{"id":"eb74f2fe-398d-4242-934d-07c396ec98ab","arxiv_id":"2412.09780","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":2,"one_line_summary":"A cryogenic optical tweezer system with a 45-50 K cold box and room-temperature high-NA objective demonstrates a 3000 s atom trap lifetime, low imaging and cooling losses, and coherent microwave and Rydberg control of single Rb atoms.","lead":"A new cryogenic optical tweezer apparatus for rubidium atoms combines a 4 K cryopumping surface, a cold box at 45-50 K, and a room-temperature high-numerical-aperture lens, achieving a 3000-second atom trap lifetime with coherent ground-state and Rydberg control. The platform addresses two scaling bottlenecks for neutral-atom quantum computers: background-gas-limited trap lifetime and blackbody-radiation-induced Rydberg decoherence.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified. The 3000 s headline is a conservative correction of a directly measured 2800 s lifetime; the claim survives even if the correction is off.","rationale":"The reader's ACCEPT verdict is appropriate. The weakest assumption identified by the reader, the 2×10−4 per-pulse correction, is indeed the most uncertain input in the lifetime analysis, but it is not load-bearing: the raw fit already gives 2800 s, and the corrected value is a conservative estimate of the vacuum-limited lifetime. A factor-of-two error in the correction changes the headline by less than 10%, within the reported uncertainties. The paper's other claims—low imaging and cooling losses at the 10−4 level and coherent ground and Rydberg control—are directly demonstrated. The Rydberg simulation and experiment mismatch is a minor discrepancy that does not affect the feasibility claim, especially given the direct room-temperature comparison. The paper is transparent about limitations, including the absence of ITO shielding and the technical-noise-limited Rydberg coherence. No load-bearing concern was identified, so the verdict should remain UNCHANGED.","tokens_in":27167,"tokens_out":12116,"duration_ms":127277,"concrete_test":"Run the variable-number-of-pulses cooling-loss protocol (Fig. 4a) with the exact lifetime-sequence settings—saturated repumping and 0.46 mK trap depth—and use the measured per-pulse loss to re-derive the corrected lifetime. If the per-pulse loss falls within a factor of about two of 2×10−4, the headline 3000 s stands; this test directly anchors the only estimated quantity in the headline claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—a 3000 s trap lifetime with low imaging and cooling losses and coherent ground and Rydberg control in a cryogenic platform—is supported by direct measurements. The raw lifetime fit is 2800+500/−400 s; the correction to 3000 s subtracts an estimated 2×10−4 per-pulse loss, which is disclosed and conservative in the sense that underestimating the pulse loss would make the true vacuum-limited lifetime longer, not shorter. The per-pulse cooling loss was independently measured as 1.1(1)×10−4 under standard conditions, and the lifetime-sequence estimate of 2×10−4 is consistent with the increased scattering from saturated repumping. Even a factor-of-two error in this correction shifts the corrected lifetime by only about 5–10%, which remains within the quoted uncertainties and preserves the qualitative claim. The Rydberg coherence data show a mild simulation versus experiment discrepancy (7.6 versus 5.4 µs), but the direct room-temperature comparison (6(1) µs) independently supports the claim that cryogenic operation does not introduce significant additional decoherence. No circularity or missing support was found in the argument.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a cryogenic optical tweezer array for 87Rb atoms that combines a 4 K cryopumping surface, a <50 K cold box surrounding the atoms, and a room-temperature high-numerical-aperture objective. The authors demonstrate a 3000 s trap lifetime (corrected from a raw exponential fit of 2800 s), measure cooling and imaging losses at the 10^-4 level, perform ground-state microwave qubit rotations, and demonstrate coherent two-photon excitation to a 70S1/2 Rydberg state. The manuscript also includes Monte Carlo simulations of defect-free array assembly and a detailed characterization of Rydberg decoherence sources.","tokens_in":27293,"tokens_out":3757,"duration_ms":38946,"significance":"If the results hold, this is a significant technical advance for neutral-atom quantum science. The combination of cryogenic environment, high optical access, and long trap lifetime addresses two known bottlenecks: background-gas-induced loss and blackbody-radiation-induced Rydberg decay. The careful methodology—maximum-likelihood lifetime fits with censoring, separate characterization of imaging and cooling losses, and a Monte Carlo decoherence