{"id":"d5f93255-2e0f-437c-9133-db216cf368ac","arxiv_id":"2501.02085","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"JWST MIRI observations of HD and H2 in protostellar outflows yield gas-phase D/H values that vary by a factor of about 4 among nearby sources and are generally a factor of 2 to 4 below Galactic chemical evolution model predictions.","lead":"Using JWST mid-infrared spectra of 10 protostellar outflows, the authors measure gas-phase deuterium-to-hydrogen ratios from HD and H2 emission lines. The values vary by a factor of about four between nearby sources and sit below Galactic chemical evolution predictions, suggesting deuterium may be locked in dust grains rather than fully in gas.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Aperture-filling assumption in §3.3 is load-bearing: HD is more compact than H2 in the maps, so [D/H] is systematically underestimated and position-dependent beam dilution can create the claimed factor-of-4 scatter.","rationale":"The paper's central claim has two parts: a factor-of-4 variation in gas-phase D/H among low-mass sources at similar Galactocentric radii, and a factor-of-2–4 deficit relative to Galactic chemical evolution models. Both parts rest on the derived column density ratio N_HD/N_H2. The reader's weakest assumption, the equal aperture-filling of H2 and HD (Section 3.3), directly undermines this ratio. The maps in the paper are internally probative: HD is morphologically associated with bright shock knots (high-J H2, [Fe I], [S I]), while the lower-J H2 lines used to define the warm component are extended. If HD is more compact than H2 in the 1\" apertures, the HD column density is diluted relative to H2, lowering [D/H] and introducing scatter that tracks the local compactness. This is a concrete, correctable systematic, and the paper's own statement that the shared solid angle cancels is incorrect when the emitting regions differ. I agree with the reader that this makes the absolute values and the inter-source scatter conditional. The paper also contains an internal inconsistency in the applied correction factor (2.45 in §3.4 vs 2.54 in Table 2 and Fig. 5), but that is a normalization detail affecting the comparison to GCE models by only ~4%, not the variation claim. The proposed test using the existing maps would settle whether beam dilution is the dominant source of the reported scatter and low values. Until then, the conditional verdict is appropriate.","tokens_in":37958,"tokens_out":8612,"duration_ms":84451,"concrete_test":"Recompute [D/H] for all detected apertures using a matched-source-size correction: for each aperture, measure the effective emitting solid angle of H2 S(1), H2 S(4), and HD R(6) from the 1\"-convolved integrated maps (e.g., by fitting a 2D Gaussian or computing the half-flux radius), then convert the observed fluxes to intensities using these per-line solid angles rather than the common 1\" aperture. Redo the rotation-diagram fits and compare the resulting [D/H] values and scatter with Table 2. If the factor-of-4 variation among low-mass sources collapses or the mean [D/H] shifts significantly upward, the aperture-filling assumption is the cause of the central claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing assumption is in Section 3.3: 'we assume that the H2 or HD emission fills the aperture' and hence 'the assumed emission solid angle is the same for both H2 and HD and thus there is no impact on any column density ratios or the derived [D/H].' This is not safe if the angular extent of HD is smaller than that of H2. The integrated line maps (Figs. 1, 2, C.1–C.8) show HD R(4)-R(6) concentrated in bright jet knots and bow shocks, while H2 S(1)-S(4) is extended over the outflow cavity. In the aperture extraction, the flux of each line is converted to intensity as F/Ω_ap, where Ω_ap is the solid angle of the 1\" radius aperture. If HD emission arises from an area Ω_HD < Ω_ap, the true intensity is F_HD/Ω_HD > F_HD/Ω_ap, so the derived HD column density is underestimated by a factor Ω_HD/Ω_ap. Since H2 is more extended, it suffers less from this beam dilution. The resulting [D/H] is therefore systematically too low by a position-dependent factor. Because Ω_HD/Ω_ap varies between apertures and sources, this can produce the observed factor-of-4 scatter without any real change in gas-phase D/H. The strong correlation of HD R(6) with the high-excitation H2 S(7) line (ρ=0.98, Fig. 6) rather than with the lower-excitation S(1) (ρ=0.79) further indicates that HD traces the compact hot knots, not the extended warm gas; using the warm H2 column in the denominator biases the ratio. The claimed factor-of-2–4 deficit relative to GCE models could thus be partly an artifact of this beam-dilution effect. The paper does not test the filling-factor assumption despite having the spatial maps needed to do so.