Researchers from the Duke Quantum Center and IonQ have demonstrated the distributed generation of a Greenberger–Horne–Zeilinger (GHZ) state across a three-node quantum network using individual trapped atomic ions. The experimental configuration consisted of three spatially separated hardware modules positioned approximately 2 meters apart, linked by 3-meter single-mode optical fibers to a centralized, free-space GHZ-state generator. The networking setup achieved remote tripartite entanglement without requiring local two-qubit gates or post-selection protocols, establishing a bounded atomic state fidelity between 0.841(17) and 0.881(17) at an entanglement generation rate of 0.095(5) s−1.
Ion-Photon Entanglement Mechanics and Centralized Interference
Each hardware node isolates a single 138Ba+ (barium) ion qubit confined within a four-rod Paul trap. A static magnetic field of 4.2446(2) G lifts the ground-state degeneracy, defining the Zeeman qubit levels ∣↓⟩ and ∣↑⟩ with a split frequency of ω0=2π×11.8964(5) MHz. A 3-picosecond laser pulse simultaneously excites the ions, prompting a spontaneous emission sequence at 493 nm that generates an entangled ion-photon state with an average fidelity of 0.983(1). The emitted single photons travel through the fiber network to the centralized generator, where wave-plate arrays align their polarizations into a shared horizontal/vertical basis before undergoing pairwise interference at polarizing beam splitters. Triple-coincidence detection across six avalanche photodiodes erases photon “which-path” information, heralding the entangled state of the target atomic memories.
Error Demographics and Optical Channel Efficiencies
An analysis of the system’s operational constraints indicates that the primary sources of state infidelity stem from polarization mixing (0.037) and spatial mode mismatch (<0.03). Motional decoherence induced by photon recoil over the 50-nanosecond heralding window contributed an additional infidelity of 0.03, while state preparation and measurement (SPAM) configurations introduced an error rate of 0.015. Net end-to-end single-photon collection efficiencies (pi) for the three individual nodes ranged from 0.0074 to 0.0145. These efficiency limits are determined by fiber coupling drift and accumulated recoil heating during repetitive 1-microsecond entanglement cycles, which currently require periodic interruptions for Doppler cooling.
Mermin Inequality Violation and Loophole Closure
The three-node network was utilized to execute a deterministic test of quantum non-locality by measuring a Mermin parameter of 3.203(45), violating the classical local hidden variable upper bound of ≤2 by 27 standard deviations. Qubit states were analyzed by driving a 1762 nm laser π-pulse to shelf the ∣↓⟩ state populations into the target 2D5/2 manifold, followed by fluorescence imaging under combined 493 nm and 650 nm illumination. Because trapped atomic ions exhibit a high state-readout efficiency (>99.7%), the experiment closed the detection loophole. This represents a verification of a multi-node Mermin inequality violation using remote, individually addressable atomic memories rather than purely photonic or ensemble-based mediums.
Architectural Framework for Distributed Quantum Scaling
The realization of fully distributed tripartite entanglement outlines an engineering pathway for modular quantum computing systems. Rather than attempting to scale processing capacities within a single monolithic vacuum chamber, this infrastructure model connects distinct, localized quantum processing nodes through photonic interconnects to distribute computational workloads. The event-ready generation of these distributed states establishes a baseline framework for multi-party cryptographic protocols, secure quantum secret sharing, and distributed quantum sensing networks using interconnected atomic nodes.
The full technical paper detailing the multi-node network architecture and statistical derivations can be reviewed here. Executive commentary on the architectural transition toward modular, fiber-linked quantum computing networks can be tracked through the strategic landscape overview here.
June 20, 2026