simulation using independently measured noise inputs—supports the central claims. The paper also provides an honest discussion of limitations, such as the conservative correction of the lifetime and the residual discrepancy between measured and simulated Rydberg coherence.","major_comments":[],"minor_comments":[{"comment":"The phrase \"an cryogenic\" appears in the abstract and in the first sentence of the Introduction; it should be \"a cryogenic\".","section":"Abstract and Introduction"},{"comment":"The abstract states \"We demonstrate a 3000 s atom trap lifetime\" without noting that this value is corrected for an estimated cooling loss; the main text discloses this clearly, but the abstract could be misinterpreted as a directly measured lifetime.","section":"Section IIIB2"},{"comment":"The caption describes the orange curves as guides for exponential decays, but the lifetimes are extracted using MLE with censoring; it may be helpful to state that the curves are not the MLE fits themselves.","section":"Figure 4(c)"},{"comment":"The phrase \"a less than45 K cold shield\" is missing a space; it should be \"a less than 45 K cold shield\".","section":"Section IIA"}],"recommendation":"accept","confidential_remarks":"This is a strong experimental paper with well-executed measurements and transparent disclosure of assumptions. The lifetime correction is small and conservative, and the Rydberg coherence data are compared directly with a room-temperature control. I see no load-bearing errors. The paper is a good fit for the journal's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a genuine platform advance, not a headline grab. The authors combine, for the first time, a 4 K cryopumping surface, a <50 K cold box, a room-temperature high-NA objective, and coherent microwave and Rydberg control in one optical tweezer array. The 3000 s trap lifetime is a corrected number; the raw exponential fit is 2800+500/−400 s, and they subtract an estimated 2×10−4 per-pulse cooling loss to get 3000 s. That correction is disclosed, quantified, and conservative: if it is off by a factor of two, the lifetime shifts by only 5–10 percent, still inside the uncertainties. So the headline is honest.\n\nWhat the paper does well: the loss accounting is unusually careful. They measure cooling loss and imaging loss separately with maximum-likelihood estimation, including censoring for the longitudinal data. The per-image loss of 3.8×10−4 at 14 ms is among the best for alkali atoms. The Rydberg coherence simulation uses independently measured phase noise, intensity noise, beam pointing, temperature, and inhomogeneities as inputs; it does not fit the data. That is a clean way to argue that cryogenic operation does not add a mysterious decoherence channel.\n\nSoft spots: minor ones. The simulation predicts a 7.6 µs Gaussian decay for the Rydberg Rabi oscillation while the data shows 5.4 µs. The paper calls this near agreement, which is a bit generous, but the room-temperature measurement of 6(1) µs independently supports the main conclusion that the cryostat is not the culprit. The lifetime correction relies on an estimated rather than directly measured per-pulse loss in the long sequence, though the estimate is consistent with the independently measured 1.1×10−4 under standard conditions. No code or raw data are shipped, but the methods are described in enough detail that a specialized group could rebuild the setup.\n\nCitation pattern looks appropriate; they compare directly to Schymik et al. (6000 s, limited optical access) and Pichard et al. (large arrays, shorter lifetime), and these comparisons are fair.\n\nBottom line: this paper deserves a serious referee. It is a well-executed apparatus paper that advances the neutral-atom platform. I would send it out. The main thing I would ask the authors in revision is to present the raw and corrected lifetimes side by side more prominently in the abstract or conclusion, so the headline is not mistaken for a directly measured 3000 s.","headline":"A well-executed cryogenic tweezer platform; the 3000 s lifetime is a disclosed correction of a 2800 s raw fit, and the paper's claims hold up.","tokens_in":27947,"tokens_out":4825,"would_cite":true,"duration_ms":44889,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper shows a cryogenic tweezer array with a 4 K cryopump and <50 K cold box reaching a 3000 s trap lifetime and coherent Rydberg control.","keywords":["cryogenic atom arrays","optical tweezers","Rydberg atoms","atom trap lifetime","cryopumping","blackbody radiation","qubit control","hydrogen pressure"],"falsifier":"Measure the per-pulse cooling loss in the exact lifetime configuration (0.46 mK trap depth, one 25 ms cooling pulse every 10 s) by comparing survival with and without the cooling pulses over the same total hold time, and compare the resulting