This experiment aims to demonstrate the quantized behavior of light by showing that a single photon cannot be detected simultaneously in both the reflected and transmitted beams of a beamsplitter. The experiment directs single photons at a beamsplitter and records detection times and correlations between the reflected and transmitted beams. If reflected and transmitted photons are never detected simultaneously, it indicates that photons behave as discrete quantized packets rather than classical waves, supporting Einstein's hypothesis. To account for detector noise, a downconversion crystal and additional "gate" detector are used to filter out unwanted events. The experimenters compute a correlation coefficient and find that it is less than one, supporting the quantum picture of photons over the classical wave description.

1. ProgressExperiment
Demonstrating the quantized behavior of light
Tammy Nguyen, Seonyeong Ha, Jack Maseberg
Introduction
[1] J. J. Thorn et. al., “Observing the quantum behavior of light
in an undergraduate laboratory”, Am. J. Phys. 72, 1210 (2004).
[2] P. Grangier, G. Roger, and A. Aspect, ‘‘Experimental
evidence for a photon anticorrelation effect on a beam splitter: A
new light on single-photon
interferences,’’ Europhys. Lett. 1, 173 (1986).0
In 1905, Einstein was able to understand the
photoelectric experiment by assuming that light was
quantized. He postulated that the energy associated
with each discreet particle of light (photon) is 𝐸 =
ℎ𝑓 =
ℎ𝑣
𝜆
, where ℎ is Planck’s constant, 𝑓 is the
frequency of the electromagnetic wave, 𝑣 is the
speed of light, and 𝜆 is the wavelength. Wave-
particle duality is now widely accepted, but it turns
out that contrary to popular belief, Einstein’s
description of the photoelectric effect does not
require the existence of quantized light particles [1].
(It is possible to explain the photoelectric results by
assuming that the detector atoms are quantized, and
not the electromagnetic field.)
In 1986 Grangier, Roger, and Aspect performed
an experiment to unequivocally demonstrate that
photons really are quantized energy packets [2].
They were able to show that “a single photon can
only be detected once.” They did this by directing
single photons (low intensity light) at a beamsplitter,
detecting and recording those reflected and
transmitted photons as a function of time. This
allowed them to look for correlations between the
reflected and transmitted signals. If a reflected and
transmitted photon could be detected simultaneously,
then that would indicate that light behaved like
classical electromagnetic waves. Conversely, if
reflected and transmitted photons can never be
detected simultaneously, then the photon must be a
quantized packet of energy as Einstein presumed.
Fig. 1: A beamsplitter followed by “transmitted” (T) and
“reflected” (R) detectors.
If we possessed ideal (perfect) single photon
detectors, then this experiment would be
straightforward. Unfortunately, detecting single
photons involves some noise, or “dark counts”. In
order to help mitigate this noise, we must complicate
the experiment by employing a down conversion
crystal and a third “gate” detector (see Fig. 2).
Figure 3 shows the three possible observations of
correlated events from the G, R, and T detectors. We
will only examine the correlation of the reflected and
transmitted events if the gate detector is triggered
(this helps to eliminate noise events). If all three
detectors record events simultaneously, then either
the photon is behaving like a classical wave with
split amplitude, or the observation is just noise. We
label the number of these coincidences as NGRT. If
just the G and R events are correlated (or the G and
T) the photon is these events as NGR (and NGT). then
observed to be acting like a discreet particle (which
is what Grangier et. al. observed).
To determine if the photon can act like a particle,
we will compute a second order correlation
coefficient g =
𝑁 𝐺𝑇𝑅 𝑁 𝐺
𝑁 𝐺𝑇 𝑁 𝐺𝑅
.
If g < 1 , then the quantized photon picture is
correct; if g ≥ 1, then the classical wave picture is
correct [1].Fig. 2: A 405 nm blue laser is directed through a
downconversion cyrstal (DCC). Two IR photons (810 nm) are
simultaneously emitted with a 3˚ half-angle, traveling to
the gate detector and the beam splitter. The half-wave
plates allow the linear polarization of the photons to be
rotated, as both the DCC and beamsplitter are sensitive to
light polarization. The T and R detectors follow the
polarizing beamsplitter.
Fig. 3: Possible outcomes for the experiment.
References
Future Work