{"id":"aea642a6-a977-466b-bd49-c9b8874f75fe","arxiv_id":"2505.13800","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"WO3 nanoflowers on Vulcan XC-72 carbon boost the two-electron oxygen reduction pathway, achieving near-100% H2O2 selectivity in alkaline medium and 862 mg/L accumulation in a gas diffusion electrode.","lead":"This paper reports that a tungsten oxide nanoflower catalyst on carbon (WO3/Vulcan XC-72) improves the selectivity of the oxygen reduction reaction toward hydrogen peroxide in both acid and alkaline media. It matters because cheap, non-precious-metal cathodes for on-site H2O2 synthesis could lower the cost of water treatment and green oxidations.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"RRDE selectivity values for 3% WO3/C in alkaline medium are underdetermined because §2.5 does not specify which ring/N was used or the ring detection potential; the near-100% value is sensitive to that choice.","rationale":"The reader's weakest assumption and my own converge on the RRDE collection configuration. I sharpen the concern in two ways: with a numerical sensitivity estimate showing how a wrong N changes a nominal 95% selectivity to about 81% or to >100%, and with the requirement that the ring potential be on the H2O2 oxidation plateau. This is the most load-bearing issue because the advertised near-100% selectivity and n≈2.1 are the quantitative core of the experimental claim. The DFT pH correction is also questionable: subtracting kT ln10 pH from free energies at pH 13 appears to improve overpotentials in a way that may not follow from the CHE model, and the DFT calculations consider only bare WO3 surfaces, not WO3-carbon synergy. However, these issues affect the mechanistic narrative rather than the primary electrocatalyst claim, so they are secondary. The GDE accumulation number is a useful independent result but has no error bars and its CE versus applied charge is not fully reconciled. Overall, the central claim is plausible and worth conditional acceptance, but the exact selectivity values need sharper reporting and re-analysis. Since the reader already recommended conditional acceptance and my concern does not move the verdict, I recommend UNCHANGED.","tokens_in":14853,"tokens_out":7486,"duration_ms":75915,"concrete_test":"Obtain the raw RRDE files (or digitize Figure 4) and recompute X_H2O2 and n for 3% WO3/C in alkaline and 1% WO3/C in acid using (i) N=0.28, (ii) N=0.21, and (iii) an independently measured N from a ferri/ferrocyanide RRDE calibration performed in the same electrolyte, rotation rate, and scan rate. Also request the ring potential used in each experiment. If the reported near-100% selectivity and n≈2.1 for 3% WO3/C in alkaline persist under the correct N and with the ring held on the H2O2 oxidation limiting plateau, the concern is resolved; if the selectivity shifts by more than 10 percentage points or exceeds 100%, the quantitative headline must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim that WO3/C approaches 100% H2O2 selectivity (n≈2.1) in alkaline medium is computed from RRDE ring and disk currents via Eq. (1). Section 2.5 lists two possible ring configurations: a gold ring with N=0.28 and a platinum ring with N=0.21, but it does not state which ring was used in 1 mol L-1 NaOH and which in 0.1 mol L-1 K2SO4 (pH 3), nor the ring potential at which Ir was collected. The same raw currents analyzed with the other N give materially different selectivities: a reported value near 95% would become about 81% if N=0.21 were used while the true N was 0.28, and would exceed 100% (unphysical) in the opposite swap. In addition, if the ring potential is not on the H2O2 oxidation limiting-current plateau, Ir is not a quantitative measure of the H2O2 flux; Pt rings can also oxidize species other than H2O2 in alkaline media, which would inflate the apparent selectivity. Conversely, H2O2 consumed or decomposed inside the porous catalyst film before reaching the ring makes Ir an incomplete measure of the flux leaving the disc. The 862 mg L-1 GDE result is independent, but it is reported for only one catalyst and its current efficiency is not fully reconciled with the chronopotentiometric charge. The concern is load-bearing because the headline 'near 100% selectivity' is the quantitative basis for claiming WO3/Vulcan is a high-performance H2O2 catalyst.