{"id":"69b3b033-1cce-4991-8cc5-9a83c61c278a","arxiv_id":"2512.05474","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A single-ion clock's collision-shift bound equals the classical Langevin collision rate reduced by a laser-decoupling suppression factor of about 0.03.","lead":"This paper derives a simple formula for the largest frequency error a background-gas collision can cause in a single-ion atomic clock. Clock teams can now estimate the bound from the standard Langevin collision rate and a small geometry-dependent suppression factor, without Monte-Carlo simulations or molecular potential curves.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (60) is not proven as a strict upper bound: the step-function replacement of the Ramsey suppression factor can understate the true velocity-averaged suppression.","rationale":"The reader's weakest assumption correctly identifies that Eq. (60) is an approximation rather than a strict inequality. My concern sharpens this: the step-function replacement of R(v) with cutoff at R=0.5 is not merely unproven as a bound, but may systematically undercount the contribution from velocities above the cutoff. Because the Langevin velocity distribution is non-zero at v=0 and decreases slowly, the tail integral could be significant, making κ larger than the assumed ≈1. This directly affects the central claim, which is that Eq. (60) gives a universal bound. The existing Lennard-Jones tests provide evidence for typical potentials but do not sample the space broadly enough to rule out counterexamples. The proposed numerical scan would settle whether the bound holds across a wide range of physically reasonable potentials. I agree with the reader's CONDITIONAL verdict: the paper's heuristic derivation is valuable and likely correct in practice, but the word 'bound' requires a more rigorous justification or qualification. My read does not change the reader's verdict, hence UNCHANGED.","tokens_in":27651,"tokens_out":7606,"duration_ms":74093,"concrete_test":"Perform a full quantum scattering calculation for a dense grid of physically reasonable ion-molecule potentials (vary C6, C8, well depth, and core radius over ranges wider than in Fig. 11, e.g., C6/C4 from 0.1 to 10, r_m from 3 to 10 a₀, and include state-dependent short-range phases), and compute the thermally averaged collision shift via Eq. (55) with the exact RSF R(v) from Eq. (4) rather than a step function. For each potential, compare |δf_c/f_c| to Eq. (60) with κ=1, v̄_c=0.034. If any computed shift exceeds this value, the claim that Eq. (60) is a bound is falsified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central formula (60) is presented as a 'bound', but its derivation in Secs. II–V replaces the actual angularly-averaged Ramsey suppression factor R(v) by a step function whose cutoff v̄_c is defined by R(v̄_c)=0.5. This is an ad hoc choice, not a conservative one: R(v) decays gradually and the step function neglects the tail v>v̄_c, which contributes positively to ⟨R⟩. For the Langevin velocity distribution (Eq. 29), which is roughly constant near v=0 and decays as exp(-v²), the tail contribution is not obviously negligible; if it is comparable to v̄_c, then κ≈1 and Eq. (60) understate the true shift. The paper's claim that 'averaging over a recoil velocity distribution will always result in an expression given in Eqs. 11' is asserted without proof; it is not an inequality. The Lennard-Jones tests (Sec. IV-D, Fig. 11) cover only a narrow family of potentials (fixed C6/C4 ratios, r_m in 4–7 a₀) and do not establish universality. A potential with a strong forward-scattering peak or a different short-range behaviour could yield a recoil-velocity PDF with more weight near v̄_c, increasing the average. Thus the 'simple bound' may not be a rigorous bound for all single-ion clocks.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes background-gas collision shifts in single-ion optical clocks. It develops a classical Langevin-scattering model and a quantum scattering treatment with an r^-4 plus hard-sphere potential and with Lennard-Jones potentials, and argues that the worst-case clock shift is approximately the classical Langevin rate multiplied by a velocity-averaged 'Ramsey suppression factor' that accounts for decoupling of the clock laser after the ion recoils. The central formula, Eq. (60), gives δf_c/f_c ≈ κ Γ_L \\bar{v}_c/(2π f_c) with κ≈1 and \\bar{v}_c≈0.034 for a representative geometry. The authors claim this obviates Monte-Carlo simulations and molecular potential energy curves, and they propose a method to measure the relevant collision rate. The paper contains both analytic classical results and a quantum treatment with WKB cross-checks, and it applies the result to a Lu+ clock and to the two-ion crystal system of Hankin et al.","tokens_in":27994,"tokens_out":4128,"duration_ms":43310,"significance":"If Eq. (60) were established as a rigorous worst-case bound, this would be a substantial practical