{"id":"8a5f5fda-09f3-4527-84af-c11fc37e3b84","arxiv_id":"2507.03804","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A combined experimental and computational analysis reveals distinct local cation environments, copper Jahn-Teller distortion, near-random cation mixing, and a roughly 2 nm local distortion coherence length in the prototype entropy-stabilized oxide J14.","lead":"This study combines neutron scattering, X-ray absorption, and computer simulations to map the local atomic-scale structure of the entropy-stabilized oxide J14. It reveals that each metal cation sits in a distinct distorted oxygen environment, with copper showing the well-known Jahn-Teller distortion and only slight signs of like-cation clustering.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Same-cation SRO claim rests on five 50-atom AIMD configurations, and the paper's own 250-atom AIMD reportedly shows no such features; finite-size sampling may be driving the signal.","rationale":"Good-faith reading: the paper's central claim is that a DFT-relaxed 50-atom SQS model, combined with AIMD and RMC+EXAFS, reveals individualized cation environments, Cu Jahn-Teller distortion, near-complete chemical homogeneity, weak same-cation SRO, and a ~2 nm lattice-distortion correlation length. For that claim to hold, the five 50-atom SQS configurations used for AIMD must be statistically representative of the random solid solution. With 10 atoms per species, the random expectation of same-cation nearest neighbors is about 2.2 per site, and five configurations are far too few to resolve a small SRO preference. The 300 K AIMD runs have low self-diffusion, so each trajectory samples essentially one fixed cation arrangement rather than configurational equilibrium. The paper honestly notes the limited number of configurations, and its own 250-atom AIMD result reportedly lacks the distinct features seen in the 50-atom runs, providing an internal check that the same-cation signal is fragile. RMC+EXAFS gives larger-cell support, but the authors themselves flag the possibility of overfitting. I therefore keep the reader's CONDITIONAL verdict; the concern sharpens the condition under which the weak-SRO claim can be accepted, namely that a larger-cell or multi-configuration AIMD analysis should confirm the 50-atom result. This is a sampling/representativeness issue, not a question of author diligence; the paper is explicitly hedged and combines multiple independent techniques, which is creditworthy.","tokens_in":15200,"tokens_out":7167,"duration_ms":79589,"concrete_test":"Re-analyze the existing 250-atom AIMD trajectory (Fig. S15) with the same partial RDF and nearest-neighbor n_AA(r)/Warren-Cowley protocol used for the five 50-atom runs. If the 250-atom cell shows no same-cation excess—or if the 50-atom excess lies within the bootstrap 95% confidence interval across the five configurations—then the weak like-cation SRO claim should be downgraded to a finite-size sampling artifact; if the 250-atom cell reproduces the excess with comparable magnitude, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The weakest link is the inference of weak same-cation short-range order from five 50-atom AIMD runs. Each 50-atom cell contains only 10 cations of each species, and at 300 K the AIMD trajectories (10 ps, last 6 ps averaged) sample local vibrations around one fixed cation arrangement, not configurational equilibrium. Under random occupancy, the expected number of same-cation nearest neighbors per cation is 12×9/49 ≈ 2.20; with only five configurations the configuration-to-configuration variance is large enough that a slight same-cation preference is not statistically distinguishable from sampling noise. The paper itself attributes non-Gaussian bond-length histograms to 'the limited number of available configurations,' and, more importantly, states that the 250-atom AIMD simulation 'did not exhibit any significant distinct features' (Fig. S15). That larger-cell null result is direct internal evidence that the 50-atom same-cation excess may be a finite-size artifact rather than genuine SRO. Because weak like-cation SRO is one of the two headline structural claims, this unresolved sampling issue is load-bearing. The RMC+EXAFS analysis does corroborate the same-cation preference, but the authors caution that RMC 'does not rule out the possibility of overfitting with a high number of cations involved,' so that channel does not independently settle the question.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines neutron total scattering, EXAFS, small-box PDF refinements, DFT-relaxed special quasirandom structures (SQS), AIMD simulations, and large-box RMC+EXAFS refinements to characterize chemical short-range order (SRO) and local lattice distortion (LD) in the entropy-stabilized rock-salt oxide (Mg0.2Co0.2Ni0.2Cu0.2Zn0.2)O (J14). The central claims are that a DFT-relaxed 50-atom SQS model fits the neutron PDF substantially better than a disordered unit-cell model, that individual cation environments are distinct with a Cu Jahn-Teller distortion, that the cation distribution is nearly chemically homogeneous, that there is a weak preference for like-cation nearest neighbors at 300 K, and that the correlation length of local lattice distortion is about 2 nm. The