{"id":"94af7a74-3fe7-4a74-bcba-439f029466ef","arxiv_id":"1908.05023","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"HD 30085 and HD 30963 show strong mercury and manganese lines and are reclassified as HgMn stars, while HD 174567 shows mild chemical peculiarities.","lead":"This paper measures the amounts of 40 chemical elements in the atmospheres of four bright B- and A-type stars and identifies two of them, HD 30085 and HD 30963, as new members of the chemically peculiar HgMn class. It matters because such peculiar abundance patterns are a direct test of atomic diffusion in stellar atmospheres.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The most load-bearing issue is the unvalidated retuning of Hg II 3983.93 gf values in Sec. 5.7.4; this compromises the quoted mercury abundance and isotope mix, though the HgMn classification itself is still supported by Mn II and other heavy-element lines.","rationale":"The reader's weakest assumption is the same one I would flag, so I agree with the CONDITIONAL verdict. The central claim that HD 30085 and HD 30963 are HgMn stars is supported by multiple independent abundance indicators (Mn II, Sr II, Y II, Zr II, plus the conspicuous Hg II 3984 feature), so the concern does not warrant rejection. However, the paper should disclose that the Hg gf values were fitted to the data and quantify how the derived Hg abundance changes under plausible alternative gf treatments. The internal inconsistencies about scandium and sulfur noted by the reader are real but secondary; they affect the tabulated abundance pattern, not the classification. Overall, the appropriate response is to keep the manuscript conditional on the authors providing the original gf values, the adjusted values, and a reanalysis or at least an error estimate that includes the tuning uncertainty.","tokens_in":61621,"tokens_out":6161,"duration_ms":72120,"concrete_test":"Recompute the Hg II 3983.93 A profile for HD 30085 and HD 30963 using the unmodified Dolk et al. (2003) hyperfine gf values, treating only the total mercury abundance (and possibly a fixed solar-like isotope mix) as a free parameter. If the profile can be fit within the noise, Table 5's gf changes are unnecessary and the quoted abundance is not unique; if it cannot, report the residual and quote the mercury abundance as model-dependent rather than measured. As a secondary check, synthesize the Hg II 6149.4749 A line in chi Lupi A with the same gf treatment and compare to the 3983.93 A abundance; this tests whether the gf tuning preserves total line strength and provides an external consistency check that is unavailable for the two new stars.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 5.7.4 and Table 5 describe how the nine Hg II hyperfine components at 3983.93 A were initially taken from Dolk et al. (2003), and then the authors 'iteratively altered' their oscillator strengths until the synthetic profile matched the observed line. Because the gf values are adjusted to the same data that are used to infer the abundance, the quoted mercury overabundance (log Hg/H = -5.12 in HD 30085, -5.31 in HD 30963) and the isotope fractions are not independent measurements. The uncertainty budget in Table 8 and Appendix A includes sigma_log gf as if the gf values were known, but the actual systematic error from hand-tuning is not propagated; the +/-0.12 dex is therefore likely an underestimate. This matters for the central claim because 'Hg' is the defining element in 'HgMn'. If the gf rescalings changed the total line opacity as well as its wavelength distribution, the mercury abundance is degenerate with arbitrary scaling. The claim is not destroyed, however: the 3983.93 A feature is strong (about 12% of the continuum), the authors show plausible contaminants contribute only 4.9 mA of the 64-70 mA observed, and with the original gf data they already obtain an enormous overabundance (~120000 solar for HD 30085). The Mn II, Sr II, Y II and Zr II overabundances are independent of this issue, so the reclassification as HgMn stars is still on solid ground. What is weakened is the quantitative mercury abundance, the isotope ratios, and the specific comparison to the cool HgMn isotope pattern in chi Lupi A.