simplification: a collision-shift estimate for any single-ion clock would reduce to a known Langevin rate and a simple geometric factor. The paper's strengths are its cross-checking of classical and quantum treatments, the explicit WKB solution of the r^-4 problem (Fig. 8), and the numerical tests against Lennard-Jones potentials (Fig. 11). The proposed shelving-based scheme for measuring the collision rate is also constructive and useful. However, the central claim is not actually proven as a bound: Section II replaces the angularly averaged Ramsey suppression factor by a step function at its 50% point, and Section V adopts κ≈1 on the basis of a specific class of potentials. The result is a well-motivated and plausible order-of-magnitude estimate, but the word 'bound' is used more strongly than the derivation supports.","major_comments":[{"comment":"The central claim that Eq. (60) is a 'bound' is not established. The derivation replaces the angularly averaged Ramsey suppression factor R(v) by a step function whose cutoff is defined by R(\\bar{v}_c)=0.5 (Section V text). This is not a conservative replacement: R decays gradually, and the tail v > \\bar{v}_c contributes to the velocity average ⟨R⟩. The assertion between Eqs. (11a) and (11b) that 'Averaging over a recoil velocity distribution will always result in an expression given in Eqs. 11' is not an inequality and is not proven. For the Langevin distribution of Eq. (29), which is finite at v=0 and decays as exp(-v^2), the tail contribution is not obviously negligible; the hard-sphere (κ=4) and Lennard-Jones tests do not bound all possible potentials. To retain the word 'bound', the authors need either a conservative tail estimate or a rigorous monotonicity argument; otherwise the c","section":"Section V, Eq. (60); Section II, Eqs. (10)-(11b)"},{"comment":"The universality implied by Eq. (60) is supported only by a narrow family of Lennard-Jones potentials: one clock state fixed at r_m=5 a0, the second varying over r_m ∈ [4,7] a0 on a contour of fixed C_6, with C_6 either ≈0.83 C_4 or ≈8.33 C_4. The statement that 'we would expect the underlying recoil velocity distribution to have the same qualitative features as shown in Sect. III, specifically that the PDF in the neighbourhood of zero is κ≈1' (end of Section IV-D) is a plausibility argument, not a proof. A potential with a strong forward-scattering peak or a different short-range behaviour could concentrate more weight near \\bar{v}_c and increase the average. The paper should either broaden the test (e.g., vary the C_4 exponent, add barriers or orientation-dependent terms, or attempt an explicit variational argument) or clearly state that Eq. (60) is an empirical estimate rather than a","section":"Section IV-D, Fig. 11 and surrounding discussion"},{"comment":"The claim that 'the exact values of p_0 and p_1 are inconsequential' (Section III-A) is weakened by the fact that the constant part of Eq. (29), 8 p_1/√π exp(-\\bar{v}^2), is precisely the term that makes the PDF nonzero at \\bar{v}=0 and leads to κ≈0.855 in Section III-B. The p_0, p_1 are fitted to an approximate interpolation of the differential cross-section, so the derivation is not fully parameter-free. The authors should show that κ remains near unity over the range of plausible (p_0,p_1) values consistent with their fit, or compute the velocity distribution directly from the quantum/WKB scattering amplitudes, to make the κ≈1 result robust.","section":"Section III-A, III-B, and Eq. (29)"}],"minor_comments":[{"comment":"The citation 'Case [?]' is unresolved and should be replaced with the appropriate reference.","section":"Section IV-A, after Eq. (46)"},{"comment":"Typos: 'power low' should be 'power law' (Section III-C) and 'approachs' should be 'approaches' (Section IV-E).","section":"Section III-C; Section IV-E"},{"comment":"The value \\bar{v}_c≈0.034 is given for one specific geometry (laser at 45° to principal axes, ω_x=ω_y=2π×500 kHz, λ=848 nm). The text says the factor is 'readily calculated', but the angular averaging of Eq. (4) is not displayed. A compact expression or a plot of \\bar{v}_c as a function of trap parameters would make the recipe in Eq. (60) directly usable.","section":"Section II, Eq. (4) and Section V"},{"comment":"The derivation of κ=4 for the hard-sphere velocity distribution of Eq. (16) is not shown. One sentence explaining how κ enters the expansion would help the reader reproduce the numbers.","section":"Section II, Eq. (11a)"},{"comment":"The sentence 'The value of \\bar{v}_c is slightly less than the point at which R≈0.5, which occurs at \\bar{v}_c≈0.039' uses the same symbol \\bar{v}_c for both the 'effective cutoff' and the 50%-threshold point, which is confusing. Please distinguish these two definitions.","section":"Section II, text after Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"The paper contains genuinely useful physics: the classical/quantum cross-check and the WKB solution are solid, and the proposed