Optical lattice clocks are statistically limited by the quantum projection noise (QPN) of uncorrelated atoms, a floor that, for an ideal unentangled ensemble, defines the standard quantum limit (SQL). Quantum non-demolition (QND) measurement of the collective atomic spin via cavity QED can prepare spin-squeezed states that beat this limit. The source demonstration (Robinson et al., Nature Physics 20, 208 (2024)) squeezed two strontium sub-ensembles in one apparatus and compared them 2.0(3) dB below the QPN level but still 0.6(3) dB above the SQL, leaving a five-item roadmap: unexplained excess anti-squeezing noise, Ramsey-contrast loss from QND scattering and transport-induced motional dephasing, insufficient atom number, cooperativity, and quantum efficiency for a clear sub-SQL operating point, asymmetric noise between the two ensembles, and scaling to clock networks. A 2025 follow-up on the same platform (Yang et al., Physical Review Letters 135, 193202 (2025)) has now resolved the central milestone: a synchronous clock comparison 2.0(2) dB beyond the SQL (after state-preparation-and-measurement correction) with about 30,000 atoms at 10−1810^{-18}-level precision. The surviving open core is threefold: converting sub-SQL squeezing into routine, systematically (accuracy-)evaluated clock operation at the 10−1810^{-18} level and below; identifying and mitigating the residual noise mechanisms (rotation-induced anti-squeezing-quadrature noise, any excess anti-squeezing beyond the quantum-efficiency prediction, asymmetric ensemble noise); and extending entanglement-enhanced comparison from sub-ensembles of one apparatus to independent, spatially separated clocks (networks).

