{"id":"c2edbd9d-a05f-417f-8471-58774a6edc80","arxiv_id":"2502.02644","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A multi-epoch survey of 128 BAF supergiants in the SMC finds very low binary fractions (25% for B5-9, 5% for AF), implying these stars did not evolve directly from main-sequence binaries.","lead":"This paper measures how often BAF-type supergiant stars in the Small Magellanic Cloud have binary companions, using nine epochs of spectroscopy for 128 stars. It finds very few binaries and argues these stars cannot be simple descendants of main-sequence binaries, which challenges standard evolutionary paths.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Comparison with MS uses B0-2 binary fraction (~75-80%) for BAF progenitors that are mostly 6-13 Msun; if late-B MS fraction is ~30-40%, 'cannot be explained by direct evolution' may collapse.","rationale":"The reader's verdict correctly flags the paper's self-admitted unsatisfactory threshold, but that concern is not the most load-bearing for the central claim. A 5 km/s threshold contaminated by alpha Cygni pulsations would cause some single stars to be misclassified as binaries; removing that contamination lowers the observed and bias-corrected fractions, making the deficit relative to MS even larger. Thus the threshold affects the exact numbers but not the direction of the conclusion. The load-bearing step is the comparison population: the paper compares BAF supergiants (mostly 6-13 Msun) to B0-2 MS stars with ~80% binary fraction. Since binary properties scale with mass, and late-B MS stars have considerably lower fractions (Dunstall et al. 2015; Moe & Di Stefano 2017), the assumed 75% may be too high by a factor of ~2. If the true MS fraction is ~30-40%, then the observed BAF fractions (<18%, ~8%) could be consistent with direct evolution, and the claim 'cannot be explained by a direct evolution from the main-sequence' would not follow. The paper's bias correction also adopts period and q distributions from O/early-B samples (Sana et al. 2025; Shenar et al. 2022b) without testing lower-mass distributions, compounding the risk. A matched-mass comparison or literature value for late-B dwarf multiplicity is absent. Therefore the CONDITIONAL verdict is appropriate, but it should hinge on redoing the MS comparison with mass-matched samples, not primarily on the threshold.","tokens_in":28439,"tokens_out":8839,"duration_ms":81070,"concrete_test":"Redo the Fig. 9 simulations using a MS binary fraction appropriate for the BAF progenitor mass range (e.g., 30–40% for late-B dwarfs, from Dunstall et al. 2015 or a mass-dependent fit from Moe & Di Stefano 2017), keeping the same detection criteria. If the red dashed lines then overlap the observed black lines at Delta3lim = 5 km/s within uncertainties, the claim that direct MS evolution is excluded would be falsified. Alternatively, restrict the BLOeM early-B comparison or use an SMC late-B dwarf sample to obtain a matched-mass MS fraction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that BAF supergiants cannot be direct MS descendants rests on the large gap between their bias-corrected multiplicity fractions (<18% for B5-9, ~8% for AF) and the main-sequence comparison fraction. However, the comparison uses Villaseñor et al. (2025) B0-2 giants/dwarfs (80±8%) and simulates 75±10% in Fig. 9, while the BAF sample is B5–F5 with initial masses 6–30 Msun and a bulk below ~13 Msun (Sec. 2). Binary fraction, period, and q distributions are strongly mass-dependent (Moe & Di Stefano 2017); late-B dwarfs have substantially lower binary fractions (e.g., ~34% for B-type stars in 30 Dor, Dunstall et al. 2015, with later types lower). If the correct MS fraction for the progenitors is ~30–40%, the bias-corrected BAF fractions become consistent with direct evolution, and the headline conclusion is unsupported. The paper does not provide a mass-matched MS sample or justify 75% for 6–13 Msun progenitors. The reader's threshold concern is real for the absolute fractions, but since pulsation contamination inflates the observed binary fraction, it makes the true fraction even lower, thus strengthening rather than weakening the 'cannot be explained by direct evolution' claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents radial-velocity (RV) monitoring of 128 B5-F5 supergiants in the Small Magellanic Cloud from the first nine epochs of the BLOeM campaign, with baselines of 30-65 days. The authors measure RV time series with cross-correlation, apply field-level epoch corrections, and classify binary candidates using a peak-to-peak threshold of 5 km s^-1 and a per-pair significance threshold of 4. They report observed binary fractions of 25±6% for B5-9 and 5±2% for A-F stars, and bias-corrected intrinsic fractions of <18% and 