シリコンリング共振器光子源における量子干渉の測定

The overall goal of this procedure is to characterize an integrated photonic photon pair source through the measurement of quantum interference. This method can help answer key questions related to integrated quantum photonics, including how to realize chip scale sources of correlated photons and integrate them into quantum integrated photonic circuits. The main advantage of this technique is that it can be applied to a wide variety of integrated quantum photonic circuits.

At the heart of the experiment is the photonic chip. This chip is approximately five millimeters on a side, and is fabricated using standard techniques. This image of the chip reveals its components.

There is a pump circuit, including the input wave guide, a ring resonator, in which photons will propagate both clockwise and counterclockwise, and a Mach-Zehnder interferometer, which is followed by output wave guides. Metal leads allow for on-chip heating, resulting in a phase shift in the interferometer. To prepare the chip for use in the circuit, polish with a chip polisher.

First use the polisher to level the chip, and make all facets orthogonal. Polish the chip with a three-micron lapping pad, in steps of about 50 microns, until within about 100 microns of the end of the polishing marks. After each 50 microns, inspect the chip to determine the remaining distance.

When there is about 100 microns remaining, change to a one-micron lapping pad. Continue to polish the chip and to monitor the progress. When there is about 20 microns remaining, change to a 0.5 micron pad.

Polish the chip further, until within 15 microns of the end of the polishing marks. At 15 microns, change the lapping pad to one with 0.1 micron roughness. Use this pad to polish the chip until only 10 microns of the polishing marks remain.

The last polishing step, with a 0.1 micron lapping pad, ensures a smooth facet. Remove the chip before cleaning it and storing it for later use. Gather the necessary equipment to prepare the optical fibers.

This includes a fiber stripper, a fiber cleaver, a fusion splicer, and a sleeve oven. Work with the three single-mode fiber pigtails, and about 20 to 30 centimeters of ultra-high numerical aperture fiber for each. To prepare one pigtail, use the fiber stripper to remove any buffer or coding from its end.

Do the same with one end of the length of ultra-high numerical aperture fiber. After cleaning the fibers, use the fiber cleaver to prepare them for fusion splicing. Next, move the fibers to the splicer.

Put the fibers in position, and properly align the cleaved ends. Enter the appropriate parameters, and perform the splice. When done, remove the spliced fibers and inspect it.

If the splice is acceptable, slide a protective sleeve over the splice site. Then place the sleeve-covered splice into the sleeve oven to permanently secure it to the fiber. Go on to produce three spliced fibers for use in the experiment.

The experiment takes place on an optical bench. On the bench are three three-axis translation stages with piezo controllers. They are positioned to allow access to the chip wave guides.

The translation stages surround the optical chip that has already been mounted on a copper pedestal. The pedestal is in contact with the thermoelectric cooler. Each translation stage has one of the prepared fibers in a V-groove, and attached with polyimide tape.

The region with the chip can be viewed using a microscope, which is equipped with both visible and infrared cameras. At this point, the fibers can be connected to the experiment instruments. Connect the chip input to the optical output of a tunable laser source via a polarization controller.

Connect each output of the chip to an optical power meter. Now, adjust the microscope position to work with the chip. Focus the microscope where the wave guides reach the edge of the chip, and use the translation stages to position the fibers near the chip edge.

Bring the fibers into view of the visible camera, and adjust their heights, so the core of each fiber is in focus. Before proceeding, ensure the horizontal position of each fiber lines up with its wave guide. Turn on the optical output of the laser, and adjust the position of the input fiber until light couples into the wave guide.

On the infrared camera, this will appear as scattering along the input wave guide. Next, tune the wavelength of the laser so the microring resonator is lit up on the infrared camera. This indicates the resonance condition has been satisfied.

Continue by manipulating the fiber positions with the micrometers, so as to maximize the output power measured by the power meters. Finely tune the fiber positions, and move each fiber slightly closer to the chip, using the piezo stage controllers. Iterate between fine-tuning all the fiber couplings and moving all fibers closer to the chip.