conclusions rest on a combination of independent experimental benchmarks and DFT-based models, with RMC fitting grounded in the experimental neutron and EXAFS data. However, several load-bearing statistical and methodological questions are not resolved, particularly the representativeness of five 50-atom AIMD configurations and the absence of a stated extraction procedure for the 2 nm correlation length.","tokens_in":15450,"tokens_out":3306,"duration_ms":41317,"significance":"If the conclusions hold, the paper offers a valuable template for joint experimental-computational analysis of local structure in high-entropy oxides, and it usefully demonstrates that small-box PDF refinement can be improved by DFT-relaxed SQS models. The simultaneous RMC fitting of neutron total scattering and multiple EXAFS edges is a technically demanding and commendable feature, as is the explicit comparison of SQS and random-distribution large supercells. The reported Cu Jahn-Teller distortion and chemically distinct local cation environments are plausible and consistent with prior EXAFS work. However, the headline claim of weak like-cation SRO is not yet established at the required statistical level, and the apparent null result from the 250-atom AIMD simulation weakens the 50-atom evidence. The paper would be a solid contribution if the SRO claim is reanalyzed with proper configuration sampling and uncertainty quantification, and if the correlation-length estimate is made reproducible.","major_comments":[{"comment":"The claim of slightly preferred same-cation A-A correlation at 300 K rests on five 50-atom AIMD configurations, each containing only 10 cations of each species and sampling local vibrations about a fixed cation arrangement rather than configurational equilibrium. Under random occupancy the expected number of same-cation nearest neighbors per cation is about 12*9/49 = 2.20, and with only five configurations the configuration-to-configuration variance is large enough that a small same-cation excess cannot be distinguished from sampling noise. The paper's own statement that the 250-atom AIMD simulation 'did not exhibit any significant distinct features' (Fig. S15) is direct internal evidence against the 50-atom result. The authors should quantify the statistical uncertainty across configurations (for example, by reporting standard errors or bootstrap confidence intervals for the partial RDFs and Warren-Cowley parameters) and reconcile the 50-atom and 250-atom results before the same-cation SRO conclusion can be accepted.","section":"III, Fig. 3B, Fig. S15"},{"comment":"The 'almost half of the Rwp value' claim is presented without any uncertainty estimate on Rwp, so it is not possible to judge whether the improvement from the unit-cell model to the 50-atom SQS model is statistically significant. The estimated correlation length of LD of about 2 nm is also asserted in the text without a defined extraction procedure: it is not stated which r-range, lattice-parameter criterion, or Rwp threshold was used to define the crossover from local to long-range behavior. The authors should provide confidence intervals for Rwp (for example, from multiple independent fits or bootstrap resampling of the PDF) and specify exactly how the 2 nm value is obtained from the box-car and varying-r-range refinements.","section":"II, Fig. 2C and 2F"},{"comment":"The DFT+U calculations and AIMD simulations use Hubbard U values of 5.0, 5.1, 4.5, and 7.5 eV for Co, Ni, Cu, and Zn, taken from ref. 84 without sensitivity tests. Since the DFT-relaxed SQS geometries and AIMD partial RDFs are the basis for the bond-length histograms and the same-cation SRO analysis, the qualitative conclusions about Cu Jahn-Teller distortion, Zn-O broadening, and like-cation preferences should be shown to be robust to reasonable variations in U, or a targeted sensitivity study should be added. At minimum, the authors should state the known dependence of these DFT results on the choice of U and discuss how it affects the structural conclusions.","section":"IV.C, Methods"},{"comment":"The RMC+EXAFS results are presented as independent corroboration of the same-cation SRO, but the authors themselves note that 'these results do not rule out the possibility of overfitting with a high number of cations involved.' The RMC fit without EXAFS produces identical Gaussian peak distributions for all A-O pairs (Fig. 4B), and only after adding EXAFS data and atom-swapping does the fit resemble the AIMD result. This suggests that the apparent SRO signal in RMC is strongly influenced by the EXAFS constraints and by the large number of cation species, and the degeneracy of possible RMC solutions should be quantified (for example, by multiple independent RMC runs with different starting configurations or by reporting the spread of Warren-Cowley parameters). Without such an analysis, the RMC channel does not independently settle the same-cation SRO question.","section":"II, Fig. 4"}],"minor_comments":[{"comment":"There are several typographical errors and inconsistent notations, including 'Tempearture' in the Figure 3 caption, 'Numercial' in the author affiliation, 'challenge OF studying SRO' in Section II, and 'results of fittingRwp' missing a space in the Rietveld discussion. These should be