Strong coupling of a reconfigurable ${}^{171}$Yb atom array to a tunable telecom-band nanofiber cavity

AI summary

Overview

Research area: Quantum physics — cavity quantum electrodynamics (cavity QED), neutral-atom quantum computing, and quantum networking hardware.

Technical level: Advanced. The paper assumes familiarity with cavity QED parameters (cooperativity, coupling rates, critical/overcoupling regimes), optical tweezer arrays, and fiber Bragg gratings.

Scope: The paper reports an experimental platform that strongly couples a reconfigurable array of tweezer-trapped ¹⁷¹Yb atoms to a nanofiber cavity operating directly on a 1389-nm telecom-band atomic transition, and demonstrates that the atom-array functionality survives next to the device.

What This Paper Is About

Neutral-atom quantum processors need a way to send quantum information between distant nodes, and the natural messenger is a photon. Doing this efficiently requires placing atoms inside an optical cavity so that emitted photons go into one well-defined mode, and it requires the photons to travel at telecom wavelengths where fiber loss is low. Most neutral-atom cavity systems operate at shorter wavelengths and reach the telecom band only via quantum frequency conversion. This paper builds a nanofiber cavity resonant with the ¹⁷¹Yb ³P₀ ↔ ³D₁ transition at 1389 nm and shows that atoms held in optical tweezers next to it reach the single-atom strong-coupling regime while keeping the imaging, trapping, and transport capabilities an atom array needs.

Key Contributions

1. First telecom-band strong coupling with neutral atoms. The authors report strong atom–cavity coupling on the ¹⁷¹Yb ³P₀ ↔ ³D₁ transition at 1389 nm, which they describe as the first neutral-atom cavity-QED system to reach the single-atom strong-coupling regime on a telecom-band atomic transition (to their knowledge).
2. Array-compatible cavity coupling. Collective coupling follows the expected √N scaling up to N = 5 atoms, and antinode coupling is homogeneous over approximately 200 μm along the nanofiber, which the authors translate into a capacity of 100 tweezer sites at 4 μm spacing (500 sites for a 1 mm uniform region).
3. In situ tunable outcoupling. Thermal tuning of the output fiber Bragg grating varies κ_ex/κ_in over nearly two orders of magnitude, spanning the undercoupled, critically coupled, and overcoupled regimes, so one device can serve multiple protocol-specific operating points.
4. Lossless integration with atom-array operations. Site-resolved imaging reaches 99.95(6)% fidelity at 1 μm from the nanofiber axis, trap lifetimes are comparable to free space, and a diameter-engineered antireflection nanofiber (599 nm) permits tweezer transport across the fiber at 4.5 μm with no measurable loss or heating.

Main Findings

• Single-atom coupling: Reflection spectroscopy of a single-sided cavity gives g/2π = 1.86(17) MHz, corresponding to an internal cooperativity C_in = g²/(2γκ_in) = 5.6(1.0).
• Shortfall from the ideal prediction: The measured coupling is below the calculated g₀/2π = 2.95 MHz (C_in,0 = 14.2), a reduction the authors attribute to thermal spread of the atom in the tweezer and imperfect clock-state preparation and depumping during the probe.
• Collective coupling: For N = 3, 4, and 5 atoms, fitted collective couplings are G/2π = 2.89(19), 3.61(20), and 3.78(19) MHz, consistent with G = G₁√N where G₁ = 1.74(5) MHz.
• Spatially resolved mode structure: Scanning the atom radially resolves the evanescent decay, and scanning along the fiber resolves the intracavity standing wave with a period of 676 nm, matching λ/(2n_eff) for n_eff = 1.027.
• Uniformity along the fiber: Peak coupling at antinodes near y₀ = −100, 0, and 100 μm is comparable; scanning electron micrographs of a separately fabricated device show less than 1% diameter fluctuation over a 1 mm nanofiber region.
• Tunable outcoupling range: Empty-cavity spectra are shown at κ_ex/κ_in = 8.03(42), 2.13(15), 0.93(9), and 0.25(6). At κ_ex ≈ 0.1κ_in, the total decay rate falls to κ_tot/2π ≈ 1.41 MHz, satisfying g > κ_tot, γ for the measured coupling.
• Coupling survives overcoupling: With N = 5 atoms at κ_ex/κ_in = 1.82(9), the fitted collective coupling is G/2π = 3.9(1.2) MHz.
• Trapping lifetime: 6.8(4) s at 1 μm from the fiber versus 6.7(4) s at 80 μm, i.e., no measurable lifetime reduction near the device.
• Imaging: Per-image survival probabilities are 98.2(4)% at 1 μm and 97.9(4)% at 80 μm with a 30-ms imaging pulse; imaging fidelity is 99.95(6)% at both positions, achieved by blocking the excitation beam along the axis orthogonal to the imaging path to suppress fiber-scattered background.
• Transport: At an optimized fiber diameter of 599 nm and a tweezer focus 4.5 μm from the fiber center, atom survival after shuttling across the fiber at 0.14 μm/μs is consistent with unity, with unchanged atomic temperature; other diameters such as 450 nm produce pronounced standing-wave modulation of the tweezer potential.
• Projected entanglement rate: Using C_in = 5.6, an atom capacity of 50 per side of the nanofiber, and a 40 μs shuttling overhead, the authors estimate 40 kHz remote entanglement generation with the time-multiplexed cavity-assisted photon-scattering protocol of Ref. [26] at a target Bell-pair fidelity of 99%.

