{"id":"21e50ecd-4998-4536-b0c1-a5fb96f9bcba","arxiv_id":"2502.13973","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Beyond-rotating-wave corrections accumulate across long composite Ramsey pulse trains, hiding the small tensor shift that the sequence was designed to measure.","lead":"The authors show that the rotating wave approximation fails for composite Ramsey sequences of radiofrequency pulses when the number of pulses exceeds a few tens. This undermines confidence in a proposed method for testing local Lorentz invariance with trapped ions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Condition (10), asserted as the threshold for tensor-shift invisibility, is the load-bearing undefended step: the paper's practical LLI conclusion depends on a scaling that is illustrated but not derived, so the verdict stays conditional.","rationale":"The central negative claim about LLI detection is not merely that RWA fails for long sequences; that part is supported by the numerical contrast between the full Hamiltonian and Eq. (9), and the per-pulse error argument is physically reasonable. The load-bearing step is the quantitative threshold (10): it converts a real dynamical effect (RWA breakdown) into a categorical statement that a small tensor shift cannot be detected. That threshold is not derived, and its τ/τ_R factor is especially suspicious because the tensor shift also accumulates in free evolution, where no Bloch–Siegert correction exists. The absence of a J=5/2 plot and of a published convergence study strengthens the need for an independent check. However, the paper is explicitly hedged in its abstract and conclusion, and the numerical evidence for RWA breakdown is credible; the concern therefore supports the existing CONDITIONAL verdict rather than rejection. The proposed scan is feasible with the authors' own solver and would discriminate between the published criterion and plausible alternatives.","tokens_in":7913,"tokens_out":6734,"duration_ms":68532,"concrete_test":"Recompute the exact dynamics of Eq. (7) for J=7/2, m′=-1/2, n=50, Ω_L/Ω_rf=100, δ=0 while varying τ_R/τ over 1, 3, 10, 30; for each value, locate the smallest κ/Ω_rf at which the exact P_{m′} deviates by 1% from the κ=0 curve. If the threshold follows the RHS of (10), including its τ/τ_R dependence, condition (10) is supported; if it is independent of τ_R or scales differently in Ω_rf/Ω_L, the practical conclusion fails. The same scan should be repeated with F=100 to confirm the Fourier truncation is not setting the threshold.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's strongest practical claim is the sentence below Eq. (10): when ℏκ/Ω_rf ≪ (τ/τ_R)(Ω_rf/Ω_L), the tensor shift κJ_z^2 has practically no observable effect, so the Ramsey sequence {(π/2,−π/2)^n, π} cannot reliably detect an LLI tensor shift. Condition (10) is introduced after Fig. 2 and supported only by numerical examples and a qualitative error-accumulation argument; no derivation from Eq. (7) is supplied. In particular, the factor τ/τ_R is not justified. The tensor shift acts during the free intervals (duration τ_R), whereas the leading Bloch–Siegert correction (11) acts only during pulses, so the relevant comparison should involve accumulated phases such as κτ_R and Ω_rf^2 τ/Ω_L, not simply κ/Ω_rf. If the true threshold is κτ_R ≲ Ω_rf/Ω_L rather than κ/Ω_rf ≲ (τ/τ_R)(Ω_rf/Ω_L), the detectability conclusion changes. The numerical support covers J=7/2 with only m′=-1/2 and m′=-7/2 and one parameter set; the J=5/2 check against Ref. [9] is mentioned but not displayed. A Magnus/Floquet expansion of Eq. (7) retaining the leading κ-dependent term would settle the scaling.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a beyond-rotating-wave-approximation (RWA) theory of composite Ramsey sequences of radiofrequency pulses acting on a Zeeman manifold with angular momentum J. Using the full Hamiltonian of Eq. (3) [or Eq. (7)] and a Fourier expansion over harmonics, the authors numerically compare the dynamics generated by the full Hamiltonian with the RWA Hamiltonian of Eq. (9) for the sequence {(π/2, −π/2)^n, π} proposed in Ref. [9]. They find that for n exceeding several tens the RWA becomes inadequate, that the small tensor shift κJ_z^2 has practically no observable effect when condition (10) holds, and that the modified resonant Hamiltonian of Eq. (11), including the Bloch–Siegert shift, still does not reproduce the exact dynamics. The paper concludes that the effectiveness of such rf composite Ramsey sequences for detecting local Lorentz invariance (LLI) violation is not fully determined and