8+9/-7%, respectively, for orbital periods up to 10^3.5 days and mass ratios q>0.1. After noting that the B5-9 fraction is contaminated by α Cygni pulsational variability, they compare the fractions with those of B0-2 main-sequence stars and red supergiants, concluding that BAF supergiants cannot be explained by a direct evolution from the main sequence and that AF supergiants are neither progenitors nor descendants of red supergiants. They also report UV-excess detections and a dearth of short-period systems.","tokens_in":28599,"tokens_out":8496,"duration_ms":76735,"significance":"Strengths: the RV calibration is careful, with Monte-Carlo tests of the cross-correlation error formula (Appendix A), explicit field-level systematic corrections (Appendix B), and public tables of RVs and stellar parameters. The paper is transparent about the unsatisfactory detection threshold and about the contamination of the B5-9 sample by intrinsic variability, and it tests the robustness of the bias correction to the mass-ratio distribution. If the multiplicity fractions survive the mass-matching concern raised below, this is the first systematic multiplicity study of late-B to F supergiants at low metallicity and would be a valuable constraint on blue supergiant evolution. However, the central evolutionary conclusion is currently supported only by comparing to an earlier-type, higher-mass main-sequence sample; if the appropriate late-B main-sequence binary fraction is 30-40%, the claimed gap could disappear.","major_comments":[{"comment":"The headline comparison uses an intrinsic main-sequence binary fraction of 75±10%, taken from Villaseñor et al. (2025) for B0-2 giants and dwarfs, but the BAF sample is B5-F5 with the bulk of its mass below ~13 M⊙ (Sec. 2). Binary fraction, period, and mass-ratio distributions are strongly dependent on initial mass (Moe & Di Stefano 2017); for example, Dunstall et al. (2015) report a ~34% binary fraction for B-type stars in 30 Dor, decreasing toward later subtypes. If the appropriate late-B main-sequence fraction is 30-40%, the bias-corrected values reported here (<18% for B5-9 and 8+9/-7% for AF) become consistent with direct evolution, and the conclusion in the abstract and Section 6 is unsupported. Please repeat the Fig. 9 calculation with a mass-matched late-B main-sequence sample, or justify the adopted 75% value for the 6-13 M⊙ progenitor range.","section":"§5.1, Fig. 9"},{"comment":"The 5 km s^-1 binary detection threshold is chosen from the observed Δ3 distribution without an independent physical criterion, a limitation the authors explicitly acknowledge (\"We acknowledge that this limit is unsatisfactory\"). The paper itself identifies α Cygni variables with Δ3 values of 6-10 km s^-1 (e.g., BLOeM 1-112, 2-093, 6-008), so the 25±6% observed binary fraction for B5-9 is an upper limit rather than a measured binary fraction. Because the contamination inflates the observed fraction, it does not by itself bias the comparison against direct evolution; nevertheless, the abstract should state that the B5-9 fraction is an upper limit, and the paper should quantify the pulsation contribution, for example by using the α Cygni flags and any photometric variability information.","section":"§4.1, Fig. 5, Table B.3"},{"comment":"The bias-correction simulations adopt the orbital period distribution from Sana et al. (2025), a companion paper in the same campaign with overlapping authorship, and the robustness test described in Section 4.1 varies only the mass-ratio distribution. Since the quoted intrinsic fractions (<18% and 8+9/-7%) depend directly on the assumed period distribution, please add a sensitivity test that varies the period-distribution exponent within its stated uncertainty (π=+0.1±0.2), or uses an independent period distribution for the relevant mass range, and report how the intrinsic fractions change.","section":"§4.1, Table 2"},{"comment":"The conclusion that AF supergiants are neither progenitors nor descendants of red supergiants is stronger than the analysis supports. In §5.2 the bias-corrected AF fraction (9%) is stated to be \"compatible at the 3-σ level\" with the RSG fraction of 18±4%, and the stronger claim is made conditional on an inferred drop in the period distribution beyond log P ≈ 3.5. The Summary and Conclusions then present this as a key conclusion without that caveat. Please either quantify the evidence for the period-distribution drop and demonstrate that the comparison can distinguish pre-RSG from post-RSG status, or soften the conclusion accordingly.","section":"§5.2, Section 6"}],"minor_comments":[{"comment":"In the abstract, \"5 kms (10 kms)\" should be \"5 km s^-1 (10 km s^-1)\" with proper superscripts and spacing.","section":"Abstract"},{"comment":"The heading \"Constrains on intrinsic variability\" should be \"Constraints on intrinsic variability.