Quantum Interference between Light Sources Separated by 150 Million Kilometers

We report an experiment to test quantum interference, entanglement, and nonlocality using two dissimilar photon sources, the Sun and a semiconductor quantum dot on the Earth, which are separated by $\sim 150\mathrm{million}$ kilometers. By making the otherwise vastly distinct photons indistinguishable in all degrees of freedom, we observe time-resolved two-photon quantum interference with a raw visibility of 0.796(17), well above the 0.5 classical limit, providing unambiguous evidence of the quantum nature of thermal light. Further, using the photons with no common history, we demonstrate postselected two-photon entanglement with a state fidelity of 0.826(24) and a violation of Bell inequality by 2.20(6). The experiment can be further extended to a larger scale using photons from distant stars and open a new route to quantum optics experiments at an astronomical scale.

Figure 1

Experimental arrangement combing a quantum-dot single-photon source and the Sun. The InAs/GaAs QD-micropillar system lies in a cryostat at 4 K and is resonantly excited by a pulsed picosecond laser (not shown) to emit single photons in pulses which are collected by a confocal setup. The single photons then go through a polarization beam splitter (PBS) and spectrally filtering by a 1 GHz FWHM etalon. In precedence of the beam splitter (BS) for two photon interference lies a half-wave-plate (HWP) which determines whether parallel polarization (indistinguishable) case or cross polarization (distinguishable) case it is in. The upper right panel shows the scanning electron microscope image of the quantum dot in a micropillar. The sunlight is collected by a simple optical setup fixed to an equatorial mount, in precedence of which a grating is placed to select out infrared band near 893 nm. These photons from the Sun are polarized at a PBS and filtered by a 1 GHz FWHM etalon. A superconducting nanowire single-photon detector (SNSPD) with time resolution of 20 ps is used to register the arrival of single photons. Detection signals are led to a time-to-digital converter (not shown) where electrically gating and coincident analysis are performed. For correlation measurement in entanglement establishment and Bell experiment, between the beam splitter and SNSPDs two polarizers are inserted (not shown).

Figure 2

Characteristics of the independent sources. (a) Intensity-correlation histogram of the pulsed RF from quantum dot obtained using a Hanbury Brown–Twiss-type setup. The second-order correlation $g_Q^{(2)}(0)=0.011(1)$ is calculated by summing all coincidence counts in the zero-delay peak divided by that of its adjacent peak. The right panel is a zooming-in plot of the coincidence counts around zero time delay (green) and 26 ns delay (purple) with log scale. (b) Intensity-correlation histogram of the pulsed RF from Purcell-enhanced QD-micropillar system in Hong-Ou-Mandel interference setup with cross polarization (red) and parallel polarization (blue). The time separation between the two single photons emitted from the single QD set to 13.2 ns (one pulse separation). The right panel is a zooming-in plot of the coincidence counts around zero time delay. A raw visibility $V_r=(P_{\text{cross}}-P_{\text{parallel}})/P_{\text{cross}}=96.36\%$, where $P_{\mathrm{cross}(\mathrm{parallel})}$ stands for the coincidence probability in the cross (parallel) polarization case, is calculated by integrating all the coincidence counts on the whole central coincidence peak. The data are normalized to the parallel-polarization situation. (c) Intensity-correlation histogram of the sunlight photons after 1 ns electrical gating obtained using a Hanbury Brown–Twiss-type setup. The second-order correlation $g_S^{(2)}(0)=1.94(13)$ is calculated from the 40-ps-long central bins counts of the zero-delay peak divided by that of its adjacent peaks.

Figure 3

HOM interference between photons from QD and the Sun. (a) Histogram of the time delays of the coincident arrivals of the two photons on the superconducting nanowire single photon detectors. Blue disks correspond to the parallel polarization (indistinguishable) situation, while red circles correspond to cross polarization situation (distinguishable). Each bin is 10 ps long. We got an overall coincidence count rate of 32 Hz for cross polarization and 24 Hz for the parallel polarization, and for the central 10 ps bin around zero time delay the coincidence count rate is 0.36 Hz for cross polarization and 0.07 Hz for parallel polarization. The solid lines are fits based on the model in [25]. The counts of parallel-polarization situation have been normalized to the counts of cross polarization situation for fair comparison. Error bars indicate the standard deviation. (b) Dependence of the raw visibility on the width of sampling delay bins centered at the zero delay. Red horizontal dashed line corresponds to the classical limit 0.5. Error bars represent statistical errors.

Figure 4

Entanglement fidelity measurement of the entangled photon pairs with no common history. (a)–(c) Polarization analysis of the entangled pairs on three different bases. The plotted data are normalized raw data without subtraction of background counts or any other correction. The bars indicate the observed fractions for the counts of specific polarization observables over the total counts of all the polarization observables. The measurements are taken on three bases ($\mathrm{HV},+\text{−},\mathrm{RL}$), respectively, to calculate the fidelity of established entanglement between photons from the Sun and the QD. Error bars represent statistical errors.

Figure 5

Correlation functions for the CHSH inequality. Labels on x-axis with the form of $({\phi }_1,{\phi }_2)$ indicate the measured polarization bases of the entangled pairs of photons from the Sun and quantum dot. Error bars represent statistical errors.