{"id":"6f20c976-12fe-4270-96e8-109d42fb26de","arxiv_id":"2608.11388","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Diffusion quantum Monte Carlo reverses the hybrid-DFT ordering in Na2-xMn3O7, making the localized oxygen-hole polaron lower in energy than the split polaron.","lead":"A quantum Monte Carlo study of Na2-xMn3O7 finds that the 'split' oxygen-hole state favored by hybrid DFT is destabilized by explicit many-body treatment, leaving a localized oxygen polaron as the lower-energy state. The result offers a benchmark for correlated oxides and warns that X-ray absorption alone cannot distinguish the two configurations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Single-determinant fixed-node bias is the load-bearing risk: a ~1 eV DMC reversal could reflect nodal flexibility differences rather than the true many-body ordering, especially since trial parameters were optimized against DMC energies.","rationale":"The reader's weakest assumption identifies the same central risk: fixed-node DMC with single-determinant Slater-Jastrow trial wavefunctions may be insufficient to resolve near-degenerate localized and split polaron states. My stress-test agrees and adds a specific mechanistic route: the trial-wavefunction parameters were themselves optimized against DMC energies, and because DMC is variational in the nodal surface, separate parameter optimization for each configuration can differentially improve one nodal surface and produce a spurious energy gap. The paper's own admission that single determinants may not be flexible enough, combined with the DMC[HSE] spin-density collapse of the split state, makes this the most load-bearing concern. The secondary concern about DFT-relaxed geometries is real but subordinate, since the energetic reversal is computed at fixed structures and could be modulated by relaxation. The reader's CONDITIONAL verdict already captures the need for multideterminant benchmarking and geometry checks, so no verdict change is needed; the concern reinforces, rather than overturns, that verdict. Credit is due for testing two trial-wavefunction classes and for reporting the HSE alpha dependence, but those checks do not isolate the fixed-node error because all trials remain single-determinant in form and share the parameter-optimization protocol.","tokens_in":12489,"tokens_out":3398,"duration_ms":34534,"concrete_test":"Run DMC with multideterminant trial wavefunctions generated from a CASSCF/CASCI expansion in the active space of O1a/O1b 2p orbitals and nearby Mn 3d orbitals, including both symmetry-equivalent localized determinants with equal weights, at the same HSE(alpha=0.15) geometries and with all trial parameters fixed before any DMC energy minimization. If the localized configuration remains lower by more than 0.5 eV with per-energy statistical error below 0.05 eV, the reversal is robust; if the gap shrinks below 0.3 eV or changes sign, single-determinant nodal bias is the cause. As a control, also rerun the existing HSE trial set with one a priori alpha for both configurations to verify that the reversal is not an artifact of DMC-driven parameter selection.","verdict_should_be":"UNCHANGED","load_bearing_attack":"To support the central claim, the DMC energy difference must be a reliable estimate of the many-body ground-state gap. The paper uses single-determinant Slater-Jastrow trial wavefunctions throughout, and fixed-node error is controlled only by these nodes. For two near-degenerate, symmetry-related oxygen-hole configurations, a single determinant cannot represent a balanced superposition of competing localizations; the paper itself concedes in Section V and Supplemental Section IX that 'single-determinant trial wavefunctions may not provide a sufficiently flexible description of the correlated many-body state.' This is not a generic worry: the DMC[HSE] spin density for the nominal split-polaron state collapses onto one oxygen site, showing that the projection is strongly influenced by the chosen node. The magnitude of the reported reversal (~1 eV) is several times the HSE preference (0.28 eV at alpha~0.15), yet no statistical or systematic error bars are reported. More importantly, Supplemental Section V states that trial parameters (alpha for HSE; U and V for PBE+U+V) were optimized using the DMC energy as a benchmark. Because DMC energies are variational with respect to the nodal surface, optimizing parameters separately for each configuration can differentially improve the localized determinant's node; the energy gap then measures variational freedom in the trial class rather than the true many-body ordering. The finite-size check (Supplemental Fig. S8) uses only one 2x2x2 