{"id":"28d5babb-9543-4a41-b5c1-88863331557d","arxiv_id":"2507.06081","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"UV photodesorption yields of SO2 ice were measured at 14-80 K (about 2.3e-4 molecules per photon for SO2 at 14 K rising to 3.8e-4 at 70 K), with a tentative SO yield near 6e-5, plus a calculated SO3 infrared band strength.","lead":"This paper reports laboratory measurements of how much sulfur dioxide (SO2) and sulfur monoxide (SO) are kicked off ice surfaces by ultraviolet light, at temperatures found in cold interstellar clouds. The yields are new inputs for astrochemical models trying to explain unexpectedly high SO2 and SO gas abundances around young stars.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The tentative SO yield depends on assigning a small m/z=48 excess to SO rather than O3; the m/z=32 non-detection is not quantitative and the O3 IR non-detection is weak at sub-ML abundances. Isotopolog experiments, which the authors recommend, would settle the carrier.","rationale":"The reader's weakest assumption is the one I would also identify. The SO2 yield is supported by direct m/z=64 measurement, duplicate experiments, and deposited data; no internal inconsistency is apparent. The SO yield, by contrast, is a residual signal after subtraction, and its carrier assignment is the single point where the central claim could fail. The paper's own caveat, 'we tentatively assigned the excess in the measured m/z=48 QMS signal to the photodesorption of SO molecules' and the statement that isotopolog experiments would be required, is an explicit limitation and should weigh in the verdict. The m/z=32 and IR arguments are suggestive but not quantitative at the relevant column densities: the required O3 amount is orders of magnitude below the IR detection limit, and no m/z=32 upper limit is given. The proposed isotopolog test is decisive and straightforward. I do not see a reason to move beyond the reader's CONDITIONAL verdict: the main SO2 numbers can be used by modelers, but the SO yield should be marked as unconfirmed pending the isotopolog measurement. The DFT SO3 band strength is secondary to the desorption claim; its validation by an in-prep reference is a separate concern and does not change this verdict.","tokens_in":23430,"tokens_out":7360,"duration_ms":76172,"concrete_test":"Repeat Experiment 1 (amorphous SO2 ice deposited and irradiated at 14 K, same lamp and fluence) using 34SO2 as the ice precursor while monitoring m/z=50 (34S16O+), m/z=66 (34SO2+), and m/z=48. Use the measured 34SO2 fragmentation pattern to subtract the 34SO2 contribution to m/z=50. A residual m/z=50 signal corresponding to roughly 6×10−5 molecules photon−1 would confirm SO photodesorption; a null residual would require O3 (or another 48-amu species), because 16O3 produced from the oxygen atoms would remain at m/z=48. In the same series, record a quantitative 3σ upper limit on the m/z=32 increment during irradiation of normal SO2 to make the O3 exclusion testable.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central SO photodesorption yield (≈6×10−5 molecules photon−1 at 14 K, ninth column of Table 1) is derived in Sect. 3.2 by subtracting the SO2 fragmentation contribution from the measured m/z=48 QMS signal and assigning the residual to SO. The two candidate carriers are SO and O3, both mass 48 amu. The evidence against O3 is (i) the absence of the 1040 cm−1 O3 IR band and (ii) the statement that no significant m/z=32 increase was observed, since O3 would produce a m/z=32 signal about five times larger than the m/z=48 excess. Both arguments are weaker than they appear. The total O3 column density needed to explain the excess is of order 6×10−5 × 3.5×1017 ≈ 2×1013 cm−2, i.e., about 0.02 ML, which is below the IR detection threshold; the absence of an IR band therefore does not exclude O3. And no quantitative upper limit or baseline noise level is reported for m/z=32, so the expected SO m/z=32 signal (only roughly two times lower than the excess) could easily be buried in the O2/background baseline. In addition, the SO quantification itself uses an electron-impact ionization cross-section and fragmentation fraction adopted from SO2 and Dols et al. (2024), contributing a systematic uncertainty not captured by the stated 35% experimental scatter. If the m/z=48 carrier is O3 rather than SO, the reported SO yield is not a measurement of SO, and the photodesorption contribution to SO gas-phase abundances in the Sect. 4.2 models would need to be removed or replaced. Because the paper itself labels this assignment tentative and calls for isotopolog experiments, the SO2 yields remain unaffected; the concern is specific to the SO claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports laboratory measurements of VUV photodesorption yields for SO2 ice and, tentatively, for SO, over a temperature range of 14–80 K, using the ISAC setup with a calibrated quadrupole mass spectrometer and simultaneous IR spectroscopy. The authors find a SO2 yield of about 2.3×10^-4 molecules per incident photon at 14 K, increasing to about 3.8×10^-4 at 70 K and then decreasing at 80 K, with a tentative SO yield of about 6×10^-5 molecules per photon at 14 K that shows no clear temperature trend. They also use DFT calculations to derive a band strength of 1.1×10^-16 cm molecule^-1 for the ~1395 cm^-1 SO3 IR feature. The photodesorption yields are then implemented in the Nautilus gas-grain chemical model for the Horsehead PDR, and the SO3 band strength is used to estimate JWST/MIRI detectability thresholds for SO3 in interstellar ices.","tokens_in":23744,"tokens_out":5591,"duration_ms":60218,"significance":"If the results hold, this paper provides the first ISM-relevant laboratory constraints on non-thermal desorption of SO2 and SO, quantities that astrochemical models currently lack. The SO2 yield is well supported by duplicate experiments, documented calibrations, and deposited data, and the temperature dependence is a useful input for models of UV-illuminated icy environments. The DFT-based SO3 band strength is a valuable addition for future JWST searches, and the synthetic ice spectra illustrate a concrete observational strategy. The paper is also transparent about the tentative nature of the SO assignment, which is appropriate given the unresolved carrier ambiguity. The main scientific impact will depend on whether the SO result can be confirmed, because the modeling conclusions in Section 4.2 use that yield as a quantitative input.","major_comments":[{"comment":"The SO photodesorption yield is obtained by subtracting the SO2 fragmentation contribution from the measured m/z=48 QMS signal and assigning the residual to SO, but O3 is not quantitatively excluded. The O3 column density needed to explain the excess is on the order of 2×10^13 cm^-2 (about 0.02 ML), which is small compared with the ~1 ML of SO3 detected and likely below the detection limit of the IR spectra; the absence of the 1040 cm^-1 O3 band therefore does not provide a quantitative upper limit. In addition, the m/z=32 argument is qualitative because no baseline noise level or upper limit is reported, even though the expected O3 m/z=32 signal would be roughly five times the m/z=48 excess. The adopted SO ionization cross-section and fragmentation fraction from SO2/Dols et al. (2024) add systematic uncertainty that is not included in the quoted 40% uncertainty. The authors correctly recommend isotopolog experiments; until such data or a quantitative O3 upper limit are available, the SO yield should be presented as an upper limit rather than a measured value.","section":"Section 3.2, Table 1, ninth column"},{"comment":"The Nautilus models treat the tentative SO yield as a quantitative input, and the conclusion that SO photodesorption is a main gas-phase source at extinctions below 5 mag depends on the m/z=48 carrier being SO rather than O3. If the carrier is O3, the SO yield would need to be removed from the model, and the predicted SO abundance and SO2/SO ratio at low extinction would change. The modeling section should either explicitly frame the SO run as a provisional scenario contingent on confirmation of the carrier, or defer the quantitative SO modeling until the identification is established.","section":"Section 4.2, Fig. 6"}],"minor_comments":[{"comment":"The text and the Figure 5 caption state that the SO photodesorption rate is calculated 'using Eq. 3', but Eq. 3 is the IR band-strength relation; the QMS calibration uses Eq. 1, and this appears to be a typographical error.","section":"Section 3.2 and Fig. 5 caption"},{"comment":"The word 'assginment' should be 'assignment' in the phrase about the SO3 IR feature.","section":"Section 3.1"},{"comment":"The reference list contains some typographical errors, including 'Dufuor' for 'Dufour' in the Le Gal et al. entry and 'Wesemberg' for what is presumably 'Wesenberg' in the 2007 paper; please verify these names.","section":"References"},{"comment":"The citation 'Le Gal et al. 20217' should be 'Le Gal et al. 2021'.","section":"Section 4.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of MNRAS and the experimental work on SO2 photodesorption is solid. The main issue is the tentative SO identification, which the authors themselves flag but nevertheless use quantitatively in the chemical modeling. I would encourage the editor to request either isotopolog confirmation or a clear re-framing of the SO yield as an upper limit, with the modeling consequences stated accordingly. The data availability and calibration documentation are strengths and should be preserved in the revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the practical takeaway: the SO2 photodesorption yield is the real result, and it looks solid. 