{"id":"596b2530-d693-4721-89e8-43f68cb10077","arxiv_id":"2504.18740","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Six z≈2 super-Eddington quasar candidates observed with LBT/LUCI show CIV outflow shifts of thousands of km/s, kinetic powers near a few percent of L_bol, and broad-line metallicities around 10 times solar.","lead":"Astronomers measured near-infrared spectra of six distant quasar candidates that likely swallow matter at extreme rates, then combined them with ultraviolet spectra to estimate black hole mass, wind power, and metal content. The data suggest these quasars drive powerful, metal-rich winds that could help shape the galaxies that host them.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Table 8 outflow energetics do not reproduce the paper's own Section 6.2 equations, and the quoted |epsilon|≈0.04 L_bol average is not supported by the tabulated values; the feedback claim is unsupported as printed.","rationale":"The paper's most novel quantitative conclusion is that the CIV outflows of these super-Eddington quasars carry ≈4% of L_bol, close to the canonical feedback threshold, and that the BLR gas is 10-20 Z_sun. The new LUCI data and the multi-component line decomposition are useful contributions, and the metallicity section is appropriately hedged (e.g., J2108's 500 Z_sun solution is explicitly flagged as needing extreme care). However, the feedback conclusion rests on Table 8, and Table 8 does not follow from the equations printed in Section 6.2. Spot checks show factor-of-8 discrepancy in Mdot and factor-of-3-to-4 discrepancies in kinetic power for representative rows. Moreover, the quoted average ≈0.04 is not the average of the tabulated kinetic power / L_bol values, which scatter around 0.2-1.4%. This is an internal inconsistency, not merely sensitivity to the assumed density, radius, metallicity, and velocity factor. It is the most load-bearing concern because it attacks the headline number before any astrophysical assumptions are debated. The proposed check is a simple spreadsheet recomputation of Table 8 from the published equations; it can settle the issue. If the table is corrected and the average remains ≈0.04, the concern is resolved; if not, the abstract and Section 7 cannot be accepted as written.","tokens_in":33521,"tokens_out":20445,"duration_ms":181116,"concrete_test":"Recompute every row of Table 8 directly from the Section 6.2 equations with the stated normalizations (L45=L_CIV/1e45, v_5000=|v_o|/5000, k=5, Z=5 Z_sun, n_9=1, r_1pc=1), then recompute the sample mean kinetic power / L_bol using the L_bol values in Table 7. Check that each tabulated Mdot, thrust, and kinetic power agrees with the equations at the 10% level and that the quoted ≈0.04 average is reproduced; if not, the feedback conclusion must be revised.","verdict_should_be":"REJECT","load_bearing_attack":"The central feedback claim depends on the CIV kinetic power being ≈0.04 L_bol. Section 6.2 states that Table 8 was derived from the displayed M17/Deconto-Machado expressions with k=5, Z=5 Z_sun, n=1e9 cm^-3, and r=1 pc. Applying those expressions to J084502 (log L_CIV=44.52, |v_o|=11293 km/s) gives Mdot_ion≈6 M_sun/yr and kinetic power≈6.3e45 erg/s, whereas Table 8 lists 50 M_sun/yr and 2.1e45 erg/s. For J093403 the same formula gives kinetic power≈2.6e44 erg/s, while Table 8 lists 8.6e44 erg/s. The table is therefore not a transparent evaluation of the stated model. In addition, using Table 8 entries and the L_bol values in Table 7 yields per-object kinetic power / L_bol ≈0.002-0.014 and a sample average ≈0.004-0.006, not ≈0.04. Because the abstract and Section 7 use the ≈0.04 value to argue that mechanical feedback is near the 5% L_bol threshold, the central quantitative conclusion is not supported by the numbers as printed; if the equations are instead the intended result, Table 8 must be corrected.