{"id":"6d949370-6e87-4c8f-886c-365a3ec60c57","arxiv_id":"2608.08000","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The epsilon polymorph of LiMnO2 has low lithium-migration barriers but only one-dimensional pathways, while the spinel form has a three-dimensional network, even though both have the same fraction of favorable 0-TM sites.","lead":"This paper simulates a newly identified low-energy form of the battery material LiMnO2 and traces how lithium ions move through it and how it releases lithium when charged. The results suggest that the spatial arrangement of favorable lithium pathways, not just their abundance, controls how fast the material transports lithium.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quasi-1D assignment rests on an unvalidated BVSE 2D/3D threshold of 2.031 eV; the only first-principles transverse hop examined (2-TM ODH, 0.99 eV) is far below that threshold and no explicit inter-chain CI-NEB calculation is reported.","rationale":"The paper presents a coherent structural counterexample: ε-LiMnO2 and spinel Li2Mn2O4 share identical first-shell 0-TM tetrahedral fractions yet are assigned different migration-network dimensionalities. Independent support exists: the two 0-TM TSH CI-NEB barriers (0.35–0.36 eV) are consistent with the AIMD apparent activation energy (0.32 eV), and the direction-resolved MSD shows preferential c-axis motion. These results support low-barrier, anisotropic hopping, but they do not by themselves prove quasi-one-dimensionality; they only show that the dominant low-barrier hops are aligned along c. The formal evidence for one-dimensionality is the BVPA threshold structure, specifically the 0.468 eV versus 2.031 eV separation in Table 1. That threshold is an empirical bond-valence estimate, not a first-principles result, and the one identified transverse-path candidate (2-TM ODH, 0.99 eV) lies more than 1 eV below the claimed 2.031 eV threshold. Either that ODH path is not the actual percolation bottleneck, or the BVSE estimate is substantially too high; in the latter case transverse percolation could become accessible at energies that are still higher than intra-chain hops but low enough to make the quasi-1D label an overstatement. A single explicit inter-chain CI-NEB calculation would resolve this. No internal inconsistency was found in the n-TM statistics, the structural comparison, or the delithiation analysis; the concern is calibration of the empirical BVSE network surrogate. The reader's CONDITIONAL verdict is therefore appropriate, and no change to that verdict is warranted; the missing inter-chain barrier calculation is a concrete condition that, once supplied, could either validate or correct the quasi-1D claim.","tokens_in":13675,"tokens_out":9693,"duration_ms":100044,"concrete_test":"Compute CI-NEB barriers for the specific ε-phase hops that BVSE identifies as opening 2D/3D percolation at 2.031 eV, i.e., at least one and preferably two symmetry-inequivalent inter-chain hops connecting adjacent 1D chains, using the same 128-atom supercell and PBEsol+U settings as in Section 3.3. If any relaxed inter-chain CI-NEB barrier falls below roughly 0.6 eV—or if a supercell kinetic Monte Carlo model built from CI-NEB barriers shows percolation in the ab-plane—then the quasi-1D assignment and the quantitative 2.031 eV threshold would need revision; if the inter-chain barriers remain above about 1.5 eV, the BVSE prediction and the central contrast with spinel are supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that ε-LiMnO2 is quasi-one-dimensional while spinel is three-dimensional—is carried by the BVPA percolation thresholds in Table 1 and Section 3.2: ε reaches 1D connectivity at 0.468 eV and 2D/3D only at 2.031 eV, whereas spinel reaches all dimensionalities at 0.495 eV. The first-principles validation in Section 3.3 consists of two 0-TM TSH hops (0.35/0.36 eV) and one 2-TM ODH hop (0.99 eV). The 0.99 eV path is described as constraining transverse transport, yet it is about 1 eV lower than the claimed 2.031 eV BVPA threshold. If the BVSE landscape overestimates the inter-chain bottleneck—for example, by not accounting for lattice relaxation around the migrating Li, entropic stabilization, or the ODH bypass of the tetrahedral center—then transverse percolation could occur at energies much closer to the intra-chain value. The AIMD direction-resolved MSD at 2000 K (Fig. 6d) shows only about a 2.3-fold anisotropy (c: 5.2 Å2 vs a: 2.3 and b: 1.9 Å2), which does not establish true one-dimensionality; such an anisotropy could also arise from a two- or three-dimensionally percolating network with direction-dependent barrier prefactors. Thus the headline assertion that identical n-TM fractions fail to predict transport dimensionality depends on the empirical BVSE threshold, which has not been checked with an explicit first-principles inter-chain migration barrier.