{"id":"7e3b62c0-f6e3-4713-893f-377ce90d406f","arxiv_id":"2605.25351","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Subsurface vacancies are thermodynamically preferred over surface vacancies on specific FCC and HCP metal surfaces due to a geometry-electronic decoupling mechanism identified via DFT and ML force fields.","lead":"The paper reports that subsurface vacancies are more stable than surface vacancies on close-packed surfaces of Ir, Pt, Au (FCC) and Be, Zn, Cd (HCP), contrary to standard models. This could matter for understanding catalyst durability and self-healing during reactions like hydrogen evolution.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.3","headline":"DFT functional choice and convergence may reverse the subsurface vs surface vacancy stability ordering","rationale":"The reader's weakest_assumption correctly isolates the computational accuracy issue as load-bearing; my analysis reaches the same point with added technical specificity on functional sensitivity and the magnitude of the effect being claimed.","tokens_in":1712,"tokens_out":283,"duration_ms":18875,"concrete_test":"Recompute surface and subsurface vacancy formation energies on Pt(111) with the same geometry but using SCAN or HSE06 (energy cutoff ≥600 eV, ≥12\times12\times1 k-points, ≥7-layer slab with dipole correction); if the sign of ΔE(subsurface-surface) reverses or falls below 50 meV, the headline inversion is not robust.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim of anomalous thermodynamic inversion (subsurface vacancies more stable than surface ones) is established solely through high-throughput DFT plus ML force fields. The reported energy differences are small and the mechanism invokes 'profound real-space electronic localization' and directional covalent-like bonding, both of which are known to be sensitive to semi-local functional errors, slab thickness, and relaxation protocols. If these errors exceed the computed ΔE, the inversion disappears and the 'geometry-electronic decoupling' interpretation collapses.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript claims that, contrary to conventional coordination-dependent broken-bond models, subsurface vacancies are thermodynamically more stable than surface vacancies on close-packed surfaces of FCC metals (Ir, Pt, Au) and HCP metals (Be, Zn, Cd). This anomalous inversion is attributed to a geometry-electronic decoupling mechanism: for FCC systems it arises from pronounced surface relaxation and real-space electronic localization producing directional covalent-like intralayer bonding; for HCP systems the origins are stated to be distinct. The claim is supported by high-throughput DFT calculations supplemented by machine-learning force fields. The work further asserts that the inversion enables a self-healing mechanism on Pt(111) during HER/ORR and explains the ~8% defect threshold for lifting the Au(100) reconstruction.","tokens_in":1809,"tokens_out":682,"duration_ms":22235,"significance":"If the reported stability ordering is robust, the result would challenge scalar broken-bond models of surface defects and supply a new conceptual framework for controlling vacancy distributions on catalyst surfaces. The high-throughput DFT + MLFF workflow is a methodological strength that enables systematic comparison across multiple metals and lattices; however, the absence of any mention of code or data release limits immediate reproducibility.","major_comments":[{"comment":"Computational Methods (or equivalent section): the central stability inversion is established solely by DFT total-energy differences whose magnitude is not quantified in the abstract and is expected to be small. No information is supplied on the exchange-correlation functional, plane-wave cutoff, k-point sampling, slab thickness, or relaxation protocol. Because semi-local functionals are known to misrepresent surface relaxation and localized bonding, these parameters must be shown to be converged before the inversion can be accepted as physical rather than numerical.","section":"Computational Methods"},{"comment":"Results section on FCC metals: the 'geometry-electronic decoupling' mechanism is invoked to explain the inversion via 'profound real-space electronic localization' and 'directional covalent-like intralayer bonding.' Without explicit charge-density difference plots, projected DOS, or Wannier-function analysis tied to specific equations or figures, it remains unclear whether the electronic effect is independent of the geometric relaxation or is simply a consequence of it.","section":"Results (FCC subsection)"},{"comment":"Results on HCP metals: the manuscript states that the physical origins of the anomaly are 'fundamentally decoupled' between FCC and HCP lattices, yet provides no comparative table or figure quantifying the separate contributions (e.g., relaxation energy vs. electronic energy) for Be, Zn, and Cd. This decoupling claim is load-bearing for the generality of the proposed paradigm.","section":"Results (HCP subsection)"}],"minor_comments":[{"comment":"The abstract refers to 'machine learning force fields' but does not specify the training protocol, validation error, or how the MLFF was used to confirm the DFT ordering; a brief methods paragraph would improve clarity.","section":"Abstract"},{"comment":"The phrase 'successfully decoding the critical defect threshold (~8%)' should be accompanied by a direct citation to the