{"id":"a562b203-db36-4761-b5c7-0199ab621964","arxiv_id":"1909.01204","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Oxygen incorporated into antimonene sheets forms stable 2D antimony oxide semiconductors with predicted band gaps of 2.0 to 4.9 eV and distinctive Raman spectra.","lead":"Using density functional theory, this paper predicts several new two-dimensional antimony oxide structures in which oxygen atoms bond inside the antimony sheet, not just on top. These predicted layers are semiconductors with band gaps between roughly 2 and 5 electron volts, and the paper gives Raman fingerprints to identify them in experiments.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 4.9 eV upper band-gap claim rests on an idealized Sb2O3 monolayer that the authors themselves report is dynamically unstable; the gap of the stable relaxed geometry is never computed.","rationale":"The reader's weakest assumption correctly identifies that the 4.9 eV band gap is computed for an idealized Sb2O3 monolayer that is itself reported to be dynamically unstable and to relax into a distorted geometry in a larger cell. This is the most load-bearing concern because the band-gap range is a central quantitative claim of the paper, and the upper endpoint is exactly the value that lacks support from a stable geometry. The concern is not merely a disagreement with consensus; it is an internal admission in the manuscript that the structure used for the headline gap is not the ground-state arrangement. The paper does provide independent support for the broader type (II) concept through energy comparisons, phonon calculations, and molecular dynamics, and the lower-band-gap structures (e.g., Sb2O2) appear stable. Therefore the appropriate verdict remains CONDITIONAL: the central concept is credible, but the claimed 2.0–4.9 eV range should be re-evaluated or re-qualified once the stable Sb2O3 monolayer geometry is properly characterized. No adjustment to the reader's verdict is needed, hence UNCHANGED.","tokens_in":8806,"tokens_out":2372,"duration_ms":22277,"concrete_test":"Take the fully relaxed, distorted Sb2O3 monolayer geometry obtained from the larger unit cell (Fig. S4) and recompute its electronic band structure with the same HSE12 functional including spin-orbit coupling, using identical numerical parameters as for the idealized structure. Report the fundamental band gap and its direct/indirect character, and compare with the 4.9 eV value. Also verify dynamical stability by computing the phonon dispersion of the distorted geometry. If the gap differs substantially (e.g., by more than 0.5 eV) or the relaxed structure remains dynamically unstable, the upper endpoint of the claimed 2.0–4.9 eV range must be revised or explicitly qualified as pertaining only to an idealized, unstable geometry.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim includes a band-gap range of 2.0–4.9 eV for type (II) single-layer antimonene oxides. The upper endpoint, 4.9 eV, is computed for the idealized Sb2O3 monolayer shown in Fig. 1(j). However, the text explicitly states that this structure 'is not fully dynamically stable and relaxes into a slightly distorted geometry if a larger unit cell is used (Fig. S4)', and the idealized geometry is retained only for convenience because the energy difference is 5 meV per formula unit. This creates a mismatch between the structure whose gap is reported and the structure that would actually form. The authors do not provide the electronic band structure of the fully relaxed distorted geometry, so the 4.9 eV value may not correspond to any physically realizable monolayer. This is load-bearing because the tunable range of 2.0–4.9 eV is a headline result, and the upper end is precisely the part that is not anchored to a stable or dynamically accessible configuration. The molecular dynamics simulations described in the text (starting from antimonene plus oxygen atoms) are said to produce structures 'resembling' Sb2O3, but the band gaps of those MD-equilibrated structures are not reported either, so the gap of the realistic product of oxidation remains unknown. The lower endpoint (2.0 eV for Sb2O2) is on a dynamically stable structure and is not affected. Thus the concern is specific: the claimed upper range is not supported by a calculation on a stable geometry, and the paper itself supplies the admission.