Experiment 1 Making discrete photon effects visible in an optical interference experiment

Experiment 2 Photons in a Digital interference experiment

Making discrete photon effects visible in an optical interference experiment.

Summary

With Young's double slit experiment for a single photon an important feature of quantum mechanics, namely the wave particle duality, can be shown.
The interference pattern which is observed at a screen placed behind two narrow slits is a proof that light can be described as a electromagnetic wave. But as the intensity of the incident light is lowered so that only one photon at the time strikes the screen, only localized impacts of individual photons are observed. After a long exposure time an interference pattern, build up from these localized impacts, slowly appears again.
From this phenomena, one concludes that light behaves simultaneously like a wave and like particles.

At the Delft University of Technology, The Netherlands, we have succeeded making this duality visible is a very simple experiment which was built by two third year students as their "First free research" project..

The theory

In a two slit experiment, when the two slits are open at the same time, a system of interference fringes are observed at a observation screen (a photographic plate, for example) placed behind the slits, such as shown in the figure below. The interference pattern observed at the screen can be explained using the concept that light is a wave, and the observed maxima and minima depend on the phase difference between the light that comes from the first and second slits. The wave theory predicts that diminishing the intensity of the incident light will cause the fringes to diminish in intensity but do not vanish.

In terms of the particle theory, one could attempt to explain the interference term as being the interaction between the photons which pass through the first and second slits. Such an explanation would lead to the conclusion that if the intensity of the source is lowered until photons strike the screen one by one, the interaction between the photons would vanish, and consequentely, the interference fringes would disappear.

What actually happens when the two-slit experiment is performed with a single photon at the time is that neither the predictions of the wave nor the particle theory are demonstrated. The following sequence of pictures show a typical result observed in a film placed behind the slits in a single-photon two-slit experiment, for increasing exposure times:

The first two pictures, observed after only a few photons strike the screen shows very localized impacts and not a very weak interference pattern. Thus, the purely wave interpretation of light is not valid. From the last few pictures, however, one can see that fringes have not disappear, and therefore, the purely particle interpretation, which relies on the presence of fringes only due to interaction between photons, is also not valid.

With the concept of wave-particle duality, one concludes that light must described as being simultaneously wave and particles.

The experiment

Because we are looking at individual photons the entire experiment has to be done in a dark room to avoid background light. Also the most sensitive part of the experiment (camera and image intensifier) are placed in a black box to avoid scattered light from the laser. The image intensifier is an essential part of this experiment. With the intensifier, every single photon are amplified by a factor of up to one million, so that the signal generated by each photon at the output of the intensifier (a phosphor screen) can be detected with a sensitive film or a CCD camera.
The light source was a 4 mW He-Ne laser with a wavelength of 632.8 nm. The laser power was lowered with a set of neutral density filters by a factor of about 10^-11, so that the flux of photons was of the order of 10⁵ photons/sec, implying that the mean path between the photons was a couple of kilometers.
The slits were made by combining a metal frame with a thin wire with a diameter of 50 mm across it and a metal plate with a single slit of 100 mm. This results in a pair of slits, each being 25 mm wide and 75 mm separation.
At about one meter behind the slits there was an image intensifier from Proxitronics (model BV2561TZ-V1N) with a variable amplification from 100 up to 1 million. The output of the intensifier was shielded to avoid a feedback from the phosphor screen back to the intensifier's input.
The output image at the phosphor screen was projected with a single lens with focal length of 50 mm and a diameter of 2 cm placed 10 cm from the intensifier and 10 cm from the film (1:1 magnification). The film was a high speed black-and-white film, type Kodak T-Max p3200. When used with its proper developer it can be pushed to ASA values of about 25,000. The film was placed in a reflex photocamera, without lenses. A long cable was used to manually control the shutter from the outside of the black box. The film was developed and the negatives were scanned so that the images could be a bit worked out with the computer for better contrast.

The results

Typical results obtained with this set-up are showed in the 5 figures below. In this example, the gain of the intensifier was 10⁵, with exposure times of 0.5;1;2;3 and 5 seconds, respectively.

From these pictures, one can clearly see that an interference pattern was build-up from individual impacts. The background noise is mainly due to the thermal noise of the image intensifier.

Ideas to improve this experiment

As a follow up experiment, we will try to use a lens with lower F number (so that more light can be collected from the phosphor screen to the camera), and maybe also replace the photocamera by a sensitive CCD camera, so that the measurements are more direct than the present experiment. In this case, the influence from the electronic noise from the CCD camera should be investigated.

Acknowledgements

We would like to acknowledge the Applied Physics Department of the Delft University of Technology for financial support and also ir. J. de Jong and drs. R.H. Beunder from the Education Section of the Applied Physics Department for their help.

This work was performed by the third year students Niels Vegter and Thijs Wendrich, under supervision of Dr. S.F. Pereira.

Comments and suggestions are WELCOME !!!!

Experiment 2

Photons in a Digital interference experiment

Experiment Setup

To generate the coherent light needed for this experiment a He-Ne laser was used, with an average wavelength of 632.8nm.

