{"id":"20eeb1a5-f04e-4a6f-8ed2-c8bc5b94aeb0","arxiv_id":"2505.21175","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"H2O2 in water systems forms by reduction of dissolved O2 at solid-water interfaces, so without O2 no H2O2 is produced, contradicting claims of air-water interface generation.","lead":"Experiments in this preprint show that hydrogen peroxide forms in water only when dissolved oxygen is present, and they trace its production to reactions on solid surfaces rather than to the droplet-air boundary. If correct, this refutes a prominent claim that microdroplets spontaneously generate hydrogen peroxide at their water-air interface.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The no-O2 control removes the oxidant, not the interface; the paper never quantifies solid-surface H2O2 production inside the actual spray, so its universal refutation of air-water mechanisms remains an inference, not a demonstration.","rationale":"The reader's weakest-assumption analysis identifies exactly the inference gap I see: the paper demonstrates that O2 is necessary and that corroding solids make H2O2 in bulk, but it does not isolate the solid-surface contribution inside the actual spray geometry. The O2-free control is powerful for showing oxygen dependence but logically cannot locate the reaction; an O2-dependent air-water mechanism survives that control. The bulk Al/Ti experiments are consistent and carefully cross-validated, but they use different materials and geometries, so they do not close the locus question. The proposed capillary-length scaling test is a feasible, direct way to separate fluid-path solid contact from droplet-phase chemistry. I do not see grounds to reject the paper: the corrosion-based explanation for halide and pH trends is credible and the negative oxygen result is an important constraint. The conditional verdict is appropriate, with the missing in-spray locus control as the main condition before the strong universal claim can be accepted.","tokens_in":13705,"tokens_out":7950,"duration_ms":98750,"concrete_test":"Run the O2-saturated 10 mM NaCl spray at the same 25 µL/min liquid flow and 100 psi N2 through three lengths of the same 0.10-mm i.d. fused-silica capillary (e.g., 5, 20, and 80 cm), collecting equal volumes in identical glassware and measuring H2O2 by HPAK cross-checked with 1H-NMR. If the yield scales with capillary length, the solid-liquid interface in the fluid path is dominant; if the yield is invariant with length, the source is downstream of the capillary and the air-water interface or collection-vial surface cannot be excluded. Include a no-sheath (no atomization) jet through the same capillary to bound the capillary-plus-vial background.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that microdroplet H2O2 forms by O2 reduction at solid-water interfaces and that the air-water interface or microdroplet geometry plays no role. The supporting evidence is (i) O2-free sprays give no H2O2, and (ii) Al/Ti immersion in air-equilibrated water gives H2O2 that tracks corrosion rates. Leg (i) only establishes that O2 is a necessary reactant; an O2-dependent interfacial mechanism, such as O2 reduction assisted by the electric fields or charge-separation effects contested here, is equally consistent with the null result. Leg (ii) demonstrates that corroding solids can produce H2O2, but the spray apparatus contains silica and glass surfaces, not the Al/Ti coupons, and no measurement in this paper quantifies their solid-water contribution inside the spray geometry itself. The statement that glass and other non-metallic materials also corrode and produce H2O2 is cited to prior work, not demonstrated at the rates, surface areas, and contact times of this spray setup. Thus the universal negative 'no O2, no H2O2, and therefore no air-water contribution' is underdetermined: if an O2-dependent air-water pathway contributes when O2 is present, the paper's refutation of George and co-workers would fail even though the oxygen requirement still holds. The missing piece is direct locus identification in the spray, not additional bulk immersion chemistry.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports experiments on spontaneous hydrogen peroxide (H2O2) production in sprayed water microdroplets and at solid–water interfaces. The authors find that removing dissolved O2 eliminates detectable H2O2 in sprays regardless of halide identity, while in O2-bearing solutions H2O2 yields depend on halide type and pH. They reproduce the halide trend (Br > Cl > I at 10 mM) and the pH dependence reported by George and co-workers, but attribute all H2O2 production to O2 reduction at solid–water interfaces driven by surface corrosion. Halide effects are explained by pitting corrosion and iodide oxidation, and pH effects by the contrasting pH-dependent corrosion of Al versus Ti. The central conclusion is the mnemonic 'no O2, no H2O2', which is presented as refuting air–water interfacial or microdroplet-geometry mechanisms.","tokens_in":13993,"tokens_out":5959,"duration_ms":65522,"significance":"If