UNIVERSITY OF ROCHESTER

THE INSTITUTE OF OPTICS

OPT 253, OPT 453, PHY 434

Lab. 2. Single Photon Interference

Instructor: Dr. Svetlana G. Lukishova

[email protected]

Fall 2008

Important Safety Tips

Laser safety
You will use red He-Ne laser (~633 nm wavelength, ~ 5 mW power) in this
lab. Despite its low output power, He-Ne lasers may still represent a
significant hazard to vision.
One should never look into the beam of any laser - especially if it is
collimated. Use an indirect means of determining proper operation such as
projecting the beam onto a white paper or laser power meter.

Equipment safety
The CCD camera used in this lab is internally cooled, and ideally works at
an internal temperature of -60 C. Never block the vent on the camera
since doing so can disrupt the cooling mechanism. The temperature inside
the camera is displayed at the bottom left of the screen when the Andor
iXon software is open. If the camera ever reaches a temperature too high
to operate safely, a buzzer will sound from inside the camera. If one ever
hears this buzzer, it is extremely important to turn off the internal cooling in
the camera and make sure that the internal fan is on and turned to High.
This can be done in the iXon software by first click on the Hardware drop
down menu, and then choosing Temperature. Turn the cooling option to
off and click Ok. Next click on Hardware again and choose Fan Control.
Make sure that High is selected and press Ok. Make sure to stop any
data collection. The camera should then be left alone to cool off before
usage continues.

PREPARATORY QUESTIONS
You should answer two sets of questions:
(1) before your first laboratory session and
(2) within each section of this Manual. All questions have a blue-color font.
Answer these questions before your first laboratory session

1. What should be a filter transmission value to attenuate a 5 mW, 633

nm laser beam to a single-photon level (e.g., for a ~ 300-m distance
between photons)?
2. Will you expect to obtain an antibunching histogram from the
attenuated laser beam?
3. Draw the interference pattern from a single slit and a double-slit.
Provide the equations for these patterns.
4. Draw the schematics of a Mach-Zender interferometer.
5. How will you align the interferometer?
6. Explain what is a which-way information.

References and recommended literature:

1. G. Greenstein and A.G. Zajonc, The Quantum Challenge, Jones and Bartlett Publ.,
Boston, 2006.
2. M.B. Schneider and I.A. LaPuma, "A simple experiment for discussion quantum
interference and which-way measurement", Am. J. Phys., 70 (3), 266-271 (2002).
3. The Feynman Lectures on Physics. Vol. 3, Chapter 1, R. P. Feynman, R. B. Leighton, &
M. Sands, Addison-Wesley, 1965.
4. Technical University of Delft, Optics Research Group::
http://www.optica.tn.tudelft.nl/education/photons.asp
5. http://www.physics.brown.edu/physics/demopages/Demo/modern/demo/7a5520.htm
6. W. Rueckner and P. Titcomb, A lecture demonstration of single photon interference,
Am. J. Phys. 64, No. 2, 184-188, Feb 1996.
7. http://www.fas.harvard.edu/~scdiroff/lds/QuantumRelativity/SinglePhotonInterference/Si
nglePhotonInterference.html

Introduction
Light, as we know it, can behave like particles or like waves under different physical conditions.
This is what is called the wave-particle duality of light. The particle nature of light is explained
by considering light as massless energetic particles called photons.
It is only recently that people have been able to generate and detect single photons. The main
driving force behind single-photon research has been the possible applications in the field of
quantum information. During the last few decades, this field has revealed a number of
possibilities where quantum physics could be used to achieve tasks that were not possible with
classical physics. Among these tasks are quantum computing, quantum teleportation, and
quantum key distribution. The latter is the most technically mature application, and has been one
of the most important motivators for the increased interest in the generation and usage of single
photons.
Interference of light is a well-studied phenomenon. One way to observe interference is by using
an interferometer. In an interferometer, light from a single source is divided into two paths that
are then recombined to produce an intensity pattern on a screen or a detector. The most
commonly used interferometers are Youngs double slit, Michelson, and Mach-Zehnder
interferometers. This Laboratory uses the Mach-Zehnder and Youngs double slit interferometers
to explain some important optical concepts, specifically the wave-particle duality of light.
Interference is usually understood in terms of waves and does not have an adequate explanation
in terms of particles. Thus, if a stream of single photons being modeled as particles were
incident on the two slits of a Youngs double slit interferometer, one might expect that the
photons would first pass through one of the two slits (or hit the screen), then travel to the screen
resulting in the superposition of the two intensity patterns formed as a result of geometry of each
slit, as in Fig. 1 below:

