{"id":"32c3dc2b-61d8-429d-b122-24f998c53644","arxiv_id":"2608.08912","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"New high-resolution spectra of Rb2 improve the Dunham constants of the B1Pi_u state and give the first measurement of its Lambda-doubling.","lead":"Researchers measured 958 high-resolution laser transitions in three isotopologues of the rubidium dimer, Rb2, refining the excited-state rotational constants and measuring its Lambda-doubling for the first time. The new constants improve the accuracy needed for ultracold chemistry and laser-cooling proposals involving Rb2.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Global fit uses model-generated 'data' reconstructed from Amiot–Vergès, so the claimed improvement over Ref. [18] is partly circular; p0 remains independently supported.","rationale":"The paper's headline has two separable parts. The first Lambda-doubling constant p0 comes from the standalone fit of 709 new transitions (Section 4, Table 4) and is not affected by the reconstructed Amiot–Vergès levels. The global-fit claim in the abstract is, however, central to the paper's stated contribution of 'improved Dunham coefficients' and 'significantly refines rotational and rovibrational coupling constants.' The reader's weakest assumption identifies exactly this circularity: the Amiot–Vergès input is generated from the model being compared. I agree. The correct response is conditional acceptance: standalone p0 and refined low-J constants are plausible, but the global improvement claim requires either the genuine measured Amiot–Vergès data or removal of the circular comparison. Thus the reader's CONDITIONAL verdict is unchanged.","tokens_in":16850,"tokens_out":9446,"duration_ms":97008,"concrete_test":"Obtain the original measured line list of Amiot–Vergès (or digitize the experimental level energies in Fig. 5 rather than regenerating them from Dunham coefficients), and rerun the identical global least-squares fit with those genuine data. If Y'11, Y'12, Y'0,2 or the quoted σ shift by more than their reported uncertainties, then the reconstructed levels are not a faithful proxy and Table 5's improvement claim fails. If the original data are unobtainable, repeat the fit on the 709 new transitions alone and show that the claimed high-v parameter changes persist; otherwise the global claim should be dropped or reworded.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 5 states that Ref. [18]'s transition frequencies are not publicly available, so the 611 'Amiot–Vergès levels' included in the global fit were extracted from Fig. 5 and reconstructed using the Dunham coefficients reported in [18]. These are therefore not measured data: they are predictions of the very model that Table 5 later compares against. Including such noiseless synthetic levels in the least-squares objective is equivalent to adding a penalty that pulls the fitted coefficients toward those of [18], so the global fit is a regularized version of the new-data fit rather than an independent combination of experimental data sets. The original work used 1186 measured levels; only 611 regenerated ones are used, with assignments 'retained as given' but no raw positions to check them. Consequently, the abstract's claim that 'a global least-squares analysis combining the new and published data yields improved Dunham coefficients' is not supported as stated. Table 5's deviations, e.g. Y'11 = +6.2±0.7% and Y'12 = -58.3±4.1%, are not independent evidence. The Section 4 determination of p0 from 709 new transitions is unaffected by this circularity, so the Lambda-doubling claim can stand separately.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports cw-diode-laser fluorescence spectroscopy of Rb₂ in a supersonic beam, resolving the R, Q, and P branches of the X¹Σ₉⁺ → B¹Πᵤ bands for ⁸⁵Rb₂, ⁸⁵Rb⁸⁷Rb, and ⁸⁷Rb₂. A total of 958 rovibronic transitions were recorded, of which 709 (excluding the v″ = 1 band) were used in a least-squares fit to determine rotational, rovibrational, and Λ-doubling parameters of the B¹Πᵤ state. The main new result is a first determination of the Λ-doubling constant p₀, with p₀ = (4.28 ± 0.21) × 10⁻⁷ cm⁻¹ in the band-origin treatment. The paper also presents a 'global' fit that combines the new transitions with excited-state levels attributed to the earlier work of Amiot and Vergès [18], and uses the resulting Table 5 to claim improved Dunham coefficients for the B state.","tokens_in":17164,"tokens_out":4519,"duration_ms":48135,"significance":"If