supercell, so size consistency of the 1 eV gap is not established. The fixed DFT geometries are a secondary issue: if the true many-body state has different local relaxations, the comparison at DFT-relaxed structures could bias it. These points make the reversal a plausible but not yet fully established many-body result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports fixed-node diffusion Monte Carlo (DMC) calculations for two competing oxygen-hole configurations in layered Na2-xMn3O7: a localized oxygen polaron and a bond-centered split polaron. Hybrid HSE calculations favor the split state by about 0.28 eV at the DMC-optimized exact-exchange fraction alpha≈0.15, whereas DMC reverses the ordering, placing the localized polaron lower by approximately 1 eV. The authors show that the reversal is reproduced with HSE- and PBE+U+V-generated trial wavefunctions, that the DMC[HSE] spin density of the nominal split state partially collapses onto one oxygen site, and that the two configurations have similar calculated O K-edge spectra, so X-ray absorption alone cannot distinguish them. They conclude that hybrid functionals can qualitatively mis-rank competing oxygen-hole localization motifs and propose Na2-xMn3O7 as a benchmark system.","tokens_in":12799,"tokens_out":6567,"duration_ms":58242,"significance":"If the energetic reversal is correct, the paper is a valuable demonstration that hybrid DFT can stabilize an intermediate oxygen-hole state that is not the many-body ground state, and it provides a clear benchmark for correlated methods. The work uses state-of-the-art DMC with correlation-consistent effective core potentials, tests two trial-wavefunction classes, employs twist averaging, and presents spin-density analyses that go beyond energy differences. The identification of a spectroscopic ambiguity for localized versus split oxygen holes is an important methodological insight. However, because the central claim rests on fixed-node DMC with single-determinant trial wavefunctions whose parameters were optimized against DMC energies, the result should be regarded as a strong conditional benchmark rather than a definitive many-body ordering.","major_comments":[{"comment":"No error bars are reported for the DMC total-energy differences. The central quantitative claim is a ~1 eV reversal, yet the stochastic uncertainty of DMC, the timestep error, and the twist-averaging/finite-size corrections are not combined into an uncertainty on ΔE. Please report the statistical error on each DMC energy and on ΔE, and state which systematic contributions have been estimated.","section":"Fig. 2(d), main text 'DMC instead predicts...'"},{"comment":"The trial-wavefunction parameters were optimized by minimizing DMC energies: alpha for HSE and U,V for PBE+U+V. Because fixed-node DMC energies are variational with respect to the nodal surface, optimizing parameters separately for the localized and split configurations can differentially improve the nodes and thereby bias the energy gap. The manuscript itself concedes that single-determinant trial wavefunctions may not provide a sufficiently flexible description. I request (i) a demonstration that ΔE remains negative when a single common parameter set is used for both configurations, (ii) a report of ΔE across the full scanned parameter range rather than at per-configuration minima, and (iii) a discussion of possible fixed-node bias, ideally with a multideterminant or orbital-optimized trial wavefunction test.","section":"Supplemental Section V, main text Section V"},{"comment":"The DMC[HSE] spin density for the nominal split-polaron state collapses onto a single oxygen site, and the paper states that the system yields two localized variants. The energy comparison is therefore not between a stable delocalized split state and a localized state, but between two states whose character changes under the many-body projection. This collapse is consistent with the fixed-node concern raised above and should be addressed explicitly: would a balanced trial wavefunction restore a split density, and if so, what would happen to the ordering? Without this, the statement that DMC 'destabilizes' the split state is ambiguous.","section":"Fig. 3(c,d), Fig. 3(e,f)"},{"comment":"The finite-size test uses only one 2x2x2 supercell, so no actual extrapolation is performed; the main text refers to 'supercell extrapolation procedures' and 'finite-size extrapolation' but only one supercell size is presented. Please provide additional supercell sizes or a quantitative estimate of the residual finite-size error.","section":"Supplemental Section VI, Fig. S8"},{"comment":"The DMC energies are evaluated at DFT-relaxed geometries