2.3e-4 molecules per photon at 14 K, rising to 3.8e-4 at 70 K, with duplicate experiments that agree within the stated scatter. The calibration chain (CO reference, QMS sensitivity, pumping speeds) is documented, and the raw data are on Zenodo. The temperature trend, including the amorphous-to-crystalline comparison, is a nice piece of work. The DFT band strength for the 1395 cm-1 SO3 feature (1.1e-16 cm molecule-1) is also new and sensible, with a density sensitivity analysis that gives a 15% uncertainty. I would trust the SO2 numbers and the band strength for modeling and for JWST ice interpretation.\n\nThe soft spot is the SO yield. The authors assign a small excess in m/z=48 to SO after subtracting SO2 fragmentation. The only other plausible carrier is O3, and the evidence against O3 is thinner than it looks. The O3 column that would explain the excess is about 0.02 ML, below the IR detection limit, so the absence of the 1040 cm-1 band does not rule it out. And the m/z=32 argument—no significant increase above baseline—is not quantitative; the expected SO signal at m/z=32 is only a factor of two below the excess, and could hide in the O2 background. The authors know this; they call the assignment tentative and ask for isotopolog experiments. That is honest, but it means the SO yield should be treated as a provisional estimate, not a measured value, until the carrier is confirmed. The adopted ionization cross-section for SO (from SO2) adds another unquantified systematic.\n\nA couple of minor things: the plotted photodesorption rates have no uncertainty envelopes, and the DFT method is validated by an in-prep reference. Neither bothers me much—the tabulated yields carry uncertainties.\n\nBottom line: this is a competent experimental paper that fills a genuine gap: first ISM-relevant VUV photodesorption yields for SO2, and first SO3 IR band strength. The SO2 result and the band strength will get used. The SO carrier question is exactly what peer review should probe. Yes, I would send it to review, with a request that the m/z=32 upper limit be stated quantitatively or an isotopolog experiment be added. I would cite the SO2 yield and the band strength; I would cite the SO yield only with a 'tentative' qualifier.","headline":"Solid SO2 photodesorption yields and a useful SO3 band strength; the tentative SO yield rides on a m/z=48 assignment that needs a quantitative check.","tokens_in":24412,"tokens_out":4630,"would_cite":true,"duration_ms":40265,"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":"This paper reports the first ISM-relevant laboratory measurement of how vacuum-ultraviolet photons release SO2 and SO from SO2 ice, giving yields of about 2e-4 and 6e-5 molecules per photon at 14 K, and a theoretical band strength for the…","keywords":["photodesorption","SO2 ice","sulfur monoxide","vacuum-ultraviolet irradiation","interstellar ices","astrophysical sulfur chemistry","IR band strength","SO3"],"falsifier":"Irradiate an isotopically labeled SO2 ice, such as 18O-substituted or 34S-substituted SO2, and watch the excess photodesorption signal around mass 48: if the excess shifts to the mass of the labeled SO product, the tentative SO assignment is confirmed, whereas if it shifts to the mass of an ozone isotopolog or remains at 48, the claimed SO yield does not measure SO. A simpler concurrent check is to search for the companion m/z=32 signal during irradiation; the paper reports none, which favors SO over O3, but an isotopolog run would settle it directly.","tokens_in":23193,"feed_emoji":"☀️","tokens_out":8622,"duration_ms":86558,"temperature":0.7,"pith_summary":"This paper measures, for the first time under interstellar-medium relevant conditions, how efficiently vacuum-ultraviolet photons release sulfur dioxide and sulfur monoxide from SO2 ice. At 14 K the SO2 photodesorption yield is about 2.3e-4 molecules per incident photon; it grows with temperature to 3.8e-4 at 70 K and then falls at 80 K, a behavior the paper ties to the amorphous-to-crystalline phase transition of the ice. A residual mass-48 signal, tentatively assigned to SO, implies a yield near 6e-5 molecules per photon at 14 K with no strong temperature dependence, and the yield's constancy with fluence points to prompt photochemical desorption of SO right after formation. These numbers matter because astrochemical models previously had no photodesorption constraints for these two