{"id":"76858d86-2f98-4dc2-a24a-5d1d3dc9f9a5","arxiv_id":"2608.07669","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Polarization of five hot, dust-obscured quasars confirms their blue excess is scattered AGN light, and the wavelength trend in one object requires graphite-dominated dust in a polar outflow.","lead":"Astronomers measured polarized ultraviolet light from five heavily obscured quasars whose blue light leaks out despite thick dust. They find the blue light is scattered by dusty, cone-shaped outflows rich in graphite, a clue to how these quasars blow away their surroundings.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'graphite-only' claim rests on an asymmetric comparison: silicate mixtures get one fiducial geometry, graphite-only dust gets a full 3-parameter grid; a silicate-rich model might match at other cone/torus angles.","rationale":"The paper's empirical core is solid: the new FORS2 polarization measurements are calibrated against standard stars, the zero-polarization standards are consistent with null, and all five BHDs show high polarization fractions. The strongest claim, however, is not just that the blue excess is scattered light; it is the more specific claim that the scatterer is a dusty polar outflow whose dust is dominated by graphites. That composition claim is the load-bearing part because it is the basis for the physical interpretation (graphite sublimation temperature) and for the predicted 1800 Angstrom feature. My concern is that the comparison establishing 'only graphites' is not a fair comparison. All non-graphite mixtures are evaluated at one fixed cone/torus geometry, while the graphite mixture is given the full geometric parameter space. The paper explicitly notes that the polarization fraction depends strongly on geometry and that psi_Cone and eta are degenerate (Figure 8). It also notes the model's idealization of uniform dust in a thin shell at the cone edge, acknowledging that real dust may be concentrated at the outflow base. Either of these alone would be a caveat; together they mean the graphite conclusion may be an artifact of the chosen comparison. This is not a claim that the authors are wrong, only that the central composition claim is less secure than the scattered-light claim. The proposed test is straightforward: re-run the refined grid for silicate-bearing mixtures. If silicates still cannot approach the observed I-band polarization at any geometry, the conclusion is strengthened; if they can, the paper needs spectropolarimetry or a physical argument for why the preferred geometry is the only plausible one. Such a test directly settles the load-bearing doubt, which is why I recommend no change to the reader's CONDITIONAL verdict.","tokens_in":25606,"tokens_out":5412,"duration_ms":55593,"concrete_test":"Run the same refined SKIRT grid used for MRN77gra (psi_Torus = 25-60 deg, psi_Cone = 20 deg - psi_Torus, eta = psi_Torus - 90 deg, step 5 deg) for at least MRN77gra+sil and one Draine (2003b) mixture (e.g., SMC bar), fitting the three W0116-0505 broad-band polarization measurements with the same Ly-alpha dilution treatment used for the graphite model. Compare the best chi-squared and the predicted I_Bessel polarization. If any silicate-bearing mixture reaches a chi-squared within ~2 of the graphite-only best fit and predicts P_I >= 14%, the 'graphite-only dust' conclusion is not secure; if silicate mixtures remain unable to exceed P_I ~ 10% at any grid point, the composition claim is robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 4.2, the mixture comparison that establishes graphites as the only viable scatterer is asymmetric. The initial broad search over mixtures is run for a single fiducial geometry (psi_Torus = 50 deg, psi_Cone = 30 deg, eta = 55-85 deg in steps of 10 deg; Figure 6 left, Table 3). Only the graphite-only MRN77gra mixture is subsequently allowed the refined 3-parameter grid (psi_Torus = 25-60 deg, psi_Cone = 20 deg - psi_Torus, eta = psi_Torus - 90 deg, step 5 deg) and an MCMC fit. Silicate-bearing mixtures (MRN77gra+sil, large-grain gra+sil, Gaskell et al. 2004, and the Draine 2003b SMC/LMC/MW mixtures) are never re-optimized over psi_Torus and psi_Cone. Since the observed polarization fraction is highly sensitive to scattering geometry - the paper itself finds tight psi_Cone-eta degeneracies (Figure 8) - a silicate-rich mixture could plausibly reproduce the W0116-0505 wavelength trend at cone/torus angles that are excluded from the comparison. The conclusion that the dust 'must be dominated by