\"","section":"§4.2"},{"comment":"The phrase \"provides constrains on the evolution\" should be \"provides constraints on the evolution.\"","section":"§5.1"},{"comment":"In the reference list, \"Bodensteiner, J., Shenar, T., Bodensteiner, J., et al. 2025\" repeats the first author; the third author should be corrected.","section":"References"},{"comment":"The phrase \"sputious offsets\" should be \"spurious offsets.\"","section":"Appendix A"},{"comment":"The instrument name \"AstroS at\" should be \"AstroSat.\"","section":"§4.4"},{"comment":"The caption says \"the size of the symbol in both panels are determined\"; this should be \"the size of the symbols in both panels is determined.\"","section":"Fig. 5 caption"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is part of a coordinated series and relies on several companion papers (Sana et al. 2025; Villaseñor et al. 2025; Britavskiy et al. 2025) for the central comparison values. I recommend that the editor check that those papers are available and refereed before accepting this manuscript, since the main conclusion depends on numbers from them. The mass-matching issue in §5.1 is the key risk: if the authors cannot provide a mass-matched main-sequence comparison, the abstract's \"cannot be explained by a direct evolution\" should be replaced with a statement restricted to evolution from early-B main-sequence stars."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper for one measurement: the first systematic multi-epoch radial-velocity survey of BAF supergiants (B5–F5) in the SMC, and it finds very few binaries. 128 stars, nine epochs from the BLOeM campaign; observed binary fractions of 25±6% (B5-9) and 5±2% (AF) at a 5 km/s threshold, with bias-corrected intrinsic fractions of <18% and 8+9/−7%. There's also a clean period-gap result: the survey would have detected short-period (5–100 d), q>0.1 companions at >90% confidence for the B5 stars, and none are there.\n\nCredit where due: the RV work is careful. Sub-km/s precisions, field-by-field corrections for FLAMES wavelength drifts, Monte Carlo validation of the error model in Appendix A, full RV tables, and detection-probability grids. They also do something rare — they visually inspect every candidate time series and admit that most flagged candidates don't show convincing orbital motion. The abstract itself concedes the true multiplicity fraction “may be significantly smaller.”\n\nThe soft spots are in the interpretation, not the measurement. First, the 5 km/s binarity threshold is arbitrary; the authors call it “unsatisfactory” in Section 4.1. Several of their own alpha Cygni variables show 6–10 km/s peak-to-peak pulsation, so a good fraction of the B5-9 candidates are probably not binaries. Note this cuts toward a lower true fraction, so it doesn't rescue the direct-evolution picture — but the absolute numbers are fuzzier than the abstract implies.\n\nSecond, and more consequential: the “cannot be explained by direct evolution” headline leans on a main-sequence comparison that is not mass-matched. Fig 9 simulates the expectations for an intrinsic MS fraction of 75±10%, taken from Villaseñor et al.'s B0-2 dwarfs/giants. The BAF sample is mostly 6–13 Msun. Late-B MS stars have substantially lower binary fractions — around 30–40% for B-type stars in 30 Dor (Dunstall et al. 2015), with Moe & Di Stefano showing the fraction falls with mass. A mass-matched comparison would shrink the gap. The AF fraction of ~8% would still sit below 30–40%, so the direction probably survives, but the quantitative force of the claim is much weaker than the abstract's phrasing.\n\nThird, the RSG step over-reaches: the paper's own comparison gives an AF fraction “compatible at the 3-sigma level” with the RSG fraction, and the “neither progenitors nor descendants” conclusion is conditional on an assumed period distribution drop. That's thin for a headline conclusion.\n\nWho's this for: anyone working on massive-star binaries or supergiant demographics. It deserves a serious referee — it's the first measurement of this population, the data are public, and the limitations are honestly flagged. Send it to review, but push for a mass-matched MS comparison and softer headline claims in the abstract and conclusions.","headline":"First systematic RV survey of SMC BAF supergiants finds a genuinely low binary fraction and a real short-period gap, but the 'no direct evolution' headline leans on a non-mass-matched MS comparison and an arbitrary 5 km/s threshold.","tokens_in":29541,"tokens_out":5854,"would_cite":true,"duration_ms":56461,"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 survey of 128 BAF supergiants in the Small Magellanic Cloud finds that very few have binary companions, challenging