{"id":"f9cdde68-330f-47ec-80e4-f8244757b589","arxiv_id":"2608.05286","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"The 3.3 um PAH feature grows relative to 7.7 and 11.3 um features at low metallicity, with secondary radiation-field-driven variations, indicating smaller PAH populations in low-metallicity environments.","lead":"Using JWST near-infrared images of 19 nearby galaxies, this paper maps the 3.3 micrometer PAH emission band and shows that its strength relative to other PAH bands depends on both gas-phase metallicity and the hardness of the local radiation field. The result sharpens how astronomers can use PAH band ratios as diagnostics of dust grain sizes and interstellar medium conditions at high spatial resolution.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The metallicity trend in 3.3/11.3 and 3.3/7.7 may be partly manufactured by the variable-B_PAH continuum subtraction, because B_PAH is itself correlated with metallicity (Table 2: rho=-0.28) and is used to build the F335M PAH maps; no fixed-B_PAH control is shown.","rationale":"The reader's weakest-assumption statement already identifies the continuum-subtraction method as the fragile step, specifically noting that environmental variation in B_PAH is partly built into the maps. My stress-test sharpens this into a concrete, testable threat to the central astrophysical claim: the metallicity trend in Figures 7 and 13 may be an artifact of using a B_PAH that is itself metallicity-dependent. This is not an accusation of error; the paper is transparent about the method and includes honest caveats. However, the two map versions compared (annular versus galaxy-wide B_PAH) do not isolate the effect, because both use environmentally measured slopes. The paper's independent support includes agreement with prior work (Whitcomb+2025, Baron+2024) and its explicit FSPS test, but neither of those controls the map-making feedback loop. Therefore the verdict should remain CONDITIONAL: the central trend is plausible and aligned with earlier results, but the paper should supply the fixed-B_PAH and shuffled-B_PAH control maps before the slope values and the 'inhibited growth' interpretation are accepted at face value. I do not move the verdict because the concern is exactly the kind of empirical robustness check that a conditional acceptance should require, and the prior literature provides some independent weight in favor of a real metallicity trend.","tokens_in":43134,"tokens_out":4437,"duration_ms":44865,"concrete_test":"Recompute the Figure 7 metallicity-binned ratios and the best-fit slopes using F335M_PAH maps generated with a single fixed B_PAH, e.g., B_PAH=1.6 (the Sandstrom+2023 value) or the luminosity-weighted sample median of 1.76, applied to every galaxy and every radial bin. If the best-fit slopes in 3.3/11.3 and 3.3/7.7 versus 12+log(O/H) remain within the quoted uncertainties of -1.00 and -0.85, the environmental-B_PAH prescription is not manufacturing the trend. As a stronger null test, shuffle B_PAH values among 1.5 kpc regions (destroying any physical correlation with metallicity) while keeping all other map-making steps identical, then remeasure the metallicity slopes; if the slopes persist after shuffling, the trend is robust to the variable-B_PAH recipe, whereas a significant flattening would indicate that the headline trend is partly built into the maps.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim relies on F335M_PAH maps constructed with Equations 3-6, where the PAH-correlated slope B_PAH is allowed to vary with environment (Section 3.5). Table 2 reports that B_PAH correlates with metallicity (Spearman rho=-0.28), sSFR (0.38), and [NII]/Halpha (-0.30). Since these quantities are mutually correlated, the low-metallicity regions that show elevated 3.3/11.3 and 3.3/7.7 ratios in Figure 7 are also the regions where the continuum-subtraction recipe itself changes. A larger B_PAH in low-metallicity regions does not merely trace PAH size; it changes the amount of F335M flux assigned to PAH emission through Equations 3-6. The quoted 2-5% differences between annular and galaxy-wide maps do not settle this, because both map versions use environmentally measured B_PAH values: the galaxy-wide version still lets B_PAH vary between galaxies, and low-metallicity galaxies like NGC 2835 and NGC 5068 have among the highest B_PAH in Table 1. Thus a metallicity trend could be imprinted on the final maps even if the intrinsic PAH population were unchanged. The FSPS comparison in Section 5.2 addresses radiation-field spectral effects on PAH emission, not