corrected lifetime with one obtained from a sequence that never applies cooling light; a discrepancy would invalidate the 3000 s value.","tokens_in":26897,"feed_emoji":"❄️","tokens_out":11354,"duration_ms":102760,"temperature":0.7,"pith_summary":"This paper reports a cryogenic platform for Rydberg atom arrays that keeps the atoms in a <50 K cold box while a 4 K surface cryopumps hydrogen, reaching a vacuum-limited atom trap lifetime of 3000 s. That long hold time lets the authors measure imaging and cooling losses at the $10^{-4}$ per-operation level, the losses that matter for assembling defect-free arrays. The same apparatus demonstrates microwave-driven ground-state qubit rotations and two-photon coherent excitation to the $70S_{1/2}$ Rydberg state with the electric field nulled. The deeper goal is a cold blackbody environment that suppresses unwanted transitions out of Rydberg states, which currently limit gate fidelity and many-body simulation.","feed_headline":"Cryogenic cold box keeps atoms trapped for 3000 seconds","feed_subtitle":"A 4 K cryopump and <50 K walls give vacuum-limited trapping plus coherent Rydberg control in one high-optical-access platform.","key_machinery":"The central object is the cold box: a <50 K differentially pumped metal enclosure surrounding the atoms, built with cold windows for laser access and a small 4 K cold finger that cryopumps hydrogen. A room-temperature high-NA objective sits outside the cold box and creates the tweezer array, while the cold box's slit and titanium pedestal limit eddy currents and thermal load, and lead-shot-damped helium lines reduce vibration to 3 nm RMS. This arrangement makes the vacuum at the atoms largely independent of the outer room-temperature chamber, and sets up the low-temperature blackbody environment that Rydberg operations will benefit from once the windows are coated with conductive ITO.","core_discovery":"The central claim is that high optical access and a cryogenic environment are not in conflict for neutral-atom arrays: a room-temperature high-NA objective can sit outside a differentially pumped <50 K cold box, and a small 4 K cryopumping surface can reduce the hydrogen background enough to give a 3000 s single-atom trap lifetime (raw exponential fit 2800 s, corrected for a per-pulse cooling loss of $2\\times10^{-4}$). With this lifetime, the authors resolve loss channels at the $10^{-4}$ level: $1.1(1)\\times10^{-4}$ loss per 25 ms cooling pulse and $3.8(4)\\times10^{-4}$ loss per 14 ms image. They also show a Ramsey dephasing time $T_2^* = 131(5)$ µs with a sub-mG magnetic-field fluctuation, and Rabi oscillations between the ground state and $70S_{1/2}$ at $2\\pi\\times3.8$ MHz with a $5.4(7)$ µs Gaussian decay that the Monte Carlo analysis attributes to laser technical noise rather than to the cryogenic environment. The cold box is expected to reduce blackbody-radiation-induced decay to nearby Rydberg states by a factor of 8 for an ideal 40 K environment, raising the calculated $70S_{1/2}$ lifetime from 150 µs at room temperature to 310 µs.","pith_inferences":["The paper's lifetime is corrected rather than directly measured; a direct comparison of cooling-pulse loss inside the 3000 s sequence would test whether the hydrogen-pressure estimate is accurate.","If the cold box temperature can be scanned, the Rydberg lifetime itself becomes a quantitative probe of the blackbody environment, a measurement the present paper leaves for future work.","The same <50 K, high-optical-access design should transfer to other trapped species, including molecules and alkaline-earth atoms, where BBR-driven rovibrational or metastable-state transitions are the analogous loss channel."],"forward_implications":["Lifetimes above about 1000 s shift the limit on assembling large defect-free arrays from background-gas collisions to imaging and cooling losses; the simulations show little further gain from 3000 s to 6000 s at the measured loss rates.","The inferred hydrogen pressure at the atoms is $7\\times10^{-13}$ mbar, which the authors argue is decoupled from the outer vacuum and limited by residual outgassing inside the cold box.","With imaging loss of $3.8(4)\\times10^{-4}$ per 14 ms image and cooling loss of $1.1(1)\\times10^{-4}$ per 25 ms pulse, an improved camera or collection efficiency could reach roughly $3\\times10^{-5}$ per image and enable about 90% defect-free probability for arrays near 1000 atoms.","Sub-mG magnetic-field stability and a $T_2$ of $0.95(5)$ ms show that the cryogenic environment is compatible with single-qubit microwave control, and the Rydberg Rabi data indicate the cryostat does not add significant decoherence on the measured timescales.","Adding ITO-coated windows should extend the platform's benefit to Rydberg operations by blocking room-temperature microwave blackbody