time = t

time > t

time >> t

Fig 1: Theoretical intensity pattern observed at distant screen caused by single photons
modeled as particles incident on two slits. Images are over increasing accumulation time.
This, however, is not the case unless one does now know which slit the single photons goes
through, referred to as which-path information. As long as one does not have which-path

6
information, the pattern seen in fig. 2 is created when a stream of single photons is incident on
the slits in a Youngs double slit interferometer (the setup described above) even though only
one photon is incident on the slits at a time. The interference pattern observed would be a
sinusoidal function, whose amplitude is modulated by an envelope function corresponding to
diffraction from a single slit, as in Fig. 2. The absence of which-path information is the
essential criterion for single-photon interference. For example, if one had a way of knowing
(even in principle) which of the two paths a photon took in a Youngs double slit interferometer,
one would not see any interference fringes. Knowing the path that a photon takes is equivalent to
performing a measurement on the photon. As dictated by quantum mechanics, performing a
measurement collapses the photons wave function, causing it to behave like a particle (a particle
cannot interfere with itself).

Fig 2: Interference pattern observed at distant screen resulting from single photons incident
on the two slits of a Youngs double slit interferometer.
But if a photon is a particle, not a wave, how can it interfere with itself? Suppose one had a
perfect two-dimensional detector. After only one single-photon event, one clearly would not
see interference fringes. Here, an event is understood to mean a single photon incident on the
double-slit. The photon would hit the detector at a single point, thus the photon behaves like a
particle. However, as one records more and more single-photon events, an interference pattern
begins to form. Fig. 3 illustrates the emergence of single-photon interference fringes. This clearly
demonstrates that the particle-like photons behave like waves. This is what is known as the
wave-particle duality.

Fig. 3: Single-photon interference over increasing time intervals.1

Associated with each point in space is the probability that a photon emerging from the doubleslit will go there. Dark fringes in Fig. 3 correspond to high probability regions and light fringes
in Fig. 3 correspond to low probability regions. The parameters of the double-slit determine the
probability distribution (i.e. the fringe pattern). The classical interference pattern produced by
light waves can therefore be regarded as the sum of a great number of single-photon interference
patterns.

This laboratory provides a visual demonstration of the appearance and

disappearance of interference fringes, both at high light and single-photon
levels, by carefully destroying and restoring which-path information using a
polarizer in a Mach-Zehnder interferometer. It also demonstrates the waveparticle duality and by observing interference fringes created by single
photons incident on a Youngs double slit interferometer.

http://www.optica.tn.tudelft.nl/education/photons.asp

Activity 1. Photon : a Particle or a Wave ?

To better understand the importance of which-path information, consider the simple example
of Youngs double slit experiment. The principles involved are the same as those in a MachZehnder interferometer.
Question 1: Sketch on the figure below the intensity pattern you would see if the top slit were
blocked.

Youngs double slit experiment with top slit blocked.

Question 2: Sketch on the figure below the intensity pattern you would see if the bottom slit
were blocked.

Youngs double slit experiment with bottom slit blocked.

9
Question 3: Sketch on the figure below the intensity pattern you would see if both the slits were
open.

Youngs double slit experiment with both slits open.