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a family of two-dimensional antimony oxide structures, termed type (II), in which oxygen atoms are incorporated into the antimonene plane with Sb–O–Sb bridge bonds, in contrast to the type (I) structures with perpendicular Sb=O double bonds. Using DFT total-energy comparisons, phonon calculations, and molecular dynamics, the authors argue that type (II) structures are significantly more stable than type (I) structures, and they report a stoichiometry-dependent band-gap range of approximately 2.0–4.9 eV for type (II) monolayers, computed with the HSE12 hybrid functional. They also provide calculated Raman-active mode frequencies for experimental identification and claim qualitative agreement with Raman spectra of oxidized few-layer antimonene reported in an unpublished companion work.","tokens_in":9097,"tokens_out":2360,"duration_ms":24286,"significance":"If the central claims hold, the paper offers a plausible structural model for the natural oxidation of antimonene, with technologically relevant tunable band gaps and clear spectroscopic fingerprints. The work is significant because it challenges the previously assumed double-bonded oxide structure and provides a concrete alternative that is more stable at the DFT level. Strengths include the internally consistent use of frozen-phonon distortions from calculated unstable modes to construct type (II) structures, the corroborating molecular dynamics simulations, and the explicit energy lowerings (about 3.9 eV for 1L and 2.7 eV for 2L) that make the type (I)-to-type (II) transformation physically plausible. The Raman mode predictions, if anchored to published experimental data, would be practically useful. However, the upper end of the reported band-gap range rests on a dynamically unstable idealized structure, and the key experimental corroboration is an unpublished manuscript with overlapping authorship, so the significance of the headline claims is currently conditional.","major_comments":[{"comment":"The central claim of a tunable band-gap range of 2.0–4.9 eV is not fully supported because the 4.9 eV value for the Sb2O3 monolayer is computed for the idealized structure of Fig. 1(j), which the authors themselves state is not fully dynamically stable and relaxes into a slightly distorted geometry in a larger unit cell (Fig. S4). The energy difference of only 5 meV per formula unit does not guarantee that the electronic structure, especially the fundamental gap, is unchanged; the band structure of the stable distorted geometry is not reported. Since the upper endpoint of the headline range is precisely the value that is not anchored to a dynamically stable structure, the authors should either compute and report the band structure of the fully relaxed geometry at the same HSE12 level, or explicitly revise the claimed range to exclude the unstable idealized case. This issue is load-bearing for the paper's main conclusion.","section":"Electronic properties, Fig. 4(c) and text near Fig. 1(j)"},{"comment":"The claimed agreement with experimental Raman spectra of oxidized antimonene relies entirely on Ref. [33], which is unpublished and shares several authors with the present manuscript. No experimental spectra, peak positions, or quantitative fitting are shown. As a result, the confirmation loop is not closed: the type (II) structures are presented as consistent with data that the reader cannot independently inspect. The authors should either include the experimental Raman data and a quantitative comparison, or restrict themselves to stating that the predicted modes fall in a plausible range and defer experimental validation to future published work.","section":"Comparison with experiment, Ref. [33] and Raman discussion"},{"comment":"The trilayer type (II) structures used to motivate the stoichiometries and to illustrate the phase transformation impose up to 17% strain on the inner non-oxidized antimonene layer, as the authors acknowledge. This large strain makes the trilayer geometries physically unrealistic as direct models of oxidized few-layer antimonene. The paper does not examine whether the type (II) motifs survive in larger commensurate supercells with lower strain, nor does it report the electronic or vibrational properties of such relaxed trilayer models. Given that the paper aims to describe few-layer antimonene oxides, this limitation deserves a quantitative treatment or an explicit caveat that only the monolayer and bilayer results are representative.","section":"Trilayer structures and strain, text near Figs. 1(e)-(i)"}],"minor_comments":[{"comment":"The abstract states that the structures range \"from topological insulators to semiconductors,\" but no type (II) structure is predicted to be a topological insulator in the manuscript; the only topological-insulator statement concerns the type (I) monolayer in Ref. [22]. The wording should be aligned with the actual content.","section":"Abstract and summary"},{"comment":"The caption says the height of the bars indicates Raman activity on a logarithmic scale, but the axis label \"log. Raman activity\" and the use of