Background

An optical atomic clock measures an ultra-stable laser against a narrow atomic transition — for 87^{87}Sr, the 1S0→3P01S_0\to 3P_0 clock transition at 698 nm — of many neutral atoms trapped in standing waves of light (an optical lattice clock). Averaging NN uncorrelated atoms reduces the statistical frequency uncertainty as 1/N1/\sqrt{N}; the corresponding floor is the quantum projection noise (QPN), and once technical noise is removed it defines the standard quantum limit (SQL) of an ideal unentangled ensemble. In a clock comparison of two sub-ensembles of NAN_A and NBN_B atoms at initial Ramsey fringe contrast CiC_i, the QPN-limited comparison stability scales as 1/(CiNA+NB)1/(C_i\sqrt{N_A+N_B}), while the SQL is the stricter, lower bound 1/Ci(NA+NB)1/\sqrt{C_i(N_A+N_B)} that treats the (1−Ci)(1-C_i) fraction of atoms lost to finite fringe contrast as no longer participating; the two limits differ by 10log⁡10(1/Ci)10\log_{10}(1/C_i) in stability — about 2.6 dB at the Ci=0.55C_i=0.55 measured by Robinson et al. Entanglement can lower this floor: a spin-squeezed state has suppressed fluctuations of the collective spin projection JzJ_z at the price of enlarged fluctuations (anti-squeezing) in the conjugate quadrature. Such states can be prepared by a quantum non-demolition (QND) measurement, in which the atoms' collective dispersive coupling to an optical cavity (cavity quantum electrodynamics, cavity QED; strength characterized by the single-atom cooperativity C=4g2/(κΓ)C=4g^2/(\kappa\Gamma) with atom-cavity coupling gg, cavity linewidth κ\kappa, and excited-state decay rate Γ\Gamma) shifts the cavity resonance by an amount proportional to the atom number, so that homodyne detection of the probe light (an interferometric measurement of its phase) measures JzJ_z with minimal back-action on the spin. Metrological usefulness is quantified by the Wineland parameter ξ2=R Ci/Cf2\xi^2 = R\,C_i/C_f^2, where RR is the spin-noise reduction relative to QPN and CiC_i, CfC_f are the Ramsey fringe contrasts (the visibility of the clock interference signal) before and after the QND measurement; ξ2=1\xi^2=1 marks the SQL and ξ2<1\xi^2<1 witnesses entanglement. Entanglement on an optical clock transition was first demonstrated in 2020 (Pedrozo-Peñafiel et al., Nature 588, 414 (2020)). The source work (Robinson et al., Nature Physics 20, 208 (2024); arXiv:2211.08621) integrated such a cavity-QED system into a competitive 87^{87}Sr optical lattice clock: 1.9×1041.9\times 10^4 atoms in a movable 1D lattice were shuttled through the cavity so that two spatially separated sub-ensembles could be independently squeezed and then synchronously compared, so that common clock-laser noise cancels in the difference. The authors measured a Wineland parameter of −1.8(7)-1.8(7) dB and a two-clock comparison 2.0(3)2.0(3) dB below the QPN level but 0.6(3)0.6(3) dB above the SQL. They explicitly identified what still stood in the way of routine entanglement-enhanced operation: (i) the measured anti-squeezing sat about 9 dB above the level predicted by the estimated total quantum efficiency Q=0.28Q=0.28 of the detection chain, an unexplained excess noise source; (ii) Ramsey contrast was lost both to free-space scattering from the QND probe and to single-particle motional effects of the optical π/2\pi/2 rotations and lattice transports; (iii) a clear operating point below the SQL would require larger atom number, higher cooperativity, and better quantum efficiency; (iv) a small asymmetric noise between the two ensembles appeared as an optimal differential estimator βD=0.907(5)≠1\beta_D=0.907(5)\neq 1, attributed to inhomogeneous ac-Stark shifts (position-dependent, light-induced shifts of the clock-transition frequency), unequal squeezing, or differential thermal motion; and (v) scaling to entangled clock networks remained, and remains, proposal-only (Polzik and Ye, Physical Review A 93, 021404 (2016)). In 2025 the same JILA platform, upgraded to a two-dimensional lattice with sideband cooling (laser cooling of the atoms close to the vibrational ground state of each lattice well) and spin-echo QND measurements (a pulse sequence that refocuses slow dephasing during probing), resolved the central milestone (Yang et al., Physical Review Letters 135, 193202 (2025)): with about 30,000 atoms, a spin-noise reduction of −7.2(1.0)-7.2(1.0) dB and metrological gain of −5.1(1.0)-5.1(1.0) dB were reached, and a synchronous comparison of two spin-squeezed ensembles achieved 8.0(2)×10−17/τ8.0(2)\times 10^{-17}/\sqrt{\tau} instability and 2.0(2)2.0(2) dB beyond the SQL after correcting for state-preparation-and-measurement (SPAM) errors, at a single-clock precision of 1.1×10−181.1\times 10^{-18}. Theoretical work has meanwhile clarified constraints on such comparisons: clock-laser noise limits synchronous comparisons of squeezed clocks (Yu, Zhang, and Chen, Physics Letters A 552, 130634 (2025)), and simulated differential comparisons of conditionally (measurement-outcome-conditioned) and unconditionally squeezed states are limited by decoherence from probe-photon quantum jumps, i.e. spontaneous photon-scattering events (Yu et al., Advanced Quantum Technologies 8, 2400491 (2025)). The classical frontier has also advanced — an unentangled Wannier-Stark lattice clock (one operated in a strongly tilted optical-lattice potential, where the atoms occupy tilt-localized Wannier-Stark states) now shows 118 s coherence and 1.5×10−181.5\times 10^{-18} instability at 1 s (Kim et al., Physical Review Letters 135, 103601 (2025)) — setting the bar any entanglement gain must beat.