Next consider Youngs double slit experiment with a bullet gun instead of a laser. (Assume that
the bullets are about the same size as the slit width).
Question 4: Sketch on the figure below the accumulation pattern (intensity pattern) of bullets
that you would see if top slit were blocked.

Youngs double slit experiment with bullets (top slit closed).

10
Question 5: Sketch on the figure below the accumulation pattern (intensity pattern) of bullets
that you would see if bottom slit were blocked.

Youngs double slit experiment with bullets (bottom slit blocked).

Question 6: Sketch on the figure below the accumulation pattern (intensity pattern) of bullets
that you would see if both the slits were open.

Youngs double slit experiment with bullets (both slits open).

SEE YOUR TA to make sure you correctly answered the above questions before
proceeding further.
Question 7: If light is nothing but a stream of particles (called photons), why doesnt a stream of
bullet give rise to an interference pattern similar to that of photons? (Take your time to answer
this question and then check with the answer below).

11

Answer: Which-path information and wave particle duality. A bullet is a massive object and
one can always find out for sure which slit or path, a bullet happens to take. Since interference is
a wave phenomenon and it is impossible for a bullet to act like a wave on a measurable scale,
one never sees an interference pattern with bullets.
On the other hand, photons are massless particles and act like waves as long as there is no
measurement performed on them. The moment a measurement is performed on the photons, they
start behaving like particles.
Making a measurement is the same as finding the path or slit that a photon happens to take. In
the above double slit experiment, one has no way of knowing which path a photon takes. So
individual photons can act like waves and interfere with themselves. In Activity 2 well actually
see how a which-path measurement destroys the interference in a Mach-Zehnder
interferometer.

12

Activity 2. Which Path Puzzle

In this activity, you will experimentally observe the effect of which-path information using a
Mach-Zehnder interferometer. The experiment is set-up as shown in the schematic diagram (Fig.
5). You will assemble this interferometer from the scratch.

Experimental Set-up: (for a description of the optical elements used, see Appendix 1)

Fig. 4: Photograph of experimental set-up (Mach-Zehnder interferometer).

Polarizer D
NPBS

mirror

screen

Path 2

Polarizer C
Polarizer A

EM-CCD
camera

Path 1
Laser
N.D.filter Spatial
filter

PBS

mirror
Polarizer B

Fig. 5: Schematic of the experiment with a Mach-Zehnder interferometer.

As shown in the schematic diagram (Fig. 5), light from the laser passes through a spatial filter,
increasing the beam diameter from a few mm to a few cm. It then enters into the PBS, which

13
transmits the horizontal polarization while reflecting the vertical polarization. A polarizer is
placed before the PBS to adjust the polarization of incoming light such that it gets equally
divided between horizontal and vertical polarizations after passing through the PBS. Polarizer B
and polarizer C are put in path 1 and path 2, respectively, to make sure that there are only
horizontally polarized photons in path 1 and only vertically polarized photons in path 2.
The beams in path 1 and path 2 are then combined at the NPBS via reflection from the mirrors.
The resulting light is observed on a screen to the right of the NPBS. Polarizer D is used to
demonstrate the appearance and disappearance of fringes, which can be observed with the naked
eye on the screen at high light level. To demonstrate the same effect at single photon level, an
EM-CCD camera is used to collect data.
Before you start the experiment, remove Polarizer D if it is there. Verify that you have only
horizontally polarized photons in path-1 and only vertically polarized photons in path-2.
1. Switch the laser on. Turn the room lights off.
2. Put Polarizer B at 90 and hold a piece of paper right after Polarizer B.
Question 8: Do you see a bright or a dark spot?

3. Put Polarizer B at 0 and hold a piece of paper right after Polarizer B.

Question 9: Do you see a bright or a dark spot?

4. Put Polarizer B back at 90.

Question 10: What do question-8 and question-9 tell you about the polarization of the photons
in path-1?

5. Now put Polarizer C at 0 and hold a piece of paper right after Polarizer C
Question 11: Do you see a bright or a dark spot?