graphites' therefore depends on the untested assumption that the graphite model's geometric advantage is not simply an artifact of giving it more freedom. The manuscript's own caveat that real dust may be concentrated at the outflow base rather than uniformly distributed in the cone shell (Section 4.2) reinforces that the fixed shell geometry is load-bearing for the composition claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents new VLT/FORS2 imaging polarimetry of three Blue Excess Hot DOGs (BHDs) in the R_Special band, new v_High and I_Bessel observations of W0116-0505, and a re-analysis of existing R_Special data for W0116-0505 and W0204-0506. It reports high polarization fractions (6-25%) in all five BHDs, confirming that the blue excess is scattered AGN light, and a rest-frame wavelength trend in W0116-0505 rising from 9.62% in v_High to 14.36% in I_Bessel. Using SKIRT radiative transfer models of a biconical dusty outflow, the paper argues that only a graphite-dominated dust mixture can reproduce this trend, and that a single torus/cone geometry with varying inclination can match all five BHDs. It also discusses the physical plausibility of graphite-rich dust in hot outflows and identifies a predicted spectral dip near 1800 Å in the scattered light.","tokens_in":25834,"tokens_out":3962,"duration_ms":41525,"significance":"If the graphite-dominated dust conclusion holds, this is an important result: it would link the scattering polarization of hyper-luminous obscured quasars to dust chemistry in radiatively driven outflows, with consequences for the evolutionary picture of Hot DOGs transitioning to unobscured quasars. The observational work is careful: polarization measurements are calibrated against standard stars and zero-polarization standards, the data are re-analyzed consistently with previous work, and the modeling is a forward radiative-transfer test using the publicly available SKIRT code without feeding the observed polarization back into the model. The all-BHD common-geometry fit is a strong, falsifiable statement. However, the central composition claim rests on a single idealized geometry and an asymmetric model comparison, so the significance is conditional on those assumptions being tested more fairly.","major_comments":[{"comment":"The conclusion that only the graphite-only MRN77gra mixture can reproduce the W0116-0505 wavelength trend is based on an asymmetric search. The initial broad mixture test fixes psi_Torus = 50 deg and psi_Cone = 30 deg and only optimizes the inclination eta (in 10 deg steps), while the subsequent refined grid (psi_Torus = 25-60 deg, psi_Cone = 20 deg to psi_Torus, eta = psi_Torus to 90 deg) and the MCMC are run only for MRN77gra. Given the tight psi_Cone-eta degeneracy shown in Fig. 8 and the strong geometry dependence of the polarization fraction, silicate-bearing mixtures might reproduce the observed three-band trend at torus/cone angles outside the fiducial geometry. The claim that the dust 'must be dominated by graphites' requires either repeating the refined grid for at least one representative silicate-rich mixture or providing an explicit argument that no geometry can make those mixtures match the I_Bessel polarization of 14.36%.","section":"§4.2, Fig. 6, Table 3"},{"comment":"The mixture comparison is carried out for a single, highly idealized dust distribution: 'all the dust in the outermost ten degrees' of a biconical shell, uniformly distributed. The paper itself later notes that the dust may in reality be concentrated at the base of the outflow. A base-concentrated or clumpy radial distribution changes the distribution of scattering angles and can alter the relative polarization of different grain compositions. Since the composition claim is the paper's central new inference, the authors should test at least one alternative radial distribution for a silicate-bearing mixture (e.g., MRN77gra+sil) before concluding that graphites are uniquely required.","section":"§4.2, Fig. 7"},{"comment":"The quantitative support for the graphite model is weaker than the qualitative language suggests. The best-fit model has chi^2 = 36 for the three W0116-0505 points, and the minimum chi^2 over the grid is 29. The paper only reaches chi^2_min = 0.6 by assuming that all Ly-alpha in the v_High band is unpolarized, a correction it then says it does not apply. As written, the model does not formally fit the data well without an additional, unmodeled dilution mechanism. The authors should report the fits with and without the Ly-alpha correction side by side, and