requires further study.","tokens_in":8195,"tokens_out":3861,"duration_ms":40869,"significance":"If the numerical result is correct, this is an important cautionary result for precision tests of local Lorentz invariance with rf composite Ramsey spectroscopy. The paper's strength is that the full and RWA Hamiltonians are independently defined and no parameters are fitted to the target conclusion; the comparison is a direct numerical test. The explicit identification of the Bloch–Siegert correction in Eq. (11) and the demonstration that this first-order correction is insufficient for long sequences are useful and falsifiable. The main quantitative condition, inequality (10), is, however, asserted from numerical observation rather than derived, and the paper's practical conclusion about undetectability of LLI tensor shifts rests on that condition. If the scaling in (10) is not correct, the negative conclusion about the sequence of Ref. [9] would not follow even though the RWA breakdown itself would remain real.","major_comments":[{"comment":"The criterion ℏκ/Ω_rf ≪ (τ/τ_R)(Ω_rf/Ω_L) is introduced after Fig. 2 as a numerically observed condition, but it is load-bearing for the paper's practical conclusion that the sequence {(π/2, −π/2)^n, π} cannot reliably detect an LLI tensor shift. No derivation from Eq. (7) is supplied, and the factor τ/τ_R is not justified: the tensor shift κJ_z^2 acts during free-evolution intervals of duration τ_R, whereas the leading non-resonant correction in Eq. (11) acts only during pulses of duration τ, so accumulated phases proportional to κτ_R and Ω_rf^2 τ/Ω_L are the more natural quantities to compare. A Magnus or Floquet expansion of Eq. (7) retaining the leading κ-dependent term is needed to establish the scaling. Without such a derivation, the detectability conclusion is conditional on a plausible but undemonstrated estimate, and this is exactly the concern raised by the stress-test note; I agree that it lands.","section":"§3, Eq. (10)"},{"comment":"The numerical method truncates the Fourier expansion at F = 50, and the text states that F = 50 'was enough with a good margin' for the curves in Figs. 2–4. No convergence test or error estimate is shown, despite the fact that the central claim is a quantitative discrepancy for n ≫ 1. The paper should display the convergence of P_{m'} as a function of F for at least one long-sequence case and for the parameter set of Fig. 3, where the dynamics are governed by a different regime of condition (10).","section":"§3, Fourier truncation (p. 2)"},{"comment":"The text asserts that 'similar radical discrepancies' are observed for other values of m′ and J, including J = 1/2, 3/2, 5/2, and that calculations for J = 5/2 using experimental data from Ref. [9] confirm good agreement with the resonant Hamiltonian. None of these cases is displayed or described quantitatively. Because the practical conclusion is meant to apply beyond J = 7/2, the paper should include at least a summary figure or table for the J = 5/2 comparison with the experimental parameters, and should specify which m′ values were scanned. Without this evidence, the generality claim is asserted rather than demonstrated.","section":"§3, Figs. 2–3 and subsequent paragraph"}],"minor_comments":[{"comment":"The title contains a typo ('pul ses' instead of 'pulses'); it should be corrected.","section":"Title"},{"comment":"The abstract states that the effectiveness of the sequences 'has not yet been fully determined', while the text below Eq. (10) makes the stronger claim that the sequence 'will not allow to reliably detect this shift'. These statements should be reconciled, because the stronger claim depends on the undemonstrated condition (10).","section":"Abstract and Conclusion"},{"comment":"The sentence 'the effectiveness of the dynamical decoupling technique ... is also not guaranteed' is vague; the paper should specify which property is not guaranteed (e.g., robustness to magnetic-field fluctuations) and what concrete test would establish it.","section":"Conclusion, final paragraph"},{"comment":"The J = 5/2 comparison against data from the Supplemental material of Ref. [9] is mentioned but not located precisely; the authors should identify the specific dataset and parameters used.","section":"Reference [9] comparison"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of physics.atom-ph and the central observation—that the RWA fails for long rf composite Ramsey sequences—is likely of interest to the community. The missing derivation of condition (10) is not merely a presentation issue; it is the pivot on which the main practical conclusion turns. If the authors cannot supply a derivation, the alternative would be to substantially expand the numerical evidence for (10) across parameter space and to soften the wording of the conclusion. I do not see grounds for rejection, but the revision needs more than cosmetic changes."