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses JWST/MIRI MRS observations of ten protostellar outflows (five low-mass, five high-mass) to measure gas-phase [D/H] from mid-IR H2 and HD rotational lines. It extracts 1\"-radius aperture spectra, builds integrated line maps, fits two-temperature H2 rotation diagrams and single-temperature HD rotation diagrams, applies extinction and ortho-to-para corrections, and multiplies the derived LTE [D/H] by a correction factor for non-LTE excitation and chemical conversion taken from Bertoldi et al. 1999. The key results are that [D/H] varies by up to a factor ~4 among the low-mass sources and that most measured values are ~(0.1-1.8) x 10^-5, i.e. a factor 2-4 below Galactic chemical evolution predictions and below local UV absorption values. The authors interpret the scatter and low values as evidence for deuterium depletion onto carbonaceous dust grains, possibly with release in shocks, and report tentative support from enhanced [D/H] where gas-phase Fe is enhanced in HH 211.","tokens_in":38314,"tokens_out":8599,"duration_ms":82981,"significance":"The observational material is valuable: this is one of the first JWST studies to map HD emission in protostellar outflows across Galactocentric radii of 4-11 kpc, and the paper provides full tabulated line fluxes and rotation-diagram fit parameters that will be useful to the community. If the derived [D/H] values are robust, the conclusion that molecular-line measurements trace gas-phase rather than total deuterium would be an important constraint for Galactic chemical evolution studies and for the deuterium-depletion debate. The strengths of the paper include a diverse sample, careful Gaussian line fitting, and explicit handling of extinction and ortho-to-para corrections. However, the central quantitative claims rest on the assumptions that H2 and HD fill the same aperture and that the warm H2 component is the correct denominator for the HD column; neither assumption is established by the data as presented, and the manuscript contains an internal contradiction in the HH 211 trend that is used to support the depletion interpretation.","major_comments":[{"comment":"The aperture-filling assumption is load-bearing. The text states that 'the assumed emission solid angle is the same for both H2 and HD and thus there is no impact on any column density ratios or the derived [D/H]'. This is only true if H2 and HD have the same spatial extent. The integrated-line maps (Figs. 1, 2, C.1-C.8) show HD concentrated in bright knots and bow shocks, while H2 emission is extended over the outflow cavities. For a compact HD source with true solid angle Ω_HD < Ω_ap, converting the measured flux to intensity as F/Ω_ap underestimates I_HD, and hence N_HD, by a position-dependent factor Ω_HD/Ω_ap, while the more extended H2 suffers less dilution. The resulting [D/H] is therefore systematically low by a source- and position-dependent factor, which can produce both the factor-of-4 scatter and the low absolute values without any real variation in gas-phase D/H. This issue must be addressed, for example by measuring the emitting solid angles of HD and H2 separately, by restricting the analysis to a common compact region, or by propagating the beam-dilution ratio as a correlated systematic uncertainty on every [D/H] value.","section":"§3.3"},{"comment":"The use of the warm H2 column as the denominator is not justified for the gas traced by HD. [D/H] is computed as 0.5 N_HD/N_warm, where N_warm corresponds to the 400-900 K H2 component determined from S(1)-S(4). However, the flux correlations in Fig. 6 show that HD R(6) is most strongly correlated with the high-excitation H2 S(7) line (ρ=0.98) and less with S(1) (ρ=0.79), indicating that the HD emission traces the compact hot component rather than the extended warm component. If the HD-bearing gas is not well mixed with the warm gas used in the denominator, the derived ratio is not a measure of D/H in a single gas parcel and will vary with the local temperature structure. Please derive [D/H] also using the hot H2 component or otherwise demonstrate that the warm H2 column is the appropriate reservoir for the HD traced by the observed R(4)-R(6) transitions.","section":"§3.4 and Fig. 6"},{"comment":"The HH 211 trend used to support the dust-depletion interpretation appears to be contradicted by the tabulated values. The text states that in HH 211 [D/H] is 'robustly lower in the bow-shock positions (apertures 1-3) than the jet positions (apertures 4-6)', and later that the data are consistent with an 'increased [D/H] in the inner jet (apertures 4-6) relative to the bow-shocks (apertures 1-3)'. Table 2 lists detected values of [D/H] = (1.30±0.49, 1.74±0.14, 1.78±0.47) x 10^-5 for apertures 1-3, while apertures 4-6 are upper limits that are mostly lower. This is the opposite trend to the one stated in the text and abstract. Please verify the aperture classification and either correct the text/abstract or explain how the stated trend is obtained.","section":"§4, §4.1, and Table 2"},{"comment":"The correction factor for non-LTE excitation and chemical conversion is both internally inconsistent and applied without a propagated uncertainty. The text says the LTE [D/H] estimates are multiplied by a factor of 2.45, whereas the Table 2 caption and Fig. 5 caption state a factor of 2.54. More substantively, the factor is imported from Bertoldi et al. (1999) for the Orion OMC-1 outflow, where the exact value depends on shock density, temperature, and dissociation fraction. Applying a single factor to all sources and positions sets the absolute scale for the claim that the observed [D/H] is a factor of 2-4 below GCE predictions. The expected spread of the correction factor should be propagated into the [D/H] uncertainties, or at least a plausible range should be quoted alongside the central values.","section":"§3.4, Table 2, and Fig. 5"}],"minor_comments":[{"comment":"The bullet states that HD column densities are '∼ 105 orders magnitude smaller' than H2; this should read 'five orders of magnitude'.","section":"Summary bullet, §6"},{"comment":"The phrase 'polycyclic aromatic hydrocarbons' is misspelled as 'polcyclic aromatic hydrocarbons' in the Introduction; please correct the typo.","section":"Abstract and Introduction"},{"comment":"There are several instances of broken spacing such as 'e ffect' and 'V oort' that should be corrected in the final typeset version.","section":"Throughout"},{"comment":"The correction factor is quoted as 2.45 in the main text and 2.54 in the table and figure captions; the inconsistency should be resolved in revision.","section":"Fig. 5 and Table 2"}],"recommendation":"major_revision","confidential_remarks":"The aperture-filling issue is the crux of the paper. If the authors can show from the maps that H2 and HD have similar intrinsic emitting sizes after PSF correction, or can provide corrected values using measured source sizes, the conclusions may survive; otherwise the factor-of-4 scatter and the low absolute [D/H] values may be artifacts. The internal contradiction in the HH 211 trend should be resolved before resubmission, since it currently undermines the main depletion-relevant claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, this is genuinely new data: the first systematic JWST/MIRI MRS census of HD in protostellar outflows, including a detection at 11 kpc and resolved maps of HD in low-mass outflows. The line fitting and rotation-diagram work is careful, and the upper limits are honestly presented. Second, the central quantitative claims—factor-of-4 variations and a factor 2-4 deficit relative to GCE models—are not supported, because the paper's aperture-filling assumption is contradicted by its own maps.