the central claim is correct, the paper would be an important experimental contribution to the ongoing debate on spontaneous H2O2 formation in microdroplets, providing evidence against air–water interfacial mechanisms and supporting a corrosion-based alternative. Strengths include dual quantification by 1H-NMR and HPAK, the O2-free control, the cross-validation of trends with corrosion current measurements, and falsifiable predictions such as the absence of H2O2 without dissolved O2. However, because the spray experiments do not directly isolate solid-surface contributions within the spray geometry itself, the universal negative claim is not fully demonstrated; the significance is therefore conditional on additional locus identification.","major_comments":[{"comment":"The O2-free spray control establishes that dissolved O2 is necessary for the H2O2 observed in sprays, but it does not identify the reaction site; an O2-dependent air–water or microdroplet pathway is equally consistent with the null result. The paper's universal claim that 'the air–water interface or the microdroplet geometry ... do not form H2O2' is therefore an inference from the bulk immersion experiments (Fig. 2) and prior work (ref 14), not a direct measurement in the spray geometry. Please add a control that varies the solid–water surface area or material inside the spray and collection train (e.g., compare silica versus PTFE capillaries, or add/remove solid beads in the collection flask) to demonstrate that the H2O2 yield scales with solid contact and not with droplet surface area.","section":"Results, Fig. 1c-d and Conclusion"},{"comment":"The pitting-consumption mechanism is invoked to reconcile the divergent ordering of H2O2 yields (Br > Cl > I at 10 mM and above) and corrosion currents (Cl > Br > I). The only direct evidence for H2O2 consumption is Fig. S6, which shows that a freshly polished Al surface consumes more H2O2 than an oxide-covered one in a 100 µM H2O2 solution; it does not quantify consumption during pitting in chloride solutions at the concentrations and times of the experiments. The reconciliation should be presented as a hypothesis and tested directly, for example by measuring H2O2 decay in pitting versus non-pitting conditions with matched solution chemistry.","section":"Results, The Curious Case of Cl- and Br-, and Fig. S6"},{"comment":"The sentence 'we compare the trends in measured H2O2 concentrations in sprays (Figure 2b) and the corrosion at the solid–water interface (Figures 3b-c)' mislabels Figure 2b, which reports bulk immersion measurements at the Al–water interface, not spray measurements; the spray data are in Figure 1c and Figure S2. This conflation obscures which dataset supports the solid-water attribution. Please correct the cross-reference and state explicitly which data come from sprays and which from bulk immersion.","section":"Results, Quantifying Oxidation Rates of Solid Surfaces"}],"minor_comments":[{"comment":"In the sentence 'demonstrating that aluminum and titanium surfaces more H2O2 in acidic and alkaline conditions', the verb 'produce' is missing.","section":"Abstract"},{"comment":"The phrase 'also noted by Georged & co-workers' contains a typo; it should read 'George'.","section":"The Curious Case of I-"},{"comment":"The references to 'Fig. 4a' in the text should be 'Fig. S4a' to match the supplementary numbering.","section":"Section S4"},{"comment":"The sentence 'The tests were carried out at a scan rate of 0.5 mV s−1 and were performed.' is grammatically incomplete; remove the trailing phrase.","section":"Methods S2.4"},{"comment":"The corrosion current values appear to mix units (mA/cm2 and µA/cm2); please verify and unify the units in the text and figure.","section":"Figure 5c-d"},{"comment":"The mnemonic 'no O2, no H2O2' should be qualified as 'under the conditions and detection limits tested' to avoid overgeneralization.","section":"Conclusion"}],"recommendation":"major_revision","confidential_remarks":"The paper is a direct continuation of the authors' earlier work (ref 14), and the field is highly polarized. The new data do extend the previous work to halides and pH, but the central missing control—quantifying the solid-surface contribution to H2O2 production inside the spray geometry itself—is the same limitation flagged in the stress-test and in the paper's own reliance on ref 14. I would encourage the editor to request this additional control before publication; without it, the universal claim overreaches. The citation list is heavily weighted toward the authors' own group and allied critiques, but this is common in this debate and is not by itself a reason for rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — this paper does something useful: it reproduces George's halide and pH trends and shows they line up with corrosion chemistry on Al and Ti. The O2-free spray control is clean and, combined with the NMR cross-checks, makes the necessity of O2 about as solid as these measurements get. The halide crossover (Cl > Br at low concentration, Br > Cl at high) and the pitting-consumption explanation are genuinely new and worth taking seriously. The triiodide observation for iodide is a nice supporting detail.