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript uses bond-valence site-energy (BVSE) and bond-valence pathway analysis (BVPA) in combination with DFT+U calculations to compare Li-ion migration topology in four LiMnO2 polymorphs, focusing on the recently predicted ε-LiMnO2 phase. The authors report that ε-LiMnO2 and the lithiated spinel have identical fractions of 0-TM, 2-TM, and 4-TM tetrahedral sites, yet the ε phase exhibits quasi-one-dimensional migration channels while the spinel is three-dimensional. CI-NEB gives 0.35–0.36 eV barriers for two 0-TM tetrahedral-site hops in ε and 0.41–0.53 eV for 0-TM hops in spinel; a 2-TM oxygen-dumbbell hop in ε has a 0.99 eV barrier. AIMD yields an apparent activation energy of 0.32 eV and shows preferential Li motion along c. Delithiation calculations connect the 0-TM connectivity to Li-site evolution and voltage steps. The main claim is that first-shell n-TM statistics alone do not determine long-range transport dimensionality.","tokens_in":14025,"tokens_out":4125,"duration_ms":38562,"significance":"If the quasi-one-dimensional classification is correct, the paper makes a conceptually useful point: identical local coordination statistics do not guarantee identical transport network dimensionality, and it provides a concrete example with a new polymorph. The work also provides a consistent set of migration barriers and an AIMD estimate that are internally consistent. However, the central claim currently rests on BVSE percolation thresholds rather than on explicit first-principles inter-chain migration barriers, and the AIMD anisotropy is modest; these weaknesses limit the strength of the conclusion as it stands.","major_comments":[{"comment":"The quasi-1D assignment for ε-LiMnO2 is based on the BVPA threshold of 2.031 eV for 2D/3D connectivity, but the only first-principles transverse hop reported (2-TM ODH, 0.99 eV) lies almost 1 eV below that threshold. Since CI-NEB includes lattice relaxation and the ODH trajectory can bypass the tetrahedral center, the true inter-chain percolation barrier may be far lower than the BVSE estimate. The authors should either compute an explicit CI-NEB barrier for a hop that connects neighboring 0-TM chains, or otherwise validate the BVSE threshold for this material, before concluding that ε-LiMnO2 is quasi-one-dimensional.","section":"§3.2, Table 1; §3.3, Fig. 4"},{"comment":"The direction-resolved MSD at 2000 K shows only a factor of ~2.3 anisotropy (c ≈ 5.2 Ų vs a ≈ 2.3 and b ≈ 1.9 Ų). Such a ratio can also arise in a 2D or 3D percolating network with anisotropic barrier prefactors or bottleneck energies, so this observation alone does not establish quasi-one-dimensional transport. Please provide a quantitative percolation analysis of the AIMD trajectories (e.g., counts of hop events along each direction) or temper the claim accordingly.","section":"§3.3, Fig. 6(d)"},{"comment":"The BVPA connectivity thresholds in Table 1 depend on the energy cutoff of 2.5 eV, which is stated to be referenced to the lowest BVSE site. The conclusion that ε is quasi-1D and spinel is 3D is sensitive to this choice and to the BVSE parameters. Please include a sensitivity analysis (e.g., thresholds at 2.0 and 3.0 eV) and clarify whether the isosurface level of -1 used in Fig. 3(a) is an absolute or relative value; otherwise the visual comparison among the four polymorphs may be misleading.","section":"§2 (Computational methods), §3.2"}],"minor_comments":[{"comment":"The abstract and introduction state that ε-LiMnO2 is a 'recently reported' polymorph; the cited work (Ref. 28) is a computational prediction, so please use 'predicted' rather than 'reported' to avoid implying experimental synthesis.","section":"§1 (Abstract and Introduction)"},{"comment":"The sentence 'the lowest-energy configuration was used subsequently' refers to Fig. S1, but the energy differences among the tested magnetic configurations are not given in the main text; please provide them (or at least a reference to the ESI table).","section":"§2"},{"comment":"The phrase 'the low one-dimensional BVPA connectivity threshold' is