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports new LUCI/LBT near-IR spectroscopy of six SDSS quasars at z ≈ 2.37–2.44 selected as candidate super-Eddington (xA) sources. The Hβ+FeII coverage is combined with SDSS UV spectra to determine systemic redshifts, decompose line profiles, estimate black hole masses and Eddington ratios, derive CIV outflow dynamical parameters (mass rate, thrust, kinetic power), estimate BLR metallicities via photoionization grids, and build a virial-luminosity Hubble diagram. The headline conclusions are that the CIV winds carry kinetic power |epsilon| ≈ 0.04 L_bol, close to the 5% L_bol feedback threshold, and that the broad-line/outflow gas is metal-rich (Z ≳ 10 Zsun), implying both mechanical and chemical feedback to the host galaxies.","tokens_in":33808,"tokens_out":10857,"duration_ms":112339,"significance":"If the quantitative results hold, the paper is a valuable addition to the study of super-Eddington quasars at the epoch of peak galaxy growth: high-z Hβ spectroscopy of xA candidates is rare, and the combination of systemic redshifts, line decompositions, accretion parameters, outflow energetics, and metallicity in one sample is useful. The detailed multi-component fits, residual/noise analysis, quality-parameter uncertainties, and publication of the IR spectra are strengths. The central quantitative claim, however, is not currently reproducible from the printed equations and tabulated values, and the near-threshold feedback conclusion therefore needs revision before the paper can be accepted.","major_comments":[{"comment":"Table 8 does not reproduce the equations printed in §6.2 under the stated normalizations. For J084502 (log L_CIV = 44.52, v_o = 11293 km/s), the displayed expressions with n = 10^9 cm^-3, r = 1 pc, Z = 5 Zsun, and k = 5 give Mdot_ion ≈ 8 × 0.331 × 2.26 ≈ 6 Msun/yr and kinetic power ≈ 6.6 × 10^43 × 0.331 × 25 × 2.26^3 ≈ 6.3 × 10^45 erg/s, whereas Table 8 lists 50 Msun/yr and 2.1 × 10^45 erg/s. For J093403 the same equations give kinetic power ≈ 2.6 × 10^44 erg/s, while Table 8 lists 8.6 × 10^44 erg/s. In addition, combining the Table 8 kinetic powers with the L_bol values in Table 7 gives |epsilon|/L_bol ≈ 0.002–0.014 per object, with a sample average near 0.004–0.006, not the ≈ 0.04 quoted in §6.2 and used in §7 to argue proximity to the 5% L_bol feedback threshold. The table must be recomputed consistently with the stated model, the input values and equations for each column must be specified, and the abstract and conclusions must be based on the corrected average ratio.","section":"§6.2 and Table 8"},{"comment":"The outflow energetics are highly sensitive to assumed parameters that are not measured: n = 10^9 cm^-3, r = 1 pc, Z = 5 Zsun, and the terminal-velocity factor k = 5. Since kinetic power scales as k^2 and r^-1, a factor-of-two change in k or in r moves |epsilon| by factors of several, enough to cross the 0.005 L_bol and 0.05 L_bol thresholds for individual objects. The paper states these assumptions but does not propagate them into uncertainties or present alternative choices. Because the near-threshold feedback conclusion depends directly on these scalings, a sensitivity analysis (or at minimum explicit conservative lower/upper bounds) should be added before the claim can be evaluated.","section":"§6.2 and Table 8"}],"minor_comments":[{"comment":"The abstract contains grammatical slips: 'The aim is obtain' should be 'The aim is to obtain', and 'on second stance' should be 'on second instance' or similar.","section":"Abstract"},{"comment":"The units in the Table 5 note are inconsistent: line fluxes are listed as 10^-17 erg s^-1 cm^-1, which should be 10^-17 erg s^-1 cm^-2.","section":"Table 5 note"},{"comment":"The text 'For J2018 the Mgii is not available' should refer to J2108 (J210831). In addition, the abstract's 'Z ≳ 10 Zsun' is based on three objects, one of which (J210831) reaches Z ≈ 500 Zsun and another (J144218) has a wide, poorly constrained 1σ range; the summary should state explicitly that the metallicity result rests on a small subset.","section":"§6.3 and Table 9"},{"comment":"The caption's sentence about blue squares and black circles is garbled ('...are in blue'); please clarify which symbol/color corresponds to the full Hβ profile and which to the BC-only FWHM.","section":"Fig. 5 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is data-rich and the central scientific question is timely, but the main quantitative claim is internally undercut by the mismatch between the equations in §6.2 and Table 8, and by the use of the unsupported ≈ 0.04 L_bol average in the abstract and conclusions. The issue appears to be fixable by recomputing the table and re-deriving the average ratio, so I recommend major revision rather than rejection. The heavy use of calibration relations from the same group (M17, Marziani et al. 2022, Floris et al. 2024) makes an explicit external consistency check desirable, though I do not see evidence of bad faith."