the standalone Section 4 analysis is accepted, the work provides a valuable high-resolution, three-isotopologue data set with dense low-J coverage of the B¹Πᵤ state, plus the first Λ-doubling constant for this state, which is relevant for spectroscopy and for applications such as laser-cooling proposals and ultracold-molecule studies. The reported residual scatter, 2.8 × 10⁻⁴ cm⁻¹, shows that the fit to the new data is internally consistent. The Section 5 global fit, however, is not an independent combination of measured data sets, and therefore the abstract's claim of improved Dunham coefficients from combining new and published data is not supported as stated. The stress-test concern about circularity lands directly on Section 5 and Table 5.","major_comments":[{"comment":"The 611 excited-state levels included in the global fit are not measured data: they were extracted from Fig. 5 of Ref. [18] and reconstructed using the Dunham coefficients reported in that same paper. These are therefore model-generated points that already encode the Amiot–Vergès parameterization. Including such noiseless pseudo-levels in the least-squares objective acts as a regularizing penalty that pulls the fitted coefficients toward the Amiot–Vergès values; Table 5's relative deviations, such as ΔY'₁₁ = +6.2 ± 0.7% and ΔY'₁₂ = −58.3 ± 4.1%, are not independent evidence of improvement. The abstract's statement that 'a global least-squares analysis combining the new and published data yields improved Dunham coefficients' is therefore not supported. The authors should either obtain and use the original measured transitions/levels of Ref. [18], or explicitly reframe Section 5 as a constrained fit that inherits the high-v' behavior from the Amiot–Vergès model and remove the claim of independent improvement.","section":"Section 5"},{"comment":"The quoted statistical uncertainties do not include any systematic error budget, despite a linewidth of about 60 MHz, a spectral step of 10 MHz, and a wavelength calibration based on a wavemeter and iodine cell. For example, Y'₀₁ is reported as (1.955797 ± 0.000018) cm⁻¹, i.e. about ±0.54 MHz; with 60 MHz-wide lines and 10 MHz steps, such a claim requires a careful discussion of peak-center determination and absolute calibration accuracy. The authors should provide an explicit error budget, including the iodine-cell calibration uncertainty, and state whether the reported 2.8 × 10⁻⁴ cm⁻¹ residual scatter is a precision or an accuracy statement.","section":"Section 4, Tables 3 and 4"},{"comment":"The global fit weights the reconstructed Amiot–Vergès levels with σ_A = 4.9 × 10⁻³ cm⁻¹ as though they were independent experimental measurements. Since these points are noiseless model outputs reconstructed from Dunham coefficients, this weighting is not a measurement-noise weighting; it is an arbitrary choice that controls how strongly the fit is pulled toward the old model. Consequently, the reported global standard deviation σ = 6.6 × 10⁻⁴ cm⁻¹ and the residual plot in Fig. 7 are not meaningful goodness-of-fit diagnostics for the combined data set, and the interpretation in terms of 'agreement' between data sets is not justified.","section":"Section 5, Fig. 7"},{"comment":"The conclusion that the global fit 'provides a more complete and accurate description of the excited-state structure' is overstated. The high-v' pure-vibrational and high-order rotational parameters in Table 5 are essentially inherited from the Amiot–Vergès model because the pseudo-levels are generated from that model. The independently supported new results are the Section 4 parameters and p₀; the manuscript should be revised so that the abstract, highlights, and conclusion are limited to what the new data actually establish unless the original data of Ref. [18] become available.","section":"Section 6 and Abstract"}],"minor_comments":[{"comment":"Many entries in Table 5 report far more digits than the uncertainties justify (e.g., Y'₁₅,₀ = −(1.0962854880 ± 0.00173806) × 10⁻²¹ cm⁻¹); the values and uncertainties should be rounded consistently to the last significant digit.","section":"Table 5"},{"comment":"The 'spectral resolution' of 3.3 × 10⁻⁴ cm⁻¹ is the 10 MHz scan step, while the observed linewidth is about 60 MHz; the manuscript should clarify how the central positions of peaks were determined and whether the quoted resolution represents the step size or the actual line-position uncertainty.","section":"Section 3"},{"comment":"The sentence 'The terms l > 1 account for high-order