{"id":"a2e798b8-dfc8-4db3-8e23-b0384e44ab95","arxiv_id":"2411.17121","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In seven of twelve mapped massive star-forming regions, the ammonia deuterium fraction decreases with increasing rotational temperature, and the highest fractionation is offset from the ammonia peak in nearly every source.","lead":"Astronomers mapped deuterated ammonia and normal ammonia in 12 massive star-forming regions and found that, within most individual clouds, the fraction of ammonia that is deuterated drops where the gas is warmer. The maps also show the most deuterated gas sits away from the ammonia peak, which the authors explain with temperature, gas-phase depletion, and carbon monoxide release from grains.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Beam-filling assumption for NH2D may produce the reported anticorrelation and peak offset; needs validation with resolved observations.","rationale":"The paper is a careful observational study with openly stated limitations, and I read it in good faith as an attempt to use single-dish mapping to compare NH2D and NH3 spatial distributions across a dozen massive star-forming regions. The strongest claims are that Dfrac(NH3) decreases with increasing Trot within seven sources and that the Dfrac peak is offset from the NH3 peak. Both claims are based on maps of N(NH2D)/N(NH3) computed under the explicit assumptions that the beam filling factor is unity (Section 2.3) and that the NH3 rotation temperature equals the NH2D excitation temperature (Section 3.2). The reader's verdict identified these same premises as the load-bearing assumptions, and I agree. My stress-test focused on the filling-factor assumption because it is the least constrained: the data do not provide any way to measure the intrinsic size of the NH2D-emitting region, and the three telescopes involved have different beam sizes, so the relative dilution is not even constant within a map after regridding. If NH2D is more compact where the temperature is higher, the systematically underestimated N(NH2D) near NH3 peaks would naturally produce the reported anticorrelation and offset. The 15% error quoted for the Tex assumption is not the dominant uncertainty; filling factors can easily change the derived ratio by a factor of several. The proposed test is feasible because several of these sources have interferometric follow-up data, and even a two-source comparison would show whether the single-dish spatial trends survive at higher resolution. Until such a check is performed, the paper's central interpretation remains conditional rather than established. I therefore keep the reader's CONDITIONAL verdict unchanged; no adjustment is needed.","tokens_in":38284,"tokens_out":3579,"duration_ms":35190,"concrete_test":"Select two or three sources (e.g., G035.19-00.74, G081.87+00.78, G111.54+00.77) with archival ALMA or VLA interferometric data covering NH2D 1_11-1_01 and NH3 (1,1)/(2,2). Measure the deconvolved source sizes of NH2D and NH3, compute the beam filling factors at the single-dish resolution, and recompute the Dfrac maps with the ratio f(NH2D)/f(NH3) applied as a correction. If the corrected maps still show the anticorrelation and peak offset, the claims are robust; if the trends weaken or vanish, they are artifacts of beam dilution. A complementary, purely analytic check: run a RADEX non-LTE grid for the NH2D 1_11-1_01 transition over plausible density and temperature ranges, and determine the filling-factor ratio needed to reproduce the observed Dfrac-Trot trend; if f(NH2D)/f(NH3) must be far below unity at the NH3 peaks, the single-dish result is not interpretable without resolved data.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claims depend on converting observed intensity ratios into column density ratios, and the conversion assumes a beam filling factor of unity (Section 2.3) for all lines. NH2D and NH3 were observed with different telescopes and beam sizes (IRAM 30 m at 28.6 arcsec versus GBT at 31.8 arcsec or Effelsberg at 42.5 arcsec) and were smoothed to a common beam, but the intrinsic source sizes are never measured. If NH2D emission is more compact than NH3 emission, the main-beam temperature of NH2D is diluted relative to NH3, so the derived N(NH2D)/N(NH3) is underestimated. The dilution factor depends on position and can vary systematically with temperature: near the NH3 peak, where the gas is hotter and denser, NH2D may be more compact, making the derived Dfrac artificially low at the peak. This would create both the apparent anticorrelation between Dfrac and Trot and the spatial offset of the Dfrac peak without any real chemical gradient. The paper's stated 15% error for assuming Tex(NH2D)=Trot (Section 3.2) does not address filling-factor dilution, which can be substantially larger and is not bounded by the