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a combined experimental and DFT study of WO3/Vulcan XC-72 as an electrocatalyst for the two-electron oxygen reduction reaction to H2O2. WO3 nanoflowers synthesized solvothermally are supported on Vulcan at 1, 3, and 5 wt%. RRDE measurements in 1 mol L-1 NaOH and 0.1 mol L-1 K2SO4 at pH 3 are used to obtain H2O2 selectivity and electron transfer numbers; the authors report near-100% selectivity and n ≈ 2.1 for 3% WO3/C in alkaline medium and about 80% selectivity for 1% WO3/C in acid. A gas diffusion electrode made from 1% WO3/C accumulated 862 mg L-1 H2O2 in 120 min at 100 mA cm-2 in pH 3 sulfate electrolyte. DFT calculations (PBE+U with an ab initio Hubbard U) using the computational hydrogen electrode on WO3 (001), (010), and (100) surfaces yield theoretical overpotentials that become smaller at high pH, which the authors use to rationalize the pH dependence of the experiments. The central claim is that WO3 modification of Vulcan XC-72 is a promising non-precious-metal catalyst for decentralized H2O2 electrosynthesis.","tokens_in":14994,"tokens_out":9437,"duration_ms":85120,"significance":"If the central numbers are correct, the work is a useful contribution to non-precious H2O2 electrocatalysis: the GDE accumulation is competitive, the catalyst preparation is simple, and the DFT part is not circular—the Hubbard U is computed ab initio, the CHE free energies are not fitted to the measured currents, and the pH correction is stated explicitly. The paper also makes falsifiable predictions about facet-dependent activity on WO3 surfaces. However, the current manuscript under-documents the RRDE conditions that produce the headline near-100% selectivity, provides no uncertainty quantification for the central quantities, and contains an apparent error in the DFT pH correction. These issues are fixable with additional reporting and re-analysis, so the contribution is promising but not yet ready in its present form.","major_comments":[{"comment":"The RRDE section specifies two possible ring configurations (gold, N = 0.28; platinum, N = 0.21) but does not state which ring was used in 1 mol L-1 NaOH and which in 0.1 mol L-1 K2SO4 at pH 3, nor the ring potential at which the ring current was collected. Because Eq. (1) divides the ring current by N, this choice materially changes the reported selectivity: a true selectivity of 95% recorded with N = 0.28 would appear as roughly 81% if N = 0.21 were the correct collection factor, and the opposite swap would produce unphysical values above 100%. Please specify the ring/electrolyte pairing, the ring potential, and whether the ring current was verified to lie on the H2O2 oxidation limiting-current plateau with no interfering oxidation currents.","section":"§2.5, Eq. (1), Fig. 4"},{"comment":"The pH correction of the CHE free energies is under-specified and appears to double-count the RHE reference. The two-electron ORR involves two proton-coupled electron transfers, so the pH term should enter as 2 kT ln 10 × pH rather than kT ln 10 × pH, and no pH correction should be applied if the overpotentials are already referenced to RHE, as in the experiments. Please state the potential reference scale used in Fig. 7 and justify the prefactor; with the current text, the reported pH-corrected overpotentials (η = 0.10 V and 0.30 V) are not uniquely determined.","section":"§3.3, Fig. 7"},{"comment":"The paper states that measurements were performed in duplicate, but no error bars or standard deviations are shown for the H2O2 selectivity or electron transfer number. The headline near-100% selectivity and n ≈ 2.1 for 3% WO3/C in alkaline medium therefore have no quantified uncertainty. Please report the spread of the duplicate (or, preferably, triplicate independent electrode) measurements and state how many independent datasets are represented in Fig. 4.","section":"§3.2, Fig. 4"},{"comment":"The current efficiency is stated as '80% around' but is never defined, and the reported 862 mg L-1 accumulation at 100 mA cm-2 is not reconciled with that number. For a 3.5 cm2 GDE at 100 mA cm-2 for 120 min in 350 mL of electrolyte, 80% current efficiency would correspond to about 1015 mg L-1 H2O2, whereas 862 mg L-1 corresponds to about 68% of the Faradaic charge. Please provide the current-efficiency formula, the charge passed, and a consistency check between Figs. 6a and 6b.","section":"§2.6, Fig. 6"},{"comment":"The DFT calculations model bare WO3 surfaces, while the experimental catalyst is WO3 supported on Vulcan XC-72, and the manuscript attributes the improved selectivity partly to oxygen functional groups and carbon defects on the support (Fig. 3). The theoretical