Telecom-Wavelength Quantum Repeater Node Based on a Trapped-Ion Processor

A quantum repeater node is presented based on trapped ions that act as single-photon emitters, quantum memories, and an elementary quantum processor. The node’s ability to establish entanglement across two 25-km-long optical fibers independently, then to swap that entanglement efficiently to extend it over both fibers, is demonstrated. The resultant entanglement is established between telecom-wavelength photons at either end of the 50 km channel. Finally, the system improvements to allow for repeater-node chains to establish stored entanglement over 800 km at hertz rates are calculated, revealing a near-term path to distributed networks of entangled sensors, atomic clocks, and quantum processors.

Figure 1

Quantum repeater concept. (a) A repeater node attempts to establish entanglement links (green lines) between its qubits (black spheres) and a qubit in each neighboring node separately, via the transmission of photons through channels on the order of the attenuation length ($L/2$). Whichever link succeeds first (left hand is shown) is stored in memory until the remaining link succeeds. The maximum link attempt rate is inversely proportional to the photon travel time [$L/(2c)$]. (b) A Bell state measurement in the repeater node establishes entanglement between the neighboring nodes.

Figure 2

Trapped-ion quantum repeater node. (a) Experimental schematic of an elementary network segment containing one repeater node. Photonic nodes $A$ and $B$ are each sent a telecom-converted [50, 51] 1550 nm photon, entangled with a different ion, via 25-km-long fiber spools. Ion-qubit readout is done via imaging ion fluorescence at 397 nm on a camera. Ion-qubit quantum logic (Q. logic) gates are done via a 729 nm laser that couples equally to the ions. AOD, acousto-optic deflector; HWP, half-wave plate; QWP, quarter-wave plate. Level scheme: $|S⟩=|4{{}^{2}S}_{1/2,m_j=-1/2}⟩$, $|S^{\prime }⟩=|4{{}^{2}S}_{1/2,m_j=+1/2}⟩$, $|P⟩=|4{{}^{2}P}_{3/2,m_j=-3/2}⟩$, $|D⟩=|3{{}^{2}D}_{5/2,m_j=-5/2}⟩$, $|D^{\prime }⟩=|3{{}^{2}D}_{5/2,m_j=-3/2}⟩$. (b) Envisioned concatenation of the network segment in (a) into a repeater chain. Photon detection heralds remote ion entanglement [52].

Figure 3

Repeater protocol results. (a) Left axes: probability distribution ($P_d$) for single-photon detection per $1\mathrm{\mu }\mathrm{s}$ time bin. Lighter and darker colored bars: loop 1 and loop 2 data, respectively. Time origins: Raman laser pulse onset. Dashed black lines: windows where DBSM is performed. Right axes: cumulative photon detection probability for loop 1 (lighter line) and loop 2 (darker line). (b) Photon detection probability (green bars) and counts (yellow bars) in attempt $k$ of loop 2. Black line (right axis): cumulative photon detection (count) probability, totaling $P_2=0.346(4)$. (c) Colored bars show measured real parts of the density matrix of the two single-photon polarization qubits shared by nodes $A$ and $B$, when ion $B$ stored a qubit in memory. Fidelities with targeted Bell states (wire mesh) are (i) ${72}_{-6}^{+7}\%$, (ii) ${67}_{-7}^{+7}\%$, (iii) ${69}_{-7}^{+6}\%$, (iv) ${0.83}_{-6}^{+6}\%$. Absolute values of imaginary parts do not exceed 0.09. Inset camera images: corresponding two-ion measurement (DBSM) outcome.

Figure 4

Quantum memory performance. Red diamonds: fidelity between the ion-photon state stored in ion $A$ after $k$ photon generation attempts on neighboring ion $B$, ${\rho }_{Aa}(k)$, and the maximally entangled state closest to ${\rho }_{Aa}(0)$. Red line: fit of the fidelity of the modeled ion-photon state (see text) stored in ion $A$, ${\tilde{\rho }}_{Aa}(t)$, to the data, yielding an ion-qubit decoherence time of $\tau =62(3)\mathrm{ms}$. Black dashed line: Bell state fidelity of the modeled photon-photon state established in attempt $k$, using perfect DBSM and feed forward (see text). Time axis: average attempt duration. Error bars represent one standard deviation of uncertainty due to Poissonian photon counting statistics.