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a detailed LTE abundance analysis of 40 elements in four late-B/early-A stars: HD 30085, HD 30963, HD 174567 and the known HgMn star χ Lupi A, using SOPHIE, DAO and FEROS spectra. Effective temperatures and surface gravities are derived from Strömgren photometry, microturbulent velocities from Fe II line dispersion, and abundances from SYNSPEC spectrum synthesis with ATLAS9/ATLAS12 model atmospheres, with NLTE corrections for Na, Mg and Ca. The central claim is that HD 30085 and HD 30963 are two new HgMn stars, based on the strong Hg II 3983.93 Å line and on large overabundances of Mn, Sr, Y, Zr, Pt and Hg together with light-element deficiencies, while HD 174567 is a mild chemically peculiar star. A reanalysis of χ Lupi A on the same temperature scale and with the same line list is used as a comparison anchor.","tokens_in":61972,"tokens_out":4183,"duration_ms":45212,"significance":"If the classification claim holds, the paper adds two new members to the small class of cool, sharp-lined HgMn stars and identifies a new mild CP star, which is valuable for studies of radiative diffusion in late-B/early-A stars. The paper has clear strengths: the HgMn classification is supported by several independent abundance anomalies (Mn II, Sr II, Y II, Zr II) that do not depend on the Hg line modeling; the comparison of all stars on a common effective-temperature scale and with the same atomic data is methodologically sound; and the list of line identifications for HD 30085 (Table 10) is a useful resource. The quantitative mercury abundance and especially the isotope fractions are, however, compromised by the iterative adjustment of the Hg II hyperfine oscillator strengths described in Section 5.7.4, so those specific numbers should not be presented as independent measurements until that issue is resolved.","major_comments":[{"comment":"The mercury abundance and isotope fractions are not independently measured. The text states that the oscillator strengths of the nine Hg II hyperfine components were 'iteratively altered' until the synthetic profile matched the observed line; these modified gf values are then used to derive log Hg/H and the isotope fractions. This is circular at the line-formation level, because any mismatch between model and observation can be absorbed into the gf rescaling. The Appendix A uncertainty budget (Eq. A1 and Table 7) treats sigma_log gf as an independent input, but for Hg II the gf values were tuned to the same data used to measure the abundance, so the quoted ±0.12–0.23 dex does not include the dominant systematic error. The authors should either use the published Dolk et al. (2003) gf values and report the resulting abundance with an explicit un-modeled systematic uncertainty, or validate any gf modification against independent data, e.g., the Hg II 6149.47 Å line that is resolved in χ Lupi A. The HgMn classification itself is not endangered, since the Mn II, Sr II, Y II and Zr II overabundances are independent of this issue.","section":"5.7.4 and Table 5"},{"comment":"The abstract's abundance-pattern summary is incorrect for scandium. The abstract lists scandium among the underabundant elements, but Section 5.2.10 states that scandium is overabundant in HD 30085, HD 30963 and HD 174567, and Table 8 gives [Sc/H] = +0.30, +0.70 and +0.11 for these three stars; only χ Lupi A is scandium-underabundant ([Sc/H] = -1.30). This is a load-bearing inconsistency in the summary of the paper's main result and must be corrected.","section":"Abstract and Section 5.2.10 / Table 8"},{"comment":"There is an internal inconsistency in the final mercury abundance for HD 30085. Section 5.7.4 reports log Hg/H ≈ -5.83 for HD 30085, while Table 8 lists log Hg/H = -5.12 for the same star and the same line. This 0.71 dex discrepancy in the defining element of the HgMn classification needs to be reconciled. In addition, the isotope fractions quoted in Section 5.7.4 are derived from equivalent-width ratios of hyperfine components whose gf values were changed by hand; given that arbitrariness, the isotope fractions should either be removed or presented only as a rough line-profile decomposition with no quantitative claim.","section":"5.7.4 and Table 8"}],"minor_comments":[{"comment":"The sulfur paragraph is self-contradictory: it reads 'Sulfur is solar in HD 30085, overabundant in HD 30963 and χ Lupi A and underabundant in HD 30963'. The last occurrence should presumably refer to HD 174567.","section":"5.2.8"},{"comment":"The header of Table 5 is malformed: 'The Hg II 3983.93 Å linelist' is merged with the column names. The table should be reformatted, and the caption should state that the listed log gf values are the authors' modified values, not the original Dolk et al. (2003) values.","section":"Table 5"},{"comment":"Several entries contain placeholder references such as 'V ALD??' and '?', and some lines are labelled 'no data' or 'nd'. These should be resolved or explicitly marked as upper limits