the continuum-subtraction systematics, so it does not mitigate this concern. A control experiment with a fixed, spectroscopically calibrated B_PAH is needed before the reported slopes (-1.00+/-0.08 for 3.3/11.3 and -0.85+/-0.08 for 3.3/7.7) can be attributed to PAH size variation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript presents an empirical method to isolate PAH-correlated emission in the JWST NIRCam F335M filter using F300M and F360M continuum subtraction, and applies it to 19 PHANGS galaxies. The authors introduce an environment-dependent slope B_PAH in the F335M/F300M versus F360M/F300M color plane, show that this slope correlates with sSFR and [NII]/Halpha, and use it to construct continuum-subtracted F335M maps. They then measure 3.3/7.7 and 3.3/11.3 micron band ratios as functions of metallicity, galactocentric radius, and optical-line ratios. The central claim is that the 3.3 micron feature strengthens relative to 7.7 and 11.3 micron at low metallicity, implying smaller PAH populations, with radiation field spectrum playing a secondary, separable role.","tokens_in":43570,"tokens_out":7448,"duration_ms":62889,"significance":"The 3.3 micron PAH feature is a promising high-resolution ISM tracer, and this is one of the first systematic, sample-wide studies using the NIRCam medium-band method. The paper is well positioned: it includes 19 galaxies, forward-modeling tests of the slope-recovery bias, an explicit treatment of the environmental dependence of B_PAH, and comparisons to established PAH models (Draine et al. 2021) and stellar population synthesis. If the metallicity trend survives a control for the continuum-subtraction systematics, it will be an important confirmation of inhibited PAH growth at low metallicity. The main limitation is the entanglement between the variable B_PAH used to construct the maps and the environmental trends the maps are used to measure.","major_comments":[{"comment":"The F335M_PAH maps used to measure the metallicity trends are produced with a region-dependent B_PAH through Eqs. (3)-(6), and Table 2 shows that B_PAH itself correlates with 12+log(O/H) (rho = -0.28, p << 0.03). The 2-5% difference between the annular and galaxy-wide map versions does not control for this, because the galaxy-wide version still uses per-galaxy B_PAH values that vary with metallicity (Table 1 lists B_PAH ~1.94 for NGC 2835 and ~1.96 for NGC 5068, two of the low-metallicity galaxies that drive the trend in Fig. 7). The paper should re-measure the slopes in Fig. 7 using a fixed, spectroscopically calibrated B_PAH (e.g., the Lai et al. 2020 value or a sample-wide constant) to show that the reported slopes (-1.00 +/- 0.08 and -0.85 +/- 0.08) are not imprinted by the subtraction recipe.","section":"Section 3.5 / Table 2 / Fig. 7"},{"comment":"The method assumes that all PAH-correlated emission in F360M scales linearly with F335M_PAH, so that a single color-color slope B_PAH captures the 3.4 micron aliphatic feature, the 3.47 micron plateau, and the PAH continuum. Section 5.1 explicitly acknowledges that these components are not separated and may vary with environment. The forward-model test in Section 3.3 is built on the same linearity assumption (model F335M from F1130W and F360M from B_PAH), so it cannot validate that assumption. A spectroscopic check (e.g., NIRSpec or PAHFIT-based synthetic photometry like Appendix E, applied to representative pixels spanning the environmental range) is needed to confirm that the F360M PAH-correlated component is proportional to F335M_PAH and to quantify how much an environmental variation in the 3.4 micron or plateau features biases the subtracted maps.","section":"Section 3.3 / Eqs. (3)-(6)"},{"comment":"The argument that metallicity-dependent radiation field spectral changes are too small to explain the observed trend rests on an FSPS calculation that is not fully specified: the text states that a 0.135 dex change in FUV-optical slope 'corresponds to' a ~1.15% change in 3.3/11.3, but the mapping from the stellar population slope to the PAH band ratio is not shown or accompanied by uncertainties. Because this calculation is load-bearing for the 'primarily PAH size' interpretation, the manuscript should either provide the synthetic-photometry details and the Draine et al. (2021) grid interpolation, or downgrade the claim to 'the stellar-population radiation field change is modest compared to the observed dynamic range'.","section":"Section 5.2"}],"minor_comments":[{"comment":"There is a duplicate word in 'a new method to to isolate the PAH emission'.","section":"Section 