for each polaron configuration; no many-body geometry relaxation or force calculation is reported. Since polaron stability is intimately tied to lattice distortion, the 1 eV reversal could be modified by geometry relaxation in the many-body state. Please state the sensitivity of ΔE to the geometries or test with DMC-based relaxation.","section":"Supplemental Section I (Computational Details)"}],"minor_comments":[{"comment":"The figure numbering is inconsistent: the main text refers to 'Supplemental Material Fig. S5' for finite-size extrapolation, while the Supplemental Material labels that figure as Fig. S8 (and Fig. S5 is the DFT+U energy landscape). Please correct the cross-references.","section":"Main text and Supplemental Material"},{"comment":"The phrase 'DMC calculations based on PBE trial wavefunctions' should read 'PBE+U+V trial wavefunctions' to match the rest of the manuscript and the parameter scans reported in Fig. 2(e,f).","section":"Main text, Fig. 3(e,f)"},{"comment":"The caption says that neither the localized nor the split configuration reproduces the experimentally observed pre-edge feature well, while the abstract emphasizes that the two configurations produce similar O K-edge features; please clarify that the similarity is between the two calculated spectra, not with experiment.","section":"Fig. 1(c) caption and abstract"},{"comment":"The manuscript alternates between Na2-xMn3O7 and NaMn3O7; please specify the exact supercell composition, number of formula units, and net charge used in the DMC calculations, since the hole count is central to the comparison.","section":"Throughout"},{"comment":"Consider reporting the DMC energy differences in a table with statistical uncertainties and the corresponding parameter values, as the figures are difficult to read at the resolution provided.","section":"Fig. 2(d-f)"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope and potentially important. My main concern is the fixed-node bias combined with variational parameter optimization against DMC energies; the current evidence does not fully support the strong 'reversal by ~1 eV' claim as stated. If the authors provide error bars, a fixed-node sensitivity test, and a more complete finite-size analysis, the paper could become acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: the central result is new. Prior HSE work on Na2-xMn3O7 predicted a bond-centered split oxygen-hole polaron; this paper's DMC calculations reverse that, putting the localized polaron lower by about 1 eV. If that holds, it is a genuinely useful benchmark for oxygen-redox materials and a clear caution about hybrid functionals in correlated oxides.\n\nWhat the paper does well: the DMC setup is state-of-the-art, with two trial-wavefunction classes (HSE and PBE+U+V) giving the same qualitative ordering. The spin-density analysis is a real addition — it shows the DMC[HSE] solution partially collapses the split state toward a localized one, and the point that O K-edge spectra cannot distinguish the two configurations is well made. The authors are also honest about their own limitations, explicitly conceding in Section V and Supplemental Section IX that single-determinant trial wavefunctions may not be flexible enough for this near-degenerate problem.\n\nNow the soft spots, in proportion. The load-bearing risk is fixed-node bias. The DMC energy is variational with respect to the nodal surface, and the paper tunes alpha, U, and V by minimizing DMC energies separately for each configuration. That can differentially improve one trial wavefunction's node, so the ~1 eV gap partly measures variational freedom in the trial class rather than the true many-body ordering. The spin-density collapse in DMC[HSE] is a concrete symptom: a single determinant cannot represent a balanced superposition of competing localizations, so the projection picks a side. Also, no statistical error bars are reported anywhere in the main text — for a reversal that is several times the HSE preference (0.28 eV at alpha~0.15), that is a real omission. The finite-size check uses a single 2x2x2 supercell, and the DMC evaluations are done at DFT-relaxed geometries. These are all acknowledged or visible in the text, but together they mean the ordering is plausible, not proven.\n\nNone of this kills the paper. The question is important, the QMC work is serious, and the authors have flagged the weakest link themselves. Who will get value from this: computational materials scientists working on oxygen redox, polaron benchmarking, and method validation; also spectroscopists who rely on X-ray absorption to infer hole localization.