sulfur molecules, and the paper shows they can account for the early gas-phase enrichment of SO and SO2 in photon-dominated regions at extinctions below about 5 magnitudes. The paper also delivers a theoretically derived band strength of 1.1e-16 cm molecule-1 for the 7.2 micron infrared feature of SO3, the main photoproduct, which will help future JWST searches for sulfur in interstellar ices.","feed_headline":"UV light frees SO2 and SO from ice at measured rates","feed_subtitle":"First interstellar-lab yields for sulfur molecules give models real numbers; new SO3 ice band strength sharpens JWST searches.","key_machinery":"The central experimental object is the calibrated quadrupole mass spectrometer signal: ion currents at m/z 64, 48, and 32 are converted into photodesorption column densities using a calibration constant derived from CO photodesorption, electron-impact ionization cross-sections, measured fragmentation fractions, QMS sensitivity as a function of mass, and species-dependent pumping speeds. The SO yield is the residual in the m/z=48 signal after subtracting the SO2 fragmentation contribution, and its tentative identification relies on the absence of a companion m/z=32 signal that an ozone contribution would produce. On the infrared side, the paper runs density-functional-theory vibrational calculations on a periodic eight-molecule amorphous SO3 ice model to sum the intensities of the antisymmetric stretching vibrations into a band strength, with the ice density assumed equal to that of liquid SO3 and a 15 percent uncertainty budget derived from density variation.","core_discovery":"On the paper's own terms, the discovery is a set of quantitative laboratory constraints: VUV-irradiated SO2 ice photodesorbs SO2 at a yield of roughly 2.3e-4 molecules per photon at 14 K, rising to 3.8e-4 at 70 K and declining at 80 K, and it also releases SO at a tentative yield of roughly 6e-5 molecules per photon at 14 K, with the SO yield insensitive to temperature and fluence. The SO2 yield is comparable to that of CO2 but one to two orders of magnitude below CO and NH3, showing that photodesorption efficiency is molecule-specific. The fluence-independence of the SO signal is interpreted as evidence for a photochemical desorption mechanism, in which SO desorbs promptly upon its formation from photodissociated SO2, rather than through an energy-transfer process that would accumulate surface photoproducts. In addition, the paper provides the first reported infrared band strength for SO3 in an amorphous ice, 1.1e-16 cm molecule-1 at about 1395 cm-1 (7.2 micron), and demonstrates that using this value rather than the SO2 band strength changes derived SO3 ice column densities by at least a factor of two.","pith_inferences":["If the SO assignment survives isotopolog tests, the photochemical desorption of a photoproduct from its parent ice would be a rare direct laboratory demonstration of a mechanism that many models invoke to explain unsaturated molecules in cold gas.","The nearly linear relationship between the computed SO3 band strength and the assumed ice density in the appendix suggests that a single experimental measurement of amorphous SO3 ice density would sharpen every JWST-derived SO3 column density; until then, the 15 percent density-induced uncertainty is the limiting error.","Because SO2 photodesorption efficiency changes with temperature and crystallinity, similar experiments on mixed ices (for example SO2 in water- or CO-dominated mantles) could reveal whether the yields measured here over- or underestimate desorption in realistic interstellar grain mantles.","The paper's model uses a static isobaric PDR structure, leaving open whether dynamical front motion would enhance the photodesorption contribution; a time-dependent PDR model with the new yields would be a natural next test."],"forward_implications":["Astrochemical models can now treat UV photodesorption of SO2 and SO with lab-measured yields instead of scaling from other molecules; the paper's models show the effect is largest at extinctions below about 5 magnitudes and in the first roughly 100 years of cloud evolution.","The strong temperature dependence of the SO2 yield, peaking near the amorphous-to-crystalline transition at 70 K, means models of warmer ice mantles should use temperature-dependent desorption rates rather than a single low-temperature value.","With the new SO3 band strength, the 7.2 micron feature becomes a quantitative tool: 5-sigma detection thresholds of about 4e15 cm-2 toward bright continua are within reach of JWST/MIRI, though methanol and continuum-subtraction contamination must be removed first.","The fluence-independence of the SO signal implies prompt photochemical