state explicitly whether the composition conclusion depends on this correction.","section":"§4.2, near chi^2 = 36"}],"minor_comments":[{"comment":"The abstract states that the three new R_Special BHDs show polarization of about 6% to 15%, but the object W0204-0506 in Table 2 shows 24.72%; consider clarifying that the 6-15% range refers only to the three newly observed targets, not the full sample.","section":"Abstract and §3.1"},{"comment":"The phrase 'the goodness-of-fit is high, with chi^2 = 36' is ambiguous; a high chi^2 usually indicates a poor fit. Please rephrase as 'the chi^2 is large' or 'the fit is poor' to avoid confusion.","section":"§4.2"},{"comment":"The column labeled 'P_ran' is used in the text but never explicitly defined at first occurrence; please define it in the table caption or in the text.","section":"Table 1"},{"comment":"The legend colors for the different dust mixtures are difficult to distinguish in grayscale; adding line styles or markers would improve readability.","section":"Fig. 6, left panel"},{"comment":"One sentence refers to 'Thompson scattering' rather than 'Thomson scattering'; please correct the typo.","section":"Appendix D"}],"recommendation":"major_revision","confidential_remarks":"The observational results are solid and the paper addresses a timely question in quasar feedback and dust evolution. The main issue is that the central composition claim (graphite-dominated dust) is currently supported by an asymmetric model comparison and a single idealized geometry. This is fixable within the manuscript's scope by re-running the refined parameter grid for at least one silicate-rich mixture and by testing a non-uniform dust distribution. I would not reject, but I would not accept without these tests. The fit statistics also need to be presented more transparently."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this paper gives you new, well-calibrated polarization measurements for three Hot DOGs with blue excess and a three-band series for W0116-0505. That part is solid and new. The further claim that the scattering dust must be graphite-dominated is plausible but not proven; the model comparison is unfair in one specific way that a referee should catch.\n\nCredit where due: the polarization fractions are carefully measured against standard stars and zero-polarization standards, and the residual systematics are below 0.3%. The wavelength trend in W0116-0505 (9.62% in v_High to 14.36% in I_Bessel, with constant angle) is statistically significant and rules out Thomson scattering and standard ISM dust mixtures in a way that is properly argued, including the appendix tests on line contamination. The SKIRT setup is described clearly, and the authors are honest about the fit quality: chi^2=36 for three points, and the need to assume unpolarized Ly-alpha or 5% line polarization to make the v_High point fit. The ensemble fit with shared geometry is a nice idea, and the paper is refreshingly explicit about its limitations.\n\nThe soft spot is the asymmetric mixture comparison. Silicate-bearing mixtures are tested at one fiducial geometry (psi_Torus=50, psi_Cone=30, eta=55-85), while only the graphite-only MRN77gra gets the full grid and MCMC. The paper itself notes the tight psi_Cone-eta degeneracy (Fig. 8), so a silicate-rich mixture might match the trend at other cone/torus angles. The stress-test note makes this point, and it holds up on reading. It does not sink the paper: the empirical result stands, and the graphite conclusion is offered as a plausible interpretation, not a claim with rigorous exclusion. But the talk of 'must be dominated by graphites' is stronger than the evidence. A referee should ask for the silicates to be re-optimized over the same grid, or better, spectropolarimetry to break the degeneracy.\n\nWho this is for: anyone working on obscured quasars, AGN feedback, or polarization as a probe of dust. It is a solid empirical contribution with an interpretive claim that will need more work. It deserves serious peer review, not a desk reject. I would send it out, but I would flag the geometry asymmetry to the referees as a specific point to chase.","headline":"Solid new polarization data confirm scattered AGN light in these Hot DOGs; the graphite-only outflow conclusion is plausible but rides on an untested geometry.","tokens_in":26598,"tokens_out":3126,"would_cite":true,"duration_ms":28727,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The blue excess of Hot Dust-Obscured Galaxies is light from the hidden quasar scattered off graphite-rich