s21_spqi

Proving Single Correlated Photons

Aidan Alioto and Coredell Coss

Introduction and Motivation:

The nature of light has been a question intriguing physicists for centuries. One might think that it’d be easy to show that photons exist, just bring a detector into a very dark room, get a laser and lower its count rates to around one photon per hour, and wait for a detection. In reality the Poisson statistics of laser light and dark noise make this proof much more difficult(1). Another complicating factor is that semi-classical theories, where light is treated as a classical wave but the detector is treated quantum mechanically(1), can also account for many of the quantum effects that arise in experiments on the nature of light(1).

In our experiment we used a Beta-Barium Borate(BBO) crystal to create pairs of entangled photons from a 405 nm blue pump laser. We used a two detector set up(Figure 1) to measure the coincidence rates and count rates of the entangled photons and we recorded this data in a LabView program which calculated a high correlation between the two arms of entangled light. We then split one of the arms once more using a polarizing beam splitter(PBS) which allowed us to use a three detector set up(Figure 2). Again measuring the coincidence and count rates, but now between the three detectors we were able to show anticorrelation between the detectors proving the existence of single photons.

After proving the existence of single photons we moved on to an efficiency analysis of the fiber optic cables involved in the experiment in the hopes that it will help future groups get the best measurements and results they can.

Theory

By shooting a 405 nm wavelength laser light through a Beta-Barium Borate crystal, the laser drop in energy and forms a cone of 810 nm wavelength photons as a result of conservation of energy and momentum. By down-converting, light becomes entangled in photon pairs. As such, we expect certain correlations to be observed. We, thus, define a helpful quantity known as the anticorrelation parameter with the following equation:

This can be written for the two-detector and three-detector systems respectively as

This parameter gives a lot of information regarding correlations. For instance, α < 1 indicates anticorrelation, α=1 indicates uncorrelation, and α>1 indicates correlation. For the case of two-detectors, we expect the two detected beams to be correlated as entanglement implies correlated light. For three-detectors, we assume α to be less than one as we suspect triple coincidence between three detectors must be zero if there are single-photons. For instance, if the light behaves quantum mechanically then it is consistent of single photons where if we have correlated light it means a single photon on one arm means a single photon on the other arm. Thus, we expect one detector on the two detector arm not to detect a photon while the other detector at that arm does when the single detector arm detects one. To simply state why we cannot due this for one or two detectors, the reason is because accidentals we cannot eliminate the possibility that we have in fact seen a single photon come from the laser. A single photon detected in one or two detectors may have been a result of random background noise or thermal fluctuations in the detector. Furthermore, a false reading in the two detector system could arise due to the Poisson statistics of the laser light.

Sketch of Two-Detector System

Sketch of Three Detector System

Experimental Apparatus

Two Detector Detailed Apparatus Diagram

Three-Detector Detailed Apparatus Diagram

Light is generated by a 405 nm "pump" blue diode laser which is reflected off two mirrors to guarantee the laser is leveled. The laser light when attenuated goes through a Neutral Density filter before all other optical devices. The laser light then gets filtered by a 405 nm HWP before hitting the center of the Beta-Barium Borate crystal. The laser light then down-converts into a cone which then goes through 810 nm Bandpass filters. In the case of three detectors, one arm also has an 800 nm HWP after the Bandpass filter. The light is then received by the collimators which focus the light into the fiber optic cables for the light to eventually be received by the Single Photon Counting Modules located in the Black Box. From here, the NEXUS 3 board processes the data from the SPCM to convert to a digital signal the computer can read. This data is then displayed via a LabView program in the form of various coincidence and count rate plots. In addition, α is displayed. Attenuation was usually used when collecting data and table below shows how effective the different attenuations when comparing to counts to unattenuated light.

An important piece of equipment that was used in both experiments was the neutral density(ND) filter which attenuates the laser light coming out of the pump laser. To better understand this process we measured the count rates of each of the ND filter settings and compared that to the unattenuated light. Our results for this are summarized in Table 1.

We hoped to help future groups who decided on doing this project by measuring the efficiency of the fiber-optic cables used in the experiment as well as comparing the SPMC’s inside the black box. We found a wide range in the fiber-optic cables with the lowest being only 46.4 percent as efficient as the highest. While the SPCM themselves were much closer in efficiency, with the lowest being only 90.7 percent as efficient as the highest.

To do the analysis we started by maximizing the count rates in the A detector and using unattenuated light we recorded the measured count rates between the four SPCMs. Following this we then reconnected to SPCM A, used the ND filter to attenuate the laser, and switched the fiber optic cables between collimator A and its cage assembly. We kept constant the cable between the cage assembly and the black box, this was Pacer cable 2 and its efficiency was not measured but future investigations should look into it. The Pacer cables proved to be much more efficient than the orange ThorLabs cables. Analysis on how the efficiency differences between the wires and SPCM affects alpha could be an area of investigation for future groups.

Results of Down Converted Light experiment

Two different data sets were taken to show the correlation between the down-converted photons coming out of the BBO crystal, both using light attenuated by the ND filter. The greatest values of alpha were found to be when the count rates of detectors A and B were about twice the background count rates, this can be seen Table 4 below. To account for the low count rates in Table 4 we needed to increase the window time to 50 seconds to average the results over a longer time period. To get the uncertainty in each of the measurements we took 5 rounds of data with the same parameters for each data set. The experiment could be improved by attenuating the light more, so that the count rates are below that shown in Table 4. This would improve alpha because the R_acc is dependent on the individual count rates of each detector, thus lowering the overall counts increases alpha.

Results of Single Photon Experiment