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the new LUCI/LBT NIR spectra are the real contribution here, and the analysis around them is careful and honestly reported. But the headline feedback claim — CIV outflows with kinetic power ≈0.04 L_bol — does not survive contact with the paper's own table and equations. I checked the stress-test arithmetic and it holds up.\n\nWhat is actually new: Hβ coverage for six z≈2–3 super-Eddington candidates, giving the first Hβ-based systemic redshifts, black hole masses, and Eddington ratios for these objects, plus a joint UV–optical look at the xA selection criteria. The multi-component fits are careful, with residual and noise analysis, and the IR spectra ship as ASCII tables, so there is reproducible value. Five of six sources meet both selection criteria; the paper is explicit that J1259 does not and that J093403's CIV is BAL-contaminated.\n\nWhere it goes wrong: Section 6.2 prints M17/Deconto-Machado-style equations with k=5, r=1 pc, n=1e9 cm^-3, Z=5 Zsun. Applying them to J084502 (log L_CIV=44.52, v_o=11293 km/s) gives Ṁ≈6 M_sun/yr and Ė≈6.3e45 erg/s. Table 8 lists 50 M_sun/yr and 2.1e45 erg/s. The table is not a transparent evaluation of the stated model. Worse, if you divide Table 8's Ė by Table 7's L_bol, per-object ratios come out ≈0.002–0.014 with sample average ≈0.005, not the ≈0.04 quoted in the abstract and Section 7. So either the table or the prose is wrong; as printed, the load-bearing 'near 5% L_bol feedback' conclusion is unsupported. This is fixable — recompute or explain the table, then rewrite the abstract to match — but it has to be fixed before the feedback claim is used.\n\nThe softer concerns are the usual ones for this program: the outflow energetics ride on adopted density, radius, and velocity factor (Ė ∝ k^2 v^3 / r), and the scaling relations come largely from the same group's earlier papers, so the zero points are not independently anchored. The metallicity side is steadier: Z≈10–20 Zsun for the three objects with full diagnostics is consistent with prior work, and the authors themselves flag the J2108 extreme value as unstable.\n\nWho it is for: AGN outflow and quasar-selection people. It deserves a serious referee, not a desk reject — the data and the selection-criteria check are worth having. My recommendation: major revision, with the outflow table/abstract inconsistency as the gating issue.","headline":"The new LBT spectra and the selection-criteria check are worth having; the headline 0.04 L_bol feedback number does not reproduce from the paper's own Table 8 and equations.","tokens_in":34380,"tokens_out":9041,"would_cite":true,"duration_ms":81481,"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":"CIV winds from six super-Eddington quasars carry about 4% of the bolometric luminosity.","keywords":["super-Eddington quasars","C IV emission line","quasar outflows","AGN feedback","broad-line region metallicity","Eddington ratio","near-infrared spectroscopy","Hubble diagram"],"falsifier":"Measure the C IV outflow radius and electron density in one of these six quasars directly, for instance by reverberation mapping of the C IV line or by using the C III]/Si III] ratio as a density diagnostic; if the emitting radius is roughly 10 pc rather than 1 pc, or the density is $10^8$ cm$^{-3}$ rather than $10^9$ cm$^{-3}$, the computed kinetic power would drop by roughly an order of magnitude, below the $0.5\\%\\,L_{\\rm bol}$ threshold for circum-nuclear feedback.","tokens_in":33291,"feed_emoji":"🌌","tokens_out":13844,"duration_ms":122721,"temperature":0.7,"pith_summary":"Six quasars at redshifts $z \\approx 2$–$3$, selected as candidates for accreting at or above the Eddington rate, were observed in the near-infrared to capture the H$\\beta$ region, with archival optical spectra providing the UV lines. The paper's central claim is that these objects combine two ingredients relevant for galaxy evolution: the C IV line traces outflowing gas whose kinetic power is close to 4% of the bolometric luminosity $L_{\\rm bol}$, near the threshold at which AGN-driven