Improving the indistinguishability of single photons from an ion-cavity system

We investigate two schemes for generating indistinguishable single photons, a key feature of quantum networks, from a trapped ion coupled to an optical cavity. Through selection of the initial state in a cavity-assisted Raman transition, we suppress the detrimental effects of spontaneous emission on the photon's coherence length, measuring a visibility of $81\left(2\right)\%$ without subtraction of background counts in a Hong-Ou-Mandel interference measurement, the highest reported for an ion-cavity system. In comparison, a visibility of 50(2)% was measured using a more conventional single-photon scheme. We demonstrate through numerical analysis of the single-photon generation process that our scheme produces photons of a given indistinguishability with a greater efficiency than the conventional one. Single-photon schemes such as the one demonstrated here have applications in distributed quantum computing and communications, which rely on high-fidelity entanglement swapping and state transfer through indistinguishable single photons.

Figure 1

(a) The experimental setup. The lasers involved in this experiment and their orientations relative to the cavity axis are depicted. A quarter-wave plate (QWP) and a polarizing beam splitter (PBS) filter the cavity emission by polarization. An 866 nm laser beam is optionally guided to the optical setup by releasing the shutter and moving a half-wave plate (HWP) into the beam path. The polarization in the delay arm may be adjusted using the polarization controller paddles (PC). The electrooptical switch (EOS) guides the photons alternately into the delay and the direct arms, such that successive photons meet at the 50:50 beam splitter (BS). A time-to-digital converter (TDC) records the photon detection times on the detectors Det 1 and 2. (b), (c) ${}^{40}{\text{Ca}}^{+}$ level scheme showing two single-photon schemes. $\mathrm{\Delta }$ represents the detuning from atomic resonance of the laser and cavity. (d), (e) The sequence of laser pulses used to generate single photons.

Figure 2

Coincidence probability density for single photons produced using the $S_{1/2}\to D_{3/2}$ scheme, normalizing the area under the curve for perpendicular polarized photons to unity. The orange circles show the reference signal obtained by fully distinguishable photons with orthogonal polarization, while the blue triangles show the coincidences between two detectors with parallel polarized photons. Error bars representing one standard deviation. The solid orange and dashed blue lines show the expected coincidence probability from numerical simulation for the orthogonal and parallel cases.

Figure 3

Temporal probability distribution of detecting single photon shown as blue dots for the $S_{1/2}\to D_{3/2}$ scheme (left) and $D_{3/2}\to D_{3/2}$ (right). To extract this plot, all the photon arrival times with respect to the sequence trigger during the measurements are sorted into 20 ns time bins and the resulting histograms are normalized to unity. The solid lines are obtained by numerical simulation.

Figure 4

Time-resolved HOM interference signal for single photons produced using the $D_{3/2}\to D_{3/2}$ scheme. As in Fig. 2, the orange dots and blue triangles show the signal obtained from orthogonal and parallel polarized photons, respectively. The solid and dashed lines are from numerical simulations of the system.

Figure 5

Simulated visibility and emission probability for the $S_{1/2}\to D_{3/2}$ and $D_{3/2}\to D_{3/2}$ schemes plotted against Raman detuning and drive laser Rabi frequency in units of the transition linewidth ${\mathrm{\Gamma }}_{SD}$. (a) $S_{1/2}\to D_{3/2}$ efficiency. (b) $D_{3/2}\to D_{3/2}$ efficiency. (c) $S_{1/2}\to D_{3/2}$ visibility. (d) $D_{3/2}\to D_{3/2}$ visibility.

Figure 6

Visibility (blue) and coincidence count probability (orange) with varying window size $T$ for the simulation of the $S_{1/2}\to D_{3/2}$ scheme. The coincidence count probability is normalized to the experimental rate. The insert demonstrates the temporal filtering; coincidences in the gray shaded area are neglected, increasing the visibility.