collision-rate measurement is interesting. My main reservation is terminological: Eq. (60) is presented as a bound, but the step-function replacement and the κ≈1 assumption are not proven conservative. If the authors reword the claims to distinguish the rigorous parts from the estimate, and add a sensitivity check of κ against the fit parameters, the paper would be a good candidate for publication. As written, the central assertion is stronger than the derivation supports, so I recommend major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This one deserves a serious referee. The genuinely new thing is Eq. (60): the collision-shift bound written as κ Γ_L v̄_c / 2π f_c, with κ≈1 and v̄_c≈0.034 for the Lu+/H2 case. That reduces what previously required Monte-Carlo simulations or molecular potential curves to a simple analytic expression. The classical and quantum treatments are cross-checked against each other and against Lennard-Jones potentials, and the corrected WKB phase for the r^{-4} plus hard-sphere problem looks like a real, checkable contribution. The discussion of glancing collisions is also the clearest I have seen, and it explains the factor-of-two discrepancies between earlier quantum and Monte-Carlo estimates rather than just noting them.\n\nThe soft spot is the word \"bound.\" The stress-test concern is real, but it is a qualification, not a collapse. Eq. (60) is derived by replacing the angle-averaged Ramsey suppression factor R with a step function cut at R=0.5, then taking κ≈1. That is well motivated for the Langevin recoil distribution, which is roughly constant near zero, but it is not a proof that the tail of R cannot push the true average above the step-function value. The paper's own assertion that averaging over a recoil-velocity distribution will always give the forms in Eqs. (11) is too quick; that is an estimate, not an inequality. The Lennard-Jones tests cover a reasonable but narrow family of potentials, so universality is plausible, not established.\n\nTwo smaller things: the Rabi-spectroscopy generalization is asserted without derivation, and there is an unresolved citation placeholder (\"Case [?]\") in the WKB discussion. Both should be fixed in revision.\n\nNone of this undermines the central physics. If the authors relabel Eq. (60) as a practical estimate rather than a proven upper bound, or actually prove the step-function replacement is conservative, the paper stands. The comparison to [6], [7], and [8] is honest and useful, and the paper will save people real time when building 10^{-19} error budgets. I would accept it for peer review and ask for someone who can check the connection formulas and the averaging carefully. I would also cite it.","headline":"A genuinely useful analytic reduction of ion-clock collision shifts, with a real but addressable gap between the word \"bound\" and the inequality actually proven.","tokens_in":28425,"tokens_out":2783,"would_cite":true,"duration_ms":33326,"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":"The paper claims that the worst-case collision shift in any single-ion optical clock is set by the classical Langevin collision rate ΓL reduced by a simple recoil-decoupling factor, giving δfc/fc = κ ΓL v̄c/(2πfc).","keywords":["collision shift","ion clock","Langevin scattering","Ramsey spectroscopy","recoil decoupling","background gas","systematic uncertainty"],"falsifier":"Measure the fractional clock shift of a single ion as a function of background H2 pressure over a range spanning, say, 1–50 nPa while independently calibrating the pressure. If the slope of shift versus ΓL exceeds κ v̄c/(2πfc) with κ≈1 by more than a factor of two, or if the shift does not scale linearly with pressure, the bound as stated is falsified. A complementary calculation would be to include the full anisotropic H2 quadrupole and tensor-polarizability terms in the recoil-velocity distribution; a value of κ that departs substantially from 1 would also break the simple formula.","tokens_in":27570,"feed_emoji":"⏱️","tokens_out":8155,"duration_ms":75957,"temperature":0.7,"pith_summary":"The paper tries to establish that for any single-ion optical clock, the largest possible frequency shift caused by a collision with a background molecule is captured by a one-line formula: take the classical Langevin collision rate ΓL and multiply it by a small, easily computed factor v̄c that describes how often the collision kick decouples the ion from the clock laser. The claimed bound is δfc/fc ≈ κ ΓL v̄c/(2πfc), with κ≈1 and v̄c≈0.034 for a typical geometry. A sympathetic reader would care because this replaces large-scale Monte-Carlo simulations and calculations of molecular potential energy curves with a simple, parameter-light estimate for a systematic that is otherwise hard to pin down. It also suggests a direct experimental way to measure the relevant collision rate.","feed_headline":"One formula bounds ion-clock collision shifts","feed_subtitle":"No Monte-Carlo runs or