results therefore do not, by themselves, establish that WO3 sites are responsible for the enhanced H2O2 selectivity of the composite. Either soften the theory–experiment confirmation claim or add calculations that include the carbon support or the WO3–carbon interface.","section":"§3.3 vs. §3.1/§3.2"}],"minor_comments":[{"comment":"The NaOH electrolyte is described as pH 14 in §2.5 and as pH 13 in the contact-angle and DFT sections; please use a single measured value.","section":"§2.5 and §3.1/§3.3"},{"comment":"The EIS frequency range '10-5 to 10-1 Hz' is likely a typographical error for 10^5 to 10^-1 Hz; please correct.","section":"§2.5"},{"comment":"Please state the sign convention for I_r and I_d; as written, the reader must assume I_d is negative and I_r positive for the formula to yield values between 0 and 1.","section":"Eqs. (1)–(2)"},{"comment":"The manuscript refers to Figures S1–S3 and Table S1, but the supplementary material is not included in the provided version; please ensure the SI is submitted and that all cited items exist.","section":"Supplementary material"},{"comment":"The GDE demonstration is performed only for 1% WO3/C; a pure Vulcan XC-72 GDE control and, ideally, an alkaline GDE test would make the comparison with the RRDE selectivity claims stronger.","section":"§2.6, Fig. 6"}],"recommendation":"major_revision","confidential_remarks":"The central idea is sound and the paper is within scope for an electrochemistry/materials journal. The main issues are documentation and consistency rather than incurable flaws; with the ring conditions, CE calculation, and DFT pH correction fixed, I would support publication. I would also ask the editor to ensure the supplementary material is complete, since several figures and a table cited in the text are not present in the submitted version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this is a solid, workmanlike paper from a group that has been systematically testing oxide-modified Vulcan for H2O2. The specific new thing is WO3 nanoflowers on Vulcan XC72, tested in both acid and alkaline media, with a gas-diffusion-electrode accumulation of 862 mg/L H2O2 at 100 mA/cm2 and current efficiency around 80%. That GDE result alone is useful for decentralized water treatment.\n\nWhat the paper does well: the synthesis and characterization are careful (XRD, Raman, XPS, contact angle), the pH comparison is a legitimate extension of their earlier Nb2O5, ceria, and MnO2 work, and the DFT on WO3 facets is independent — the Hubbard U is computed ab initio, the CHE free energies are not fit to the measured currents, and the pH correction uses a standard formula. So circularity is low, exactly as the reader's report says.\n\nThe main soft spot, which the stress-test flags, is real. Section 2.5 lists a gold ring with N=0.28 and a platinum ring with N=0.21 but never states which ring was used in 1 mol/L NaOH vs 0.1 mol/L K2SO4 at pH 3, nor the ring detection potential. The near-100% selectivity for 3% WO3/C in alkaline medium is computed from Eq. (1) and is sensitive to that choice. Swapping the collection efficiency changes a reported value near 95% to roughly 81%, or gives an unphysical >100% in the opposite direction. So the qualitative claim that WO3 improves selectivity probably survives, but the headline quantitative numbers are underdetermined. Related issues: no error bars on the central electrochemical numbers, and the DFT overreaches slightly — it models bare WO3 surfaces, not the WO3/carbon interface, so saying theory confirms the synergistic effect is too strong.\n\nMinor stuff: the text has typos and some ambiguous phrasing, but nothing that obscures the methods.\n\nWho is this for? People working on carbon/oxide hybrids for H2O2 electrosynthesis, especially the GDE community. It deserves a serious referee. The core experimental claim is likely reproducible and useful, but the RRDE ring specification and error bars need to be addressed before publication. I'd send it to peer review with a request for major revision.","headline":"A useful extension of the group's oxide-on-carbon H2O2 work with a strong GDE result, but the RRDE selectivity numbers are under-specified and need fixing before I'd trust the near-100% claim.","tokens_in":15754,"tokens_out":1490,"would_cite":true,"duration_ms":15087,"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":"Adding WO3 nanoflowers to Vulcan XC-72 shifts the oxygen reduction reaction toward two-electron H2O2 production, with near 100% selectivity for a 3% loading