awkward; consider 'the low 1D connectivity threshold'.","section":"§3.3"},{"comment":"The production time of 20 ps per temperature is short for a quantitative Arrhenius analysis; please state whether longer simulations or multiple independent runs were used to estimate statistical errors.","section":"Fig. 6"},{"comment":"When comparing calculated voltages to the experimental spinel plateaus, please clarify the normalization per Mn and specify the Li concentration ranges in the figure caption.","section":"§3.4"},{"comment":"The sources for the ortho and layered barriers are given as Refs. 45, 66, 67; please ensure the barrier values are consistent with those references and note the functional/method used there.","section":"Table 2"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the journal's scope and the central concept is interesting, but the quasi-1D claim as presented goes beyond what the first-principles data support. If the authors can provide an explicit inter-chain migration barrier or substantially soften the conclusion, the paper could become acceptable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a look. The paper gives the first transport and delithiation characterization of the low-energy ε-LiMnO2 polymorph, and its central observation is solid: ε and lithiated-spinel have identical n-TM fractions (12.5% 0-TM, etc.) yet different spatial arrangement of those motifs. That geometric fact alone disproves the naive idea that first-shell statistics determine transport dimensionality. The authors show this clearly with BVSE maps and Table 1, and they support it with CI-NEB barriers (0.35/0.36 eV for 0-TM TSH in ε vs 0.41/0.53 eV in spinel). The AIMD activation energy of 0.32 eV is consistent with those barriers. The paper is well-written, includes proper stability checks (phonons, elastic constants, 500 K AIMD), and honestly ends with an experimental call. No circular reasoning or obvious citation problems.\n\nThe main soft spot is the quasi-1D label. The BVPA thresholds put 2D/3D percolation in ε at 2.031 eV, but the only transverse hop actually computed with CI-NEB is a 2-TM ODH at 0.99 eV. That's a large gap. If BVSE overestimates the inter-chain bottleneck—because lattice relaxation, entropic stabilization, or the ODH bypass is not fully captured—then transverse percolation could occur well below 2 eV, making the material anisotropic 3D rather than quasi-1D. The AIMD direction-resolved MSD at 2000 K shows only about 2.3x anisotropy along c; that is consistent with a strongly anisotropic network but not proof of true 1D percolation. I would want an explicit inter-chain CI-NEB calculation, or at least a careful discussion of why the 0.99 eV path does not contribute to percolation. This concern does not kill the paper's main message—connectivity matters beyond n-TM fractions—but it does mean the specific \"quasi-1D\" designation for ε-LiMnO2 should be treated with caution. The delithiation section is a smaller concern: sampling is limited, and the voltage steps from DFT+U are preliminary.\n\nWho this is for: computational battery-materials people working on metastable cathodes and transport descriptors. It deserves a serious referee. I would suggest asking for an inter-chain hop calculation, a clearer justification of the percolation threshold, and longer or lower-temperature AIMD before acceptance.","headline":"Solid geometric insight—identical n-TM fractions don't determine connectivity—but the 'quasi-1D' label is overreaching given that the only computed transverse hop is 0.99 eV versus a BVSE 2D/3D threshold of 2.031 eV.","tokens_in":770,"tokens_out":685,"would_cite":true,"duration_ms":32702,"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":"The paper demonstrates that the spatial connectivity of 0-TM tetrahedral intermediates, not merely their abundance, determines lithium migration dimensionality and delithiation voltage in ε-LiMnO2.","keywords":["Li-ion batteries","LiMnO2 polymorphs","0-TM tetrahedral sites","migration-network dimensionality","bond-valence site energy","quasi-one-dimensional diffusion","delithiation voltage","DFT+U"],"falsifier":"Compute or measure the transverse (in-plane) migration barriers directly: CI-NEB on the candidate 2-TM transverse hop, or anisotropic tracer diffusion on oriented samples, would settle it. A transverse barrier below about 0.35 eV, or comparable in-plane diffusivity, would contradict the quasi-one-dimensional assignment; the paper's own AIMD shows appreciable but smaller a/b