arbitrary units are potentially confusing. Please clarify the normalization and the meaning of the y-axis.","section":"Fig. 3 caption"},{"comment":"The phrase \"Sb2O\" is used both as a stoichiometric label and as a structure name; for example, \"1L, type (II)\" is labeled Sb2O in Fig. 1(f), which is consistent but not defined in the main text. A sentence defining the notation (number of Sb and O atoms per unit cell of one outer layer) would help.","section":"Text near Fig. 1"},{"comment":"The manuscript cites \"Sec. 5 of the Supplemental Material\" for several points, but the Supplemental Material itself is not provided to the reader in the main text and the structure of that document is not described. Please ensure the supplementary document is available and that all cross-references (Fig. S4, S5-S7, S8-S12) are correct.","section":"References"},{"comment":"There are minor typographical errors, such as \"stochiometry\" in the summary and \"computational resources used for the calculations\" being a sentence fragment. The manuscript should be carefully proofread.","section":"General text"}],"recommendation":"major_revision","confidential_remarks":"The core scientific idea—that oxygen incorporation into the antimonene plane is thermodynamically favored over double-bonded oxygen—is plausible and well supported by the energy and phonon data. However, the headline band-gap range is compromised by the unstable Sb2O3 monolayer, and the experimental confirmation depends on an unpublished manuscript from the same group. The authors can likely fix both issues with additional calculations and by presenting the published experimental data, which is why I recommend major revision rather than rejection. I also note that the novelty of the structural search is moderate: the paper uses a direct frozen-phonon construction rather than an unbiased global search, so the claim that these are the ground-state oxide structures should be tempered."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the type-II antimonene oxide family: oxygen bridging two antimony atoms in Sb-O-Sb chains, rather than the Sb=O double-bonded type-I picture from Ref. [22]. The paper makes a coherent case that these chainlike structures are more stable, with energy lowerings of ~3.9 eV (1L) and 2.7 eV (2L) against type-I, negative phonon frequencies for the type-I monolayer, and MD runs that spontaneously form Sb-O-Sb chains. The Raman spectra and phonon-mode assignments are a useful, concrete experimental handle, and the authors are honest that type-I is metastable at best. I give them credit for that: the structural argument is internally consistent and the frozen-phonon route from instability to a new structure is sensible, not circular. The Sb2O2 result, a direct 2.0 eV gap on a dynamically stable geometry, looks solid. My main reservation is exactly the one you flagged: the 4.9 eV gap for Sb2O3 is computed on the idealized monolayer of Fig. 1(j), which the text admits relaxes into a slightly distorted geometry in a larger cell. The energy penalty is only 5 meV per formula unit, so the idealized structure is not a crazy approximation, but the gap of the relaxed geometry is never computed. Since the tunable 2.0–4.9 eV range is the headline claim, the upper endpoint being unanchored to a stable geometry is a real soft spot, and it is load-bearing in the abstract and summary. I also wish the MD-equilibrated structures' band gaps had been reported, since those are what would actually form in air. Minor point: the confirmation via Ref. [33] is an unpublished, same-group report, so that experimental agreement is suggestive, not independent. The paper is a standard DFT pipeline with no error bars and no deposited input files, but the methods (HSE12, phonons, MD) are appropriate and the results are clearly presented. Overall the central type-II stability concept holds up; the issue is a specific missing calculation, not a fundamental flaw. Who is this for? Anyone working on pnictogen 2D materials, oxidation/passivation of antimonene, or 2D oxide heterostructures. It deserves a serious referee: the structural prediction is new and potentially useful, and the authors can fix the gap question in revision. I would not cite the 4.9 eV endpoint as is, but I would cite the type-II structures and the 2.0 eV Sb2O2 gap. Recommend: send to peer review, ask for the relaxed Sb2O3 geometry's band structure and MD-equilibrated gaps.","headline":"A credible DFT prediction of chainlike type-II antimonene oxides with a tunable 2.0–4.9 eV gap range, but the 4.9 eV upper endpoint rests on a dynamically unstable idealized geometry and needs a follow-up on the relaxed structure.","tokens_in":9634,"tokens_out":679,"would_cite":true,"duration_ms":7738,"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":"Antimonene