Methodology in Plain English

The device is an ordinary silica optical fiber pulled down to a waist about 550 nm across, with two fiber Bragg gratings written into the standard fiber on either side so that the waist region forms a small Fabry–Pérot cavity resonant near 1389 nm. One grating is highly reflective (R₁ ∼ 99.98%) and the other is the output coupler (R₂ ∼ 99.5%), giving a 3 cm-long cavity with a 3.17 GHz free spectral range, an internal loss rate κ_in/2π = 1.28(6) MHz, and an intrinsic finesse of 1240(60). A portion of the cavity light leaks out of the fiber surface as an evanescent field, and atoms placed in that field interact with the cavity mode.

Ytterbium-171 atoms are loaded into a 759-nm optical tweezer array, using light-assisted collisions to ensure single occupancy, then moved to 1 μm from the fiber axis and optically pumped into the ³P₀ clock state. To measure coupling, the team shines a weak 1-ms probe through the output grating and records the reflected photons with a superconducting nanowire single-photon detector, fitting the resulting spectra with a single-sided cavity input–output model. Because a tweezer can place an atom anywhere, the same atom doubles as a scanner that maps out the cavity mode in space. Separately, the output grating is heated to shift its stopband, which changes how strongly the cavity leaks out without changing the cavity resonance or internal loss much. Finally, the team characterizes the array operations next to the device: holding atoms and imaging them, and shuttling a tweezer straight across the fiber on a second, nearly identical setup with a 599-nm waist.

Why This Matters

Impact on research: The work removes a piece of overhead from neutral-atom quantum networking. Instead of generating photons at a visible or near-infrared atomic wavelength and converting them to telecom, the photon is emitted directly into the telecom band from a metastable qubit state that is already compatible with high-fidelity Rydberg gates and erasure-conversion logical qubits. It also extends nanofiber-cavity strong coupling, previously shown only with single trapped atoms, to a reconfigurable array.

Real-world applications:

• Distributed quantum computing: Remote entanglement between separate neutral-atom processors at rates relevant to fault-tolerant architectures.
• Quantum repeaters and long-distance links: Direct telecom-band photons avoid conversion loss in silica fiber.
• Secure quantum communication: Entanglement distribution between distant nodes underpins device-independent and other entanglement-based protocols.
• Networked quantum sensing: Entangled nodes can improve sensitivity in distributed metrology.
• Programmable photonic interconnects: The tunable outcoupling lets one cavity serve switching/routing, photon-scattering gates, or high-efficiency emission without hardware changes.

Industry relevance: The work comes largely from Nanofiber Quantum Technologies, Inc. (NanoQT) with co-authors at Waseda, RIKEN RQC, NICT, and Oxford, and includes a patent application (U.S. No. 19/793,903) on the lossless transport scheme. The fiber-native, lithography-free cavity geometry is directly fiber-coupled, which matters for manufacturability of networked quantum hardware.

Future Directions

1. Close the gap to the ideal coupling. Improved atom cooling and shorter probe duration should recover most of the deviation from g₀/2π = 2.95 MHz and C_in,0 = 14.2.
2. Push cooperativity higher. A shorter tweezer wavelength such as 486 nm would allow a 450 nm atom–surface distance and raise C_in to 40; reduced internal loss through improved fabrication would then raise it to 65.
3. Extend the uniform array beyond the current field of view. The sub-percent diameter variation over 1 mm suggests the 200 μm uniformly coupled region, presently limited by the objective's field of view, can be extended, and the 40 kHz estimate can be revisited with channel multiplexing across multiple cavities.
4. Demonstrate the full protocol. Actually generating remote entanglement with the time-multiplexed cavity-assisted photon-scattering scheme, and integrating the networking layer with a computing zone via the proposed two-layer transport architecture, remain to be shown.

Target Audience

Experimental quantum-optics and cavity-QED researchers; neutral-atom quantum computing groups building interconnects; photonics engineers working on fiber cavities, fiber Bragg gratings, and nanophotonic waveguides; and quantum-networking hardware developers who need telecom-compatible atom–photon interfaces. Readers without a background in cooperativity, cavity outcoupling regimes, or optical tweezer arrays will find the parameter-level detail dense.

Authors’ abstract

Nanophotonic cavities provide strong atom-photon coupling in a compact, directly fiber-coupled geometry, while a reconfigurable array of $^{171}$Yb atoms combines high-fidelity gates with a metastable qubit state coupled directly to a telecom-band transition, making the two a natural interface between neutral-atom quantum processors and telecom quantum networks. Here, we realize strong atom-cavity coupling on the ${}^{3}\mathrm{P}_0 \leftrightarrow {}^{3}\mathrm{D}_1$ transition of ${}^{171}\mathrm{Yb}$, establishing, to our knowledge, the first neutral-atom cavity-QED system to reach the single-atom strong-coupling regime on a telecom-band atomic transition. Reflection spectroscopy of a single-sided cavity yields a single-atom internal cooperativity of $C_{\mathrm{in}}=5.6(1.0)$ and collective coupling following $\sqrt{N}$ scaling up to $N=5$ atoms, with homogeneous coupling at antinodes over $\sim$200 $\mathrmμ$m along the nanofiber. Thermal tuning of a fiber Bragg grating provides in situ control of the cavity outcoupling rate over two orders of magnitude from the undercoupled to the overcoupled regime, relevant for various atom-cavity protocols. We confirm compatibility with reconfigurable atom arrays by demonstrating 99.95(6)\% imaging fidelity and trap lifetimes comparable to free-space array, at 1 $\mathrmμ$m from the nanofiber axis. Moreover, the diameter-engineered antireflection design of the nanofiber enables tweezer-based atom transport across the nanofiber with no measurable loss or heating at 4.5 $\mathrmμ$m above the nanofiber, supporting a two-layer architecture for the nanofiber-integrated atom array. From the demonstrated atom-cavity coupling, atom capacity, and cavity mode profile, we project time-multiplexed remote atom-atom entanglement generation at 40 kHz, with prospects for further scaling through technical enhancements and channel multiplexing.