anharmonic corrections' is imprecise: l = 2 and higher terms in X'' are centrifugal-distortion terms, not anharmonic corrections; the wording should be corrected.","section":"Eq. (1)"},{"comment":"The caption of Fig. 6 says that the 112 red squares are 'levels experimentally accessed in the present work for ⁸⁵Rb₂', but Table 2 shows additional levels for ⁸⁵Rb⁸⁷Rb and ⁸⁷Rb₂; the figure should either include all isotopologues or explain why only one isotopologue is shown.","section":"Section 5, Fig. 6"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the circularity in Section 5: the 'published data' are not raw data but levels reconstructed from the Amiot–Vergès Dunham coefficients, so the global fit is effectively a regularized refit to that model. The p₀ result and the Section 4 analysis appear to stand independently, but the abstract and conclusions must be revised to avoid claiming an independent global improvement. The authors should be asked to either obtain the original data from Ref. [18] or substantially reframe Section 5 as a constrained combination and present Section 4 as the primary new determination."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Paul—\n\nThe piece of work worth knowing here is the standalone analysis of 709 new transitions: it gives the first Lambda-doubling constant p0 for the B1Pi_u state of Rb2 and improves the low-J rotational constants by roughly an order of magnitude over Amiot–Vergès. That part is solid and stands on its own. The section that should make you hesitate is the 'global fit' in Section 5. The 611 Amiot–Vergès 'levels' fed into that fit were not measured points; they were regenerated from the Dunham coefficients of Ref. [18] itself, because the original data were not available. So the global fit is effectively a regularized refit of the new data against the old model, and the Table 5 deviations (e.g., Y'11 +6.2%, Y'12 -58%) are not independent evidence. The authors are transparent about this in Section 5, but the abstract and conclusion overstate what the global analysis establishes.\n\nA couple of softer concerns. The quoted uncertainties on the fitted constants look tight (Y'01 at 1.8e-5 cm^-1, p0 at 0.21e-7 cm^-1) relative to the 60 MHz linewidth, 10 MHz step, and the 2.8e-4 cm^-1 fit residual. With no systematic error budget, I'd treat those errors as lower bounds. Also, the abstract's 'spectral resolution 3.3e-4 cm^-1' is actually the scan step; the observed linewidth is 60 MHz. That matters for how the resolution claim should be read.\n\nThe p0 determination is the genuinely new result. It comes from fitting P/R vs Q branches, and the magnitude (~10^-7 cm^-1) is plausible for a heavy dimer. But the uncertainty should be re-examined; at J=50 the p0 splitting is ~5e-4 cm^-1, comparable to the residual scatter, so 5% accuracy may be optimistic.\n\nBottom line: this deserves a serious referee, but the global-fit claim needs to be decoupled from the standalone results. I'd ask the authors to either obtain the original Amiot–Vergès data or present the global fit strictly as a consistency check, and to provide a systematic error budget. The line lists and reconstructed-level lists are in the supplement, which is good; at least they shipped the data. If I were handling it, I'd send it out. The new data and p0 are useful to the Rb2/ultracold-chemistry crowd; the circular global fit is a fixable presentation problem, not a fatal one.","headline":"New high-resolution Rb2 B1Pi_u data and a first p0 determination, but the global fit rests on model-reconstructed 'data' and the error bars likely undercount; worth a careful referee.","tokens_in":17712,"tokens_out":3952,"would_cite":true,"duration_ms":35059,"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":"A 958-transition laser study of three Rb2 isotopologues determines the first Lambda-doubling constant for the B^1Πu state and refines its Dunham coefficients.","keywords":["high-resolution laser spectroscopy","supersonic molecular beam","rubidium dimer","Lambda doubling","Dunham coefficients","global least-squares fit","rovibronic transitions","B 1Pi_u state"],"falsifier":"Recover the original transition frequencies of the older dataset and repeat the same weighted global fit with the 709 new transitions; if the resulting $Y'_{k,l}$ and $p_0$ move outside their quoted one-standard-deviation uncertainties, the reconstruction-based claims are not supported.","tokens_in":16691,"feed_emoji":"⚛️","tokens_out":11323,"duration_ms":106431,"temperature":0.7,"pith_summary":"This