oxidation is predicted to yield stable, tunable semiconducting oxide layers.","keywords":["antimonene","antimony oxide","two-dimensional materials","density functional theory","band gap engineering","Raman spectroscopy","oxidation","heterostructures"],"falsifier":"Compute the HSE12 band structure of the fully relaxed distorted Sb$_2$O$_3$ monolayer (the larger-unit-cell geometry the paper mentions): a gap far from 4.9 eV would remove the upper end of the claimed range. Experimentally, measuring the optical absorption edge of an oxidized single-layer antimonene flake and finding no absorption onset between 2.0 and 4.9 eV would contradict the semiconductor claim.","tokens_in":8590,"feed_emoji":"⚛️","tokens_out":8758,"duration_ms":79131,"temperature":0.7,"pith_summary":"Using density functional theory, the paper predicts a family of two-dimensional antimony oxide layers in which oxygen atoms are incorporated into the antimonene plane, each bonded to two antimony atoms in Sb-O-Sb chains (type II structures). It argues that these are markedly more stable than the previously assumed double-bonded Sb=O structures (type I), by about 3.9 eV for a monolayer and 2.7 eV for a bilayer, and that every type-II monolayer is a semiconductor with a stoichiometry-dependent gap between roughly 2.0 and 4.9 eV. The paper further computes Raman-active phonon modes, giving fingerprints to identify each oxide phase experimentally, and reports that the predicted Sb$_2$O$_3$ modes fit Raman spectra measured on oxidized few-layer antimonene. A sympathetic reader would care because this turns oxidation from a degradation problem into a possible knob: controlled oxidation could tune antimonene from semimetallic few-layer behavior to semiconducting layers spanning the visible-to-ultraviolet range, and might naturally form semiconductor/semimetal heterostructures.","feed_headline":"Oxygen woven into antimonene makes stable semiconducting layers","feed_subtitle":"DFT predicts gaps from 2.0 to 4.9 eV, with Raman fingerprints to identify each oxide phase.","key_machinery":"The central object is the type (II) antimonene oxide monolayer: a two-dimensional lattice in which every oxygen bridges two antimony atoms, forming chains of alternating Sb and O that are cross-linked by Sb-Sb or Sb-O-Sb bonds, in place of the perpendicular Sb=O double bonds of type (I). The argument is carried by density functional theory: hybrid-functional (HSE12) band-structure calculations set the gaps, phonon calculations supply the Raman-active frequencies, and the type (II) geometries are generated by displacing atoms along unstable phonon modes of type (I) structures (a frozen-phonon approach) and corroborated by 300 K molecular dynamics in which oxygen atoms enter the antimonene plane. The electronic-structure mechanism is the oxygen content: adding oxygen raises the gap from 2.0 to 4.9 eV and changes the gap from direct to indirect, so the degree of oxidation acts as the tuning parameter.","core_discovery":"The central claim is that air-oxidized antimonene is not a random oxide coating but a family of ordered two-dimensional crystals whose properties follow from stoichiometry and bonding. Specifically, the paper states that all type (II) single-layer antimonene oxides it presents are semiconductors, with direct and indirect band gaps from approximately 2.0 eV (Sb$_2$O$_2$, direct at the zone edge) to 4.9 eV (Sb$_2$O$_3$, indirect), and that these chainlike oxides are energetically preferred over the perpendicular double-bonded type (I) geometry by roughly 3.9 eV (monolayer) and 2.7 eV (bilayer). Type (I) Sb$_2$O$_2$ is instead predicted to be a topological insulator with a small 168 meV gap once spin-orbit coupling is included, matching an earlier proposal. The paper concludes from molecular dynamics that oxygen spontaneously inserts into the antimonene sheet at room temperature, and that the resulting Sb$_2$O$_3$-like layers account for the Raman spectra observed on oxidized few-layer antimonene.","pith_inferences":["Inference: the oxygen-insertion motif may extend to other pnictogen monolayers such as arsenene or bismuthene, which could form analogous stable oxide layers instead of degrading; this is not asserted by the paper.","Inference: the upper bound of 4.9 eV rests on an idealized Sb$_2$O$_3$ monolayer that relaxes into a slightly distorted structure in larger cells, so real oxidized flakes may show a range or a lowered effective gap; measuring the optical absorption edge of single oxidized flakes would test this.","Inference: the distinct Raman frequencies suggest a practical oxidation monitor, watching the disappearance of the type (I) 823 cm$^{-1}$ mode and growth of type (II) modes, which the paper does not explicitly propose."],"forward_implications":["Type (I) double-bonded antimonene oxides are metastable at best: relaxation to type (II) lowers the monolayer energy by about 3.9 eV and the bilayer energy by about 2.7 eV.","Oxidation state controls the electronic character: type (II) Sb$_2$O$_2$ is a direct-gap semiconductor near 2.0 eV, Sb$_2$O$_3$ is an indirect-gap semiconductor near 4.9 eV, and few-layer antimonene itself is semimetallic.","Raman spectroscopy can distinguish the phases: type (I) shows a high-frequency Sb=O stretch near 823 cm$^{-1}$, while type (II) modes fall below about 600 cm$^{-1}$, with Sb$_2$O$_3$ matching published spectra of oxidized few-layer antimonene.","Natural heterostructures should form: semimetallic antimonene layers sandwiched between semiconducting oxide layers, because antimony reacts readily with oxygen.","Stoichiometries beyond Sb$_2$O$_3$ are unlikely to be stable as 2D layers, since the antimony atoms available for oxygen bonding are exhausted."],"supporting_citations":[{"why":"Supplies the prior type (I) fully oxidized monolayer with Sb=O double bonds, whose 168 meV topological-insulator gap the paper reproduces and whose geometry seeds the type (II) structures.","marker":"[22]"},{"why":"Provides the experimental Raman spectra of oxidized few-layer antimonene against which the predicted Sb$_2$O$_3$ type (II) modes are matched and type (I) is ruled out.","marker":"[33]"},{"why":"Defines the HSE12 hybrid functional used for all band-structure and band-gap calculations.","marker":"[34]"},{"why":"Reference GW value of about 2.4 eV for the monolayer antimonene gap, the baseline against which oxidation-induced gap changes are discussed.","marker":"[21]"},{"why":"Documents the stoichiometries and polymorphism of bulk antimony oxides, motivating the search for multiple 2D oxide structures.","marker":"[31]"}],"fun_headline_variants":["DFT predicts ordered 2D antimony oxides with tunable gaps","Antimonene oxide family: from topological insulator to semiconductor","Oxygen insertion yields stable 2D Sb-O layers with Raman fingerprints","2D antimony oxides: gaps from 2.0 to 4.9 eV predicted"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The 4.9 eV upper end of the claimed band-gap range is computed for an idealized Sb$_2$O$_3$ monolayer that the paper itself says is not fully dynamically stable and relaxes into a slightly distorted geometry in a larger unit cell, so the claim depends on that idealized geometry being representative of the real oxidized layer.","fun_headline_variants_meta":{"raw":{"variants":["DFT predicts ordered 2D antimony oxides with tunable gaps","Antimonene oxide family: from topological insulator to semiconductor","Oxygen insertion yields stable 2D Sb-O layers with Raman fingerprints","2D antimony oxides: gaps from 2.0 to 4.9 eV predicted"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000165,"raw_usage":{"total_tokens":1205,"prompt_tokens":857,"completion_tokens":348,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":473,"completion_tokens_details":{"reasoning_tokens":267}},"tokens_in":473,"tokens_out":348,"duration_ms":3993,"temperature":1.0,"reasoning_tokens":267,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:24:27.790829+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the HSE12 band structure of the fully relaxed distorted Sb$_2$O$_3$ monolayer (the larger-unit-cell geometry the paper mentions): a gap far from 4.9 eV would remove the upper end of the claimed range. Experimentally, measuring the optical absorption edge of an oxidized single-layer antimonene flake and finding no absorption onset between 2.0 and 4.9 eV would contradict the semiconductor claim.","supporting_citations":[{"cited_title":"Zhang, W","cited_arxiv_id":null,"evidence_quote":"Supplies the prior type (I) fully oxidized monolayer with Sb=O double bonds, whose 168 meV topological-insulator gap the paper reproduces and whose geometry seeds the type (II) structures."},{"cited_title":"Assebban, C","cited_arxiv_id":null,"evidence_quote":"Provides the experimental Raman spectra of oxidized few-layer antimonene against which the predicted Sb$_2$O$_3$ type (II) modes are matched and type (I) is ruled out."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the HSE12 hybrid functional used for all band-structure and band-gap calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reference GW value of about 2.4 eV for the monolayer antimonene gap, the baseline against which oxidation-induced gap changes are discussed."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents the stoichiometries and polymorphism of bulk antimony oxides, motivating the search for multiple 2D oxide structures."}],"review_version":1}