in alkaline medium and 862 mg/L accumulated in a gas-diffusion-electrode…","keywords":["oxygen reduction reaction","hydrogen peroxide electrosynthesis","tungsten trioxide","Vulcan XC-72","two-electron ORR","rotating ring-disk electrode","gas diffusion electrode","DFT overpotential"],"falsifier":"Measure the actual collection efficiency of the gold and platinum rings in O2-saturated 1 M NaOH and 0.1 M K2SO4 (pH 3) with a known one-electron redox couple, then recompute the Figure 4 selectivity; if the near-100% value for 3% WO3/C drops below about 90% or the electron number rises above about 2.3, the central claim fails.","tokens_in":14516,"feed_emoji":"🧪","tokens_out":7267,"duration_ms":59233,"temperature":0.7,"pith_summary":"The paper argues that decorating Vulcan XC-72 carbon with small amounts of solvothermally synthesized WO3 nanoflowers redirects the oxygen reduction reaction (ORR) toward the two-electron pathway that produces hydrogen peroxide instead of the four-electron pathway that produces water. In alkaline solution, the 3% WO3/C catalyst reaches near 100% H2O2 selectivity over a wide potential range with about 2.1 electrons transferred; in acid, the 1% WO3/C catalyst reaches about 80% selectivity. A gas diffusion electrode made from 1% WO3/C accumulated 862 mg/L of H2O2 after 120 minutes at 100 mA/cm2 in pH 3 sulfate electrolyte. The authors attribute the improved selectivity to extra oxygen functional groups, better hydrophilicity, and a synergistic effect of the WO3 nanoflowers, with DFT calculations showing low overpotentials on the (001) and (010) WO3 surfaces.","feed_headline":"Tungsten oxide on carbon nears 100% H2O2 selectivity","feed_subtitle":"A 3% WO3/Vulcan electrode makes peroxide almost exclusively in base; a gas-diffusion version produced 862 mg/L.","key_machinery":"The central object is the WO3 nanoflower-decorated Vulcan XC-72 electrode, evaluated by rotating ring-disk electrode (RRDE) voltammetry. Selectivity and electron-transfer number are computed from the ring and disk currents as $X_{H_2O_2} = \\frac{2 I_r/N}{-I_d + I_r/N}$ and $n = 2(X_{H_2O} + 1)$, using a constant collection efficiency $N = 0.28$ (gold ring) or $N = 0.21$ (platinum ring). On the theory side, the argument is carried by the computational hydrogen electrode model, in which the Gibbs free energy of the *OOH intermediate, with an ideal value near $4.2 \\pm 0.2$ eV, is the activity descriptor; overpotentials are read from free-energy diagrams on the (001), (010), and (100) surfaces.","core_discovery":"The central claim is that WO3/Vulcan XC-72 is a non-precious electrocatalyst with higher selectivity for the two-electron oxygen reduction to H2O2 than plain Vulcan in both acidic and alkaline media, with the strongest effect at 3% loading in alkaline medium. The paper further claims that the pH dependence reflects a mechanistic change: in alkaline medium the rate-determining step is a non-coupled proton transfer, while in acid a proton-coupled electron transfer dominates and favors deeper reduction to water. DFT with the computational hydrogen electrode identifies the (001) and (010) monoclinic WO3 surfaces as the active ones, with a theoretical overpotential as low as 0.10 V at pH 13 on (010) and 0.23 V at pH 0, while the (100) surface binds the *OOH intermediate too strongly and is easily poisoned.","pith_inferences":["The facet-dependent DFT results suggest a testable design rule: WO3 morphologies that expose more (010) or (001) surface per gram should outperform nanoflowers in alkaline H2O2 yield; comparing nanorods, nanosheets, and nanoflowers would settle this.","Because the RRDE geometry data are reported in alkaline medium while the GDE electrolysis was run in acid, an explicit bridge experiment—GDE electrolysis in 1 M NaOH—would show whether the near-100% selectivity survives at practical current densities.","The DFT pH correction is a simple free-energy shift of $kT\\ln 10 \\times \\mathrm{pH}$; a fuller treatment with explicit water or a microkinetic model could change the predicted overpotential ordering, especially on the (100) surface."],"forward_implications":["If correct, WO3/Vulcan is a low-cost, scalable cathode material for decentralized H2O2 electrosynthesis, an alternative to noble-metal catalysts.","Alkaline conditions are the more favorable operating regime: the highest two-electron selectivity appears at a large potential range with low overpotential.","The 1% WO3/C gas diffusion electrode at 100 mA/cm2 yields 862 mg/L H2O2 with about 80% current