analysis. The paper explicitly flags the unity filling-factor assumption but does not quantify its impact, so the most load-bearing unverified step remains the one that turns observed line ratios into the physical quantity used for every conclusion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"Using IRAM 30 m maps of ortho-NH2D 1_11^s-1_01^a together with GBT and Effelsberg maps of NH3(1,1) and NH3(2,2), the paper derives the deuterium fractionation Dfrac(NH3) at roughly 30-42 arcsecond resolution in 12 late-stage massive star-forming regions. The authors report a decreasing Dfrac(NH3) with increasing rotational temperature within seven individual sources, no clear global trend when all 12 sources are combined, and a spatial offset between the NH3 peak and the position of highest Dfrac(NH3). They interpret these results with gas-grain chemical model calculations over a grid of temperature, density, cosmic-ray ionization rate, and time, concluding that temperature and environmental evolution drive the NH2D distribution.","tokens_in":51,"tokens_out":8798,"duration_ms":141878,"significance":"If the central claims survive scrutiny, this paper provides a comparatively large single-dish mapping sample connecting NH2D/NH3 to temperature within individual regions, extending earlier interferometric case studies. The authors make a useful effort to tabulate per-position rotational temperatures, column densities, and Dfrac values, and they use an established public chemical network with a documented parameter grid rather than fitting the model to the observed Dfrac values. They also state their key assumptions openly, including the unity beam filling factor and the assumed equality between Trot and Tex(NH2D). The main risk is that the conversion from observed line intensities to the physical ratio N(NH2D)/N(NH3), which underlies every conclusion, depends on assumptions about source size and excitation that are not yet quantitatively bounded.","major_comments":[{"comment":"The assumption that the beam filling factor is unity for all lines (Section 2.3) is load-bearing because NH2D and NH3 were observed at different native resolutions and no source-size measurement is presented. If NH2D emission is more compact than NH3 emission, particularly near the NH3 peak, the NH2D main-beam temperature is diluted relative to NH3, and the derived Dfrac(NH3) would be artificially low at the peak. This mechanism alone could generate both the within-source anticorrelation in Figure 4 and the spatial offset between Dfrac and NH3 peaks in Figures E1/E2. The statement in Section 3.2 that the Tex(NH2D)=Trot assumption gives at most a 15% error for Tex>10 K does not cover filling-factor dilution or subthermal excitation below 10 K, both of which can be substantially larger. Please quantify the sensitivity of the results to a range of plausible source-size and excitation profiles, or explicitly restrict the claims to quantities that are robust under such variations.","section":"§2.3 and §3.2"},{"comment":"The Dfrac values for G031.28+00.06 are internally inconsistent by roughly an order of magnitude. Table C1 lists Dfrac(NH3)×100 in the range 0.062-0.13, while Section 3.4.3 states a range from 0.009±0.001 to 0.013±0.002. Computing the ratio from the tabulated N(NH3) and N(NH2D) columns in Table C1, for example 0.69×10^15 cm^-2 and 0.81×10^13 cm^-2 in the first row, gives Dfrac≈1.17×10^-2, i.e., about 1.17%, not 0.07%. Because G031.28+00.06 is one of the seven sources used to support the anticorrelation claim, the tables and text must be reconciled before the statistical conclusions can be accepted.","section":"Table 3, Table C1, §3.4.3"},{"comment":"The classification of the seven sources as showing a clear anticorrelation between Dfrac(NH3) and Trot(2,2;1,1) is not supported by any quantitative measure. No correlation coefficient, significance test, or explicit definition of a 'clear' trend is given, and some sources assigned to the 'no' group have comparable dynamic ranges, such as G015.03-00.67 with Dfrac varying from 0.005 to 0.012 over Trot 19.5-24.8 K and G081.75+00.59 with Dfrac varying from 0.022 to 0.068. Additionally, the rows in Table C1 are averages over 3×3 or 2×2 pixel boxes, so adjacent rows are not independent samples; the effective number of independent data points is smaller than the number of rows. Please provide a correlation statistic or a pre-specified criterion and account for spatial autocorrelation.","section":"§4.1, Table 3"},{"comment":"The paper's offset claim is stated in inconsistent strengths. The abstract says the region of highest Dfrac is 'offset from the NH3 peak in each source'; Section 3.3 states this for the six sources in which the NH3 and NH2D morphologies differ; Section 4.4 notes that in G075.76+00.33 the highest Dfrac and the highest rotation temperature occur at