{"id":"5f4aab10-a3ce-4d35-93bb-1e7ff46e7e05","arxiv_id":"2607.22307","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Zn(hba)'s correlated disorder is governed by one-dimensional ligand ordering and ice-like rules, captured by a Truchet-tile model whose Monte Carlo simulations reproduce the measured diffuse X-ray scattering.","lead":"Researchers redescribed the disordered crystal structure of the metal-organic framework Zn(hba) using X-ray data and showed that its disorder follows 'Truchet tile' patterns, like a tiling puzzle where tile orientations store information. The work suggests this puzzle language may help describe disorder in other framework materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"MC validation is partly circular: T_MC and the stacking fault are set from the same diffuse data used as confirmation, and agreement is only visual; local-rule transfer from Li(inox) remains an unverified expectation.","rationale":"The paper does several things well: it reports new diffraction data, uses rocking curves to argue that the l-odd features are diffuse rather than Bragg, and constructs a concrete Truchet/Voronoi model that makes a non-obvious prediction (subextensive entropy/size-dependent disorder). The average-structure reinterpretation is plausible and could stand even if the disorder model is wrong. However, the central claim requires that a single model quantitatively accounts for three independent observables. The evidence for that claim is currently weaker than the prose suggests. The MC inputs are set using the same diffuse features later used for validation, and the comparison is qualitative. This is a methodological circularity, not a question of intent. Even the local bonding rules — the foundation of the model — are inferred from Li(inox) and labelled an 'expectation' in the text; the assertion that polarity reversal is impossible 'without either breaking the two-short-two-long rules or heavily distorting' is a geometric plausibility argument, not a derived proof or a measured local-structure result. If the local rules are wrong, the whole Truchet architecture collapses; if they are right but the agreement is overfit, the central claim is still unproven. The reader's weakest_assumption targeted the local-rule transfer; I partially agree, but I view the quantitative/circular validation as the more decisive gap because it directly concerns the claimed 'accounts for' relation. A hold-out test that fits only part of the data and predicts the rest would settle this. If that test passes, the model earns its conclusion; if it fails, the verdict should remain conditional at best. Since the reader already requested quantitative comparison and disentangling fitted inputs, my read does not move the verdict.","tokens_in":13379,"tokens_out":10486,"duration_ms":92940,"concrete_test":"Hold-out cross-validation: fit T_MC, stacking-fault density, and the harmonic force constant using only the average P4_2/mmc structure and the (hk0) diffuse plane; then compute the predicted (h0l) scattering and compare line profiles along [50l] (or a normalized intensity residual) with the measured data without any further adjustment. If the odd-l diffuse maxima are reproduced, the model is not merely fitting the validation data; if they are not, the central claim is overfit.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The core claim — that the Truchet-tile model 'accounts simultaneously for the local chemistry, the crystallographic average structure, and the observed diffuse scattering' — is not yet supported by an independent, quantitative test. In the Monte Carlo section the model inputs are fixed using the observations it is then said to reproduce: T_MC = 2J is chosen to give 'weak antipolar correlations implied by experiment,' and an odd number of layers inserts a stacking fault 'to mimic the ~50 Å correlation length identified above.' The agreement shown in Fig. 9 is visual; no residual, R-factor, or line-profile comparison is reported. The local-rule input from Li(inox) is also presented as an expectation ('Our expectation is that the disordered Zn(hba) structure contains the same local bonding rules...'), with the two-short/two-long and uniform-chain-polarity rules asserted rather than directly measured in Zn(hba). Therefore the successful reproduction of the diffuse scattering could be an overfitting artifact rather than evidence for the Truchet/ice description. The average-structure collapse in Fig. 9(b) is likewise shown but not quantitatively compared with the refined P4_2/mmc model.