efficiency, a practical level for electro-Fenton water treatment.","The identification of (001) and (010) as the active WO3 facets implies that morphology control of the oxide can further tune selectivity."],"supporting_citations":[{"why":"supplies the computational hydrogen electrode method used for the free-energy diagrams and overpotential values","marker":"[28]"},{"why":"sets the ideal *OOH binding energy near 4.2 eV as the selectivity descriptor for the two-electron ORR","marker":"[62]"},{"why":"gives the ab initio procedure for determining the Hubbard U used in the DFT calculations","marker":"[27]"},{"why":"supports the claim that oxygen functional groups on carbon surfaces enhance two-electron ORR selectivity","marker":"[34]"},{"why":"provides evidence that electrolyte pH changes the ORR mechanism and selectivity on metal-free catalysts","marker":"[17]"},{"why":"explains the proton-affinity argument for why acidic conditions reduce H2O2 selectivity on carbon catalysts","marker":"[3]"},{"why":"provides the gas-diffusion-electrode fabrication and current-efficiency methodology used for the H2O2 accumulation runs","marker":"[23]"},{"why":"previous GDE work with oxide-modified Vulcan that the hot-pressed electrode preparation is based on","marker":"[12]"},{"why":"the periodic DFT code used to build and relax the WO3 slab models","marker":"[24]"}],"fun_headline_variants":["WO3/C catalyst hits 862 mg/L H2O2","Nanoflower WO3 enhances H2O2 yield in ORR","Tungsten oxide on carbon ups peroxide selectivity","pH-dependent ORR: WO3/C makes H2O2 selectively"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results stand on the assumption that the ring-disk collection efficiency N is the same constant in every measurement, but the paper never specifies which ring was used for each electrolyte, so a wrong or drifting N would change every selectivity number.","fun_headline_variants_meta":{"raw":{"variants":["WO3/C catalyst hits 862 mg/L H2O2","Nanoflower WO3 enhances H2O2 yield in ORR","Tungsten oxide on carbon ups peroxide selectivity","pH-dependent ORR: WO3/C makes H2O2 selectively"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000673,"raw_usage":{"total_tokens":3097,"prompt_tokens":1013,"completion_tokens":2084,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":629,"completion_tokens_details":{"reasoning_tokens":2011}},"tokens_in":629,"tokens_out":2084,"duration_ms":16833,"temperature":1.0,"reasoning_tokens":2011,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:10:27.127944+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the actual collection efficiency of the gold and platinum rings in O2-saturated 1 M NaOH and 0.1 M K2SO4 (pH 3) with a known one-electron redox couple, then recompute the Figure 4 selectivity; if the near-100% value for 3% WO3/C drops below about 90% or the electron number rises above about 2.3, the central claim fails.","supporting_citations":[{"cited_title":"Enabling direct H2O2 production through rational electrocatalyst design","cited_arxiv_id":null,"evidence_quote":"sets the ideal *OOH binding energy near 4.2 eV as the selectivity descriptor for the two-electron ORR"},{"cited_title":"High-efficiency oxygen reduction to hydrogen peroxide catalysed by oxidized carbon materials","cited_arxiv_id":null,"evidence_quote":"supports the claim that oxygen functional groups on carbon surfaces enhance two-electron ORR selectivity"},{"cited_title":"Effect of pH on the Activity of Platinum Group Metal-Free Catalysts in Oxygen Reduction Reaction","cited_arxiv_id":null,"evidence_quote":"provides evidence that electrolyte pH changes the ORR mechanism and selectivity on metal-free catalysts"},{"cited_title":"Origin of Selective Production of Hydrogen Peroxide by Electrochemical Oxygen Reduction","cited_arxiv_id":null,"evidence_quote":"explains the proton-affinity argument for why acidic conditions reduce H2O2 selectivity on carbon catalysts"},{"cited_title":"Hydrogen peroxide electrosynthesis: A comparative study employing Vulcan carbon modification by different MnO2 nanostructures","cited_arxiv_id":null,"evidence_quote":"provides the gas-diffusion-electrode fabrication and current-efficiency methodology used for the H2O2 accumulation runs"},{"cited_title":"Using a novel gas diffusion electrode based on Vulcan XC-72 carbon modified with Nb2O5 nanorods for enhancing H2O2 electrogeneration","cited_arxiv_id":null,"evidence_quote":"previous GDE work with oxide-modified Vulcan that the hot-pressed electrode preparation is based on"}],"review_version":1}