motion, so the decisive test is quantitative comparison of the transverse jump barrier.","tokens_in":13459,"feed_emoji":"🔋","tokens_out":8010,"duration_ms":76918,"temperature":0.7,"pith_summary":"This paper establishes that a material's lithium-conduction character is set by how favorable local tetrahedral environments are wired together, not just by how many of them exist. Focusing on the low-energy ε-LiMnO2 polymorph, the authors show that although ε-LiMnO2 and the lithiated-spinel phase Li2Mn2O4 have identical fractions of O4 tetrahedral-site types (12.5% 0-TM, 75% 2-TM, 12.5% 4-TM), their low-energy lithium migration networks are topologically different: the ε phase conducts along quasi-one-dimensional chains, while the spinel conducts in three dimensions. They support this with bond-valence migration maps, climbing-image nudged elastic band barriers (0.35–0.36 eV for the ε-phase 0-TM hops versus 0.41–0.53 eV in spinel), and ab initio molecular dynamics that gives a 0.32 eV apparent activation energy and direction-resolved displacements favoring the c axis. This matters because the widely used first-shell n-TM classification, which counts face-sharing transition-metal neighbors around a tetrahedral intermediate, is the standard heuristic for screening cathode materials; the paper shows it is incomplete. Correctly predicting transport and voltage behavior in manganese-rich, cobalt-free cathodes requires adding higher-shell coordination and tetrahedral connectivity to the picture.","feed_headline":"ε-LiMnO2 moves Li along 1D chains despite spinel-like local sites","feed_subtitle":"Identical 0-TM tetrahedral fractions yield different networks, so connectivity must join local motifs in cathode design.","key_machinery":"The central object is the O4 tetrahedral intermediate labeled by its n-TM count, the number of face-sharing Mn neighbors, together with the spatial graph of 0-TM tetrahedra that forms when these intermediates are connected through shared faces. The paper uses bond-valence site-energy and bond-valence pathway analysis to map the low-energy lithium network and assign dimensionality, then climbing-image nudged elastic band calculations to get quantitative single-vacancy barriers, and ab initio molecular dynamics with direction-resolved mean-squared displacements to test the predicted anisotropy at finite temperature. The key comparison is between ε-LiMnO2 and the lithiated spinel, which share the same tetrahedral-type fractions but differ in the connectivity of 0-TM sites and in the composition of the next-nearest corner-sharing octahedral shell (eight MnO6 plus four LiO6 in ε versus twelve MnO6 in spinel).","core_discovery":"The central claim is that in ε-LiMnO2 the 0-TM tetrahedral motifs—the sites where lithium can hop with low Li–Mn repulsion—assemble into quasi-one-dimensional chains along the crystallographic c direction, whereas the same motifs in the lithiated-spinel polymorph form a three-dimensional percolating network. Consequently, identical local n-TM statistics do not determine long-range transport dimensionality. The paper further claims that the ε-phase 0-TM hops have lower barriers (0.35 and 0.36 eV) than the spinel's 0-TM hops (0.41 and 0.53 eV), correlating with a lower Mn density in the next-nearest corner-sharing octahedral shell, and that during delithiation the edge-sharing arrangement of 0-TM tetrahedra in ε-Li0.5MnO2 leads to short Li–Li distances (2.150 Å), keeping lithium in off-center octahedral positions and producing a three-stage voltage profile (3.40, 4.08, and 4.50 V) instead of the spinel's two-plateau behavior.","pith_inferences":["The paper does not directly measure transverse barriers, but if a future CI-NEB calculation found an in-plane hop below about 0.35 eV, the quasi-1D label would need to be revised to accommodate a second transport channel.","The connectivity hierarchy should transfer to other rocksalt-derived cathodes: a high-throughput screen that computed migration-network dimensionality thresholds for many cation orderings could rank candidate materials by predicted rate capability more faithfully than 0-TM fractions alone.","The edge-sharing 0-TM geometry that raises Li–Li repulsion in ε-LiMnO2 may be a general sign that tetrahedral Li will be destabilized; using shortest Li–Li distance among 0-TM sites as a screening descriptor could help avoid voltage surprises.","The next-nearest corner-sharing shell result suggests that substituting cations in the octahedral shell around a 0-TM site