with a consistent notation.","section":"Table 8"},{"comment":"The captions and axis labels should be checked: Figure 7 is captioned as synthesizing 'Hg II 3983.87 Å' while the text and Table 5 refer to 3983.93 Å, and Figure 4/5 captions should identify the plotted wavelength ranges more clearly.","section":"Figures 4, 5, 7, 8"},{"comment":"The evolutionary tracks are said to include microscopic diffusion while the isochrones are 'retrieved for the current solar composition'; the text should clarify whether the adopted diffusion treatment matters for the quoted masses and ages.","section":"4.1.2"}],"recommendation":"major_revision","confidential_remarks":"The central classification claim is defensible and I do not recommend rejection, but the paper currently presents the mercury abundance and isotope fractions as quantitative results even though they are obtained by adjusting the line-formation input gf values to the observed profile. This is the main correctness risk and it affects the headline element. The scandium mismatch between abstract and Table 8 is an easily fixable but embarrassing inconsistency. I would ask the authors to either validate the gf modifications with independent data or re-frame the mercury results with an explicit dominant systematic uncertainty, and to correct the internal Hg abundance discrepancy."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a useful census paper. HD 30085 and HD 30963 get their first real 40-element abundance patterns, HD 174567 gets a first detailed look, and the comparison to chi Lupi A on a common temperature scale is a nice consistency move. The classification of the two stars as HgMn is probably right, even though the headline mercury numbers should not be taken at face value.\n\nWhat is actually new: the abundance patterns for two previously barely studied late B stars, plus a long line identification table for HD 30085. The spectra are high S/N and the analysis is mostly standard LTE spectrum synthesis with ATLAS9/ATLAS12 and SYNSPEC49, with NLTE corrections for Na, Mg, and Ca. The Mn II, Sr II, Y II and Zr II overabundances are large and based on several lines; those alone support the HgMn label. The paper also re-derives chi Lupi A abundances on the same scale, making the comparison internally consistent.\n\nThe main soft spot is the mercury analysis. Section 5.7.4 is explicit: the nine Hg II hyperfine components at 3983.93 Å were taken from Dolk et al., then the oscillator strengths were iteratively altered until the synthetic profile matched the observed line. The resulting Hg abundance and isotope fractions are therefore not independent measurements. Since Hg is the defining element in the class name, the abstract and conclusion should either quote the pre-tuning abundance or flag the post-tuning value as model-dependent. The uncertainty budget in Table 7 includes sigma(log gf) as if the gf values were known; the actual systematic error from hand-tuning is not propagated, so the quoted +/-0.12 dex is likely too small.\n\nThere are also internal inconsistencies. The abstract says scandium is underabundant, but Table 8 and Section 5.2.10 say scandium is overabundant in HD 30085 and HD 30963, with only chi Lupi A underabundant. Section 5.2.8 says sulfur is both overabundant and underabundant in HD 30963 in consecutive sentences. These look like editing slips, but they need fixing before publication.\n\nOne more minor point: many of the 40 abundances are single-line estimates or upper limits. The authors are usually honest about this, but the abstract's '40-element' framing oversells the secure sample. Data and line list access would help future work.\n\nBottom line: the classification is likely robust, but the quantitative mercury is not. That is exactly the kind of issue peer review should catch. Send it to an expert referee; if the authors disclose the Hg gf adjustments, correct the inconsistencies, and soften the Hg claims, this becomes a citable contribution to the HgMn census.","headline":"Two new HgMn stars have solid Mn/Sr/Y/Zr support, but the quoted mercury abundance is not independently measured because the Hg II gf values were tuned to fit the line.","tokens_in":62508,"tokens_out":3015,"would_cite":true,"duration_ms":33336,"reading_group":"maybe","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 two stars previously classified as normal, HD 30085 and HD 30963, are actually mercury-manganese (HgMn) stars, and that HD 174567 is a mildly chemically peculiar star.","keywords":["chemically peculiar stars","HgMn stars","abundance analysis","mercury-manganese stars","radiative diffusion","spectral