6, item 1"},{"comment":"'Humbolt Research Award' should be 'Humboldt Research Award'.","section":"Acknowledgments"},{"comment":"The table header 'Spearman (rho,p)' followed by two numeric columns is ambiguous; the reader must infer which column corresponds to 3.3/11.3 and which to 3.3/7.7, so the header should explicitly label both ratio columns.","section":"Table 3"},{"comment":"Equation references in Appendix B ('Equation B1', 'Equation B2') are styled inconsistently with the numbered equations in the main text; unify the formatting.","section":"Appendix B / main text"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid empirical contribution with a well-documented method, but the stress-test concern about the variable B_PAH is real and lands on the central claim. I do not see circularity in the sense that the ratios are not fitted to the same data points that define them, but the subtraction recipe and the environmental driver are entangled through B_PAH. If the authors can show that the metallicity trend in Fig. 7 persists when B_PAH is fixed, I would support acceptance; without that control, the reported slope values are not yet robust."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the take: this is a competent and genuinely useful paper, and the central metallicity trend in 3.3/11.3 and 3.3/7.7 is probably real. It deserves a serious referee, but not a rubber stamp.\n\nWhat's actually new: the paper shows that the F335M/F300M vs F360M/F300M PAH-correlated slope B_PAH is not constant. It rises with sSFR and falls with [NII]/Halpha, and the paper gives usable fitting equations for continuum subtraction. That is a practical advance for the growing JWST 3.3 micron mapping program. The 19-galaxy sample, the forward-model bias correction, and the care taken to compare against Sandstrom et al. 2023 and Lai et al. 2020 are all real strengths. The comparison to Draine et al. 2021 models is honest, including the admission that the models do not fully reproduce the observed radiation-field trends.\n\nNow the soft spots. The stress-test worry is that because B_PAH is environmentally variable and correlated with metallicity (Table 2, rho = -0.28), the continuum-subtraction recipe could imprint the metallicity trend on the final maps. I checked the sign, and I don't think that specific mechanism works. Equations 3-6 imply that a larger B_PAH subtracts more continuum and reduces F335M_PAH. Low-metallicity regions have larger B_PAH, so the variable-B map should suppress the low-metallicity 3.3/11.3 ratio, not enhance it. In other words, the correction likely weakens the reported trend rather than creating it. That is not a full exoneration, though: the authors do not show a fixed-B_PAH control, and they should. The 2-5% map differences between annular and galaxy-wide versions do not settle the question, because both versions let B_PAH vary either within or between galaxies.\n\nMinor issues: the paper ships no code or reproducibility artifact, which is a real gap for a method paper even though the equations are detailed. The p-values reported as \"much less than 0.03\" are sloppy; exact values or a description of the resampling should be in the paper. The F1130W dust continuum is not subtracted, and the authors lean on an in-prep reference for the <20% estimate, which is acceptable but should be firmed up in revision.\n\nCitation pattern looks fine. The self-citations are to the direct methodological predecessors and are used correctly.\n\nBottom line: send it to a referee. After a moderate revision — fixed-B control, exact p-values, code release — this will be a standard reference for 3.3 micron PAH mapping.","headline":"Solid, useful methods-plus-survey paper; the metallicity trend is probably real, but the variable-B_PAH continuum subtraction and the trend are entangled enough that the authors should show a fixed-B control before the slopes are taken at face value.","tokens_in":44391,"tokens_out":3853,"would_cite":true,"duration_ms":40346,"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 paper shows that the 3.3 µm PAH feature strengthens at low metallicity because PAHs are smaller, not just hotter.","keywords":["polycyclic aromatic hydrocarbons","3.3 micron PAH feature","PAH size distribution","metallicity","radiation field hardness","JWST NIRCam","continuum subtraction","PAH band ratios"],"falsifier":"Take a low-metallicity galaxy or region and obtain a JWST NIRSpec spectrum across 3.2–3.6 µm alongside F300M/F335M/F360M photometry; if the aliphatic and plateau features do not scale with the 3.3 µm feature