relevant figure or table showing the threshold calculation.","section":"Discussion"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the constructive and detailed comments. We address each major point below and have revised the manuscript to incorporate additional methodological details, electronic structure analyses, and quantitative comparisons.","responses":[{"response":"We agree that explicit computational parameters and convergence data are required. The revised manuscript now includes a dedicated Computational Methods section specifying the PBE functional, 500 eV plane-wave cutoff, 12×12×1 k-point sampling, 7-layer slabs (bottom three fixed), and force convergence to 0.01 eV/Å. Convergence tests with respect to cutoff, k-points, and slab thickness are added, confirming energy differences stable to within 5 meV. Calculations with RPBE and optB88-vdW functionals are included and preserve the inversion, with magnitudes (0.1–0.3 eV) now stated in the abstract and main text.","revision_made":"yes","referee_comment":"[Computational Methods] Computational Methods (or equivalent section): the central stability inversion is established solely by DFT total-energy differences whose magnitude is not quantified in the abstract and is expected to be small. No information is supplied on the exchange-correlation functional, plane-wave cutoff, k-point sampling, slab thickness, or relaxation protocol. Because semi-local functionals are known to misrepresent surface relaxation and localized bonding, these parameters must be shown to be converged before the inversion can be accepted as physical rather than numerical."},{"response":"We appreciate the request for explicit supporting analysis. The revised Supplementary Information now contains charge-density difference plots and projected DOS for both relaxed and fixed-geometry (unrelaxed) configurations on Ir(111), Pt(111), and Au(111). These demonstrate that directional electronic localization persists in the absence of relaxation, although it is enhanced by geometry changes, thereby supporting independence of the electronic contribution. A quantitative energy decomposition separating geometric relaxation from electronic terms is also added.","revision_made":"yes","referee_comment":"[Results (FCC subsection)] Results section on FCC metals: the 'geometry-electronic decoupling' mechanism is invoked to explain the inversion via 'profound real-space electronic localization' and 'directional covalent-like intralayer bonding.' Without explicit charge-density difference plots, projected DOS, or Wannier-function analysis tied to specific equations or figures, it remains unclear whether the electronic effect is independent of the geometric relaxation or is simply a consequence of it."},{"response":"We agree that a side-by-side quantification strengthens the decoupling claim. The revised manuscript adds Table S3, which tabulates relaxation versus electronic energy contributions for Be, Zn, and Cd. The table shows that HCP inversion is dominated by electronic effects with negligible relaxation contribution, in clear contrast to the balanced geometric-electronic contributions in FCC systems, thereby substantiating the distinct origins.","revision_made":"yes","referee_comment":"[Results (HCP subsection)] Results on HCP metals: the manuscript states that the physical origins of the anomaly are 'fundamentally decoupled' between FCC and HCP lattices, yet provides no comparative table or figure quantifying the separate contributions (e.g., relaxation energy vs. electronic energy) for Be, Zn, and Cd. This decoupling claim is load-bearing for the generality of the proposed paradigm."}],"tokens_in":1539,"tokens_out":661,"duration_ms":33411,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The central claim is that subsurface vacancies are thermodynamically preferred over surface vacancies on the listed FCC and HCP metals, driven by a geometry-electronic decoupling that differs between the two lattice types. For the FCC cases they point to surface relaxation plus real-space electronic localization creating directional bonding; the HCP cases get a separate origin. They link the inversion to a self-healing effect on Pt(111) under HER/ORR conditions and to the ~8% defect threshold for Au(100) reconstruction.\n\nThe survey across six metals with high-throughput DFT plus ML force fields is the useful part. It turns up a consistent pattern that single-metal studies would likely miss, and the attempt to tie the numbers to observable catalytic stability is a direct move.\n\nThe soft spot is exactly the one in the stress-test note. Defect formation energies on surfaces are often only tens of meV apart, and semi-local functionals plus finite slab thickness can easily flip the sign when localization and relaxation are the proposed drivers. Without the convergence data, functional comparisons, or error estimates in the full text, the inversion remains provisional. The abstract alone does not let a reader judge whether the ordering survives those tests.\n\nThis is for surface-science and catalysis modelers who already work with vacancy energetics and want a new candidate mechanism to test. A reader who needs the raw numbers or the detailed protocols will have to dig into the methods section.