dust in a polar outflow.","keywords":["Hot Dust-Obscured Galaxies","blue excess","AGN scattering","imaging polarimetry","dusty outflows","radiative transfer","graphite dust","quasar feedback"],"falsifier":"A spectropolarimetric observation of W0116-0505 across the rest-frame 0.15-0.3 μm window would test the predicted 1800 Å dip in the scattered-light SED and whether the polarization angle stays constant into the blue; absence of the dip, or detection of a silicate-related feature in the scattered component, would disfavour the graphite-only outflow model.","tokens_in":25274,"feed_emoji":"✨","tokens_out":12004,"duration_ms":103052,"temperature":0.7,"pith_summary":"The paper reports imaging polarimetry of five Blue Excess Hot DOGs, hyper-luminous quasars so dusty that their ultraviolet light is normally hidden. All five show linear polarization fractions of 6-25%, the signature of scattered light, confirming that the blue excess is AGN light scattered into our line of sight rather than star formation or a second unobscured quasar. In W0116-0505 the polarization rises with rest-frame wavelength, which rules out Thomson scattering and ordinary interstellar dust. Using radiative transfer, the authors find that a dusty bi-conical polar outflow explains the data only if its dust is graphite-dominated. If this is right, the blue excess is a direct probe of the quasar winds that eventually clear the obscuring dust, and the scattering dust itself records the high temperatures of the outflow's inner regions.","feed_headline":"Graphite dust, not ordinary ISM dust, causes hot DOGs' blue excess","feed_subtitle":"Five obscured quasars betray hidden engines through polarized light scattered by graphite dust in their outflows.","key_machinery":"The machine is a set of SKIRT Monte Carlo radiative transfer models with a fixed geometry: an optically thick dust torus of half-opening angle $\\psi_{\\rm Torus}$, a biconical polar outflow of half-opening angle $\\psi_{\\rm Cone}$, and all scattering dust placed in the outermost ten degrees of the cone, with an optical depth $\\tau=0.1$ at rest-frame 5500 Å. The dust grid compares silicate+graphite MRN77 mixtures (sizes $a=0.005$-$0.25\\,\\mu$m, power-law exponent $-3.5$), large-grain variants, the Draine (2003b) SMC/LMC/MW mixtures, and a graphite-only version. The comparison selects the graphite-only MRN77gra as the only mixture that fits the wavelength dependence; it also predicts a dip near 1800 Å in the scattered spectrum that the authors report as tentatively visible in three targets.","core_discovery":"The central claim is that the scattered light in BHDs comes from dust in a polar outflow, and that this dust is dominated by graphites. The evidence has two parts. First, all five BHDs observed in the R Special band are polarized at 6.35-24.72%, and W0116-0505's polarization rises from 9.62% in v High to 14.36% in I Bessel with a constant angle; this wavelength dependence rules out electron scattering and ISM-like dust. Second, SKIRT radiative transfer simulations of an accretion disk plus torus plus biconical outflow reproduce the observations only for a graphite-only MRN77 grain mixture, with best-fit torus and cone half-opening angles near 50 degrees and inclination 78 degrees; the same geometry with only inclination varied fits all five targets. The authors take the graphite requirement as evidence that the scattering dust formed or survived near the sublimation region, where graphite's higher sublimation temperature ($\\sim$1800 K) lets it outlive silicates ($\\sim$1200 K).","pith_inferences":["If the 1800 Å dip is confirmed by spectroscopy, it would give a redshift-independent marker of graphite-bearing outflow dust that works even for unresolved quasars, complementing mid-IR silicate features.","The same polarization diagnostic could be applied to other obscured quasar populations, such as extremely red quasars (ERQs) and heavily reddened quasars (HRQs), to test whether their lower scattered-light fractions come from lower dust columns or from different viewing angles.","A direct radiative-transfer experiment with clumpy dust concentrated near the base of the outflow would test whether the graphite-only conclusion is an artifact of the uniform outermost-shell assumption.","The temperature argument suggests a selection effect: quasars whose winds carry dust past the graphite sublimation radius should preferentially show this blue excess, so surveying Hot DOGs without blue excess for their scattered-light fraction could test that prediction."],"forward_implications":["All