radiation, increasing Rydberg lifetime and suppressing many-body loss channels."],"supporting_citations":[{"why":"This reference reports the prior 6000 s single-atom cryogenic tweezer lifetime with limited optical access, the benchmark this design combines with high-NA access.","marker":"[62]"},{"why":"This reference demonstrates large cryogenic tweezer arrays but with line of sight to room-temperature surfaces, the limitation this cold box removes.","marker":"[63]"},{"why":"This reference provides the longest room-temperature tweezer lifetime (1400 s) and a low per-image imaging loss benchmark used for comparison.","marker":"[24]"},{"why":"This reference shows that imaging loss limits defect-free array size, motivating the high-lifetime, low-loss measurements here.","marker":"[56]"},{"why":"This reference supplies the rubidium-hydrogen collisional loss cross section used to convert the lifetime into an equivalent hydrogen pressure of $7\\times10^{-13}$ mbar.","marker":"[85]"},{"why":"This reference supplies the two-photon 'inverted scheme' for exciting $^{87}$Rb to the $70S_{1/2}$ Rydberg state through the $6P_{3/2}$ intermediate state.","marker":"[90]"},{"why":"This reference provides the gray-molasses loading scheme used to achieve over 80% single-atom loading probabilities across the tweezer array.","marker":"[75]"}],"fun_headline_variants":["3000-second atom trap lifetime at cryogenic temperatures","High optical access cryostat holds atoms 3000 seconds","Cold box and high-NA objective give 3000s atom lifetime","Cryogenic environment enables 3000s trap and Rydberg control","Atoms trapped 50 minutes in cryogenic optical tweezers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The 3000 s lifetime is a corrected number: the raw fit is 2800 s, and the correction assumes a per-pulse cooling loss of $2\\times10^{-4}$ estimated from scattered-photon counts rather than measured directly in the long lifetime sequence; if that estimate is wrong, both the lifetime and the inferred hydrogen pressure shift.","fun_headline_variants_meta":{"raw":{"variants":["3000-second atom trap lifetime at cryogenic temperatures","High optical access cryostat holds atoms 3000 seconds","Cold box and high-NA objective give 3000s atom lifetime","Cryogenic environment enables 3000s trap and Rydberg control","Atoms trapped 50 minutes in cryogenic optical tweezers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001565,"raw_usage":{"total_tokens":6304,"prompt_tokens":1052,"completion_tokens":5252,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":668,"completion_tokens_details":{"reasoning_tokens":5163}},"tokens_in":668,"tokens_out":5252,"duration_ms":34816,"temperature":1.0,"reasoning_tokens":5163,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T16:43:22.529045+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the per-pulse cooling loss in the exact lifetime configuration (0.46 mK trap depth, one 25 ms cooling pulse every 10 s) by comparing survival with and without the cooling pulses over the same total hold time, and compare the resulting corrected lifetime with one obtained from a sequence that never applies cooling light; a discrepancy would invalidate the 3000 s value.","supporting_citations":[{"cited_title":"Timmermans, Degenerate fermion gas heating by hole creation, Physical Review Letters87, 240403 (2001)","cited_arxiv_id":null,"evidence_quote":"This reference reports the prior 6000 s single-atom cryogenic tweezer lifetime with limited optical access, the benchmark this design combines with high-NA access."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"This reference demonstrates large cryogenic tweezer arrays but with line of sight to room-temperature surfaces, the limitation this cold box removes."},{"cited_title":"De Léséleuc, D","cited_arxiv_id":null,"evidence_quote":"This reference provides the longest room-temperature tweezer lifetime (1400 s) and a low per-image imaging loss benchmark used for comparison."},{"cited_title":"Norciaet al., Iterative Assembly of 171 Yb Atom Ar- rays with Cavity-Enhanced Optical Lattices,PRX Quan- tum 5, 030316 (2024)","cited_arxiv_id":null,"evidence_quote":"This reference shows that imaging loss limits defect-free array size, motivating the high-lifetime, low-loss measurements here."},{"cited_title":"Efron and R","cited_arxiv_id":null,"evidence_quote":"This reference supplies the rubidium-hydrogen collisional loss cross section used to convert the lifetime into an equivalent hydrogen pressure of $7\\times10^{-13}$ mbar."},{"cited_title":"Shadmanyet al., Cavity QED in a high NA resonator, Science Advances11, eads8171 (2025)","cited_arxiv_id":null,"evidence_quote":"This reference supplies the two-photon 'inverted scheme' for exciting $^{87}$Rb to the $70S_{1/2}$ Rydberg state through the $6P_{3/2}$ intermediate state."}],"review_version":1}