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a redetermination of the average crystal structure of the disordered metal–organic framework Zn(hba) from single-crystal X-ray diffraction, explicitly separating Bragg reflections from structured diffuse scattering. The authors reinterpret the previously reported doubled-c-axis model with partially occupied sites as a mis-assignment of diffuse scattering as Bragg intensity, and propose instead a higher-symmetry P4_2/mmc average structure. Using crystal-chemical arguments transferred from the ordered analogue Li(inox), they develop a Truchet-tile description in which Zn nodes and hba linkers obey two-short/two-long Zn–O bonding and uniform ligand polarity along each chain. They then generate atomistic configurations via a two-step Monte Carlo simulation and calculate diffuse scattering that is compared visually with experiment. The central claim is that the same model simultaneously accounts for the local chemistry, the crystallographic average structure, and the observed diffuse scattering, and that this Truchet-tile formalism may apply generally to correlated disorder in framework materials.","tokens_in":13679,"tokens_out":3328,"duration_ms":32069,"significance":"If the central claim is upheld, this would be an important demonstration that the Truchet-tile concept extends beyond TRUMOF-1 and that a previously published doubled-c-axis structure was an artifact of misinterpreting diffuse scattering as Bragg scattering. The manuscript has real strengths: a fresh single-crystal dataset, a quantitative rocking-curve distinction between Bragg and diffuse features, an ordered analogue (Li(inox)) that provides physically motivated local rules, a second ordered analogue (Zn(hca)) that supports the idea of chemical control over order/disorder, and the use of 3D-ΔPDF to visualise correlations. The model Hamiltonian is transparent and the simulation details are described sufficiently for reproduction. However, the validation is currently qualitative: the Monte Carlo parameters are set from the same diffuse data that the simulation is said to reproduce, and agreement is shown visually, not through a residual or line-profile analysis. The local-rule transfer from Li(inox) is stated as an expectation rather than directly measured in Zn(hba). These issues are load-bearing for the paper's strongest claim, so the present version is defensible but not yet con","major_comments":[{"comment":"The Monte Carlo validation is partly circular as presented. T_MC=2J is chosen because it gives 'weak antipolar correlations implied by experiment,' and an odd number of layers introduces 'a single stacking fault ... mimicking the ~50 Å correlation length identified above.' The agreement in Fig. 9(a) is visual only; no quantitative residual, R-factor, or line-profile comparison is reported. Please provide such a comparison (e.g. integrated diffuse intensity along [h00] and [00l], or an R-factor on the diffuse scattering volume) and show the sensitivity of the simulated patterns to T_MC and to the stacking-fault construction. This is necessary to support the central claim that the same model accounts for the observed diffuse scattering.","section":"Monte Carlo simulations and diffuse scattering calculations, Eq. (1) and Methods"},{"comment":"The local-rule transfer from Li(inox) is explicitly presented as an expectation ('Our expectation is that the disordered Zn(hba) structure contains the same local bonding rules...'). The two-short/two-long coordination rule and the uniform chain-polarity rule are the load-bearing constraints that generate the one-dimensional order and ice-like correlations. The average structure being 'consistent with' these rules is not the same as establishing that they hold locally, because the refined occupancies and displacement ellipsoids are compatible with many local configurations. The authors should either provide direct experimental or computational evidence for these rules in Zn(hba) (e.g. EXAFS, total scattering, or DFT on representative local fragments) or demonstrate that the diffuse scattering is uniquely sensitive to the Li(inox)-derived rules by allowing variations in the MC constraints","section":"Results, 'Average structure: then and now' and 'Voronoi decomposition and Truchet tiling'"},{"comment":"The statement that collapsing the MC supercell 'recovers the experimentally-determined average structure' is supported only by a visual inset. A quantitative comparison is needed: for example, the refined fractional occupancies of the four Zn positions and the two hba orientations in the collapsed MC cell versus the P4_2/mmc refinement, and a comparison of positional/thermal parameters. Without this, the simultaneous consistency with the average structure is not quantitatively established.","section":"Monte Carlo simulations and diffuse scattering calculations, Fig. 9(b)"}],"minor_comments":[{"comment":"The author names contain stray spaces ('T ristan', 'Y evheniia'); these should be cleaned up.","section":"Author list"},{"comment":"Reference 