five BHDs can be described by one shared torus-plus-outflow geometry with only the viewing inclination changing, arguing that BHDs form a single evolutionary population.","The rise in polarization with rest wavelength becomes a clean diagnostic separating dusty outflow scattering from Thomson scattering or host-galaxy ISM scattering in obscured quasars.","A graphite-dominated scattering dust implies the dust has been processed near the sublimation zone of the outflow, which constrains outflow temperatures and the survival of grain species.","The models predict a dip near 1800 Å in the scattered-light spectrum, with tentative evidence already present in three of the five objects; spectroscopy can confirm or reject it.","In sources with resolved polarization gradients (W0019-1046 and W0204-0506), the gradient direction and angle can be combined with [O iii] kinematics to map the outflow geometry in three dimensions."],"supporting_citations":[{"why":"Measured the R Special polarization of W0116-0505 (about 11%) and framed the two scenarios, smooth ISM or dusty polar outflow, that this paper discriminates.","marker":"A22"},{"why":"Measured 24.72% polarization in W0204-0506, resolved the polarized emission along an outflow, and first found in radiative transfer that the scattering dust must be graphite-rich.","marker":"A25"},{"why":"Defined the BHD sample and showed their UV excess is best modeled as roughly 1% of the obscured quasar luminosity scattered into view.","marker":"Assef et al. (2016)"},{"why":"Provided the homogeneous SED fits, black-hole masses, and Eddington ratios for the five targets used to argue BHDs form a single class.","marker":"Li et al. (2024)"},{"why":"Supplied the Milky Way, LMC, and SMC dust mixtures that stand in for standard ISM dust and fail to match the wavelength dependence.","marker":"Draine (2003b)"},{"why":"Gave the power-law grain-size distribution used to define the MRN77 dust mixtures in the radiative transfer grid.","marker":"Mathis et al. (1977)"},{"why":"Fixed the 51/49 silicate-to-graphite ratio and size normalizations that separate the MRN77gra+sil from MRN77gra mixtures.","marker":"Weingartner & Draine (2001)"},{"why":"Provided the SKIRT Monte Carlo radiative transfer code that computes the scattered polarization fraction and angle for each geometry.","marker":"Camps & Baes (2015, 2020)"},{"why":"Detected the powerful [O iii] outflows in several BHDs, motivating the biconical polar-outflow scattering geometry.","marker":"Finnerty et al. (2020)"},{"why":"Supplied the accretion-disk SED and the AGN torus setup used as the illumination source in the SKIRT simulations.","marker":"Stalevski et al. (2012)"}],"fun_headline_variants":["Graphite dust in outflows causes hot DOGs' blue excess","Polarized light shows graphite dust in hot DOG winds","Hot DOGs' blue glow comes from graphite dust scattering","Graphite-rich outflows scatter light in blue excess hot DOGs","Blue excess in hot DOGs traced to graphite dust in outflows"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the scattering dust sits uniformly in the outermost ten degrees of a smooth biconical shell; the paper itself notes that real outflow dust may be concentrated at the base of the outflow, and if it is clumpy, base-concentrated, or differently sized, silicate-bearing mixtures might also fit.","fun_headline_variants_meta":{"raw":{"variants":["Graphite dust in outflows causes hot DOGs' blue excess","Polarized light shows graphite dust in hot DOG winds","Hot DOGs' blue glow comes from graphite dust scattering","Graphite-rich outflows scatter light in blue excess hot DOGs","Blue excess in hot DOGs traced to graphite dust in outflows"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00026,"raw_usage":{"total_tokens":1683,"prompt_tokens":1135,"completion_tokens":548,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":751,"completion_tokens_details":{"reasoning_tokens":460}},"tokens_in":751,"tokens_out":548,"duration_ms":5173,"temperature":1.0,"reasoning_tokens":460,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T00:25:18.722208+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A spectropolarimetric observation of W0116-0505 across the rest-frame 0.15-0.3 μm window would test the predicted 1800 Å dip in the scattered-light SED and whether the polarization angle stays constant into the blue; absence of the dip, or detection of a silicate-related feature in the scattered component, would disfavour the graphite-only outflow model.","supporting_citations":[],"review_version":1}