corrected.","section":"Throughout"},{"comment":"The explanation that the broader Ni/Zn cation clouds are due to 'smallest size of Ni and shortest bond of Zn' is somewhat speculative as written; the connection between ionic radius, bond length, and the density-map shape should be justified more explicitly or softened.","section":"II, Fig. 5D"},{"comment":"The statement that all cation-oxygen pairs 'do not exhibit an ideal Gaussian distribution' is attributed to 'the limited number of available configurations,' but the same limited-configuration issue also affects the bond-length histograms used to support the Cu Jahn-Teller and Zn broadening claims. The authors should separate the finite-size effect from the physical broadening when interpreting these histograms.","section":"II, Fig. 2H"},{"comment":"The RMC methodology description would benefit from stating which constraints were active in each run (BVS, atom swapping, EXAFS) and how many independent RMC runs were performed, since the reproducibility of the Warren-Cowley parameters depends on these choices.","section":"IV.B, RMC methods"}],"recommendation":"major_revision","confidential_remarks":"The paper is a competent and broad experimental-computational study of a well-known material, and the experimental dataset is substantial. The central difficulty is that the weak like-cation SRO claim, which is one of the two headline structural conclusions, is not statistically supported as presented, and the 250-atom AIMD null result is a red flag that the 50-atom result may be a finite-size artifact. The other main structural claims (Cu Jahn-Teller distortion, distinct local cation environments, near-complete chemical homogeneity) are much better supported and would likely survive a revision. I do not see grounds for rejection, but the required reanalysis of the SRO evidence and the correlation-length extraction is substantial enough to warrant a major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a solid, multi-technique local-structure study of the canonical entropy-stabilized oxide J14. What is genuinely new: first Warren–Cowley parameters for J14, an estimated ~2 nm correlation length for lattice distortion, and a SQS-DFT small-box fitting route that roughly halves the low-r Rwp compared to a unit-cell model. The Cu Jahn–Teller distortion and near-random cation mixing were reported before, but the combination of methods—DFT-relaxed SQS, AIMD, simultaneous RMC fitting of neutron total scattering and EXAFS—is a useful contribution to the field.\n\nThe soft spots are real, and one is load-bearing. The weak like-cation SRO claim rests on five 50-atom SQS configurations, each with only ten cations per species. At 300 K those AIMD runs sample vibrations around a fixed cation arrangement, not configurational equilibrium. Under random occupancy the expected same-cation count per cation is about 2.2; with five configurations the sampling variance is large enough that a slight preference is not statistically distinguishable from noise. The paper itself reports that the 250-atom AIMD simulation \"did not exhibit any significant distinct features\" (Fig. S15), which is direct internal evidence that the 50-atom signal may be a finite-size artifact. The RMC+EXAFS channel does show a similar same-cation preference, but the authors explicitly warn that the RMC results \"do not rule out the possibility of overfitting with a high number of cations involved,\" so it does not independently settle the question. The ~2 nm correlation length is asserted without a defined extraction procedure or error bars, and the Rwp improvements are quoted without uncertainties. These are quantitative gaps, not fatal flaws—the central picture of individualized cation environments and Cu JT distortion holds up—but the SRO and correlation-length numbers should not be taken as settled.\n\nThe DFT+U dependence on Hubbard U values borrowed from earlier work is a minor concern here because the DFT models are validated against external neutron/EXAFS data via the fits; the calculated band gap comparison is a side remark, not a core claim.\n\nWho is this for? Researchers working on local structure in high-entropy oxides and disordered rocksalt materials, and method developers combining SQS-DFT with total scattering. It deserves a serious referee, not a desk reject. In revision I would ask for a defined extraction of the correlation length, error estimates on the Rwp values, and a demonstration that the 50-atom SRO preference survives larger cells or proper configurational averaging. With those additions the quantitative claims would be citable. My own verdict would be conditional, but this is a worthwhile paper to engage with.","headline":"A competent multi-method local-structure study of J14 whose quantitative claims on weak like-cation SRO and a ~2 nm correlation length outrun the supporting statistics; worth refereeing, but those numbers should be treated as provisional until better grounded.","tokens_in":16070,"tokens_out":2386,"would_cite":true,"duration_ms":28632,"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":"DFT-relaxed random-alloy models reveal that the five-cation oxide J14 is chemically homogeneous yet carries element-specific local distortions and a weak tendency for like cations to