winds are thought to regulate host star formation, and the broad-line gas is metal-rich, around $10$–$20\\,Z_\\odot$. The authors therefore argue that highly accreting quasars at the peak epoch of galaxy growth can deliver both mechanical and chemical feedback. The same data let the authors check that the UV line-ratio criterion for identifying super-Eddington candidates agrees with the optical Fe II criterion in most cases, and that H$\\beta$-based distance estimates for these objects follow the standard cosmic expansion curve.","feed_headline":"Quasar winds approach the 5% feedback threshold at cosmic noon","feed_subtitle":"Six super-Eddington quasars at z~2-3 show C IV winds at ~4% of Lbol and gas at 10-20 times solar metallicity.","key_machinery":"The load-bearing tool is multi-component spectral decomposition: each line is fitted with a Lorentzian broad component kept at the rest frame plus a blueshifted skewed Gaussian component, so the C IV blueshifted component isolates the wind from the virialized gas. Outflow dynamical parameters then come from a published framework for collisionally excited lines in photoionized gas, which expresses the ionized-gas mass, mass outflow rate, thrust, and kinetic power as functions of line luminosity, outflow velocity, and assumed density, radius, metallicity, and terminal-to-observed velocity ratio. The terminal velocity is taken as five times the observed blueshift, following a simple radiation-pressure wind model. For metallicity, the machinery is a grid of photoionization models predicting ten line-intensity ratios such as C IV/He II, Al III/C IV, and Fe II/H$\\beta$ as functions of hydrogen density, ionization parameter, and metallicity; the best metallicity is the one minimizing $\\chi^2$ over the grid. This decomposition is what connects an observed line profile to a physical feedback power.","core_discovery":"Anchoring the quasar rest frame with H$\\beta$ shows that C IV $\\lambda$1549 in most of the super-Eddington candidates is dominated by a blueshifted component with velocities of roughly $-2000$ to $-5000$ km/s, while H$\\beta$, Mg II, and Al III stay symmetric and near the rest frame. Interpreting this blueshifted C IV component as a radiation-driven wind, the paper derives ionized-gas masses, outflow rates, thrusts, and kinetic powers under explicit assumptions of density, radius, metallicity, and terminal-velocity factor; the resulting kinetic power is, on average, about $0.04\\,L_{\\rm bol}$, close to the 5% threshold invoked for AGN feedback on the host galaxy. A grid of photoionization models fitted to ten UV and optical diagnostic ratios returns broad-line metallicities of roughly $10$–$20\\,Z_\\odot$ for the objects with the best data. The paper concludes that super-Eddington quasars at $z\\approx2$–$3$ are capable of both mechanical and chemical feedback on their hosts, and that low-ionization lines such as H$\\beta$ remain reliable virial tracers even in these extreme accretors.","pith_inferences":["The largest lever on the feedback number is the assumed terminal-velocity factor $k=5$: kinetic power scales as $k^2 v^3$, so a calibration of the true acceleration law (from line-profile shape or multi-epoch monitoring) could move the result across the $0.5$–$5\\%\\,L_{\\rm bol}$ range.","If the high metallicities are real, super-Eddington quasar winds should leave detectable abundance imprints in circumgalactic gas at $z\\approx2$; deep surveys of C II or CO emission around such quasars could test this.","The paper's single-epoch estimates could be turned into direct measurements by reverberation mapping of C IV in one or two of these objects, replacing the assumed radius of 1 pc and density of $10^9$ cm$^{-3}$ with measured values.","The H$\\beta$ Hubble-diagram agreement hints that a larger near-infrared sample of super-Eddington quasars could become a standalone distance ladder at $z>1$, but the current five-object sample is too small to establish this."],"forward_implications":["If the average kinetic power of about $0.04\\,L_{\\rm bol}$ holds, these super-Eddington quasars sit close to the 5% $L_{\\rm bol}$ threshold at which AGN outflows are expected to affect the black-hole–host-galaxy relation.","Broad-line metallicities of about $10$–$20\\,Z_\\odot$ imply the same winds carry metal-enriched