\n\nThe soft spot is the locus of production. The no-O2 control removes the oxidant but not the air-water interface, so it does not by itself rule out an O2-dependent air-water mechanism. The bulk immersion experiments show that corroding Al and Ti make H2O2, but the spray rig uses silica and glass, not those coupons. The paper cites prior work for glass corrosion, but it does not quantify the solid-surface contribution inside the spray geometry. So the universal negative—'no O2, no H2O2, and therefore no air-water role'—is an interpretation that goes beyond the data. A reader who believes in an electric-field or ion-separation mechanism can accept the O2 requirement and still say the air-water interface contributes when O2 is present. To close that gap, the authors would need to vary the solid surface area/material in the spray itself or use a non-corroding surface and show zero H2O2.\n\nI also noticed the paper gives no error bars or replicate counts in the main figures. The trends are large enough that I suspect they are real, but the quantitative claims—like the crossover concentrations—deserve more than a single curve. That is fixable with statistics.\n\nOverall: the paper is an honest, well-executed challenge to a prominent claim, with several new observations that stand on their own. The central interpretation is underdetermined, but not in a way that makes the work unserious. It deserves a real peer review, and the referees should push for direct locus identification and error bars. I'd bring it to a reading group.","headline":"A solid experimental argument that O2 is necessary for H2O2 in sprays, but the claim that solid surfaces are the only source depends on inference, not direct measurement in the spray.","tokens_in":14497,"tokens_out":2265,"would_cite":true,"duration_ms":24378,"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 argues that spontaneous hydrogen peroxide in water microdroplets is produced by reduction of dissolved oxygen at solid-water interfaces, not at the air-water interface or by microdroplet geometry; removing oxygen removes the…","keywords":["hydrogen peroxide","microdroplets","air-water interface","solid-water interface","dissolved oxygen","corrosion","pitting","pH dependence"],"falsifier":"The claim would be falsified by a single clean experiment in which H2O2 is detected in sprayed microdroplets after all dissolved O2 is removed to below 0.01 mg/L and all solid-water contacts are made from inert, non-corroding materials, or by detecting H2O2 in the presence of O2 when the collected droplets demonstrably never touched a wetted solid surface.","tokens_in":13502,"feed_emoji":"💧","tokens_out":6952,"duration_ms":68058,"temperature":0.7,"pith_summary":"This paper argues that the long-reported spontaneous formation of hydrogen peroxide in sprayed water microdroplets does not happen at the air-water interface or because of microdroplet geometry. Instead, the authors claim, the peroxide is produced by reduction of dissolved oxygen at solid-water interfaces that are inevitably present in the apparatus; in their experiments, removing O2 eliminates H2O2 regardless of added salts. They reproduce the previously reported halide and pH trends but attribute them to corrosion of metal surfaces: chloride, bromide, and iodide differ in how strongly they pit aluminum, and Al versus Ti corrosion depends oppositely on pH. If right, the electric-field and ion-separation mechanisms proposed for microdroplet H2O2 formation would be replaced by a mundane surface-oxidation pathway.","feed_headline":"No O2, no H2O2: droplets get peroxide from solid surfaces","feed_subtitle":"Halide and pH effects track corrosion of the apparatus, so degassing removes the signal.","key_machinery":"The load-bearing mechanism is oxygen reduction at a solid-water interface, coupled to oxidation (corrosion) of that solid: O2 is the electron acceptor and the solid is the electron donor, yielding H2O2. Halide ions enter through corrosion chemistry—Cl- pits the passive oxide film and then consumes H2O2 on fresh metal, I- scavenges H2O2 to triiodide, Br- sits between—so the reported halide order is a corrosion signature. pH enters only through its effect on which solid corrodes faster: Al oxidizes faster in acid, Ti in base. Quantification relies on HPAK fluorometry cross-validated by 1H-NMR, and corrosion rates are measured by potentiodynamic polarization.","core_discovery":"The paper's central claim is that spontaneous H2O2 production in water microdroplets is not an interfacial water phenomenon: it is oxygen-reduction chemistry at solid-water interfaces. In sprays, O2-free solutions produced no H2O2 regardless of halide identity, while air-saturated solutions did, with yields ordered by the halide's effect on corrosion of the aluminum surface. Separate immersion experiments on Al and Ti showed