synthesis","HD 30085","HD 30963"],"falsifier":"A laboratory measurement of the nine Hg II hyperfine component oscillator strengths at 3983.93 Å would settle the mercury claim: if independent gf values, used in the same synthesis, cannot reproduce the observed flat-bottomed profile with an overabundance near $10^5$ solar, the mercury abundance and isotope fractions as quoted would be wrong.","tokens_in":61423,"feed_emoji":"⭐","tokens_out":6056,"duration_ms":51140,"temperature":0.7,"pith_summary":"The paper argues that two stars previously classified as normal late-B/early-A stars, HD 30085 and HD 30963, are actually mercury-manganese (HgMn) stars, a class of chemically peculiar stars whose surface abundances are shaped by radiative diffusion. Using high-resolution spectra and synthetic-spectrum fits to about forty elements, it finds large overabundances of manganese, strontium, yttrium, zirconium, platinum, and mercury, and underabundances of helium, magnesium, scandium, and nickel, matching the HgMn pattern. It also finds that HD 174567, previously used as a normal comparison star, is mildly chemically peculiar. If the classification holds, the census of HgMn stars gains two members, and the two stars become additional laboratories for studying how atomic diffusion produces abundance anomalies in slow rotators.","feed_headline":"Two 'normal' stars are actually mercury-manganese stars","feed_subtitle":"Abundance patterns of HD 30085 and HD 30963 match the HgMn class, with mercury overabundant by a factor near 10^5.","key_machinery":"The analysis rests on LTE spectrum synthesis with model atmospheres (computed with ATLAS9 and ATLAS12) and the SYNSPEC49 code, fitting carefully selected unblended lines of forty elements. The load-bearing objects are the Hg II line at 3983.93 Å, modeled with nine hyperfine components from isotopes Hg 196 through Hg 204, and the Mn II lines with published hyperfine structure; the mercury oscillator strengths were iteratively adjusted to reproduce the observed flat-bottomed profile, yielding an isotope mix dominated by the heaviest isotopes. The abundance patterns are compared element by element with $\\chi$ Lupi A and the normal star $\\nu$ Cap, and the global pattern is interpreted as the result of radiative diffusion in a slow rotator.","core_discovery":"HD 30085 and HD 30963 are two new HgMn stars. The evidence is their effective temperatures ($\\sim11300$ K and $\\sim11480$ K), low projected rotational velocities (26 and 37 km s$^{-1}$), and abundance patterns that reproduce the characteristic HgMn signature: light elements depleted, iron-peak and heavy elements overabundant, with mercury overabundant by factors near $10^5$ relative to solar. Their patterns align with that of the well-known HgMn star $\\chi$ Lupi A and fall within the compiled abundance range of confirmed HgMn stars. HD 174567 shows milder anomalies and is classified as a new mild chemically peculiar star, possibly a cool and mild HgMn star.","pith_inferences":["A direct check of the mercury result would be new laboratory oscillator strengths for the nine Hg II components; without that, the extreme mercury overabundance and the heavy-isotope mix remain dependent on the model.","The same line-by-line synthesis could be run on other slow rotators from the sample used here; the paper implies the bright-end census of HgMn stars may still be incomplete.","The intermediate pattern of HD 174567 leaves open whether mild HgMn stars form a continuum with normal stars or a separate regime; the paper does not settle that.","A higher-resolution, higher-signal-to-noise observation of the Hg II profile in both stars, ideally over several epochs, could test whether the flat core and isotope fractions are stable or variable."],"forward_implications":["The census of known HgMn stars grows by at least two, both relatively bright and observable from the northern hemisphere.","The two stars can be added to samples used to test radiative-diffusion models of HgMn abundance anomalies, especially the correlation between effective temperature and mercury isotope fraction.","HD 174567 should no longer be treated as a normal comparison star in abundance studies; its mild anomalies need to be accounted for.","The heavy-isotope dominance suggested for mercury in both stars, if confirmed, links them to the cool HgMn stars rather than the hotter ones.","The line identifications and abundances for HD 30085 provide an empirical linelist for future analyses of similar stars."],"supporting_citations":[{"why":"Supplies the sample of slowly rotating A0-A1V stars, the projected rotational