paper reports rotationally resolved laser spectra of rubidium dimers in a supersonic beam, covering 958 transitions in the $X^1\\Sigma_g^+\\to B^1\\Pi_u$ bands for the three natural isotopologues at a spectral resolution of $3.3\\times10^{-4}\\ \\mathrm{cm^{-1}}$. By combining these measurements with the previously published dataset in a weighted global least-squares fit, the authors obtain an improved low-$J'$ description of the excited $B$ state: the rotational and rovibrational coupling Dunham coefficients gain roughly an order of magnitude in uncertainty. They also report the first determination of the $\\Lambda$-doubling constant, $p_0 = (4.28\\pm0.21)\\times10^{-7}\\ \\mathrm{cm^{-1}}$. If the result holds, the paper converts a coarse but widely used $B$-state model into a benchmark for low-angular-momentum spectroscopy of Rb2.","feed_headline":"958 resolved lines pin down Rb2's B-state Lambda-doubling","feed_subtitle":"New beam spectra merged with older data sharpen rotational constants for all three isotopes.","key_machinery":"The argument rests on the Dunham expansion of the excited-state energy with an explicit $\\Lambda$-doubling term, $$E'(i)(v',J',\\pm)/hc = \\sum_{k,l} Y'_{kl}\\,\\$rho_i^{{k/2+l}}$(v'+\\tfrac12)^k [J'(J'+1)-1]^l \\pm \\tfrac12\\left[p_0\\$rho_i^{{1/2}}$X' + q_0\\rho_i X'^2\\right],$$ where $X'=J'(J'+1)-1$ and $\\rho_i$ is the reduced-mass ratio of each isotopologue. The $\\pm$ term separates the parity components of the $\\Pi$ state and carries the new $\\Lambda$-doubling physics; $p_0$ is the zeroth-order constant extracted by the fit. The same expansion, with ground-state coefficients held fixed from an earlier direct-potential analysis, turns every measured line into a constraint on the excited-state parameters. A weighted global least-squares fit then combines the low-$J'$ information from the new beam spectra with the broad vibrational coverage of the older dataset.","core_discovery":"The central claim is that a global analysis of new high-resolution beam spectra together with the earlier Fourier-transform dataset yields the most accurate low-$J'$ description of the $B^1\\Pi_u$ state of Rb2 available. The paper determines the first $\\Lambda$-doubling constant of this state, $p_0 = (4.28\\pm0.21)\\times10^{-7}\\ \\mathrm{cm^{-1}}$, and it sharpens the rotational and rovibrational coupling constants by roughly an order of magnitude compared with the earlier values. The new data consist of 958 assigned transitions in the $X^1\\Sigma_g^+\\to B^1\\Pi_u$ bands for $^{85}$Rb$_2$, $^{85}$Rb$^{87}$Rb, and $^{87}$Rb$_2$; for the $v''=1\\to v'=1$ band of $^{87}$Rb$_2$ no transitions could be assigned. The authors emphasize that the purely vibrational coefficients remain anchored by the older data, since only $v'=1$ and $v'=2$ are directly sampled here.","pith_inferences":["Beyond the paper: if the new $p_0$ survives independent checks, it becomes a quantitative benchmark for ab initio calculations of $\\Lambda$-doubling in the Rb2 $B$ state, a quantity that electronic-structure methods rarely predict to better than tens of percent.","Beyond the paper: because the fit residuals against the older dataset are largest at the lowest $v'$ levels, a direct potential-curve fit of the $B$ state may eventually absorb the small systematic deviations that a Dunham expansion leaves behind.","Beyond the paper: extending the beam spectra to $v'=3$ and beyond, and assigning the $v''=1\\to v'=1$ band of $^{87}$Rb$_2$, would test whether the improved constants extrapolate outside the two vibrational levels fitted here."],"forward_implications":["Low-$J'$ term values of the $B^1\\Pi_u$ state can now be predicted for all three isotopologues with uncertainties near $10^{-4}\\ \\mathrm{cm^{-1}}$, an order-of-magnitude improvement over the earlier model in the probed region.","The measured $\\Lambda$-doubling $p_0$ fixes the parity splitting for the $v'=1$ and $v'=2$ levels, removing a free parameter from double-resonance and ionization schemes that pass through this state.","Improved values of $Y'_{0,1}$, $Y'_{1,1}$, and $Y'_{0,2}$ make the $B$ state a more reliable intermediate level for resonantly enhanced two-photon ionization and for proposed laser-cooling routes for Rb2.","The global fit preserves the high-$v'$ vibrational description of the older dataset while reducing the combined residual scatter to $\\sigma = 6.6\\times10^{-4}\\ \\mathrm{cm^{-1}}$."],"supporting_citations":[{"why":"Supplies the previous $B$-state Dunham model