6. Put Polarizer C at 90 and hold a piece of paper right after Polarizer C.

14
Question 12: Do you see a bright or a dark spot?

7. Put polarizer C 2 back at 0.

Question 13: What do question-11 and question-12 tell you about the polarization of the photons
in path-2?

Polarization of photons in path 1: ____________________________

Polarization of photons in path 2: ____________________________

SEE YOUR INSTRUCTOR to make sure you have the right polarizations.

Prediction:
Now you know the polarization states of photons in both paths. Suppose you had a device that
measured polarization (instead of the EM-CCD camera). If this device detected a horizontal
photon, you would know for sure that it came through path 1 and if it detected a vertical photon,
you would know for sure that it came through path 2. The photons which-path information is
completely known.
Question 14: Do you expect to see interference fringes with the present configuration of the
Mach-Zehnder interferometer? Why?

Check your prediction: Observe the intensity pattern on the screen.

Question 15: Did your prediction match with the observation?

15

Prediction:
Let us assume that we put a polarizer after the NPBS and set it at 45. By putting a polarizer at
45, one will obtain only 45-polarized photons at the detector. The 45 polarized photons
incident on the detector could then have taken either of the two paths. There is no way to tell
which path a photon takes. So by putting a polarizer at 45, one essentially destroys the whichpath information.
Question 16: Do you expect to see interference fringes if you introduced a polarizer at 45?
Why?

Check your prediction: Put polarizer D in after the NPBS and set it at 45. Observe the
intensity pattern on the screen.
Question 17: Did your prediction match with the observation?

See your TA if you dont see any fringes in this activity. The set-up may
need to be aligned.

16

Activity 3. Single Photon Interference (Mach-Zehnder

Interferometer)
In the previous activity, we found that any attempt to get which-path information destroys the
wave nature of the photons and therefore, the interference pattern. Removal of any possibility of
determining this information brings the fringes back. All this was done at a high light level.
In this activity, you will perform the same experiment, but at the single photon level. Obtaining
similar results as before would confirm that a single photon does, in fact, interfere with itself and
the interference fringes at high light level are just the combined effect of many single-photon
interference patterns.
What we ideally want is a single photon source, so we can be certain that there is only one
photon in the interferometer at a time. It seems difficult to have this kind of source, but it is not
entirely the case.
One photon at a time is a relative concept. Our eye has a resolution time of 0.1 seconds. So if
photons were coming at an interval of 0.1 seconds, our eye would call it one photon at a time.
With the advent of technology we now have detection system (including the counter-timer
computer card) with nanosecond resolution time. So for this detection system, photons coming at

nanosecond intervals would still be considered as one photon at a time.

Question 18: The EM-CCD camera we are using is a single photon detector that has a resolution
time of about 100 nanoseconds. If photons are coming at the rate of 100,000 photons/second,
how many resolution-time periods (on an average) would the detector have to wait before
detecting a photon?

Here we are talking about the probability that the detector would detect only one photon at a time. There is a
finite probability with the attenuated laser pulses of detecting two photons at a time. However this probability is
much smaller than the probability of detecting just one photon at a time.

17
As you saw in the above calculation, the detector, on average, detects a photon every couple of
hundred resolution times. So we are at a one photon at a time level, in the sense that there is
only one photon in the interferometer at a given time. You will now replace the screen with an
EM-CCD camera and view the fringe pattern. We can now check to see if the results of Activity
2 can be repeated with single photons.
1. There should be two cables going to the CCD camera, a data cable and a power supply
cable. Make sure that both cables are properly connected. The power cable simply plugs
into the back of the camera whereas one must press in the sides of the silver data cable
when attaching or removing it.
2. Turn off the lights if you have not done so.
3. Double click the Andor iXon icon located on the desktop of the computer attached to the
camera. On the toolbar, click Hardware, and a drop down menu will appear. Choose
Shutter Control. Choose CLOSED For Background and click ok.
Next choose
Acquisition from the toolbar, and another drop down menu will appear. Choose Setup
Data Type. Choose Counts (Bg corrected per second) and click ok. Again open the
Acquisition drop down menu, and choose Setup Acquisition. Locate the field labeled
Exposure time, and change the current value to 0.1, then click ok.
4.