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper. First, it identifies a real, under-appreciated failure mode: in composite Ramsey sequences of rf pulses with hundreds or thousands of pulses, the rotating wave approximation breaks down because non-resonant contributions accumulate. That matters because the published LLI-test proposals (Shaniv et al. PRL 2018, Dreissen et al. Nat. Comm. 2022) were justified within the RWA. Second, the paper's practical conclusion about detectability of a tensor shift rests on condition (10), which is asserted from numerical observation rather than derived. The qualitative argument is plausible, but the inequality deserves a rigorous derivation before the conclusion is treated as settled.\n\nWhat is genuinely new: the paper goes beyond the standard Bloch-Siegert shift, shows that adding the first non-resonant correction (Eq. 11) still does not reproduce the exact dynamics for n >> 1, and provides a simple per-pulse error estimate (Omega_rf/Omega_L per pulse) that explains why the discrepancy appears after tens of pulses. The numerical examples with J=7/2 are clear, and the claim that the RWA overestimates the effect of a small kappa term is well illustrated by the difference between panels (a)-(b) and (c)-(d) in Fig. 2.\n\nSoft spots, in proportion. The main one is condition (10). It is the load-bearing criterion for the statement that an LLI tensor shift cannot be seen, and it is introduced as an empirical scaling. I would want a Magnus or Floquet expansion from Eq. (7) that produces this threshold. That said, the reader's stress-test worry that the correct comparison should involve accumulated phases kappa*tau_R and Omega_rf^2*tau/Omega_L actually reduces to (10) up to a factor pi, so the threshold itself does not look wrong; what is missing is the derivation, not the physics. Minor: the Fourier truncation at F=50 is asserted to be sufficient with no convergence study, and the J=5/2 comparison against the Sr+ data is described but not shown. Neither undermines the central message, but both are easy to address.\n\nWho this is for: anyone working on rf dynamical decoupling, atomic clocks with rf dressing, or Lorentz-violation searches with trapped ions. It deserves a serious referee. The flaws are fixable, and the claim, if it survives, changes how the community reads previous experimental bounds. Send it to review.","headline":"A credible numerical demonstration that the rotating wave approximation fails for long rf composite Ramsey sequences, with the key threshold condition stated but not derived.","tokens_in":8751,"tokens_out":3761,"would_cite":true,"duration_ms":33288,"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 argues that the rotating wave approximation, which is standard for optical transitions, fails for composite Ramsey sequences of radiofrequency pulses once the sequence contains several tens of pulses.","keywords":["composite Ramsey sequences","rotating wave approximation","radiofrequency pulses","local Lorentz invariance","Zeeman structure","Fourier analysis","Bloch-Siegert shift","tensor shift"],"falsifier":"For parameters satisfying condition (10), compute the exact final population $P_{m'}$ versus $n$ for the sequence $\\{(\\pi/2,-\\pi/2)^n,\\pi\\}$ with a known tensor shift just above the bound; the paper predicts the curve is indistinguishable from $\\kappa=0$ at large $n$, whereas the resonant Hamiltonian predicts a growing separation. If the exact curves separate visibly, the criterion is wrong.","tokens_in":7758,"feed_emoji":"⚛️","tokens_out":6311,"duration_ms":57269,"temperature":0.7,"pith_summary":"The paper argues that the rotating wave approximation, which is standard for optical transitions, fails for composite Ramsey sequences of radiofrequency pulses once the sequence contains several tens of pulses. The non-resonant parameter for rf fields is only about $10^{-2}$ to $10^{-3}$, so each pulse leaves a small correction that accumulates over many pulses. For the specific sequence $\\{(\\pi/2,-\\pi/2)^n,\\pi\\}$ proposed to search for Lorentz-invariance-violating tensor shifts $\\kappa \\hat{J}_z^2$, the exact dynamics differ radically from the resonant approximation for large $n$. Moreover, when condition (10), $\\hbar\\kappa/\\Omega_{\\rm rf} \\ll (\\tau/\\tau_R)(\\Omega_{\\rm