could tune migration barriers within the same n-TM class, an avenue the authors mention but do not explore computationally."],"forward_implications":["The standard 0-TM descriptor cannot be used alone to rank cathode materials; a material with many favorable local sites may still be a poor long-range conductor if those sites do not percolate.","ε-LiMnO2, despite its low migration barriers, is predicted to have limited rate capability because its quasi-one-dimensional chains are susceptible to local blocking; the practical benefit of the low barrier may not be realized without connectivity.","During delithiation, the same connectivity effect shifts lithium-site preferences: ε-LiMnO2 keeps Li off-center in octahedra at Li0.5MnO2 while the spinel stabilizes tetrahedral Li, and the voltage steps reflect this difference.","Screening of rocksalt-derived cathodes should include connectivity descriptors such as dimensionality thresholds and chain length, not just tetrahedral-site statistics."],"supporting_citations":[{"why":"Supplies the ε-LiMnO2 structure, its low-energy relative ordering, and magnetic ground state that the paper builds on.","marker":"[28]"},{"why":"Introduced the 0-TM concept showing that low face-sharing TM counts favor Li migration through tetrahedral intermediates.","marker":"[31]"},{"why":"Establishes that migration-network dimensionality and alternative-path availability matter for long-range transport.","marker":"[32]"},{"why":"Provides the orthorhombic-LiMnO2 migration barriers and the DFT+U treatment used for comparison.","marker":"[45]"},{"why":"Supplies the bond-valence site-energy method used to map low-energy lithium regions.","marker":"[53]"},{"why":"Supplies the bond-valence pathway analysis used to infer connectivity thresholds and dimensionality.","marker":"[54]"},{"why":"Provides the climbing-image nudged elastic band method used to compute quantitative migration barriers.","marker":"[56]"},{"why":"Documents the lithiated-spinel structure and its octahedral-to-tetrahedral Li rearrangement during delithiation.","marker":"[27]"},{"why":"Provides the experimental spinel voltage plateaus that the ε-phase three-stage voltage profile is compared against.","marker":"[34]"},{"why":"Distinguishes oxygen-dumbbell and tetrahedral-site hop mechanisms used to interpret the migration pathways.","marker":"[65]"}],"fun_headline_variants":["ε-LiMnO2: same local sites, but 1D Li chains beat 3D spinel network","Li mobility in ε-LiMnO2: 1D chains, lower barriers, different voltage steps","0-TM motifs form 1D chains in ε-LiMnO2, not spinel's 3D network","ε-LiMnO2's 1D Li pathways yield lower barriers than 3D spinel","Same local Li sites, different connectivity: ε-LiMnO2 goes 1D"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole quasi-one-dimensional picture rests on the bond-valence energy maps being a trustworthy stand-in for the actual lithium migration barriers: if the transverse paths that the maps place above 2 eV actually have true barriers comparable to or lower than the 0.35 eV chain hops, then ε-LiMnO2 would support meaningful three-dimensional transport despite being labeled quasi-1D.","fun_headline_variants_meta":{"raw":{"variants":["ε-LiMnO2: same local sites, but 1D Li chains beat 3D spinel network","Li mobility in ε-LiMnO2: 1D chains, lower barriers, different voltage steps","0-TM motifs form 1D chains in ε-LiMnO2, not spinel's 3D network","ε-LiMnO2's 1D Li pathways yield lower barriers than 3D spinel","Same local Li sites, different connectivity: ε-LiMnO2 goes 1D"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000668,"raw_usage":{"total_tokens":3144,"prompt_tokens":1137,"completion_tokens":2007,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":753,"completion_tokens_details":{"reasoning_tokens":1877}},"tokens_in":753,"tokens_out":2007,"duration_ms":13326,"temperature":1.0,"reasoning_tokens":1877,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:33:33.287398+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute or measure the transverse (in-plane) migration barriers directly: CI-NEB on the candidate 2-TM transverse hop, or anisotropic tracer diffusion on oriented samples, would settle it. A transverse barrier below about 0.35 eV, or comparable in-plane diffusivity, would contradict the quasi-one-dimensional assignment; the paper's own AIMD shows appreciable but smaller a/b motion, so the decisive test is quantitative comparison of the transverse jump barrier.","supporting_citations":[],"review_version":1}