The goal is to have the fibers pressed firmly against the sides of the chip, with the measured power maximized. The next step is to characterize the dispersion. Begin the characterization by tuning the polarization controller to maximize the power reading on the power meters.

Now, scan the tunable laser over the wavelength range of interest to find the transmission spectrum. Extract the bandwidth of each resonance, and use the information to find the group indices and corresponding uncertainties. Next, identify the wavelengths of the two pump lasers by finding two resonances that have an odd number of resonances between them.

Knowing these wavelengths allows determination of biphoton wavelength. To test whether these three wavelengths are consistent with spontaneous four-wave mixing, plot the group index versus the wavelength. In this case, the blue points are the group indices.

The red shading corresponds to the uncertainty of the group indices, as a result of the bandwidth of each resonance. The green horizontal line extends between the candidate pump laser wavelengths. Since the line is entirely within the shaded region, the pump and biphoton wavelengths can be used for the experiment.

Once the probe wavelengths have been determined, create the final experiment setup. This has two tunable laser sources, one for each of the pump laser wavelengths. The laser outputs each go to separate polarization controllers.

From there, the two laser outputs are combined in a fiber combiner. Next to there is a series of fiber-based notch filters. These filters allow passage of the pump wavelengths, but they achieve approximately 120 decibels of attenuation of the biphoton wavelength.

The output of this filter goes into the photonic chip. On each output, after the chip, there is a series of bandpass filters. These filters attenuate the pump wavelengths by about 150 decibels, but pass the biphoton wavelengths.

The rejected photons from each set of filters are sent to a dedicated power meter. The output from each of the fiber-based filters goes to a dedicated single photon detector. Each of the single photon detectors provides input to a coincidence correlator.

The phase shifter for the Mach-Zehnder interferometer is an on chip resistive heater. Connect a computer-controlled current driver to the chip's contact pads to generate heat when the voltage is set. For two photon interference measurements, start with the pump lasers at the chosen wavelengths.

Monitor the power meters to ensure each laser is tuned to its resonance, and the power is maximized. Next, monitor the coincidence counts at the correlator. As indicated in this figure, find the peak of the data and integrate over an approximately 220 picosecond window, centered on it.

Keep track of the coincidence counts until there is a total of at least 100. This indicates that a sufficient integration time has passed. Now, turn to the computer to set the voltage control for the phase shifter at zero volts.

Once the phase shift is set, go to one of the tunable lasers and scan over the entire wavelength range. Use the power meters for the rejected pump photons to identify the location of the previously-selected resonances which may have drifted. Set the pump lasers to match the previously-chosen resonances.

It is important to follow the chosen resonances over time, rather than the wavelengths. is heated, the ring is also heated, but much less efficiently. This shifts the resonances to longer wavelengths.

Collect the resulting data from the time correlator using the previously-chosen integration time. This includes the number of photons counted by each detector in the coincidence counts. After collecting the data, adjust the voltage control of the phase shifter, and increment it by five millivolts.

Repeat scanning the laser and collecting count data until the desired range of voltages is covered. These classical light interference patterns were obtained using the testing setup by collecting individual photon counts as a function of the relative phase between the two paths. In addition to the measured data, represented by circles and diamonds, the solid lines are fits to the data.

The numbers represent the calculated visibility. Coincidence correlation measurements show the quantum interference of the entangled photons. Note that the oscillation is at twice the frequency of the classical pattern.

The orange curve is from a test of photon origin that requires entangled photons to be generated at a wavelength not supported by the ring. It demonstrates the coincidences are from photons generated in the ring. This data is from six experiments in which resonance pairs are symmetric in frequency, about the desired biphoton residence.

Each set of data demonstrates a period of half that of the relative phase. Once mastered, this technique can be done in 10 to 15 hours, if it is performed properly. The total time is primarily determined by the resolution of the phase shifter voltage increment and the associated integration time of each biphoton coincidence measurement.

While attempting this procedure, it is important to remember to take your time while optimizing the couplings of the chip. If not done properly, the fibers may not be stable during the measurements. After watching this video, you should have a good understanding of how to prepare and test integrated photonic photon sources.