MFW I realized most quantum experiments in introductory...

MFW I realized most quantum experiments in introductory books are just lies to children, they're thought experiments at best. The literal setups as described either never work in practice, or don't really prove their results due to loopholes.

An interferometer with a single photon is the worst offender. Ordinary light sources like lamps or lasers can never work as single-photon sources. Even if you add a huge attenuator and make the light intensity lower than the energy of a single photon, there's still a non-negligible probability that it will emit multiple photons. In fact all experiments before the 1980s had this loophole. Creating true "anti-bunching" single-photon sources requires esoteric techniques. Still, this probability can only approach but not reach 0% in practical devices.

For the same reason, the BB84 quantum key distribution protocol often don't work in practice in its original form. Source or detector imperfections make the original protocol often totally broken. https://physics.unm.edu/Courses/Deutsch/Phys566F19/Lectures/Aspect%20the%20first%20single%20photons%20sources.pdf

10.2 FEEBLE LIGHT VS. SINGLE PHOTON

10.2.1 In Search of Feeble Lightâs Wave-Like Properties:

A Short Historical Review

[...]

This pioneering experiment was followed by a long series of diffraction [27] and interference [28-33] experiments with light emitted by strongly attenuated ordinary light sources, mostly discharge lamps, so [...] It was thus concluded that "there was only one photon at a time in the interferometer," and the observation of fringes was then considered a demonstration that "a photon interferes with itself." [...]

In the 1970s, the general wisdom was then that “single-photon wave-like behavior” had been experimentally demonstrated. However, revisiting that question in the early 1980s, we realized that, according to the formalism of modern quantum optics as developed by Glauber [12,34,35], none of the experiments cited above could be considered a demonstration of single particle interference, because in none of these experiments the light used could be considered as a single-photon wavepacket. This led us to perform the experiments of [5], presented in Sections 10.3 and 10.4. In the rest of this section, we use the formalism of quantum optics to highlight the difference between single-photon wavepackets and all the types of light used in the experiments above.

The probability of a double detection is thus strictly null in the case of a single-photon wavepacket. This property ("anticorrelation") is not surprising if one remembers that the number of photons is a good quantum number, and its value is 1. Since a photodetection amounts to destroying a photon, there is no photon left to allow for a second detection.

In contrast, in a semi-classical wavepacket the number of photons is not a good quantum number, since |Ψ_qc⟩ is not an eigenstate of N-hat , and the probability to have two photons is not null. It is therefore not surprising that one can have two photodetections.

This difference allows one to make an experimental distinction between a true single-photon wavepacket, and a quasi-classical wavepacket, even when attenuated enough that the average number of photons is much less than 1. One can then ask: can such a difference be observed, when one takes into account experimental inefficiencies and noise? We will see in Section 10.3 that it is indeed possible to establish a quantitative criterion that renders the distinction presented above fully operational, leading to practical tests in realistic experiments. But before addressing that question, we will ask, still from a theoretical point of view, whether various kinds of strongly attenuated light beams may exhibit an anomalously small rate of double photodetection, by comparison to what is expected for a classical wave.

10.2.8 Conclusion: What is Single-Photon Light?

In this section, we have shown that a genuine single-photon wavepacket, i.e., a one-photon state emitted at a well-known time, exhibits a characteristic behavior, the fact that it cannot be detected jointly by two photodetectors (anticorrelation). Such a behavior is not expected in the case of an attenuated beam from a classical lamp, including the case of a discharge lamp where one has single-photon wavepackets shorter than the average time between the wavepackets. The paradoxical behavior in the latter case is resolved when we realize that the problem is the fact that one has no information about the time when any single-photon wavepacket is emitted. We can thus conclude that it is not enough to have single-photon wavepackets to have single-photon light. We need in addition to know at which time each single-photon wavepacket is emitted [40]. This is the case for the two types of sources described below: heralded single-photon sources on the one hand, and on-demand single-photon sources on the other hand.

Like 16 Aug 2022 at 10:41 | Open on cybre.space