14 is incomplete (journal volume and article number are missing). Also, please check the spelling in Reference 19: 'Mémoir' → 'Mémoire' and 'combinasions' → 'combinaisons'.","section":"References"},{"comment":"The comparison between experimental and calculated diffuse scattering would be aided by using the same colour scale and orientation in the two figures. Currently the experimental data are shown in (hk0) and (h0l), while the calculated pattern is shown in (hk0) and (0kl); please make the plane labelling consistent.","section":"Fig. 3 and Fig. 9"},{"comment":"The claim that 'Our results do not depend on the specific value of this constant' is not demonstrated. Even if the qualitative pattern is robust, a brief sensitivity test (e.g. k = 1, 3, 10 eV Å⁻²) would make the claim credible.","section":"Monte Carlo simulations, Eq. (2)"},{"comment":"The statement about subextensive entropy and crystal-size effects is interesting, but the sentence 'We have not yet explored this point experimentally' is an explicit limitation. Consider softening the claim or moving it more clearly into the outlook section.","section":"Conclusions"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the journal's scope and the Truchet-tile reinterpretation is scientifically interesting. The main gap is the absence of a quantitative validation of the Monte Carlo model against the diffuse scattering and average structure; this is fixable in revision and does not require new conceptual work. I see no reason for rejection, provided the authors address the circularity and quantitative-comparison concerns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is worth reading. It does something genuinely useful: it reinterprets the Zn(hba) structure, showing that the previously reported doubled-c cell with partial occupancies is likely an artifact of treating diffuse scattering as Bragg intensity. The new average structure in P4_2/mmc with c≈6 Å is supported by well-resolved rocking curves that distinguish Bragg from diffuse. The Truchet-tile model, with its mapping to square ice and the subextensive entropy point, is elegant and chemically motivated. The connection to Li(inox) is plausible, and the local two-short/two-long rule is physically reasonable.\n\nThe soft spots are real but not fatal. The Monte Carlo validation is more illustrative than quantitative. T_MC = 2J is chosen to give 'weak antipolar correlations implied by experiment,' and the odd number of layers is introduced to mimic the ~50 Å correlation length measured from the same diffuse scattering. So the model is not independently predicted; it is parameterized to reproduce the observations. The comparison in Fig. 9 is visual — no residuals or line-profile fits. You could imagine a different local-rule model also giving 'lumpy rods' and half-integral maxima. Also, the transfer of Li(inox) bonding rules to Zn(hba) is stated as an expectation. That is a reasonable chemical guess, but it isn't directly measured (e.g., by total scattering or EXAFS). If that rule were wrong, the ice-like correlations would collapse.\n\nThat said, the average-structure correction alone is a solid contribution. The authors are transparent about the unverified size-effect claim ('We have not yet explored this point experimentally'), and they do not overclaim DFT-level accuracy. The paper is a good candidate for peer review. A serious referee should ask for a quantitative diffuse-scattering comparison — even a simple R-factor along a few reciprocal lines would help — and a clearer statement of which parameters are fitted versus predicted. The Li(inox) transfer could be probed by showing the local coordination is consistent with the refined displacement parameters or by a small systematic study with related ligands.\n\nRecommendation: engage with the paper. Send it to referees. It deserves a proper review, but the authors will need to strengthen the evidence before it is convincing to a skeptical crystallographer.","headline":"Plausible and elegant structural reinterpretation of Zn(hba) with a Truchet/ice model, but the MC validation is partly circular and visual; deserves review with a request for quantitative comparison.","tokens_in":14159,"tokens_out":2330,"would_cite":true,"duration_ms":23133,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper argues that the apparently random disorder in the metal-organic framework Zn(hba) is actually governed by ice-like coordination rules and one-dimensional ligand order, and that a Truchet-tile model captures the material's local c","keywords":["Zn(hba)","Truchet tiling","correlated disorder","diffuse scattering","metal–organic framework","square ice","Monte Carlo simulation","Voronoi decomposition"],"falsifier":"Decisive experiment: a high-dynamic-range single-crystal diffraction measurement at