gas, making chemical enrichment of the host and circumgalactic medium a corollary of mechanical feedback.","The UV and optical selection criteria agree in roughly 75%–90% of cases, supporting the use of UV line ratios to find super-Eddington quasars when H$\\beta$ is unavailable.","Because H$\\beta$ and Mg II remain symmetric while C IV is strongly blueshifted, H$\\beta$-based black hole masses and Eddington ratios stay trustworthy in these extreme accretors.","The H$\\beta$ line-width distance moduli of the extreme quasars follow the standard cosmological curve, strengthening the proposal that low-ionization line widths can serve as distance indicators at $z>1$."],"supporting_citations":[{"why":"Supplies the equations that convert C IV line luminosity and blueshift into ionized-gas mass, mass outflow rate, thrust, and kinetic power used in Table 8.","marker":"M17"},{"why":"Recomputes the C IV-to-ionized-gas conversion for the softer SED of high-luminosity quasars, updating the M17 formulae adopted here.","marker":"Deconto-Machado et al. (2024)"},{"why":"Sets the roughly 5% $L_{\\rm bol}$ kinetic-power threshold against which the measured winds are compared.","marker":"Di Matteo et al. (2005)"},{"why":"Provides the simple radiation-pressure wind model from which the terminal-velocity factor $k=5$ is taken.","marker":"Netzer & Marziani (2010)"},{"why":"Defines the super-Eddington (xA) UV and optical selection criteria and the virial-luminosity relation used for the Hubble diagram.","marker":"Marziani & Sulentic (2014)"},{"why":"Supplies the H$\\beta$ black-hole-mass scaling tailored to high accretion rates and the high-metallicity expectation for xA sources.","marker":"Floris et al. (2024)"},{"why":"Establishes that C IV width is not a virial mass estimator, motivating the rest-frame anchoring with H$\\beta$ and the decomposition into a blueshifted wind component.","marker":"Sulentic et al. (2007)"}],"fun_headline_variants":["CIV winds in super-Eddington quasars reach 4% Lbol","Extreme quasars blow metal-rich winds near feedback threshold","z~2 quasars: CIV blueshifts imply ~4% Lbol kinetic power","Super-Eddington quasars show metal-rich outflows at 4% Lbol","H-beta anchors CIV winds in hyper-accreting quasars"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The outflow dynamical parameters assume the C IV blueshifted component is a wind with fixed electron density $n=10^9$ cm$^{-3}$, radius $r=1$ pc, abundance $Z=5\\,Z_\\odot$, and terminal velocity $k=5$ times the observed shift; since kinetic power scales as $k^2 v^3$, a different radius, density, or velocity factor would move the derived feedback power below or above the $0.5$–$5\\%\\,L_{\\rm bol}$ band.","fun_headline_variants_meta":{"raw":{"variants":["CIV winds in super-Eddington quasars reach 4% Lbol","Extreme quasars blow metal-rich winds near feedback threshold","z~2 quasars: CIV blueshifts imply ~4% Lbol kinetic power","Super-Eddington quasars show metal-rich outflows at 4% Lbol","H-beta anchors CIV winds in hyper-accreting quasars"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000652,"raw_usage":{"total_tokens":3035,"prompt_tokens":1039,"completion_tokens":1996,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":655,"completion_tokens_details":{"reasoning_tokens":1894}},"tokens_in":655,"tokens_out":1996,"duration_ms":13837,"temperature":1.0,"reasoning_tokens":1894,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:11:20.119285+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the C IV outflow radius and electron density in one of these six quasars directly, for instance by reverberation mapping of the C IV line or by using the C III]/Si III] ratio as a density diagnostic; if the emitting radius is roughly 10 pc rather than 1 pc, or the density is $10^8$ cm$^{-3}$ rather than $10^9$ cm$^{-3}$, the computed kinetic power would drop by roughly an order of magnitude, below the $0.5\\%\\,L_{\\rm bol}$ threshold for circum-nuclear feedback.","supporting_citations":[{"cited_title":"Exploring the links between quasar winds and radio emission along the Main Sequence at high redshift","cited_arxiv_id":"2408.06260","evidence_quote":"Recomputes the C IV-to-ionized-gas conversion for the softer SED of high-luminosity quasars, updating the M17 formulae adopted here."}],"review_version":1}