the same position; and Conclusion item 2 restricts the offset to the seven sources with different morphologies. The abstract and the body need to agree on the qualified version of this claim.","section":"Abstract; §3.3; §4.4; §5"}],"minor_comments":[{"comment":"The uncertainty on the lower end of the Dfrac range is printed as '0.018 ±0.2'; this is presumably a typo for ±0.002 and should be corrected.","section":"§3.4.7"},{"comment":"The caption says 'The cosmic-ray ionization rate is set to nH2=10^6 cm^-3'; this should read 'The proton density is set to nH=10^6 cm^-3'.","section":"Figure 8 caption"},{"comment":"The text states that the nine GBT sources were regridded to a beam size of 30 arcsec, while Section 2.1.2 gives the GBT beam as 31.8 arcsec at 23 GHz; please clarify whether a 30 arcsec Gaussian beam or the actual 31.8 arcsec beam was adopted.","section":"§2.3"},{"comment":"The column heading 'Dfrac(NH3)×100' is used inconsistently with the values in Table C1 and with the body text; define the unit once and apply it uniformly across all tables and text.","section":"Table 3"}],"recommendation":"major_revision","confidential_remarks":"The reader's stress-test concern about the beam filling factor lands: it is the most serious obstacle to accepting the observational claims, because it can mimic both the anticorrelation and the spatial offset. The numerical inconsistency for G031.28+00.06 is also a concrete, checkable problem that should be resolved by the authors before the statistical claims are evaluated further. I do not see grounds for rejection at this stage, since the dataset and the model grid are useful and the beam-filling issue could in principle be addressed with additional analysis or by weakening the claims appropriately. I would recommend sending the revised manuscript to a referee with radiative-transfer expertise in addition to the present referee."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this is the first resolved single-dish comparison of NH2D and NH3 across a dozen massive star-forming regions, and it finds what earlier single-pointing surveys missed: within individual sources, the ammonia deuterium fraction tends to drop as the rotation temperature rises, and the Dfrac peak is usually offset from the NH3 peak. The tabulated data support these qualitative trends. But the quantitative strength of the claim is not as solid as the abstract suggests.\n\nWhat the paper does well: it exploits archival NH3 maps and its own IRAM NH2D maps to produce a multi-region dataset, with full tables of column densities and temperatures. The authors are candid about the main limitations—only one NH2D line, no measured excitation temperature, no early-stage sources, no ionization tracers. The chemical model is a published network run across a reasonable parameter grid, not fitted to the observed Dfrac values. The negative result across the full sample is reported honestly.\n\nWhere it gets soft: the Dfrac maps assume a beam filling factor of unity for all lines, and the analysis never measures source sizes. If NH2D emission is more compact than NH3, particularly near warm peaks, the derived Dfrac is underestimated there, which would create both an anticorrelation with Trot and an apparent spatial offset without any real chemistry. The paper's 15% error estimate for Tex(NH2D)=Trot does not cover this, and the stress test is right that it is the most load-bearing unverified step. Also, the seven-source anticorrelation is identified by eye and never quantified with a correlation coefficient or a fitting statistic; the five \"no trend\" sources are explained away as resolution-limited, which may be true but is post hoc.\n\nI would not call this a fatal flaw. The qualitative morphology differences are real, and interferometric case studies already see offsets. But the claim that Dfrac decreases with temperature within sources is currently supported more by visual inspection than by statistics, and the filling-factor issue could change the answer.\n\nThis paper deserves a serious referee. The fix is straightforward: add a quantitative correlation test, and either estimate the filling-factor bias with a simple source-size model or explicitly state the range of Dfrac values consistent with plausible dilution. Then it will be a genuinely useful reference for the subfield.","headline":"A useful resolved survey of ammonia deuteration across 12 massive clumps, with plausible temperature trends that rest on an unverified beam-filling assumption.","tokens_in":39123,"tokens_out":2330,"would_cite":true,"duration_ms":23215,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper shows that within massive star-forming regions, ammonia's deuterium fraction falls as temperature rises, with the deuteration peak displaced from the ammonia peak.","keywords":["deuterium