velocities, and the radial velocities used for HD 30085 and HD 174567.","marker":"Royer et al. (2014)"},{"why":"First reported the Hg II line and strong Mn II lines in HD 30085 and gave preliminary Mn, Fe and Hg overabundances, the starting point this paper extends.","marker":"Monier et al. (2015)"},{"why":"Provides the previous high-resolution optical abundance study of chi Lupi A that the paper rederives on a consistent temperature scale for comparison.","marker":"Wahlgren et al. (1994)"},{"why":"Catalogued HD 174567 as a normal comparison star from IUE spectra, the status the paper revises.","marker":"Smith & Dworetsky (1993)"},{"why":"Provides the recent abundance analysis of nu Cap, the bona-fide normal late-B star used as the comparison baseline.","marker":"Monier et al. (2018)"},{"why":"Supplies the nine Hg II hyperfine transitions from isotopes Hg 196 to Hg 204 used to model the mercury line.","marker":"Dolk et al. (2003)"},{"why":"Supplies the hyperfine structures used for the Mn II lines that establish the manganese overabundance.","marker":"Holt et al. (1999)"},{"why":"Provides the compilation of HgMn abundance patterns that the paper uses to place HD 30085 and HD 30963 inside the class.","marker":"Ghazaryan & Alecian (2016)"}],"fun_headline_variants":["From normal to HgMn: two stars reclassified","Mercury overabundant by 10^5 in new HgMn stars","Two misclassified stars are new HgMn members","HD 30085 and HD 30963 join the HgMn class","Mercury-manganese class gains two new members"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The mercury result depends on the strengths assigned to the nine components of the mercury line, which the authors adjusted until the computed line matched the observed one; if those strengths are not physically real, the mercury overabundance and isotope mix are not independently measured, though the other element anomalies still point to HgMn.","fun_headline_variants_meta":{"raw":{"variants":["From normal to HgMn: two stars reclassified","Mercury overabundant by 10^5 in new HgMn stars","Two misclassified stars are new HgMn members","HD 30085 and HD 30963 join the HgMn class","Mercury-manganese class gains two new members"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001222,"raw_usage":{"total_tokens":5080,"prompt_tokens":1058,"completion_tokens":4022,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":674,"completion_tokens_details":{"reasoning_tokens":3939}},"tokens_in":674,"tokens_out":4022,"duration_ms":28369,"temperature":1.0,"reasoning_tokens":3939,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:25:24.131378+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A laboratory measurement of the nine Hg II hyperfine component oscillator strengths at 3983.93 Å would settle the mercury claim: if independent gf values, used in the same synthesis, cannot reproduce the observed flat-bottomed profile with an overabundance near $10^5$ solar, the mercury abundance and isotope fractions as quoted would be wrong.","supporting_citations":[{"cited_title":"2014, , 562, A84","cited_arxiv_id":null,"evidence_quote":"Supplies the sample of slowly rotating A0-A1V stars, the projected rotational velocities, and the radial velocities used for HD 30085 and HD 174567."},{"cited_title":"2015, , 577, A96","cited_arxiv_id":null,"evidence_quote":"First reported the Hg II line and strong Mn II lines in HD 30085 and gave preliminary Mn, Fe and Hg overabundances, the starting point this paper extends."},{"cited_title":"M., Adelman , S","cited_arxiv_id":null,"evidence_quote":"Provides the previous high-resolution optical abundance study of chi Lupi A that the paper rederives on a consistent temperature scale for comparison."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Catalogued HD 174567 as a normal comparison star from IUE spectra, the status the paper revises."},{"cited_title":"2018, , 854, 50","cited_arxiv_id":null,"evidence_quote":"Provides the recent abundance analysis of nu Cap, the bona-fide normal late-B star used as the comparison baseline."},{"cited_title":"M., & Hubrig , S","cited_arxiv_id":null,"evidence_quote":"Supplies the nine Hg II hyperfine transitions from isotopes Hg 196 to Hg 204 used to model the mercury line."},{"cited_title":"A., Scholl , T","cited_arxiv_id":null,"evidence_quote":"Supplies the hyperfine structures used for the Mn II lines that establish the manganese overabundance."},{"cited_title":"& Alecian , G","cited_arxiv_id":null,"evidence_quote":"Provides the compilation of HgMn abundance patterns that the paper uses to place HD 30085 and HD 30963 inside the class."}],"review_version":1}