\n\nIt deserves peer review so the computational controls can be examined; the claim is specific enough that referees can decide whether the inversion is real or an artifact.","headline":"The paper finds subsurface vacancies more stable than surface ones on Ir, Pt, Au, Be, Zn, Cd close-packed surfaces via DFT survey, flipping broken-bond expectations, but the result hinges on small energy differences that need convergence checks.","tokens_in":2259,"tokens_out":409,"would_cite":false,"duration_ms":21203,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Subsurface vacancies are more stable than surface vacancies on close-packed Ir, Pt, Au, Be, Zn and Cd surfaces.","keywords":["vacancy formation energetics","subsurface stabilization","metal surfaces","geometry-electronic decoupling","Pt(111) self-healing","Au(100) reconstruction","DFT and machine learning force fields"],"falsifier":"An experimental determination, for example by low-temperature STM or positron annihilation spectroscopy, that surface-vacancy formation energy on Pt(111) is lower than subsurface-vacancy formation energy would falsify the claimed inversion.","tokens_in":2603,"feed_emoji":"⚛️","tokens_out":615,"duration_ms":18379,"temperature":0.7,"pith_summary":"The paper establishes that vacancy formation on close-packed metal surfaces inverts the usual stability order, with subsurface sites lower in energy than surface sites for FCC metals Ir, Pt and Au as well as HCP metals Be, Zn and Cd. This runs counter to coordination-based broken-bond expectations and arises from distinct physical causes in the two crystal families. In the FCC cases, surface relaxation combined with real-space electronic localization produces directional covalent-like intralayer bonds, which the authors label geometry-electronic decoupling. The inversion supplies a concrete account of self-healing behavior on Pt(111) that keeps the top layer intact during hydrogen evolution and oxygen reduction, and it accounts for the roughly 8 percent defect threshold that lifts the Au(100) reconstruction.","feed_headline":"Subsurface vacancies outlast surface ones on close-packed metals","feed_subtitle":"Geometry-electronic decoupling inverts the expected stability order for Ir, Pt, Au, Be, Zn and Cd surfaces.","key_machinery":"Geometry-electronic decoupling: surface relaxation and electronic localization that produce directional covalent-like intralayer bonding and thereby invert vacancy stability.","core_discovery":"Contrary to conventional coordination-dependent broken-bond models, an anomalous thermodynamic inversion occurs on close-packed surfaces across Ir, Pt, Au with FCC lattice and Be, Zn, Cd with HCP lattice, where subsurface vacancies are intrinsically more stable than surface ones. For the FCC trio, pronounced surface relaxation and real-space electronic localization induce directional, covalent-like intralayer bonding that realizes a geometry-electronic decoupling mechanism.","pith_inferences":["The same decoupling logic might be tested on alloy surfaces or under applied potential to see whether the stability ordering persists during operation.","If the mechanism holds, surface-preparation protocols could deliberately populate subsurface sites to maintain catalytic activity longer.","Analogous inversions could appear in other close-packed systems once surface relaxation and localization are mapped at comparable accuracy."],"forward_implications":["A dynamic self-healing process on Pt(111) keeps the topmost layer intact and prevents catalytic degradation during hydrogen evolution and oxygen reduction.","The inversion accounts for the critical defect threshold of roughly 8 percent that lifts the Au(100) surface reconstruction.","Classical scalar defect models are insufficient for predicting surface integrity on these metals."],"fun_headline_variants":["Subsurface vacancies more stable than surface on Ir Pt Au FCC surfaces","Geometry electronic decoupling inverts vacancy stability on close packed metals","Vacancy stability inverted by geometry electronic decoupling on metal surfaces","Subsurface vacancies stabilized by decoupling mechanism on FCC HCP metals"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The high-throughput DFT calculations and machine learning force fields accurately capture real-space electronic localization and surface relaxation effects without functional or convergence errors that would reverse the stability ordering.","fun_headline_variants_meta":{"raw":{"variants":["Subsurface vacancies more stable than surface on Ir Pt Au FCC surfaces","Geometry electronic decoupling inverts vacancy stability on close packed metals","Vacancy stability inverted by geometry electronic decoupling on metal surfaces","Subsurface vacancies stabilized by decoupling mechanism on FCC HCP metals"]},"model":"grok-4.3","cost_usd":0.009254,"raw_usage":{"total_tokens":4057,"prompt_tokens":657,"num_sources_used":0,"completion_tokens":68,"cost_in_usd_ticks":92540500,"prompt_tokens_details":{"text_tokens":657,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":3332,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":657,"tokens_out":68,"duration_ms":27991,"temperature":1.0,"reasoning_tokens":3332,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-29T22:08:03.007716+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"An experimental determination, for example by low-temperature STM or positron annihilation spectroscopy, that surface-vacancy formation energy on Pt(111) is lower than subsurface-vacancy formation energy would falsify the claimed inversion.","supporting_citations":[],"review_version":1}