fractionation","ammonia","NH2D","massive star-forming regions","rotation temperature","gas-grain chemical model","interstellar chemistry","molecular clouds"],"falsifier":"Observe one of the seven anticorrelating sources (for example G081.87+00.78) with an interferometer at sub-arcsecond resolution and detect a second NH$_2$D transition to measure its excitation temperature; if the deuterium fraction at the NH$_3$ peak is not lower than at the offset position once beam dilution and excitation are corrected, then the claimed anticorrelation and peak offset would be artifacts of the single-dish assumptions.","tokens_in":37986,"feed_emoji":"🔭","tokens_out":15954,"duration_ms":124241,"temperature":0.7,"pith_summary":"The paper argues that the deuteration of ammonia in massive star-forming regions is regulated by local temperature and by the chemical evolution of the core. Using maps of ortho-NH$_2$D, NH$_3$(1,1), and NH$_3$(2,2) toward twelve late-stage regions, it finds that in seven sources the ammonia deuterium fraction $D_{\\rm frac}(\\mathrm{NH_3}) = N(\\mathrm{NH_2D})/N(\\mathrm{NH_3})$ decreases as the ammonia rotation temperature rises, and that in essentially every source the highest deuterium fraction is offset from the NH$_3$ peak. When all twelve sources are combined, the temperature trend weakens, which the paper explains as the result of source-to-source differences in density, cosmic-ray ionization rate, and evolutionary stage. A gas-grain chemical model that varies these parameters reproduces a general decline of $D_{\\rm frac}(\\mathrm{NH_3})$ with temperature while also showing why a single global trend should not be expected. The result matters because spatially resolved deuteration maps could become a practical thermometer and a tracer of core evolution in massive star formation.","feed_headline":"Ammonia deuteration falls where massive clumps warm up","feed_subtitle":"Maps of NH2D and NH3 in 12 late-stage regions show the deuteration peak sits away from the hot core.","key_machinery":"The central quantity is the deuterium fraction $D_{\\rm frac}(\\mathrm{NH_3}) = N(\\mathrm{NH_2D})/N(\\mathrm{NH_3})$, built from LTE column densities of the singly deuterated isotopologue NH$_2$D (in its ortho spin state) and of NH$_3$. The temperature axis is the ammonia rotation temperature $T_{\\rm rot}(2,2;1,1)$ derived from the NH$_3$(2,2)/(1,1) line ratio, and the argument assumes this temperature also holds for NH$_2$D and that both molecules fill the 28.6-42.5 arcsecond beams. The chemical mechanism invoked is the standard deuterium chemistry in which H$_2$D$^+$ and CH$_2$D$^+$ donate deuterium to ammonia; because the H$_2$D$^+$/H$_3^+$ ratio falls sharply above roughly 15 K, and because CO released from grains at higher temperature consumes H$_3^+$, the NH$_2$D/NH$_3$ ratio declines as the gas warms. A gas-grain chemical model varying temperature (6-60 K), density ($10^4$-$10^7$ cm$^{-3}$), cosmic-ray ionization rate, and timescale (0.1-1 Myr) is used to show this decline while demonstrating that the other parameters introduce scatter large enough to hide the temperature trend in a mixed sample.","core_discovery":"Within a sample of twelve late-stage massive star-forming regions mapped in ortho-NH$_2$D $1_{11}^s-1_{01}^a$ at 85.9 GHz and NH$_3$(1,1)/(2,2) at 23.7 GHz, the paper reports that the deuterium fraction of ammonia anticorrelates with ammonia rotation temperature inside individual clouds. The anticorrelation is clear in seven sources (G031.28+00.06, G034.39+00.22, G035.19-00.74, G081.87+00.78, G109.87+02.11, G111.54+00.77, G121.29+00.65), with $D_{\\rm frac}(\\mathrm{NH_3})$ falling by factors of several as $T_{\\rm rot}(2,2;1,1)$ climbs across its roughly 13-27 K range within a source. The paper also finds that the location of peak deuterium fractionation is offset from the NH$_3$ peak in almost all sources, most dramatically in G081.87+00.78 where the maximum $D_{\\rm frac}=0.086\\pm0.007$ lies $45''$ east and $9''$ north of the NH$_3$ peak. The offset is interpreted as the combined effect of higher temperatures near the NH$_3$ peak, gas-phase depletion of NH$_2$D as the core evolves, and CO released from grains suppressing the H$_2$D$^+$ pathway that forms NH$_2$D. The paper argues that the absence of a global anticorrelation across all twelve sources is expected because density, cosmic-ray ionization rate, and chemical timescale differ from source to source.","pith_inferences":["An untested extension of the paper's logic is that the same anticorrelation should appear for other deuterated molecules such as DCO$^+$, DCN, or N$_2$D$^+$ when mapped at comparable angular resolution, since they share the H$_2$D$^+$-based deuteration chemistry.","The offset pattern implies a testable prediction for higher-resolution observations: the single-dish peak of $D_{\\rm frac}$ is probably a