according to the paper's $B_{\\rm PAH}$ prescription, the continuum-subtracted F335M maps are biased and the metallicity trend could shift. Alternatively, if a radiation field model with fixed PAH sizes can reproduce the full increase in 3.3/11.3 with decreasing metallicity, the size interpretation would fail.","tokens_in":42876,"feed_emoji":"🌌","tokens_out":8245,"duration_ms":65692,"temperature":0.7,"pith_summary":"Using JWST NIRCam images of 19 nearby galaxies, this paper argues that the balance between the 3.3 µm PAH band and the 7.7 and 11.3 µm bands is set primarily by gas-phase metallicity, with the radiation field spectrum playing a smaller, separable role. The authors develop a continuum-subtraction procedure for the F335M filter and find that the slope describing PAH-correlated color depends on environment, most strongly on specific star formation rate. After subtracting the continuum they find both the 3.3/7.7 and 3.3/11.3 ratios rise as metallicity falls, across the whole sample. They interpret this as evidence that low-metallicity regions host smaller and more neutral PAH grains, consistent with inhibited growth rather than preferential destruction. If true, PAH band ratios are usable size diagnostics only when both metallicity and radiation field hardness are controlled.","feed_headline":"Low metal abundance shrinks galaxy PAH grains, JWST finds","feed_subtitle":"In 19 nearby galaxies, 3.3/11.3 µm ratios rise with falling metallicity, pointing to inhibited growth.","key_machinery":"The central object is the color-color slope $B_{\\rm PAH}$, the slope of $F335M/F300M$ versus $F360M/F300M$ in PAH-dominated pixels selected by $F300M/F1130W<0.1$ and $F1130W>3$ MJy/sr. This slope parameterizes how much PAH-correlated emission contaminates the F360M continuum band, and the paper measures it per galaxy and per 1.5 kpc region, finding values from about 1.6 to above 2 that track specific star formation rate and the optical line ratio $[\\mathrm{NII}]/\\mathrm{H}\\alpha$. The slope is inserted into a linear continuum relation derived from earlier F335M subtraction work to make continuum-subtracted F335M maps, which are then ratioed against stellar-subtracted F770W and F1130W maps. The second load-bearing ingredient is a grid of dust models with small, standard, and large PAH size distributions, low, standard, and high ionization, and several radiation field spectra; this grid lets the authors read ratio changes as PAH size changes rather than radiation temperature changes.","core_discovery":"The paper's central claim is that the relative strength of the 3.3 µm PAH emission compared to longer-wavelength PAH bands increases toward low gas-phase metallicity, and the dominant cause is a shift to smaller PAH populations rather than a change in the spectrum of the heating radiation. The observed slopes are $y=-1.00\\pm0.08\\,x+7.07\\pm0.69$ for $F335M_{\\rm PAH}/F1130W$ (3.3/11.3) and $y=-0.85\\pm0.08\\,x+5.97\\pm0.71$ for $F335M_{\\rm PAH}/F770W_{\\rm ss}$ (3.3/7.7), with $x=12+\\log(\\mathrm{O/H})$. Stellar population synthesis models are used to show that metallicity-driven changes in the radiation field spectrum would change the 3.3/11.3 ratio by only about 1.15%, too small to explain the trend. Binning by position in a Baldwin-Phillips-Terlevich (BPT) diagram separates a metallicity sequence in star-forming regions from a radiation-field-hardness sequence in LINER/AGN regions, so both effects operate, but the metallicity-to-size link dominates the 3.3/11.3 ratio.","pith_inferences":["A testable implication left implicit is that the weaker F360M PAH-correlated carriers (aliphatic 3.4 µm feature, 3.47 µm plateau, and PAH continuum) are suppressed in active star formation; NIRSpec spectra of high-sSFR PAH-dominated pixels should show a lower 3.4/3.3 ratio than quiescent pixels.","If $B_{\\rm PAH}$ correlates with specific star formation rate, applying a single global slope to distant galaxies with different sSFR distributions could mimic or hide a metallicity trend; the paper's [NII]/Hα and WISE-based prescriptions give a way to propagate that systematic.","The nearly vertical motion in 3.3/11.3 versus 7.7/11.3 space at low metallicity suggests the 3.3/11.3 ratio is a sensitive small-PAH fraction indicator even when 7.7/11.3 changes little; measuring it in galaxies below the metallicity of the current sample could test whether the trend saturates.","Because the same $B_{\\rm PAH}$ recipe is used to build the maps, any environmental variation in the F360M contamination is partly built into the reported ratios; comparing the