Next, again choose Acquisition, and choose Take Background. The camera should
make a clicking sound, and a new window will appear on the screen which displays the
cameras current view. It should resemble static. The purpose of taking a background is
to eliminate as much background noise in the image displayed by the camera as possible.

Note: The acquisition time will have to be changed depending on the intensity of the pattern
incident on the EM-CCD camera. The gain can also be adjusted if needed. Keep in mind,
however, that when changing the acquisition time, a new background should be taken (to be
subtracted).
5. Replace the screen used previously for viewing the fringe pattern with the EM-CCD
camera. Click the icon with a picture of a video camera to display a live video feed from
the camera. Use the image to align the camera such that an acceptable image is visible.
6. When this is the case, stop the live video input and insert the number of ND filters you
calculated previously to be necessary to reach the single photon level. Click on the
picture of a camera to take an image of the fringe pattern present. Again, changing the
exposure time and gain may be necessary to receive good results.
7. Change the polarized to 0 and again image the resulting fringe pattern using the camera.

18

Activity 4. Single Photon Interference in Youngs

Double Slit Interferometer
In this experiment, you will observe the interference of single photons further exemplifying the
wave-particle duality. The interference patterns will be examined with and without knowledge
of which-path information. The general setup for a Youngs double slit interferometer is as
follows:

Screen or
Detector

Double
Slit

Monochromatic
Source

d (slit width)

(slit separation) a

L
Fig. 6: General Youngs Double Slit Interferometer.
The setup to be used in this experiment can be seen schematically below:

Circular
Aperture

Red Band Pass

Filter

HeNe
Laser

EM-CCD
camera
Spatial
Filter

ND
Filters

Double
Slit
Fig. 7: Schematic Diagram.

y
z

19
1. Given that the fringe separation in the interference pattern (y) can be approximated by
the following equation:

calculate the distance from the double-slit to the EM-CCD camera that would give you a
fringe separation of 2 mm. The double-slit is illuminated by a HeNe laser ( = 632.8 nm),
the slit separation is 90 m, and the slit width is 10 m. If any neutral density filters are
mounted in front of the spatial filter remove them. Then place a piece of paper at the
distance you calculated and observe the interference pattern.
2. Using a power meter, measure the unattenuated laser power at the distance found in step
1. (Make sure the brightest fringes are incident on the power meter head and the room
lights are off when performing this measurement). For single-photon interference you
want to attenuate the laser power to a single-photon level. For this purpose you need to
make a distance between photons larger than a distance between the laser and a doubleslit, e.g., ~300 m (~ 106 photons per second).
The following equations will be useful in your calculation:

Power =

Energy
Time

E photon = h

Be sure you take into account that the active area of the camera is different from the area
of the power meter head. (Active area of camera is 8.2 x 8.2 mm).
Since too much light can damage the camera, have your TA check your calculations before
you proceed.
3. After the TA has approved your selection of neutral density filters, mount them at the
output of the spatial filter as (Fig. 7). Position the EM-CCD camera at the distance you
calculated in step 1. Turn off the lights and make sure that the camera is properly aligned
by viewing the image of the fringes in the live video mode of the camera. Recall that
changing the acquisition time and the gain may be necessary and that a new background
must be taken when these parameters are changed.
4. When the camera is properly aligned, take multiple exposures of the image with smaller
and smaller acquisition times.
5. Shut down the iXon software, turn off and unplug the camera, and turn off the laser.

Reference:
R.P.Feynman, Lectures on Physics, Volume III, Chapter 1.

20

Appendix 1
(i)

Non-Polarizing Beam Splitter (NPBS): This beam splitter splits the intensity of
light, transmitting 50% of the light while reflecting 50% of the light as shown in
picture.