cluster.","keywords":["entropy-stabilized oxides","short-range order","local lattice distortion","special quasirandom structures","pair distribution function","reverse Monte Carlo","ab initio molecular dynamics","Jahn-Teller distortion"],"falsifier":"A decisive check would be to repeat the AIMD and RMC analyses with a larger set of independent 50-atom SQS configurations or with a single 250-atom SQS and ask whether the like-cation nearest-neighbour enrichment and the Mg-O peak discrepancy persist; if they vanish, the claimed weak SRO is a finite-size artifact. Alternatively, chemically specific probes such as Mg K-edge EXAFS or solid-state NMR could directly test the predicted Mg-O environment without relying on the SQS representation.","tokens_in":14955,"feed_emoji":"🔬","tokens_out":5367,"duration_ms":56720,"temperature":0.7,"pith_summary":"This paper sets out to determine whether the prototypical entropy-stabilized oxide J14 (Mg0.2Co0.2Ni0.2Cu0.2Zn0.2O) is truly random at the atomic scale or carries hidden chemical short-range order and local lattice distortion. The authors argue that a 50-atom special quasirandom structure relaxed by density functional theory reproduces the measured neutron pair distribution function far better than a simple mixed-cation unit cell, cutting the low-r fit residual roughly in half. Combining small-box pair distribution function refinement, ab initio molecular dynamics, and reverse Monte Carlo fits of neutron total scattering plus EXAFS, they conclude that J14 is nearly chemically homogeneous yet hosts individualized cation environments, a Cu2+ Jahn-Teller distortion, a weak preference for like-cation neighbours, and lattice distortion correlated over about 2 nm. If correct, this picture matters because local order, even when weak, reduces the configurational entropy that is supposed to stabilize these materials and shapes their transport, magnetic, and thermal properties.","feed_headline":"50-atom models halve local-structure fit error in a five-cation oxide","feed_subtitle":"Neutron scattering plus DFT reveals weak same-cation clustering and a ~2 nm distortion length scale in J14.","key_machinery":"The load-bearing object is the 50-atom special quasirandom structure (SQS): a small periodic supercell engineered to reproduce the pair correlation functions of a random cation distribution, then fully relaxed with DFT (PBEsol+U). This DFT-relaxed SQS serves three roles at once: it supplies a physically meaningful structural model for small-box fits of the neutron PDF, its relaxed geometries seed AIMD trajectories that expose cation-specific bond statistics, and its partial pair correlations guide interpretation of reverse Monte Carlo fits of neutron total scattering and EXAFS. Supporting quantities include the nearest-neighbour function $n_{ij}(r)$ and Warren-Cowley short-range-order parameters, which quantify the tendency of like cations to sit near each other, and the RMC 'point cloud' folded into a unit cell to visualize anisotropic atomic displacement.","core_discovery":"The central claim is that the local structure of J14 is not a simple random rocksalt solid solution: each of the five cations maintains its own preferred bond environment while the overall lattice stays a single disordered phase. Evidence comes from DFT-relaxed SQS models, whose partial A-O pair correlations reveal a bimodal Cu-O distribution from Jahn-Teller distortion and a broadened Zn-O distribution, and from AIMD and RMC+EXAFS analyses showing a slight enrichment of identical-cation nearest neighbours together with near-complete chemical homogeneity. The authors estimate the correlation length of local lattice distortion at roughly 2 nm from box-car and varying-r refinements, and note that distinct cation 'clouds' in the RMC configurations indicate each cation type responds differently to the surrounding disorder. The paper presents these conclusions as the first simultaneous multi-technique account of SRO and LD in an ESO, achieved by using SQS as a computationally efficient bridge between theory and total-scattering experiments.","pith_inferences":["Editorial extension: A natural next test is to compare these conclusions against chemically specific probes such as Mg K-edge EXAFS or solid-state NMR, which would directly verify the Mg-O peak discrepancy between AIMD and RMC.","Editorial extension: The same 50-atom SQS workflow could be applied to temperature-dependent datasets: the AIMD result that the Cu Jahn-Teller distortion fades by 1500 K predicts a measurable temperature evolution in the PDF first peak that experiments could confirm.","Editorial extension: The weak like-cation preference, if real, may be amplified in other ESO compositions with larger cation size or charge contrasts, and the Warren-Cowley framework used here could be extended to a multicomponent definition separating size-driven from chemistry-driven ordering.","Editorial extension: The authors' suggestion to feed combined scattering-plus-model results into machine-learning libraries implies that future SRO detection may be automated, but only if training data includes the multi-technique redundancy this paper shows is necessary."],"forward_implications":["If J14 is nearly homogeneous but with weak like-cation SRO, the configurational entropy stabilization