\n\nThe problem sits in Section 3.3. The paper assumes H2 and HD each fill the 1-arcsecond extraction aperture, so the same solid angle is used for both and the flux ratio F_HD/F_H2 becomes the column density ratio. But the maps (Figs. 1, 2, C.1-C.8) show HD concentrated in bright knots and bow shocks, while low-J H2 fills the outflow cavity. When HD is more compact than the aperture, its true intensity is F_HD/Omega_HD, larger than F_HD/Omega_ap, so the derived HD column is too low. H2 is less affected, so [D/H] is systematically low by a position-dependent factor. That alone can manufacture the claimed scatter and the deficit relative to expectations. The strong correlation of HD R(6) with H2 S(7) (rho = 0.98) rather than S(1) (rho = 0.79) reinforces that HD traces the compact hot knots, not the warm extended component used as the denominator. The paper states that the solid-angle choice cannot impact any column density ratios, which is only true if both species actually fill the aperture; the maps show they do not. The same bias makes the inner-Galaxy upper limits (e.g. IRAS 18089-1732 at 0.71e-5) over-strict, weakening the claimed radial gradient.\n\nSmaller issues: the text says the Bertoldi correction factor is 2.45, while Table 2 and figures use 2.54; and no uncertainty is propagated from that factor, though the paper does flag it in the Discussion. The HH 211 non-detections in the jet (apertures 4-6) are also interpreted as locally higher D/H, but those apertures sit on the compact jet core, where the same filling bias is strongest.\n\nWhat is good: the paper is a solid observational step, presents a valuable dataset, is honest about many caveats, and the qualitative suggestion that outflows may not trace total D/H is plausible. But the quantitative numbers should not be used as they stand. A re-analysis using source sizes measured from the maps, or at least a quantitative sensitivity test, is needed. This deserves serious peer review: the data are new, the maps permit the correction, and the question matters. I would expect a major revision, not acceptance as is.","headline":"New JWST/MIRI HD census with a load-bearing aperture-filling problem that its own maps contradict.","tokens_in":39032,"tokens_out":6310,"would_cite":false,"duration_ms":57583,"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":"JWST observations of protostellar outflows show gas-phase deuterium varying by a factor of four and running 2–4 times below predicted total D/H, suggesting much of the Galaxy's deuterium is locked in carbonaceous dust.","keywords":["astrochemistry","stars: formation","Galaxy: abundances","ISM: jets and outflows","infrared: ISM","deuterium abundance","protostellar outflows","JWST MIRI/MRS"],"falsifier":"Re-extract the spectra with apertures matched to the HD-bright knots instead of a fixed 1-arcsecond aperture and check whether [D/H] climbs toward the local-disk value; if it does, the low values are a filling-factor artefact rather than dust depletion. Alternatively, observe a supernova remnant with MIRI/MRS and test whether complete grain destruction raises gas-phase [D/H] to the level predicted by Galactic chemical evolution models.","tokens_in":37705,"feed_emoji":"🔭","tokens_out":10534,"duration_ms":94410,"temperature":0.7,"pith_summary":"This paper reports JWST mid-infrared measurements of the gas-phase deuterium-to-hydrogen ratio in protostellar outflows, obtained from H$_2$ and HD rotational lines observed with MIRI/MRS toward ten young stellar objects at Galactocentric radii from 4 to 11 kpc. Its central finding is that gas-phase [D/H] varies by up to a factor of four even among low-mass sources formed from gas expected to share nearly the same total deuterium abundance, and that detected values and useful upper limits mostly sit at or below about $1.0\\times10^{-5}$, a factor of 2–4 below the total [D/H] inferred from local ultraviolet absorption lines and predicted by Galactic chemical evolution models. The authors conclude that these observations do not trace the total deuterium reservoir, and propose that a substantial part of the Galaxy's deuterium is locked up in carbonaceous dust grains, released only where shocks destroy the grains. A sympathetic reader would care because the result turns protostellar outflows into a new test bed for deuterium depletion and, with modelling, a possible probe of Galactic chemical evolution.","feed_headline":"JWST finds gas deuterium in outflows 2-4x below