desorption, so SO is injected into the gas phase at the moment of its formation, which couples the photochemistry of the ice directly to the gas-phase sulfur budget."],"supporting_citations":[{"why":"Supplies the vacuum-ultraviolet absorption cross-section of SO2 ice used to estimate that roughly 30 percent of incident photons are absorbed in the experiments.","marker":"Holtom et al. (2006)"},{"why":"Provides the SO2 ice density and IR band strengths used to convert infrared absorbances to ice column densities and to correct the literature absorption cross-section.","marker":"Yarnall & Hudson (2022)"},{"why":"Documents the amorphous-to-crystalline transition temperature of SO2 ice near 70 K, which the paper uses to interpret the temperature and structure dependence of the photodesorption yields.","marker":"Schriver-Mazzuoli et al. (2003a)"},{"why":"Provides electron-impact ionization and dissociative-ionization cross-sections for SO, adopted as the fragmentation pattern needed to identify the excess mass-48 signal.","marker":"Dols et al. (2024)"},{"why":"Supplies the ozone mass spectrum used to argue that the excess mass-48 signal is not due to O3 because the companion mass-32 signal would have been clearly detectable.","marker":"Herron & Schiff (1956)"},{"why":"Gives the quadrupole-mass-spectrometer calibration method and the equation that converts measured ion currents into photodesorption column densities.","marker":"Martín-Doménech et al. (2015)"},{"why":"Provides the only previously reported photodesorption yield for a sulfur-bearing molecule (H2S), used as a comparison point and as an input in the chemical modeling.","marker":"Fuente et al. (2017)"}],"fun_headline_variants":["Measured yields: UV liberates SO2, SO from ice","SO2 ice photodesorption yields: from lab to models","UV photodesorption of SO2 ice: quantitative yields","New SO3 ice band strength sharpens JWST searches","Lab yields for SO2 and SO from UV-irradiated ice"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The tentative SO yield rests on the assumption that the excess mass-48 signal, after removing the SO2 fragmentation contribution, is photodesorbing SO rather than ozone or another 48-amu molecule; the paper itself states that isotopolog experiments would be required to confirm this.","fun_headline_variants_meta":{"raw":{"variants":["Measured yields: UV liberates SO2, SO from ice","SO2 ice photodesorption yields: from lab to models","UV photodesorption of SO2 ice: quantitative yields","New SO3 ice band strength sharpens JWST searches","Lab yields for SO2 and SO from UV-irradiated ice"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000336,"raw_usage":{"total_tokens":1946,"prompt_tokens":1114,"completion_tokens":832,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":730,"completion_tokens_details":{"reasoning_tokens":746}},"tokens_in":730,"tokens_out":832,"duration_ms":8764,"temperature":1.0,"reasoning_tokens":746,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:12:06.495438+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Irradiate an isotopically labeled SO2 ice, such as 18O-substituted or 34S-substituted SO2, and watch the excess photodesorption signal around mass 48: if the excess shifts to the mass of the labeled SO product, the tentative SO assignment is confirmed, whereas if it shifts to the mass of an ozone isotopolog or remains at 48, the claimed SO yield does not measure SO. A simpler concurrent check is to search for the companion m/z=32 signal during irradiation; the paper reports none, which favors SO over O3, but an isotopolog run would settle it directly.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the vacuum-ultraviolet absorption cross-section of SO2 ice used to estimate that roughly 30 percent of incident photons are absorbed in the experiments."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the SO2 ice density and IR band strengths used to convert infrared absorbances to ice column densities and to correct the literature absorption cross-section."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides electron-impact ionization and dissociative-ionization cross-sections for SO, adopted as the fragmentation pattern needed to identify the excess mass-48 signal."},{"cited_title":"& Schiff H.I., 1956, J","cited_arxiv_id":null,"evidence_quote":"Supplies the ozone mass spectrum used to argue that the excess mass-48 signal is not due to O3 because the companion mass-32 signal would have been clearly detectable."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the only previously reported photodesorption yield for a sulfur-bearing molecule (H2S), used as a comparison point and as an input in the chemical modeling."}],"review_version":1}