Problem Statement

Can quantum-non-demolition spin squeezing be converted into entanglement-enhanced operation of an optical clock as a working frequency standard below the standard quantum limit — routinely operated, systematically evaluated, and ultimately extended to comparisons between independent clocks? The headline milestone of the source roadmap, a clock comparison operating below the SQL, has been achieved on a single apparatus (Yang et al., Physical Review Letters 135, 193202 (2025): 2.0(2)2.0(2) dB beyond the SQL after correction for state-preparation-and-measurement (SPAM) errors at 10−1810^{-18}-level precision). The open core, after the roadmap items resolved by later work, is: (1) Routine sub-SQL operation with accuracy evaluation — demonstrate that a spin-squeezed clock's stability gain below the SQL persists in sustained clock operation (not a single drift-corrected dataset), and evaluate its systematic (accuracy) uncertainty at the 10−1810^{-18} level and below to the standard of unentangled optical lattice clocks. (2) Residual noise mechanisms — identify and mitigate the mechanisms that still cap the metrological gain: noise from the anti-squeezed quadrature (the conjugate spin component with enlarged fluctuations) mapped in by imperfect optical π/2\pi/2 rotations, any excess anti-squeezing beyond the level predicted by the measured quantum efficiency of the detection chain (the ~9 dB unexplained excess of the source demonstration), and small asymmetric technical noise between the two ensembles (βD≠1\beta_D\neq 1). (3) Scaling beyond one apparatus — extend entanglement-enhanced comparison to independent, spatially separated clocks (clock networks), preserving the independence and coherence of the compared ensembles, for which only theoretical proposals exist. These three components jointly serve one objective — making entanglement a routine instrument capability of precision timekeeping rather than a laboratory demonstration — and an answer may resolve any component, with the resolution of each judged against the corresponding part of the objective. For the two components whose definitive resolution would require new apparatus data — (1) and (3) — an answer may take the form of a quantitative protocol, design, or error-budget analysis grounded in the cited experiments, and need not contain new experimental data.

References

  1. John M. Robinson, Maya Miklos, Yee Ming Tso, Colin J. Kennedy, Tobias Bothwell, Dhruv Kedar, James K. Thompson, Jun Ye. Direct comparison of two spin-squeezed optical clock ensembles at the 10^-17 level. Nature Physics 20, 208–213 (2024). DOI: 10.1038/s41567-023-02310-1. https://doi.org/10.1038/s41567-023-02310-1 (preprint arXiv:2211.08621, https://arxiv.org/abs/2211.08621)
  2. Y. A. Yang, Maya Miklos, Yee Ming Tso, Stella Kraus, Joonseok Hur, Jun Ye. Clock Precision beyond the Standard Quantum Limit at 10^-18 Level. Physical Review Letters 135, 193202 (2025). DOI: 10.1103/6v93-whwq. https://doi.org/10.1103/6v93-whwq (preprint arXiv:2505.04538, https://arxiv.org/abs/2505.04538)
  3. Deshui Yu, Shougang Zhang, Jingbiao Chen. Clock laser noise constraining synchronous comparison of spin-squeezed optical clocks. Physics Letters A 552, 130634 (2025). DOI: 10.1016/j.physleta.2025.130634. https://doi.org/10.1016/j.physleta.2025.130634
  4. Deshui Yu, Xiaobo Xue, Jia Zhang, Shougang Zhang, Jingbiao Chen. Differential Frequency Comparison of Conditionally and Unconditionally Spin-Squeezed States of an Optical Lattice Clock. Advanced Quantum Technologies 8, 2400491 (2025). DOI: 10.1002/qute.202400491. https://doi.org/10.1002/qute.202400491
  5. Kyungtae Kim, Alexander Aeppli, William Warfield, Anjun Chu, Ana Maria Rey, Jun Ye. Atomic Coherence of 2 Minutes and Instability of 1.5×10^-18 at 1 s in a Wannier-Stark Lattice Clock. Physical Review Letters 135, 103601 (2025). DOI: 10.1103/3wtv-sty2. https://doi.org/10.1103/3wtv-sty2
  6. E. S. Polzik and Jun Ye. Entanglement and spin squeezing in a network of distant optical lattice clocks. Physical Review A 93, 021404 (2016). DOI: 10.1103/PhysRevA.93.021404. https://doi.org/10.1103/PhysRevA.93.021404
  7. Edwin Pedrozo-Peñafiel, Simone Colombo, Chi Shu, Albert F. Adiyatullin, Zeyang Li, Enrique Mendez, Boris Braverman, Akio Kawasaki, Daisuke Akamatsu, Yanhong Xiao, Vladan Vuletić. Entanglement on an optical atomic-clock transition. Nature 588, 414–418 (2020). DOI: 10.1038/s41586-020-3006-1. https://doi.org/10.1038/s41586-020-3006-1
  8. Katherine Wright. Squeezing More out of Atomic Clocks. Physics 17, 8 (2024). DOI: 10.1103/physics.17.8. https://doi.org/10.1103/physics.17.8