annular and galaxy-wide map versions brackets this systematic."],"forward_implications":["The 3.3/11.3 µm ratio is not a pure PAH-size tracer: at fixed high metallicity it responds to radiation field hardness, while across the full metallicity range the size signal dominates.","Continuum-subtracted 3.3 µm maps should use an environment-dependent $B_{\\rm PAH}$; using a single fixed slope changes F335M PAH fluxes by up to 21% in high-specific-star-formation regions.","Lower-metallicity galaxies such as those reaching $12+\\log(\\mathrm{O/H})\\lesssim8.4$ should show enhanced 3.3/7.7 and 3.3/11.3 ratios, matching the outlying radial profiles of IC 5332, NGC 2835, and NGC 5068.","The metallicity trend supports the inhibited-growth scenario for PAH formation, in which the average PAH size decreases where carbon abundance limits growth, rather than destruction of small grains.","In AGN/LINER regions, the radiation field remains important: the 7.7/11.3 ratio shifts toward neutral PAHs or older stellar populations, and removing AGN hosts weakens that shift."],"supporting_citations":[{"why":"Supplies the original F335M continuum color relation and the method for isolating the 3.3 µm feature that this paper adapts.","marker":"T. S. Y. Lai et al. (2020)"},{"why":"Provides the color-color slope formalism for removing PAH contamination in F360M, which is re-measured as $B_{\\rm PAH}$ here.","marker":"K. M. Sandstrom et al. (2023)"},{"why":"Supplies the dust model grids with varying PAH size, charge, and radiation field used to interpret the band ratios.","marker":"B. T. Draine et al. (2021)"},{"why":"Provides the inhibited-growth model of metallicity-dependent PAH size that the observed trends are compared with.","marker":"C. M. Whitcomb et al. (2024)"},{"why":"Establishes the correlation between optical line ratios and PAH band ratios that this paper extends to 19 galaxies.","marker":"D. Baron et al. (2024)"},{"why":"Supplies the gas-phase metallicity maps used to bin the PAH ratios.","marker":"T. G. Williams et al. (2022)"},{"why":"Supplies the 1.5 kpc region properties such as sSFR and gas fractions used for environmental correlations.","marker":"J. Sun et al. (2022)"},{"why":"Provides the updated data reduction and F335M/F1130W maps used in the analysis.","marker":"R. Chown et al. (2025a)"}],"fun_headline_variants":["Metallicity shrinks PAH grains in nearby galaxies","Low metal means smaller PAHs, JWST finds","JWST: Metallicity dominates PAH size variation","PAH grain size tied to metal abundance","3.3/11.3 µm PAH ratio rises with low metallicity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The method assumes that the F335M/F300M versus F360M/F300M colors of PAH-dominated pixels follow a single straight line in each region whose slope $B_{\\rm PAH}$ fully describes the PAH-correlated and stellar emission, so any F360M emission component (aliphatic features, 3.47 µm plateau, PAH continuum) that does not scale linearly with the 3.3 µm PAH emission is correctly removed.","fun_headline_variants_meta":{"raw":{"variants":["Metallicity shrinks PAH grains in nearby galaxies","Low metal means smaller PAHs, JWST finds","JWST: Metallicity dominates PAH size variation","PAH grain size tied to metal abundance","3.3/11.3 µm PAH ratio rises with low metallicity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000447,"raw_usage":{"total_tokens":2367,"prompt_tokens":1166,"completion_tokens":1201,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":782,"completion_tokens_details":{"reasoning_tokens":1120}},"tokens_in":782,"tokens_out":1201,"duration_ms":10782,"temperature":1.0,"reasoning_tokens":1120,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T16:14:43.528609+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a low-metallicity galaxy or region and obtain a JWST NIRSpec spectrum across 3.2–3.6 µm alongside F300M/F335M/F360M photometry; if the aliphatic and plateau features do not scale with the 3.3 µm feature according to the paper's $B_{\\rm PAH}$ prescription, the continuum-subtracted F335M maps are biased and the metallicity trend could shift. Alternatively, if a radiation field model with fixed PAH sizes can reproduce the full increase in 3.3/11.3 with decreasing metallicity, the size interpretation would fail.","supporting_citations":[{"cited_title":"PHANGS-ML: dissecting multiphase gas and dust in nearby galaxies using machine learning","cited_arxiv_id":"2402.04330","evidence_quote":"Establishes the correlation between optical line ratios and PAH band ratios that this paper extends to 19 galaxies."}],"review_version":1}