\n\nMy recommendation: yes, send it to referees. It deserves careful peer review, and a good referee report should ask for reported error bars, a second supercell size or a stronger twist-grid test, and ideally a backflow or multideterminant check before the 1 eV reversal is taken as the accepted ordering. As it stands, treat it as a strong conditional result.","headline":"The DMC reversal of the HSE split-polaron preference in Na2-xMn3O7 is a new and plausible result, but the single-determinant fixed-node limitation and the absence of reported error bars keep it a conditional benchmark rather than a settled ordering.","tokens_in":13423,"tokens_out":1932,"would_cite":true,"duration_ms":21328,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.38.-k","71.15.-m","71.28.+d"],"model":"deepseek-v4-flash","headline":"Diffusion quantum Monte Carlo reverses hybrid DFT's energetic ordering of oxygen-hole polarons in Na2-xMn3O7, stabilizing a localized polaron by about 1 eV.","keywords":["oxygen holes","polarons","hybrid DFT","diffusion quantum Monte Carlo","electron correlation","Na2Mn3O7","ligand holes","fixed-node approximation"],"falsifier":"Run fixed-node DMC with multideterminant trial wavefunctions (e.g., generated from complete-active-space or selected-CI calculations) for both polaron configurations in NaMn3O7; if the split-polaron state becomes lower in energy by more than about 0.5 eV, the claimed reversal is an artifact of the fixed-node approximation.","tokens_in":12307,"feed_emoji":"⚛️","tokens_out":3612,"duration_ms":56968,"temperature":0.7,"pith_summary":"This paper claims that the bond-centered split oxygen-hole polaron predicted by hybrid DFT in layered Na2-xMn3O7 is not a genuine many-body ground state. Using fixed-node diffusion quantum Monte Carlo (DMC), the authors find that the localized oxygen polaron is lower in energy by about 1 eV, reversing the hybrid DFT ordering. The reversal holds for trial wavefunctions generated from both HSE and PBE+U+V calculations. Because localized and split configurations yield similar O K-edge spectra, the paper argues that X-ray absorption cannot distinguish them and that explicit many-body benchmarks are necessary.","feed_headline":"QMC overturns hybrid DFT: oxygen holes localize","feed_subtitle":"Explicit many-body treatment stabilizes the localized polaron by about 1 eV over the split state HSE favors.","key_machinery":"The key machinery is fixed-node diffusion quantum Monte Carlo (DMC) with single-determinant Slater-Jastrow trial wavefunctions, evaluated at DFT-relaxed geometries. DMC provides an explicit many-body treatment of electron correlation and yields variational energies with respect to the trial nodal surface. The comparison relies on twist-averaged finite-size extrapolation and on varying the exact-exchange fraction (for HSE) and Hubbard U and inter-site V parameters (for PBE+U+V) to optimize trial wavefunctions. Many-body spin densities, obtained via mixed-estimator extrapolation, serve as the observable fingerprint distinguishing localized from split polarons.","core_discovery":"The central claim is that explicit many-body treatment destabilizes the split-polaron state in Na2-xMn3O7: while hybrid DFT (HSE) favors a split oxygen-hole polaron across all tested exact-exchange fractions, DMC predicts the localized-polaron configuration to be lower in energy by approximately 1 eV, reversing the ordering by more than 1 eV relative to HSE. The reversal is robust to the class of trial wavefunctions used. Many-body spin densities additionally show that the nominal split state partially collapses toward a localized polaron, indicating that the split state is not a stable many-body configuration. The paper therefore establishes Na2-xMn3O7 as a benchmark where hybrid functionals qualitatively mis-rank competing oxygen-hole localization motifs.","pith_inferences":["Beyond the paper, hybrid functionals used to model battery cathodes or electrocatalysts may systematically overpredict delocalized ligand-hole intermediates; re-examining other oxygen-redox materials with DMC could reveal a broader failure mode.","The spin-density collapse of the split state suggests that spin-resolved X-ray or magnetic measurements could distinguish polaron characters where total charge densities look alike, a testable experimental extension.","The sensitivity of DFT+U+V results to the O-O inter-site interaction parameter hints that a functional correcting inter-site correlation might reproduce the DMC ordering without the cost of QMC, pointing a direction for functional development."],"forward_implications":["If the DMC ordering is correct, hybrid DFT can qualitatively mis-rank competing oxygen-hole localization motifs in correlated oxides when localization, hybridization, and