Figure A1.1: Nonpolarizing Beam Splitter

(i)

Polarizing Beam Splitter (PBS): This beam splitter does not split the intensity of
light rather it splits the polarization of light. PBS transmits the horizontal polarization
while reflecting the vertical polarization, as shown below:

Figure A1.2: Polarizing Beam Splitter.

(ii)

Laser: A Helium-Neon (He-Ne), ~ 5mW, 633 nm laser is used in this experiment.

(iii)

Spatial Filters: This is introduced to increase the beam size so that interference can
be observed over a wider area. It consists of a microscope objective, a pinhole and a
lens.

21

Figure A1.3: Spatial filter

22

Appendix 2. Previously completed lab report

The goal of this experiment was to investigate the phenomenon of single photon interference.
This was performed by analyzing the interference patterns created by both Mach-Zehnder and
Youngs double slit interferometers. Interference patterns created at high intensity levels and at
the single photon level were recorded respectively using cooled EM-CCD camera. They were
then compared and used to demonstrate that single photons, barring any knowledge of whichpath information, interfere with themselves and that interference patterns at higher intensity
levels are actually the sum of a large number of single photon interference patterns. These
results clearly exemplified the wave-particle duality of photons.
1. INTRODUCTION AND THEORETICAL BACKGROUND
Central to understanding the nature of light is its description as photons, which are governed by
wave-particle duality. This description dictates that under different circumstances, photons will
act either as particles, which do not experience interference, or waves, which do experience
interference. Further, despite perhaps being counterintuitive, when single photons act as waves,
they interfere with themselves. These properties of light were examined in this experiment using
a Mach-Zehnder and Youngs double slit interferometer (see Fig.1R and Fig. 2R).

An

interferometer is a setup which divides the emitted light from a single source (typically coherent)
into two paths, and then recombines the two paths to create an interference pattern. Interference
patterns generated by multiple and single photon light levels were imaged and compared to
demonstrate that the interference pattern generated at the single photon level, given a longer and
longer collection time, will eventually resemble that created at high light levels.
A 633 nm Helium-Neon (He-Ne) laser was used as the illumination source in both parts of
this experiment. High light levels were easily achieved by using little or no neutral density
filters. To achieve a single photon source using the He-Ne laser, enough neutral density filters
had to be placed in the beam path so that the CCD camera used for imaging the interference
patterns only detected one photon per resolution-time period, or that there is only 1 photon in the
length of interferometer at each time. To reach this criterion, two neutral density filters of
attenuation totaling ~10-7 were used together were necessary given the 10 MHz refresh rate of

23
the CCD camera used to achieve a single photon source and 1 photon per 1 meter at any given
time.
One necessary criterion for single photon interference is the absence of which-path
information. Knowledge about a photon, such as the path which it follows, requires performing
a measurement on that photon. However, quantum mechanics dictates that if one is performs
any sort of measurement on a photon, its wave function collapses, and the photon then acts like a
particle. In the part of this experiment utilizing a Mach-Zehnder interferometer, single photon
interference was first observed without any knowledge of which-path information,
demonstrating that photons can act as waves and interfere with themselves. Which-path
information was then acquired during observation, causing the photons to act as particles and
thus destroying the preciously observed interference pattern.
Single photon research is becoming increasingly important greatly due to its application
in the growing quantum information field. Insight into the nature of single photons as well as the
creation of single photon sources has application pertaining to quantum computing, quantum
teleportation, and quantum key distribution. Applications such as these rely on quantum physics
since classical physics offers no explanation for many of the interesting phenomena involved,
such as entanglement of quantum states.
2. PROCEDURE

Non-polarizing
Beam Splitter

Mirror
Path 2 (Vertical
Polarization)

Screen

Polarizer D

Polarizer C

Polarizer B

Spatial
Filter
He-Ne Laser

Path 1 (Horizontal
Polarization)

Mirror
Neutral
Density
Filter(s)