and the 611 reconstructed rovibrational levels that the global fit merges with the new data; all claimed improvements are measured against it.","marker":"[18]"},{"why":"Fixes the ground-state $X^1\\Sigma_g^+$ Dunham coefficients used to convert measured line positions into excited-state term values.","marker":"[26]"},{"why":"Establishes the supersonic-beam $B$--$X$ fluorescence method and provides the earlier $\\Lambda$-doubling estimate that the new measurement supersedes.","marker":"[17]"},{"why":"Gives an independent set of $B$-state coefficients used as cross-checks for the band origins and rotational constants in Tables 3 and 4.","marker":"[19]"},{"why":"Implements the nonlinear least-squares routines that produce the reported coefficients, uncertainties, and residuals.","marker":"[24]"},{"why":"Provides the atomic masses used to compute the reduced-mass scaling factors $\\rho_i$ in the isotopologue Dunham expansions.","marker":"[25]"},{"why":"Supplies the nuclear-spin statistical weights used to confirm rotational assignments through the odd--even intensity alternation.","marker":"[27]"}],"fun_headline_variants":["First Rb2 B-state Lambda-doubling from 958 lines","Three Rb2 isotopes yield precise Lambda-doubling","High-res beam spectroscopy refines Rb2 rotational constants","958 transitions sharpen Rb2 B-state constants"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The global-fit conclusions rest on treating the 611 excited-state levels reconstructed from the published figure and coefficients of the older study as faithful stand-ins for the original measured levels; if that reconstruction is biased, the claimed refinements inherit the bias.","fun_headline_variants_meta":{"raw":{"variants":["First Rb2 B-state Lambda-doubling from 958 lines","Three Rb2 isotopes yield precise Lambda-doubling","High-res beam spectroscopy refines Rb2 rotational constants","958 transitions sharpen Rb2 B-state constants"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000961,"raw_usage":{"total_tokens":4169,"prompt_tokens":1095,"completion_tokens":3074,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":711,"completion_tokens_details":{"reasoning_tokens":3010}},"tokens_in":711,"tokens_out":3074,"duration_ms":24884,"temperature":1.0,"reasoning_tokens":3010,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:20:29.918548+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recover the original transition frequencies of the older dataset and repeat the same weighted global fit with the 709 new transitions; if the resulting $Y'_{k,l}$ and $p_0$ move outside their quoted one-standard-deviation uncertainties, the reconstruction-based claims are not supported.","supporting_citations":[{"cited_title":"Amiot and J","cited_arxiv_id":null,"evidence_quote":"Supplies the previous $B$-state Dunham model and the 611 reconstructed rovibrational levels that the global fit merges with the new data; all claimed improvements are measured against it."},{"cited_title":"Direct potential fit analysis of theX1Σ+ g state of Rb2: Nothing else will do!Journal of Chemical Physics, 113, 8 2000","cited_arxiv_id":null,"evidence_quote":"Fixes the ground-state $X^1\\Sigma_g^+$ Dunham coefficients used to convert measured line positions into excited-state term values."},{"cited_title":"Highresolutionspectroscopyin supersonic nozzle beams: TheRb2 B1Πu−X 1Σ+ g band system.Chemical Physics, 54:21–31, 12 1980","cited_arxiv_id":null,"evidence_quote":"Establishes the supersonic-beam $B$--$X$ fluorescence method and provides the earlier $\\Lambda$-doubling estimate that the new measurement supersedes."},{"cited_title":"Drozdova, X","cited_arxiv_id":null,"evidence_quote":"Gives an independent set of $B$-state coefficients used as cross-checks for the band origins and rotational constants in Tables 3 and 4."},{"cited_title":"Virtanen and SciPy 1.0 Contributors","cited_arxiv_id":null,"evidence_quote":"Implements the nonlinear least-squares routines that produce the reported coefficients, uncertainties, and residuals."},{"cited_title":"Atomic weights of the elements: Review 2000 (iupac technical report).Pure Appl","cited_arxiv_id":null,"evidence_quote":"Provides the atomic masses used to compute the reduced-mass scaling factors $\\rho_i$ in the isotopologue Dunham expansions."},{"cited_title":"Aldegunde and Jeremy M","cited_arxiv_id":null,"evidence_quote":"Supplies the nuclear-spin statistical weights used to confirm rotational assignments through the odd--even intensity alternation."}],"review_version":1}