predictions","feed_subtitle":"The measured [D/H] varies by a factor of four, hinting that deuterium is stored in carbon dust grains.","key_machinery":"The analysis rests on rotation diagrams built from the rotational lines H$_2$ S(1)–S(8) and HD R(4)–R(9) detected with the JWST MIRI/MRS medium-resolution spectrometer. For optically thin, LTE gas, plotting $\\ln(N_u/g_u)$ against $E_u/k$ gives the column density from the intercept and the excitation temperature from the slope, and the ratio $N_{\\rm HD}/(2N_{\\rm H_2})$ yields [D/H]. The H$_2$ diagrams are fitted with a warm plus hot component and an ortho-to-para ratio correction, the same extinction correction derived from the H$_2$ S(1)–S(4) lines is applied to HD, and the LTE [D/H] is multiplied by 2.45 to account for non-LTE excitation and for chemical conversion of HD to atomic D, following the cited Orion outflow analysis. The fixed 1-arcsecond extraction aperture is assumed to be equally filled by H$_2$ and HD, so the ratio is taken to be independent of source size.","core_discovery":"On the paper's own terms, the discovery is that gas-phase [D/H] derived from HD and H$_2$ rotation diagrams in protostellar outflows is systematically low and variable: detected values run from about $0.12\\times10^{-5}$ to $2.27\\times10^{-5}$ after the standard correction, with a factor-of-four spread among low-mass sources that cannot be explained by astration, and most measurements fall a factor of 2–4 below the total [D/H] expected from Galactic chemical evolution. HD emission is concentrated in the same high-velocity jet knots and bow shocks as high-excitation H$_2$, [S I], and [Fe I] lines, and in the HH 211 outflow the inner-jet positions with higher gas-phase iron also show higher [D/H]; this spatial pattern is the paper's tentative evidence that deuterium sequestered in dust returns to the gas when shocks destroy grains. The paper concludes that rotational-line observations of outflows are sensitive only to gas-phase deuterium, not the total reservoir, and that significant depletion onto carbonaceous dust is the plausible explanation.","pith_inferences":["An untested consequence of the shared-solid-angle assumption is that any difference in how fully H$_2$ and HD fill the aperture translates directly into an offset in [D/H]; re-extracting on apertures matched to the HD knots would test whether part of the factor-of-four scatter is spatial resolution rather than dust depletion.","If deuterium depletion onto carbonaceous grains is real, the low [D/H] values from ultraviolet H I/D I and D I/O I absorption may reflect the same grain reservoir, which would change the reading of the local-disk abundance debate.","A clean extension would be MIRI/MRS observations of a supernova remnant, where complete grain destruction should push gas-phase [D/H] up toward the Galactic chemical evolution prediction; the paper notes an earlier HD detection in such a remnant.","The strong correlation between HD R(6) line flux and high-excitation H$_2$ and [S I] suggests that HD could serve as a tracer of dust-destructive shocks, making deeper [D/H] maps a way to map dust processing across an outflow."],"forward_implications":["Most gas-phase [D/H] values from this paper and earlier molecular-line work sit at or below $1.0\\times10^{-5}$, so any observationally grounded estimate of total deuterium in the disk must budget for deuterium hidden in dust.","Because the spatial pattern in HH 211 ties higher [D/H] to positions with higher gas-phase iron, stronger shocks that destroy more dust should show systematically higher gas-phase [D/H] than quiescent outflow gas.","The detection at 11 kpc and the constraining upper limits in the inner Galaxy mean that, once depletion is modelled, JWST can sample [D/H] across the Galactic disk rather than only in the local neighbourhood.","The 2.45 correction for non-LTE excitation and chemical conversion is applied to every value; without it the discrepancy with Galactic chemical evolution models would be even larger."],"supporting_citations":[{"why":"Supplies the non-LTE and chemical-conversion correction factor (2.45) applied to all [D/H] values and the earlier Orion outflow HD detection this work extends.","marker":"Bertoldi et al. 1999"},{"why":"Establishes the anti-correlation