rf}/\\Omega_L)$, holds, the small tensor shift has practically no observable effect, so the sequence cannot reliably detect an LLI-violating shift even if it exists. The paper therefore leaves the viability of such rf Ramsey tests open and calls for analysis beyond the rotating wave approximation.","feed_headline":"Rotating-wave approximation breaks down in long rf pulse runs","feed_subtitle":"Non-resonant errors accumulate after tens of pulses, masking the tiny tensor shift that Lorentz tests look for.","key_machinery":"The central object is the composite Ramsey sequence $\\{(\\varphi_1,\\varphi_2,\\dots,\\varphi_N)^n,\\varphi_{\\rm fin}\\}$, made of an initial $\\pi/2$ pulse, $n$ identical boxes of $N$ $\\pi$ pulses, and a final $\\pi/2$ pulse, with free evolution times $\\tau_R$ between pulses. The carrying mechanism is Fourier analysis of the time-dependent Schr\\\"odinger equation in the rotating frame: after the unitary transformation $a_m(t)=\\tilde a_m(t)e^{-im\\nu t}$, the Hamiltonian splits into resonant terms and rapidly oscillating counter-rotating terms proportional to $e^{\\pm 2i\\nu t}$. The resonant approximation keeps only the former; the full treatment expands the wave function in harmonics and keeps all of them. This reveals a per-pulse non-resonant error of order $\\Omega_{\\rm rf}/\\Omega_L$, which is the mechanism behind the breakdown. The small tensor shift $\\kappa \\hat{J}_z^2$ enters through condition (10), which sets the scale below which the shift is hidden by the non-resonant dynamics.","core_discovery":"On its own terms, the paper's central discovery is that the dynamics of a multilevel atom driven by a long composite Ramsey sequence of rf pulses are governed by non-resonant contributions that the rotating wave approximation discards, and these contributions change the outcome qualitatively once the number of pulses exceeds several tens. Concretely, for the sequence $\\{(\\pi/2,-\\pi/2)^n,\\pi\\}$ with $J=7/2$ and parameters $\\Omega_L/\\Omega_{\\rm rf}=100$, the exact solution of the full Hamiltonian (3) (computed by Fourier expansion over harmonics up to $F=50$) differs drastically from the resonant Hamiltonian (9) for $n\\gg 1$, whereas the two agree for small $n$. The accumulated error is of order $\\Omega_{\\rm rf}/\\Omega_L$ per pulse, so with $\\Omega_{\\rm rf}/\\Omega_L\\sim 10^{-2}$–$10^{-3}$, breakdown appears after tens to hundreds of pulses. Under condition (10), the tensor shift $\\kappa \\hat{J}_z^2$ has essentially no visible effect in the exact dynamics, in contrast to the resonant prediction where it grows with $n$. The paper concludes that the rf Ramsey method of Ref. [9] cannot reliably detect an LLI-violating tensor shift, and that the general effectiveness of dynamical decoupling for rf pulses needs separate justification beyond the rotating wave approximation.","pith_inferences":["A testable extension is to scan the ratio $\\Omega_{\\rm rf}/\\Omega_L$ and the number of boxes $n$ to locate the crossover where exact and RWA dynamics diverge; the paper's qualitative argument predicts the crossover near $n\\sim \\Omega_L/\\Omega_{\\rm rf}$.","Condition (10), if correct, implies a sensitivity floor for this class of rf composite-pulse LLI searches: any tensor shift below $\\hbar\\Omega_{\\rm rf}^2 \\tau/(\\Omega_L \\tau_R)$ is invisible, which is worse than the sensitivity predicted by the resonant Hamiltonian.","The breakdown mechanism is generic for any multilevel system driven by many pulses with a small but finite counter-rotating field, so the same Fourier-harmonic method could be applied to other rf metrology schemes, such as magic rf dressing in optical clocks.","Because the paper's conclusion about condition (10) rests on numerical examples and a qualitative error-accumulation argument, a rigorous analytic estimate of the accumulated non-resonant error as a function of $n$ would either confirm the condition or reveal a different scaling; such a derivation is the natural next step."],"forward_implications":["If the paper is right, the rotating wave approximation cannot be used to design or predict composite rf Ramsey sequences with more than a few tens of pulses; exact or beyond-RWA propagation is required.","The specific sequence $\\{(\\pi/2,-\\pi/2)^n,\\varphi_{\\rm fin}\\}$ cannot serve as a reliable detector of an LLI-induced tensor shift when condition (10) holds, because the shift's effect is masked by non-resonant dynamics.","Agreement between resonant theory and experiment for the $^{88}$Sr$^+$ data with $n$ up to 55 does not validate the method for LLI searches, because that agreement arises