19, 2021 at 4:31
  • $\begingroup$ Creation/annigilation operators can be viewed as the coefficients in the expansion of the electromagnetic field in eigenmodes (which are usually taken to be plane waves in high theory, but are cavity or waveguide eigenmodes in applied research). $\endgroup$
    – Roger V.
    Apr 19, 2021 at 8:31
  • $\begingroup$ @ChiralAnomaly The QFT of the standard model has axiomatically point particles in the table . en.wikipedia.org/wiki/Standard_Model .also the single particle detectors have experimentally defined footprints of particles en.wikipedia.org/wiki/… , with localized dxdydzdt when detected. This is your personal opinion . $\endgroup$
    – anna v
    Apr 19, 2021 at 10:29
  • 1
    $\begingroup$ @annav More succinctly: the popular idea of a photon as a pointlike object is an example of a hidden variables interpretation. Quantum theory works just fine without any such interpretation. Physics is an experimental discipline, and a theory's job is to reproduce the experimental facts using as few assumptions as possible. We should not burden a theory with extra ideas that don't improve its agreement with experiment. $\endgroup$ Apr 19, 2021 at 13:46
  • 1
    $\begingroup$ @annav I very much respect the fact that you are an experimentalist. My perspective comes from trying to pay close attention to what experiments actually say, and trying not to go beyond that. Everything revolves around experiment, and you are a great asset to this site! $\endgroup$ Apr 19, 2021 at 16:49

1 Answer 1


Firstly, the optical field can have a spatial distribution, even though we observe them as point particles called photons. The spatial distribution is the wave functions that gives us a probability amplitude to observe a photon at a certain location. For multi-photon state, the situation can become a bit more complicated when we start to make correlated observations, because of the possibility for entanglement. The wave function of the photon represents its spatiotemporal properties.

For the waveguide structure, one can assume that it behaves as a beamsplitter. If we can assume that the probability for a photon to jump to the other guide is equal to the probability to stay in the guide in which it arrived, then one should be able to observe Hong-Ou-Mandel (HOM) interference, which is a form of two-photon interference. The effect of the HOM interference depends on the spatiotemporal properties of the wave functions. In a single-mode guide, the spatial properties are fixed and only the temporal properties would play a role.

If the spatiotemporal properties in the two guides are exactly the same, one would find that both photons would always exist from the same guide. On the other hand, if they have different spatiotemporal properties, one can find that two photons can exit from the two different guides. There combined state would be an entangled state, the singlet Bell state. Due to the entanglement, this state is not a classical state and can therefore not be described by the classical optics.

Not sure I've answered all your questions. Please let me know if I need to add anything.


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.