Polarizer A

Polarizing
Beam
Splitter

Fig. 1R. Mach-Zehnder Interferometer

CCD
Camera

24
Image at CCD array

CCD
Camera

Cone of light due to diffraction

Screen with two slits
(greatly exaggerated for clarity)

Spatial
Filter
He-Ne Laser

Mirror
Neutral
Density
Filter(s)

Fig. 2R Youngs Double Slit Interferometer (The slits used had 10 m width and
were separated by 90 m. Observation plane was located at the 20 cm from the
slits plane).
1. The He-Ne laser supply was turned on and the power output was measured using a
power meter. The numbers of photons/meter was calculated. Using this calculated
value and the given refresh rate of the CCD camera used, the number of neutral
density filters required to achieve a single photon source was calculated (see below
for values). These filters were set aside and not yet used in the system.
2. The CCD camera was placed at the output of the Mach-Zehnder interferometer setup.
3.

Polarizer A at the input of the Mach-Zehnder interferometer was oriented at 45 to

assure that the polarizing beam splitter directed an equal number of horizontally
polarized photons down path 1 and vertically polarized photons down path 2.

25
4. Polarizer B in the Mach-Zehnder interferometer was then oriented to assure the
photons traveling down path 1 were indeed horizontally polarized.

Orienting

polarizer B at 90 provided maximum transmission, whereas orienting it at 0

provided minimum transmission.

Thus it was certain that path 1 contained

horizontally polarized photons.

5. Polarizer C in the Mach-Zehnder interferometer was then oriented to assure the
photons traveling down path 2 were indeed vertically polarized. Orienting polarizer
C at 0 provided maximum transmission, whereas orienting it at 90 provided
minimum transmission. Thus it was certain that path 2 contained vertically polarized
photons.
6. Polarizer D at the output of the Mach-Zehnder interferometer was then set at 45, and
the alignment of the entire system was checked by assuring that fringes were visible
in the combined path on the screen.
7. Polarizers B and C in the Mach-Zehnder interferometer were then adjusted to provide
maximum contrast in the interference pattern.
8. With polarizer D at 45 and the screen and filters removed, the interference fringes
were viewed using the CCD camera.
9. The filters (total attenuation = 1.41* 10-7) were placed in front of the laser and
diminished the He-Ne laser enough that the single photon level was achieved.
10. With polarizer D at 45 and the screen still removed, the interference fringes created
were again viewed using the CCD camera. Fringes were visible because orienting
polarizer D at 45 eliminates any mechanism for establishing which-path
information.
11. The interference pattern created by single photons was recorded for increasingly
smaller acquisition times and with varying amounts of electron multiplier gain on the
CCD camera (sees images below).
12. Polarizer D was then oriented at 0, and it was observed the interference fringes that
were visible disappeared.

This was because orienting polarizer at 0 provides

which-path information by only allowing vertically polarized photons (path 2) to

reach the CCD camera.

26
13. Polarizer D was then oriented at 90. Again, the fringe pattern present on the CCD
camera disappeared since which-path information was established by assuring that
only horizontally polarized photons from path 1 reach the CCD camera.
14. Polarizer D was then oriented at 45. The interference pattern was once again visible
because which-path information was lost. Orienting polarizer D at 45 allows for
both horizontally and vertically polarized photons to pass, and thus the photons
reaching the CCD camera could have come from path 1 or path 2. Thus one cannot
attain which-path information.
15. The neutral density filters removed, the CCD camera was then moved to the Youngs
Double slit interferometer setup.
16. The position of the camera was adjusted until the fringe pattern created by the slits
was centered on the CCD array.
17. The neutral density filters were then replaced (total attenuation = 1.41* 10-7) to
establish single photon source conditions.
18. The acquisition time was then decreased while the electron multiplier gain increased
on the CCD camera. A series of images was collected showing the interference
pattern created by the slits with fewer and fewer photons. The comparison of these
images shows how the interference pattern that results is actually the result of the
summation of single photon interference patterns.
3. RESULTS AND ANALYSIS
To achieve single photon source conditions: The measured He-Ne laser power output was 1.896
mW. To achieve single photon source conditions, the desired criterion is 1 photon per meter
within the optical system. The calculated number of photons/m given the power measurement is
2.012 * 107 photons/m. Thus an attenuation of ~10-7 is needed to achieve single photon source
conditions. Two neutral density filters, with attenuations of 1.24 * 10-3 and 1.17 * 10-4 were
used, thus providing a total signal attenuation of 1.41* 10-7.