blend of unresolved cold cores, each with its own temperature-deuteration gradient, rather than a single smooth ring.","A direct way to check the interpretation would be to observe a second NH$_2$D transition in one of the anticorrelating sources; measuring its excitation temperature independently would confirm whether the assumed $T_{\\rm ex}=T_{\\rm rot}$ is the source of the trend or a genuine chemical effect."],"forward_implications":["Within a given massive star-forming clump, NH$_2$D/NH$_3$ can be used as a temperature indicator, with cooler pixels showing systematically higher deuterium fractionation.","Single-pointing surveys that pool many regions will systematically miss the temperature dependence, because variations in density, ionization rate, and age between clouds scatter the relation.","The spatial offset between the NH$_3$ peak and the NH$_2$D fractionation peak implies that the coldest, most deuterium-enriched gas is not the densest gas, and that core-scale chemical evolution separates the two.","If the model's chemistry is right, the anticorrelation should be strongest in regions where CO has been recently released from grains, and weakest in young, CO-poor cores."],"supporting_citations":[{"why":"Supplied the IRAM 30 m ortho-NH2D maps that this work compares with ammonia maps.","marker":"Li et al. 2024"},{"why":"Provided the spectral-fitting code used to derive ammonia rotation temperatures and optical depths.","marker":"Lu et al. 2015"},{"why":"Provided the GBT NH3(1,1) and NH3(2,2) maps for three of the twelve sources.","marker":"Urquhart et al. 2015"},{"why":"Provided the RAMPS NH3 data for two of the twelve sources.","marker":"Hogge et al. 2018"},{"why":"Provided the KEYSTONE NH3 data for four of the twelve sources.","marker":"Keown et al. 2019"},{"why":"Provided the gas-grain chemical model used to simulate deuterium fractionation under varying physical parameters.","marker":"Majumdar et al. 2017"},{"why":"Showed that the H2D+/H3+ ratio drops sharply above about 15 K, the chemical basis for lower deuteration at higher temperature.","marker":"Caselli et al. 2008"},{"why":"Modelled Dfrac(NH3) increasing away from the core center, used to interpret the observed spatial offsets.","marker":"Sipilä et al. 2015b"},{"why":"Reported lower Dfrac(NH3) toward young stellar objects than prestellar cores, an interferometric precedent for the temperature dependence.","marker":"Busquet et al. 2010"},{"why":"Found lower deuterium fractionation in the warmer hot core of Orion KL than in the compact ridge, supporting the same trend.","marker":"Neill et al. 2013"}],"fun_headline_variants":["Ammonia deuteration dips as temperature rises in massive clumps","Deuteration peak sits off-center from hot ammonia cores","Warm cores lower ammonia deuteration, peak shifts away","NH2D peaks are offset from hot NH3 cores","Deuteration falls with warmth in massive clump cores"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result depends on assuming that the NH$_2$D emission fills the telescope beam and that its excitation temperature equals the ammonia rotation temperature; if the NH$_2$D is actually colder or more compact than the beam, then the derived deuterium fraction near the NH$_3$ peak would be too low, which could create both the anticorrelation and the observed offsets by itself.","fun_headline_variants_meta":{"raw":{"variants":["Ammonia deuteration dips as temperature rises in massive clumps","Deuteration peak sits off-center from hot ammonia cores","Warm cores lower ammonia deuteration, peak shifts away","NH2D peaks are offset from hot NH3 cores","Deuteration falls with warmth in massive clump cores"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00085,"raw_usage":{"total_tokens":3805,"prompt_tokens":1165,"completion_tokens":2640,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":781,"completion_tokens_details":{"reasoning_tokens":2557}},"tokens_in":781,"tokens_out":2640,"duration_ms":17355,"temperature":1.0,"reasoning_tokens":2557,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:33:10.293214+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observe one of the seven anticorrelating sources (for example G081.87+00.78) with an interferometer at sub-arcsecond resolution and detect a second NH$_2$D transition to measure its excitation temperature; if the deuterium fraction at the NH$_3$ peak is not lower than at the offset position once beam dilution and excitation are corrected, then the claimed anticorrelation and peak offset would be artifacts of the single-dish assumptions.","supporting_citations":[{"cited_title":"2010, , 517, L6, 10.1051/0004-6361/201014866","cited_arxiv_id":null,"evidence_quote":"Reported lower Dfrac(NH3) toward young stellar objects than prestellar cores, an interferometric precedent for the temperature dependence."}],"review_version":1}