argument must be refined: entropy is high but not maximal, and local enthalpy contributions are present.","The SQS-based workflow can be transferred to other compositionally complex oxides and high-entropy ceramics, giving a standard recipe for extracting SRO and LD from total-scattering data without requiring classical interatomic potentials.","The roughly 2 nm distortion correlation length sets a length scale below which property models, for thermal conductivity, ionic transport, or magnetism, must treat local disorder explicitly rather than through an averaged lattice.","The distinct Cu Jahn-Teller and Zn broadened distributions suggest element-specific functional roles, such as Cu acting as a local structural probe and Zn contributing bond-length heterogeneity.","The finding that RMC on total scattering alone cannot distinguish similar cations implies that future experimental studies of such materials should routinely combine total scattering with element-selective EXAFS constraints."],"supporting_citations":[{"why":"Introduces J14 as the prototypical entropy-stabilized oxide and defines the composition and single-phase rocksalt structure that this study re-examines.","marker":"[1]"},{"why":"Provides the prior EXAFS analysis of J14 that established the Cu2+ Jahn-Teller distortion and the average local cation environments used as a starting point.","marker":"[29]"},{"why":"Supplies the earlier combined RMC and EXAFS methodology for pyrochlore high-entropy oxides and identifies the challenge of similar-cation systems that this paper addresses.","marker":"[22]"},{"why":"Defines the special quasirandom structure concept that the paper uses to construct small supercells mimicking random cation disorder.","marker":"[77]"},{"why":"Provides the mcsqs utility and ATAT toolkit used to generate the SQS configurations of J14.","marker":"[78,79]"},{"why":"Supplies the Hubbard U parameters for Co, Ni, Cu, and Zn used in the DFT and AIMD calculations.","marker":"[84]"},{"why":"Describes the combined total-scattering and EXAFS reverse Monte Carlo fitting approach in the RMCProfile package that the paper applies.","marker":"[51,52]"},{"why":"Defines the correlation functions, including the nearest-neighbour function $n_{ij}(r)$, used to quantify cation ordering in the SQS and RMC models.","marker":"[41]"}],"fun_headline_variants":["J14 oxide's local structure: weak cation clustering, 2-nm distortion","DFT SQS plus neutron scattering decode five-cation oxide J14","Entropy-stabilized oxide shows bimodal Cu-O from Jahn-Teller effect","Local lattice distortion in J14 spans 2 nm, neutrons and DFT agree"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The five 50-atom SQS configurations are assumed to be statistically representative of the true random solid solution, so the observed same-cation pair preferences and bond-length distributions reflect real short-range order rather than finite-size sampling noise or artifacts of the chosen PBEsol+U functional and its Hubbard U parameters.","fun_headline_variants_meta":{"raw":{"variants":["J14 oxide's local structure: weak cation clustering, 2-nm distortion","DFT SQS plus neutron scattering decode five-cation oxide J14","Entropy-stabilized oxide shows bimodal Cu-O from Jahn-Teller effect","Local lattice distortion in J14 spans 2 nm, neutrons and DFT agree"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000216,"raw_usage":{"total_tokens":1421,"prompt_tokens":926,"completion_tokens":495,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":542,"completion_tokens_details":{"reasoning_tokens":410}},"tokens_in":542,"tokens_out":495,"duration_ms":6141,"temperature":1.0,"reasoning_tokens":410,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:02:03.791423+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be to repeat the AIMD and RMC analyses with a larger set of independent 50-atom SQS configurations or with a single 250-atom SQS and ask whether the like-cation nearest-neighbour enrichment and the Mg-O peak discrepancy persist; if they vanish, the claimed weak SRO is a finite-size artifact. Alternatively, chemically specific probes such as Mg K-edge EXAFS or solid-state NMR could directly test the predicted Mg-O environment without relying on the SQS representation.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the prior EXAFS analysis of J14 that established the Cu2+ Jahn-Teller distortion and the average local cation environments used as a starting point."},{"cited_title":"Jiang , author C","cited_arxiv_id":null,"evidence_quote":"Supplies the earlier combined RMC and EXAFS methodology for pyrochlore high-entropy oxides and identifies the challenge of similar-cation systems that this paper addresses."},{"cited_title":"Zunger , author S.-H","cited_arxiv_id":null,"evidence_quote":"Defines the special quasirandom structure concept that the paper uses to construct small supercells mimicking random cation disorder."},{"cited_title":"Rak , author C","cited_arxiv_id":null,"evidence_quote":"Supplies the Hubbard U parameters for Co, Ni, Cu, and Zn used in the DFT and AIMD calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the correlation functions, including the nearest-neighbour function $n_{ij}(r)$, used to quantify cation ordering in the SQS and RMC models."}],"review_version":1}