All-photonic entanglement swapping with remote quantum dots

Entanglement swapping is a protocol that details how to create entanglement between previously uncorrelated particles. Its all-photonic version - mediated by the interference of photon pairs generated by separate quantum systems-finds disparate applications in quantum networks. So far, all-photonic entanglement swapping between remote systems has been implemented only using sources that operate probabilistically. However, the scaling up of quantum networks requires deterministic quantum emitters that do not suffer from a trade-off between degree of entanglement and photonpair generation rate. Here, we demonstrate all-photonic entanglement swapping using photon-pairs generated by two separate GaAs quantum dots. The emitters are deterministically embedded in hybrid semiconductor-piezoelectric devices that make the entangled-photons from two dissimilar quantum dots nearly identical. Entanglement swapping is demonstrated with a fidelity as high as 0.71(2), more than 10 standard deviations above the classical limit. The experimental data are quantitatively explained by a theoretical model that also suggests how to boost the protocol performances. Our work opens the path to the exploitation of quantum dot entangled-photon sources in quantum repeater networks.

💡 Research Summary

Entanglement swapping is a cornerstone protocol for quantum networks, allowing two distant parties to become entangled without any direct interaction. While all‑photonic swapping has been demonstrated in many platforms, all previous implementations relied on probabilistic photon‑pair sources such as spontaneous parametric down‑conversion, which suffer from a fundamental trade‑off: increasing the pair‑generation probability inevitably raises multi‑pair emission, degrading the fidelity of the swapped entanglement. Deterministic emitters that produce on‑demand, high‑quality entangled photon pairs are therefore essential for scalable quantum repeaters.

In this work the authors report the first all‑photonic entanglement swapping performed with two independent quantum‑dot (QD) sources. The QDs are GaAs structures fabricated by droplet‑etching and embedded in circular Bragg resonators (CBRs) that provide a Purcell enhancement of ≈10 for the biexciton (XX) and ≈8 for the exciton (X) transitions, shortening the radiative lifetimes to 5–25 ps. Each QD‑CBR chip is mounted on a micro‑machined PMN‑PT piezoelectric actuator. By applying a static electric field across the actuator, a controllable strain is induced in the semiconductor membrane, allowing fine tuning of the X and XX transition energies. The authors demonstrate that the energy mismatch between two nominally different QDs (≈120 µeV for X, ≈160 µeV for XX) can be reduced to sub‑µeV levels, well within the homogeneous linewidth, by adjusting the strain.

The entangled photon pairs emitted by each QD are characterized by full quantum‑state tomography, yielding entanglement fidelities of 0.90 ± 0.02 and 0.91 ± 0.01 with respect to the ideal |Φ⁺⟩ Bell state. The indistinguishability of photons from the two remote sources is assessed via Hong‑Ou‑Mandel (HOM) interference. After strain tuning, the raw two‑photon interference visibilities are V_X = 0.43 ± 0.01 for the exciton photons and V_XX = 0.46 ± 0.03 for the biexciton photons; after correcting for detector imperfections and residual g^(2)(0), the visibilities improve to ≈0.48 and ≈0.51, respectively. These values represent a five‑fold improvement over previous QD‑in‑CBR experiments and are comparable to bulk‑QD results, confirming that the cavity does not degrade photon quality. The remaining visibility limitation is attributed to residual time‑energy entanglement in the cascade and charge‑noise‑induced dephasing, which could be mitigated by faster detectors or temporal post‑selection.

For the swapping experiment, the authors implement a Bell‑state measurement (BSM) on either the X or the XX photons from the two QDs, using a 50:50 beam splitter, polarizing beam splitters, and four single‑photon detectors. Because linear optics can unambiguously identify only two of the four Bell states, the protocol succeeds with a 50 % intrinsic probability. When a successful BSM is registered, the remaining photons (the partner X or XX photons) are projected into the corresponding Bell state. The authors perform both configurations: (i) swapping on the XX photons while using the X photons for the BSM, and (ii) the reverse.

The swapped entanglement between the distant photons is quantified by measuring the fidelity of the reconstructed two‑photon density matrix. The reported swapped fidelity is 0.71 ± 0.02, which exceeds the classical bound of 0.5 by more than ten standard deviations. This value is in excellent agreement with a theoretical model that incorporates the measured source fidelities, HOM visibilities, detector efficiencies, and the negligible multi‑pair emission probability inherent to deterministic QDs. The model further predicts that by increasing the Purcell factor (e.g., via tighter cavity confinement), improving strain‑tuning precision, and reducing charge noise, the swapped fidelity could be pushed beyond 0.85.

Overall, the paper demonstrates that deterministic, on‑demand entangled photon sources can be engineered to be spectrally indistinguishable despite originating from distinct nanostructures. The combination of strain‑tuning, high‑Q nanophotonic cavities, and careful device fabrication enables high‑quality Bell‑state measurements between remote emitters. This achievement paves the way for scalable quantum‑repeater architectures where multiple swapping nodes are concatenated, as well as for photonic quantum‑computing schemes that rely on high‑fidelity entanglement swapping. Future work will likely focus on integrating fast feed‑forward electronics for real‑time BSM outcomes, extending the technique to larger networks of quantum dots, and exploring hybrid interfaces with quantum memories to realize fully functional quantum repeaters.