Coulomb interactions compete on similar energy scales.","The low-energy oxygen-hole state in Na2-xMn3O7 after Na removal is a localized polaron, not a split polaron, with consequences for the redox mechanism and reversible capacity of this cathode material.","Conventional O K-edge X-ray absorption signatures are insufficient to identify the microscopic character of oxygen-hole states; spin-resolved probes or many-body benchmarks are required.","DMC provides a practical energetic benchmark for defect states in correlated oxides, with finite-size and trial-wavefunction uncertainties controlled enough to discriminate competing polarons."],"supporting_citations":[{"why":"Identifies Na2-xMn3O7 as an oxygen-hole system and provides the HSE prediction of the split polaron that the paper overturns.","marker":"[24]"},{"why":"Defines the HSE hybrid functional used to generate trial wavefunctions and the DFT description that favors the split state.","marker":"[33]"},{"why":"Provides the DFT+U+V method with inter-site interactions used for the alternative trial-wavefunction class.","marker":"[42]"},{"why":"Supplies the fixed-node DMC formalism and the understanding of nodal-surface errors that the accuracy argument relies on.","marker":"[43]"},{"why":"Documents the QMCPACK implementation used for the DMC calculations.","marker":"[40]"},{"why":"Supplies the twist-averaging technique used to reduce finite-size errors in the DMC energy comparison.","marker":"[45]"},{"why":"Provides the correlation-consistent effective core potentials used for the atoms, which underpin the DMC energies.","marker":"[47]"}],"fun_headline_variants":["QMC reverses hybrid DFT: oxygen holes localize","Many-body effects destabilize split oxygen-hole polaron","Split state loses to localized polaron in Na2Mn3O7","Hybrid DFT wrong: QMC stabilizes localized oxygen hole","Oxygen-hole polaron: local beats split in QMC"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The verdict rests on single-determinant fixed-node DMC trial wavefunctions faithfully bracketing the two near-degenerate hole states, and on the DFT-relaxed geometries remaining valid for the many-body ground state.","fun_headline_variants_meta":{"raw":{"variants":["QMC reverses hybrid DFT: oxygen holes localize","Many-body effects destabilize split oxygen-hole polaron","Split state loses to localized polaron in Na2Mn3O7","Hybrid DFT wrong: QMC stabilizes localized oxygen hole","Oxygen-hole polaron: local beats split in QMC"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000712,"raw_usage":{"total_tokens":3175,"prompt_tokens":888,"completion_tokens":2287,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":504,"completion_tokens_details":{"reasoning_tokens":2203}},"tokens_in":504,"tokens_out":2287,"duration_ms":39935,"temperature":1.0,"reasoning_tokens":2203,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:12:02.953848+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run fixed-node DMC with multideterminant trial wavefunctions (e.g., generated from complete-active-space or selected-CI calculations) for both polaron configurations in NaMn3O7; if the split-polaron state becomes lower in energy by more than about 0.5 eV, the claimed reversal is an artifact of the fixed-node approximation.","supporting_citations":[{"cited_title":"Abate, C","cited_arxiv_id":null,"evidence_quote":"Identifies Na2-xMn3O7 as an oxygen-hole system and provides the HSE prediction of the split polaron that the paper overturns."},{"cited_title":"Scuseria, and Matthias Ernzer- hof","cited_arxiv_id":null,"evidence_quote":"Defines the HSE hybrid functional used to generate trial wavefunctions and the DFT description that favors the split state."},{"cited_title":"Extended DFT + U + V method with on-site and inter-site elec- tronic interactions.Journal of Physics: Condensed Mat- ter, 22(5):055602, 2010","cited_arxiv_id":null,"evidence_quote":"Provides the DFT+U+V method with inter-site interactions used for the alternative trial-wavefunction class."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the fixed-node DMC formalism and the understanding of nodal-surface errors that the accuracy argument relies on."},{"cited_title":"Beaudet, Anouar Benali, M","cited_arxiv_id":null,"evidence_quote":"Documents the QMCPACK implementation used for the DMC calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the twist-averaging technique used to reduce finite-size errors in the DMC energy comparison."},{"cited_title":"Chandler Bennett, Cody A","cited_arxiv_id":null,"evidence_quote":"Provides the correlation-consistent effective core potentials used for the atoms, which underpin the DMC energies."}],"review_version":1}