This laser had an output power of about 5mW, many orders of magnitude higher than the desired low light level. A series of optical filters was placed in front of the laser to reduce it's power.

This experiment dealt with extremely low light levels, which meant that even small amounts of background radiation could disrupt the measurements. During the design phase of the setup this difficulty was dealt with by placing the intensifier and CCD camera inside an outer black box. The coherent beam of light from the laser could enter the box through a small hole, after being reflected by a 4% reflecting mirror. This way the majority of background photons would be kept out of the box and could not reach the sensitive equipment.

However, even this outer black box did not offer enough protection to allow clear measurements. The entire experiment had to be performed in a darkroom, with all the lights and the computer monitor turned off. Even during the cool-down period of the monitor and lights right after they had been turned off, they still emitted large amounts of background radiation, which was easily detectable.

Directly behind the small hole through which the laser beam entered the outer box, the double slit was placed. This consisted of a thin metal plate with two parallel slits, each with a width of 25um and seperated by 50um from inner edge to inner edge.

The interference pattern caused by the double slit was projected onto a mirror which could be finely adjusted, allowing for very precise positioning of the pattern on the intensifier.

The intensifier multiplied incident photons by a variable amount, producing between 30,000 and 500,000 photons at the output for a single detected photon. This factor could be adjusted by varying the voltage supplied to the intensifier.

The pattern displayed on the output screen of the intensifier was projected onto the camera's CCD chip by a lens. The lens used had a very short focal length, to maximize the efficiency of photon transfer between the intensifier and the CCD camera.

The camera had a built in lens which had to be removed to uncover the actual CCD chip. This way the lens described above could focus the interference pattern directly onto the chip.

Our results

The recordings have been obtained at different circumstances. A part of the recordings were obtained when the voltage over the intensifier was turned way down around 3V, where the amplified single photon impacts were only barely visible on the CCD camera. The rest of our recordings were obtained when the intensifier was maxed up to 4.4V, which was as far as we could push it, without feedback blinding the CCD. Both these series were obtained using different amounts of optical filters, and with various exposure times.

Large amplification

When the voltage over the intensifier was very high, the amplification was about 500,000. Making the exposure time very short, it was possible to see single photons on the recordings and it seemed as if they were distributed randomly on the recordings. When the exposure time was made longer the interference pattern became visible. The longer the exposure time, the better this pattern could be seen. This same effect occurred when several recordings made at a short exposure time were graphically added together. Unfortunately, while doing this, some hindrances of the CCD-camera occurred in the summed up picture.

An other disturbance which occurred was that at this amplification there was a lot of noise in the recordings. This was mainly due to the amplification-loop inside our experiment, which causes the photons that come out of the intensifier to bounce back into it, creating a self sustaining loop.


Single measurement, short exposure time.	Summed measurements, short exposure time.	Single measurement, long exposure time.

Low amplification

The next serie of recordings was made when the amplification was about 30,000. Because at this amplification the noise is a very big problem, the recordings had to be modified. For this, a Gaussian operator and a threshold algorithm were applied.

When the exposure time was again made longer, the interference pattern was clearly visible.

The advantage of using low amplification was the absence of a large amount of photon noise.


Before processing.	After processing.

Comparing theory with results

Having these recordings, one could now compare the results with the theory. The points or photons on the recordings could be counted and compared to the number of photons calculated using the theory. When the number of photons was counted, it turned out to be about 7 times the amount calculated using the theory. There are some explanations why this occurred: maybe when two photons hit the same spot on the camera, these two photons are counted only once or maybe somewhere in this experiment a loss of photons wasn't taken into account. Maybe one of these reasons can't explain the large factor, but a combination of such factors might be able to.

Conclusions

Several conclusion can be drawn from this experiment. First of all, it shows that a combination of the intensifier and CCD camera was able to visualize the single photons and the interference pattern.

It can also be concluded that a long exposure time is preferred over summation, because the visual artifacts of the CCD camera were not as obvious during long exposure images.

When a high amplification is used, the photons are clearly visible, but there is a lot of disturbance because of the big amount of noise. On the other hand, using a low amplification, the amount of noise is reduced, but it is hard to distinguish the photons from the background, so the recordings have to be processed.

Another possible conclusion is that the results didn't match the theory. The amount of photons counted and the amount of photons calculated differed by a factor 7.

Ideas to improve this experiment

Because the quantumefficiency of the intensifier is strongly related to the wavelength of incident light, in another experiment a blue Argon laser could be used, which has a wavelength of 488 nm. Doing this, the quantumefficiency increases about four fold.

Another problem is the transmission of the slits. When this is calculated, using the assumption that only the part of the laserbeam which falls on an opening is fully transmitted and the power of the laserbeam is uniformly distributed, a transmission of 0.032 is obtained. But, when this is measured, the transmission turns out to be 0.244.

A possible explanation is that the slits have a complex 3 dimensional construction that causes very complex diffraction.

Acknowledgements

We would like to thank the Applied Physics Department of the Delft University of Technology for their support during this experiment.

This work was performed by the third year students Christo Butcher and Wouter Naber, under supervision of Dr. S.F. Pereira.