between gas-phase [D/H] and refractory-element depletion along UV sightlines that motivates the dust-depletion interpretation.","marker":"Linsky et al. 2006"},{"why":"Provides the depletion-corrected local-disk total [D/H] of $2.0\\times10^{-5}$ used as the reference for comparing molecular-line values.","marker":"Prodanović et al. 2010"},{"why":"Gives the primordial [D/H] of $2.58\\times10^{-5}$ from Big Bang nucleosynthesis and the cosmic microwave background used in the comparison figures.","marker":"Cyburt et al. 2016"},{"why":"Supplies the Galactic chemical evolution model gradient that the measured values fall 2–4 times below.","marker":"Romano et al. 2006"},{"why":"Provides the cosmological zoom-in simulation prediction for total [D/H] across a Milky Way-type disk used for comparison.","marker":"van de Voort et al. 2018"},{"why":"Supplies the HH 211 shock modelling and gas-phase iron abundances that link higher [D/H] to dust destruction.","marker":"Caratti o Garatti et al. 2024"},{"why":"Contributes the earlier ISO HD detection in the Orion Bar and the comparison [D/H] value from a photodissociation region.","marker":"Wright et al. 1999"}],"fun_headline_variants":["JWST: Outflow deuterium dips 2-4x below predictions","Deuterium in protostellar outflows hints at dust sequestration","Gas-phase deuterium varies by 4x in JWST outflow study","JWST reveals deuterium might hide in carbon dust","Outflow deuterium measurements challenge galactic evolution models"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that H$_2$ and HD emission fill the same 1-arcsecond extraction aperture to the same degree, so their column-density ratio is unaffected by source size; if HD is actually more compact than H$_2$, the derived [D/H] is systematically low and the scatter between positions could be partly artificial.","fun_headline_variants_meta":{"raw":{"variants":["JWST: Outflow deuterium dips 2-4x below predictions","Deuterium in protostellar outflows hints at dust sequestration","Gas-phase deuterium varies by 4x in JWST outflow study","JWST reveals deuterium might hide in carbon dust","Outflow deuterium measurements challenge galactic evolution models"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000223,"raw_usage":{"total_tokens":1574,"prompt_tokens":1179,"completion_tokens":395,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":795,"completion_tokens_details":{"reasoning_tokens":308}},"tokens_in":795,"tokens_out":395,"duration_ms":4020,"temperature":1.0,"reasoning_tokens":308,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:15:12.324535+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-extract the spectra with apertures matched to the HD-bright knots instead of a fixed 1-arcsecond aperture and check whether [D/H] climbs toward the local-disk value; if it does, the low values are a filling-factor artefact rather than dust depletion. Alternatively, observe a supernova remnant with MIRI/MRS and test whether complete grain destruction raises gas-phase [D/H] to the level predicted by Galactic chemical evolution models.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the non-LTE and chemical-conversion correction factor (2.45) applied to all [D/H] values and the earlier Orion outflow HD detection this work extends."},{"cited_title":"H., Fields , B","cited_arxiv_id":null,"evidence_quote":"Gives the primordial [D/H] of $2.58\\times10^{-5}$ from Big Bang nucleosynthesis and the cosmic microwave background used in the comparison figures."},{"cited_title":"2006, , 369, 295","cited_arxiv_id":null,"evidence_quote":"Supplies the Galactic chemical evolution model gradient that the measured values fall 2–4 times below."},{"cited_title":"2018, , 477, 80","cited_arxiv_id":null,"evidence_quote":"Provides the cosmological zoom-in simulation prediction for total [D/H] across a Milky Way-type disk used for comparison."},{"cited_title":"P., Kavanagh , P","cited_arxiv_id":null,"evidence_quote":"Supplies the HH 211 shock modelling and gas-phase iron abundances that link higher [D/H] to dust destruction."},{"cited_title":"M., van Dishoeck , E","cited_arxiv_id":null,"evidence_quote":"Contributes the earlier ISO HD detection in the Orion Bar and the comparison [D/H] value from a photodissociation region."}],"review_version":1}