molecular potential curves needed: Langevin rate times a small recoil-suppression factor does it.","key_machinery":"The Ramsey suppression factor (RSF) is the central object: for a given recoil velocity it is the angular average of the product of Bessel functions that determines how much the velocity kick reduces the coupling of the clock laser in the second Ramsey pulse. The paper shows that this factor can be replaced, for bounding purposes, by a step function that cuts off at the recoil speed where the RSF has fallen to 50%, giving v̄c≈0.034 for the standard trap geometry. The other half of the machinery is the recoil-velocity distribution for Langevin scattering, which is derived classically and shown to be approximately constant near zero, so that the only remaining parameter is κ≈1. Together they co","core_discovery":"On the paper's own terms, the central discovery is that a collision shift bound is determined by the classical Langevin collision rate reduced by a factor that quantifies how the ion's recoil motion decouples the clock laser during a Ramsey interrogation. The authors derive a Ramsey suppression factor from the Bessel-function average over recoil directions, show that the recoil-velocity distribution for Langevin scattering is nearly flat near zero (so a single number κ≈1 captures it), and approximate the suppression by a step function whose cutoff is v̄c≈0.034. Combining these pieces gives Eq. (60). They further demonstrate, using hard-sphere, Lennard-Jones, and fully quantum treatments of t","pith_inferences":["Beyond the paper: the same formula could be used to choose a more favorable trap geometry—orienting the clock laser and choosing trap frequencies so that the Ramsey suppression factor falls at smaller recoil speeds would tighten the bound without new physics.","Beyond the paper: because the bound is linear in ΓL, a two-pressure differential measurement of the clock frequency would serve as a clean test of both the rate dependence and the assumed κ≈1.","Beyond the paper: applying this to ion crystals requires the 'crystal recoils as a whole' approximation, which the paper offers only as a rough estimate; an experiment with a two-ion crystal could verify or correct that extrapolation."],"forward_implications":["A single-ion clock's collision-shift bound can be computed from the pressure (via ΓL) and a geometry-dependent v̄c, with no Monte-Carlo simulation and no molecular potential curves.","Earlier Monte-Carlo and quantum-potential estimates that differed by roughly √2 are reconciled once the recoil-decoupling suppression is included.","The relevant collision rate can be measured by shelving the ion and observing whether it fails to reshelve; the resulting rate slightly overestimates ΓL but is still usable for a bound.","For 176Lu+ with a 300 K H2 background, the bound is ≈6.2×10−21 per nPa, and it is insensitive to interrogation time over practical ranges."],"fun_headline_variants":["Simple bound for ion-clock collision shifts","Ion-clock shifts from Langevin rate alone","Collision shift bound without Monte-Carlo","One factor tames ion-clock collisions","Recoil factor simplifies ion-clock shift bound"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the effect of the recoil velocity on the Ramsey signal is well represented by a step-function cutoff at the speed where the suppression factor has dropped to 50%, together with a recoil-velocity distribution that is roughly flat near zero (κ≈1); if the true distribution vanished linearly at zero, the suppression would scale as v̄c^2 and be about 30 times smaller, so the numerical bound would change.","fun_headline_variants_meta":{"raw":{"variants":["Simple bound for ion-clock collision shifts","Ion-clock shifts from Langevin rate alone","Collision shift bound without Monte-Carlo","One factor tames ion-clock collisions","Recoil factor simplifies ion-clock shift bound"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00034,"raw_usage":{"total_tokens":1669,"prompt_tokens":660,"completion_tokens":1009,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":404,"completion_tokens_details":{"reasoning_tokens":958}},"tokens_in":404,"tokens_out":1009,"duration_ms":7854,"temperature":1.0,"reasoning_tokens":958,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T18:22:24.674303+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the fractional clock shift of a single ion as a function of background H2 pressure over a range spanning, say, 1–50 nPa while independently calibrating the pressure. If the slope of shift versus ΓL exceeds κ v̄c/(2πfc) with κ≈1 by more than a factor of two, or if the shift does not scale linearly with pressure, the bound as stated is falsified. A complementary calculation would be to include the full anisotropic H2 quadrupole and tensor-polarizability terms in the recoil-velocity distribution; a value of κ that departs substantially from 1 would also break the simple formula.","supporting_citations":[],"review_version":1}