H2O2 yield tracks oxidation rate: Al corrodes faster in acid and yields more H2O2 there; Ti corrodes faster in base and yields more H2O2 there. The authors therefore attribute the halide and pH trends reported in prior droplet studies to surface corrosion chemistry, not to charge separation or alkalinity at the air-water interface.","pith_inferences":["If the mechanism generalizes, other reported 'spontaneous' microdroplet oxidations—beyond H2O2—may also be driven by trace corrosion or dissolution of solids rather than by interfacial electric fields.","The paper implies a practical design rule for microdroplet experiments: choose wetted materials by their corrosion resistance in the intended solution, and degas when trying to isolate interfacial effects.","A quantitative model combining pitting potential, corrosion current, and H2O2 consumption kinetics could predict the full concentration-dependent crossover (Cl > Br at low salt, Br > Cl at high salt) from electrochemical data alone.","The triiodide signature offers a cheap diagnostic: in iodide solutions, UV absorption at roughly 350 nm should track H2O2 production, providing a cross-check beyond fluorescence assays."],"forward_implications":["Any claimed air-water or microdroplet H2O2 generation should be re-examined with dissolved O2 removed and with every solid contact surface identified and passivated.","H2O2 yields in sprayed droplets should depend on the material of the capillary, tubing, collection vessel, and substrate, not just on solution composition.","Halide trends in H2O2 yield are not evidence for ion-specific interfacial chemistry; they track pitting propensity and H2O2 consumption rates.","pH trends in H2O2 yield are not universal; the same solution pH can increase or decrease yield depending on the solid surface.","Oxygen-free control experiments become a screening test for any purported interfacial oxidation reaction in water."],"supporting_citations":[{"why":"Prior demonstration that H2O2 forms at liquid-solid interfaces via O2 reduction; provides the baseline mechanism the paper extends.","marker":"[14]"},{"why":"The halide-dependent H2O2 microdroplet dataset the paper reproduces and reinterprets through pitting corrosion.","marker":"[30]"},{"why":"The pH-dependent H2O2 claim at air-water interfaces that the paper contests with Al and Ti experiments.","marker":"[31]"},{"why":"The related claim of spontaneous OH radical formation at aqueous droplet interfaces, part of the interfacial-chemistry position being refuted.","marker":"[29]"},{"why":"The original report of spontaneous H2O2 generation from aqueous microdroplets that started the debate.","marker":"[1]"},{"why":"The NMR protocol used for unambiguous low-concentration H2O2 quantification.","marker":"[34]"},{"why":"Provides the pitting corrosion understanding used to explain Cl- versus Br- versus I- behavior at high salt concentrations.","marker":"[35]"},{"why":"Source for the Henry's law dissolved O2 concentration in air-equilibrated water.","marker":"[33]"},{"why":"Earlier work showing ultrasonication artifacts in microdroplet H2O2 studies, supporting the solid-surface explanation.","marker":"[16]"}],"fun_headline_variants":["No O2, no H2O2: peroxide comes from surfaces, not droplets","Droplet peroxide is surface chemistry, not water magic","Halide and pH effects traced to corrosion, not air-water interface","Spontaneous H2O2 needs O2 and a metal surface, study finds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the absence of H2O2 in O2-free sprays, together with bulk immersion experiments on Al and Ti, proves that all H2O2 in the sprays came from solid-water interfaces rather than from an oxygen-dependent air-water pathway.","fun_headline_variants_meta":{"raw":{"variants":["No O2, no H2O2: peroxide comes from surfaces, not droplets","Droplet peroxide is surface chemistry, not water magic","Halide and pH effects traced to corrosion, not air-water interface","Spontaneous H2O2 needs O2 and a metal surface, study finds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000539,"raw_usage":{"total_tokens":2666,"prompt_tokens":1103,"completion_tokens":1563,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":719,"completion_tokens_details":{"reasoning_tokens":1495}},"tokens_in":719,"tokens_out":1563,"duration_ms":11285,"temperature":1.0,"reasoning_tokens":1495,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:33:14.527853+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The claim would be falsified by a single clean experiment in which H2O2 is detected in sprayed microdroplets after all dissolved O2 is removed to below 0.01 mg/L and all solid-water contacts are made from inert, non-corroding materials, or by detecting H2O2 in the presence of O2 when the collected droplets demonstrably never touched a wetted solid surface.","supporting_citations":[{"cited_title":"K.; Walker, K","cited_arxiv_id":null,"evidence_quote":"The original report of spontaneous H2O2 generation from aqueous microdroplets that started the debate."}],"review_version":1}