only when the controlled tensor shift is large enough to violate condition (10).","Dynamical decoupling's ability to suppress magnetic-field noise in the rf regime is not guaranteed by the two-level rotating-wave justification and must be re-examined.","For rf transitions, unlike optical ones, the non-resonant parameter cannot be made arbitrarily small: increasing $\\Omega_L$ enlarges the second-order Zeeman shift of the same tensor form, and decreasing $\\Omega_{\\rm rf}$ increases relative field fluctuations."],"supporting_citations":[{"why":"Proposes the composite Ramsey sequence and the use of the tensor shift $\\kappa \\hat{J}_z^2$ to search for LLI violation; this is the method the paper re-examines beyond the rotating wave approximation.","marker":"[9]"},{"why":"Reports the experimental implementation of composite-pulse Ramsey spectroscopy with thousands of rf pulses, the regime where the paper shows RWA becomes unreliable.","marker":"[10]"},{"why":"Presents a robust and scalable rf spectroscopy scheme in first-order magnetically sensitive states, providing a further resonant-approximation basis that the paper calls into question.","marker":"[17]"},{"why":"Introduces the Bloch-Siegert shift, the first non-resonant correction that the paper includes in the modified resonant Hamiltonian (11) and then shows to be insufficient.","marker":"[18]"},{"why":"Establishes dynamical decoupling within the rotating wave approximation for two-level systems; the paper notes this justification does not automatically transfer to rf sequences.","marker":"[11]"}],"fun_headline_variants":["Non-resonant effects break RWA in long rf pulse runs","Tens of rf pulses invalidate rotating-wave approximation","RWA breakdown masks tiny Lorentz shift in rf Ramsey tests","Long rf pulse trains: RWA fails, Lorentz signal hidden"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's key assumption is that condition (10) correctly identifies when the tensor shift becomes unobservable; this criterion is supported by numerical examples and a rough per-pulse error estimate rather than a rigorous proof.","fun_headline_variants_meta":{"raw":{"variants":["Non-resonant effects break RWA in long rf pulse runs","Tens of rf pulses invalidate rotating-wave approximation","RWA breakdown masks tiny Lorentz shift in rf Ramsey tests","Long rf pulse trains: RWA fails, Lorentz signal hidden"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000666,"raw_usage":{"total_tokens":3041,"prompt_tokens":952,"completion_tokens":2089,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":568,"completion_tokens_details":{"reasoning_tokens":2020}},"tokens_in":568,"tokens_out":2089,"duration_ms":15425,"temperature":1.0,"reasoning_tokens":2020,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T11:59:28.037735+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For parameters satisfying condition (10), compute the exact final population $P_{m'}$ versus $n$ for the sequence $\\{(\\pi/2,-\\pi/2)^n,\\pi\\}$ with a known tensor shift just above the bound; the paper predicts the curve is indistinguishable from $\\kappa=0$ at large $n$, whereas the resonant Hamiltonian predicts a growing separation. If the exact curves separate visibly, the criterion is wrong.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proposes the composite Ramsey sequence and the use of the tensor shift $\\kappa \\hat{J}_z^2$ to search for LLI violation; this is the method the paper re-examines beyond the rotating wave approximation."},{"cited_title":"Indeed, our calculations for J = 5/ 2, using experimental data from the Supplemental materials to Ref","cited_arxiv_id":null,"evidence_quote":"Reports the experimental implementation of composite-pulse Ramsey spectroscopy with thousands of rf pulses, the regime where the paper shows RWA becomes unreliable."},{"cited_title":"Sanner, N","cited_arxiv_id":null,"evidence_quote":"Presents a robust and scalable rf spectroscopy scheme in first-order magnetically sensitive states, providing a further resonant-approximation basis that the paper calls into question."},{"cited_title":"Shaniv, R","cited_arxiv_id":null,"evidence_quote":"Introduces the Bloch-Siegert shift, the first non-resonant correction that the paper includes in the modified resonant Hamiltonian (11) and then shows to be insufficient."},{"cited_title":"parameter of non-resonant con- tributions","cited_arxiv_id":null,"evidence_quote":"Establishes dynamical decoupling within the rotating wave approximation for two-level systems; the paper notes this justification does not automatically transfer to rf sequences."}],"review_version":1}