the idea that these stars evolve directly from main-sequence binaries.","keywords":["BAF supergiants","binary fraction","radial velocity variability","Small Magellanic Cloud","BLOeM survey","blue supergiant evolution","alpha Cygni variables","stellar multiplicity"],"falsifier":"If a longer-baseline survey of the same stars uncovers many binary systems with orbital periods of 5–100 days and mass ratios above 0.1, the claimed dearth of short-period binaries would be falsified; conversely, if the 'binary' RV signals in the B5–9 sample are shown to be coherent with photometric pulsation periods of a few days, the 25% observed fraction would largely evaporate.","tokens_in":28090,"feed_emoji":"⭐","tokens_out":10798,"duration_ms":89020,"temperature":0.7,"pith_summary":"This paper reports a multi-epoch spectroscopic survey of 128 late-B to F-type supergiants (6–30 $M_\\odot$) in the Small Magellanic Cloud, designed to measure how many of these stars have binary companions. Using radial velocities from nine epochs, it finds an observed binary fraction of 25±6% for the B5–9 supergiants and 5±2% for the A–F supergiants at a $5\\,\\mathrm{km\\,s^{-1}}$ variability threshold, falling to 10±4% and 0% at a $10\\,\\mathrm{km\\,s^{-1}}$ threshold. After correcting for the short observational baseline with Monte Carlo simulations of binary populations, the intrinsic multiplicity fraction is less than 18% for the late-B stars and about 8% for the A–F stars for orbital periods up to $10^{3.5}$ days and mass ratios $0.1<q<1$. These numbers are far below the multiplicity fractions of roughly 70–80% measured for main-sequence O and B stars, which the authors take as evidence that BAF supergiants cannot be explained by direct evolution from the main sequence. The paper further argues, from a comparison with red supergiant multiplicities, that the A–F supergiants are neither progenitors nor descendants of red supergiants.","feed_headline":"Survey finds supergiants are mostly single in the SMC","feed_subtitle":"A 128-star radial velocity survey finds only 5–25% of BAF supergiants have companions, far below main-sequence rates.","key_machinery":"The argument is carried by two tools. The first is the peak-to-peak radial velocity statistic $\\Delta_3$, the maximum difference between any two measured velocities for a star, combined with a detection threshold $\\Delta_{3,\\rm lim}=5\\,\\mathrm{km\\,s^{-1}}$ and a 4-$\\sigma$ significance criterion on pairwise velocity differences; this decides which stars are counted as binary candidates. The second is a Monte Carlo simulation of synthetic binary populations, with assumed orbital period, mass-ratio, eccentricity and mass distributions, that is used to correct the observed detections for orbital configurations the nine-epoch, 30–65 day baseline would miss. The simulation yields the bias-corrected intrinsic multiplicity fractions and also quantifies the survey's detection probability as a function of period and mass ratio, showing that short-period systems (5–100 days) with $q>0.1$ would have been detected at high confidence if they existed.","core_discovery":"The central claim is that BAF supergiants in the Small Magellanic Cloud have a notably low multiplicity fraction compared with their main-sequence predecessors, and that this low fraction rules out a direct redward evolutionary path for these stars. The evidence is the measured peak-to-peak radial velocity variability $\\Delta_3$ across nine epochs: only 13 of 128 stars, after excluding one Cepheid, exceed a $5\\,\\mathrm{km\\,s^{-1}}$ threshold at 4-$\\sigma$ significance, giving observed binary fractions of 25±6% for B5–9 and 5±2% for A–F supergiants. A Monte Carlo simulation that accounts for the limited baseline and the large stellar radii of supergiants yields bias-corrected intrinsic fractions of less than 18% for B5–9 and $8^{+9}_{-7}$% for A–F supergiants, for periods up to $10^{3.5}$ days and mass ratios $0.1<q<1$. The authors also show that most of the detected 'binary' candidates have radial velocity behaviour consistent with α Cygni-type pulsation rather than orbital motion, so the true multiplicity may be even lower. They conclude that these stars are either effectively single or products of binary interactions such as mergers, and that the A–F supergiants in particular are not connected to the red supergiant phase.","pith_inferences":["The paper's documented pulsation amplitudes (several α Cygni variables show 6–10 $\\mathrm{km\\,s^{-1}}$ peak-to-peak variations) mean that the $5\\,\\mathrm{km\\,s^{-1}}$ threshold probably overestimates the B5–9 binary fraction; if so, the intrinsic B5–9 fraction could be close to the A–F value, strengthening the paper's main conclusion but weakening its quantitative 18% upper limit as a measurement.","A longer baseline (the full 25 epochs of the survey) could discriminate between pulsation and orbital motion in the flagged candidates; one testable prediction is that the A–F supergiant binary fraction should remain near zero at any threshold, whereas the B5–9 fraction should drop as the threshold rises.","If the low binary fraction is a consequence of low metallicity in the SMC, similar surveys in the LMC or the Milky Way should find a higher BAF supergiant multiplicity, since metallicity affects mass loss and thus the evolution and survival of binary systems.","The UV-excess companions (three candidates, possibly with 15–20 $M_\\odot$ secondaries) hint that a hidden population of long-period, massive companions may survive among BAF supergiants; quantifying this population would require combining the RV data with astrometric or photometric monitoring."],"forward_implications":["If the central claim is correct, most BAF supergiants must have formed through evolutionary channels that leave them single, such as stellar mergers or birth as effectively single stars, rather than by direct evolution of main-sequence binaries.","The A–F supergiants are unlikely to be pre- or post-red-supergiant objects, which narrows the possible explanations for their position in the Hertzsprung–Russell diagram.","The absence of short-period (5–100 day) binaries with $q>0.1$ imposes a constraint on binary population synthesis models: any channel producing such companions in this mass range would contradict the observations.","The very low number of binary candidates with convincing orbital solutions implies that the true multiplicity fraction could be below 15% for both B5–9 and A–F supergiants, though not zero, given the UV-detected companions and long-period trends seen in a handful of stars."],"supporting_citations":[{"why":"Provides the main-sequence B-star binary fraction (80±8%) against which the BAF supergiant fractions are compared.","marker":"Villaseñor et al. (2025)"},{"why":"Supplies the Monte Carlo simulation method used to correct observed detections for observational biases.","marker":"Sana et al. (2013a)"},{"why":"Gives the red supergiant multiplicity fraction in the SMC that the AF supergiants are compared with.","marker":"Patrick et al. (2022)"},{"why":"Documents the α Cygni variable pulsation amplitudes that can mimic or hide binary motion in late-B to A supergiants.","marker":"Kaufer et al. (1997)"},{"why":"Provides the B0-3 supergiant binary fraction and comparison sample for the RV variability analysis.","marker":"Britavskiy et al. (2025)"},{"why":"Underpins the binary population assumptions used in the bias-correction simulations.","marker":"Sana et al. (2012)"}],"fun_headline_variants":["SMC supergiants: mostly single, few binaries","Low binarity among BAF supergiants in SMC","Supergiant evolution challenged by sparse binaries","BLOeM survey: SMC supergiants are loners","BAF supergiants in SMC challenge binary model"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the $5\\,\\mathrm{km\\,s^{-1}}$ radial-velocity threshold separates orbital motion from intrinsic pulsation; the paper explicitly calls this limit unsatisfactory and shows late-B supergiants can pulsate with 6–10 $\\mathrm{km\\,s^{-1}}$ peak-to-peak amplitudes.","fun_headline_variants_meta":{"raw":{"variants":["SMC supergiants: mostly single, few binaries","Low binarity among BAF supergiants in SMC","Supergiant evolution challenged by sparse binaries","BLOeM survey: SMC supergiants are loners","BAF supergiants in SMC challenge binary model"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000581,"raw_usage":{"total_tokens":2867,"prompt_tokens":1209,"completion_tokens":1658,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":825,"completion_tokens_details":{"reasoning_tokens":1575}},"tokens_in":825,"tokens_out":1658,"duration_ms":15293,"temperature":1.0,"reasoning_tokens":1575,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T11:33:53.933993+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If a longer-baseline survey of the same stars uncovers many binary systems with orbital periods of 5–100 days and mass ratios above 0.1, the claimed dearth of short-period binaries would be falsified; conversely, if the 'binary' RV signals in the B5–9 sample are shown to be coherent with photometric pulsation periods of a few days, the 25% observed fraction would largely evaporate.","supporting_citations":[{"cited_title":"1997, , 320, 273","cited_arxiv_id":null,"evidence_quote":"Documents the α Cygni variable pulsation amplitudes that can mimic or hide binary motion in late-B to A supergiants."},{"cited_title":"2025, accepted","cited_arxiv_id":null,"evidence_quote":"Provides the B0-3 supergiant binary fraction and comparison sample for the RV variability analysis."}],"review_version":1}