27

Attenuation: 1.41* 10-7, Acquisition time: 1s,

gain: 0x, Polarizer D = 45
Comment: Fringes clearly visible

Attenuation: 1.41* 10-7, Acquisition time: 100s

gain: 255x, Polarizer D = 45
Comment: Fringe pattern with small acquisition time

Attenuation: 1.41* 10-7, Acquisition time: 100ms,

gain: 0x, Polarizer D = 90
Comment: Fringes destroyed

Attenuation: 1.41* 10-7, Acquisition time: 60s,

gain: 255x, Polarizer D = 45
Comment: Fringe pattern with smaller acquisition time

28

Attenuation: 1.41* 10-7, Acquisition time: 20s,

gain: 255x, Polarizer D = 45
Comment: Fringe pattern with smallest acquisition time, single photon interference creating pervious fringe pattern

Fig. 3R. Mach-Zender Interferometer Images (see also images on page 27).

Attenuation: None, Acquisition time: 600ms, gain:0x

Attenuation: None, Acquisition time: 500ms, gain:0x

29

Attenuation: 1.41* 10-7, Acquisition time: 3s, gain:255x

Attenuation: 1.41* 10-7, Acquisition time: 2s, gain:100x

Attenuation: 1.41* 10-7, Acquisition time: 2s, gain:150x

Attenuation: 1.41* 10-7, Acquisition time: 100ms, gain:255x

Fig. 4R. Youngs Double Slit Interferometer Images (see also page 29).

30

Fig. 5R. Intensity cross sections of interference pattern

Comment: One can see that with decreasing acquisition times (a-c) the fringe pattern created by single photon interference still
exists. Thus, plot (a) can be considered a summation of many plot (c)s.

4. DISCUSSION AND CONCLUSION

The data collected in this experiment generally followed the predictions presented and served as
a validation of the underlying theory.

The images collected using the Mach-Zehnder

interferometer clearly demonstrate that when which-path information is present by introducing

polarizer D at either 0 or 90, the interference pattern that was present before is completely
destroyed. In addition, as the acquisition time is decreased further and further, one can see that
the interference pattern present becomes decreasingly visible, however, the areas where single
photons strike are still consistent with the prior pattern. Thus one can see that interference
pattern at higher light levels in simply a sum of single photon interference patterns. This was
also the case with the images attained using the Youngs double slit interferometer: one can see
that the interference pattern created by the interference of single photons will create the
interference pattern characteristic of greater light levels given enough time.

31
There were two properties of the images collected which were not accounted for
previously and require explanation. Firstly, on many of the above images, small dots and
blotches are visible. This is most likely due to the fact that the vacuum seal on the CCD
camera used is failing, and therefore small water droplets may form on the inside of the lens
system within the cooled camera. Also, some of the larger blotches visible in the MachZehnder interferometer images are due to imperfections in the glass of the polarizers used.
Higher quality polarizers were not used due to price limitations.
Secondly, there is a series of finer fringes within the larger fringe pattern on the images
collected from the Youngs double slit interferometer. Calculations showed that it may be due to
interference of the beam diffracted on the double slits and a plane wave. This plane wave was
formed by reflections from the metal and the glass surfaces of the screen. (Two slit pattern on the
screen was made by a photolithography on a glass substrate).

Acknowledgements: Following students contributed to this Manual: A.K. Jha, L. Elgin, S.

White, Z. Shi, H. Shin.