l = odd in the (h0l) plane. If the features at l = (2n+1)/2 remain broad with a ~50 Å correlation length, the Truchet-tile model is supported; if they sharpen to instrument-limited Bragg peaks as the crystal grows, they are true superstructure reflections and the c-doubled model stands. Independently, Zn K-edge EXAFS or total-scattering PDF would show whether Zn–O bonds are two-short, two-long or four nearly equal.","tokens_in":13299,"feed_emoji":"🧊","tokens_out":6158,"duration_ms":49969,"temperature":0.7,"pith_summary":"Zn(hba), a metal–organic framework long thought to contain partially occupied sites and a doubled c-axis, actually owes its apparent messiness to correlated, not random, disorder. This paper shows that the material's diffuse X-ray scattering is governed by two local rules: every Zn2+ ion forms two short and two long bonds to its four ligand donors, and every row of hba2− ligands has a uniform polarity. These rules force one-dimensional ordering along each ligand chain and weak antipolar correlations between chains—exactly the two-in-two-out constraint of the square-ice model. A Truchet-tile representation, built by decorating Voronoi cells of the average structure, turns these rules into matching conditions, and Monte Carlo simulations produce models that reproduce both the average crystal structure and the diffuse scattering. If this model is right, the previously published c-doubled, partially occupied structure is a misinterpretation of diffuse scattering as Bragg intensity.","feed_headline":"Disordered MOF decoded as a Truchet tiling","feed_subtitle":"Ice-like rules and one-dimensional ligand order explain the crystal's diffuse scattering.","key_machinery":"The key mechanism is the Voronoi decomposition of the average crystal structure into node and linker tiles, followed by chemical decoration that lowers tile symmetry. The decorated tiles—four orientations for the Zn node, two for the hba linker—must pack according to matching rules derived from the two-short-two-long coordination and uniform chain polarity. These rules are equivalent to the two-in-two-out constraint of the square-ice model, which forces correlations that are strong along one dimension but only weak (antipolar) between rows; Monte Carlo simulation of these tile packings, with harmonic bond springs for displacements, yields configurations whose computed diffuse scattering matc","core_discovery":"The central claim is that the correlated disorder in Zn(hba) is a physical realisation of a Truchet tiling governed by ice-like coordination rules. Each Zn2+ centre is displaced along one of four ⟨110⟩ directions and connects to two carboxylate (short) and two phenoxide (long) donors; each chain of hba ligands must have a single polarity because reversing polarity would break the two-short-two-long rule or distort the coordination. These constraints yield four distinguishable node-tile orientations and two linker-tile orientations, and the matching rules between tiles generate one-dimensional order along ligand rows, weak antipolar order between rows, and a ~50 Å correlation length along the","pith_inferences":["We infer that the Truchet-tile formalism is likely to apply beyond these two frameworks: any network whose average node and linker sites have higher point symmetry than the chemical fragments themselves could hide analogous tile structure, which would be revealed by the same combination of Voronoi decomposition and diffuse-scattering analysis.","The 'barcode' picture suggests a concrete information-storage scheme: if edge orientations determine the bulk pattern, then controlling the ligand orientation at crystal edges during growth could program a specific global configuration, effectively writing data into the crystal.","A testable extension: total scattering or pair-distribution-function analysis of Zn(hba) should show the two-short-two-long Zn–O bond distribution; if instead a four-equal-bond environment is observed, the transfer of local rules from the ordered analogue fails.","The ~50 Å correlation length along c implies that nanocrystals of Zn(hba) should appear nearly ordered; comparing diffraction from microcrystals and nanocrystals would test the subextensive-entropy prediction directly."],"forward_implications":["The average structure of Zn(hba) is the higher-symmetry P42/mmc cell halved along c; the previously reported c-doubled structure with partially occupied sites is an artifact of integrating diffuse scattering as Bragg intensity.","Because the disorder follows square-ice rules, its configurational entropy is subextensive: the number of configurations grows with the square root of the number of nodes, so a crystal's information content is carried by binary 'barcodes' along the a and b axes.","Subextensive entropy makes the degree of disorder size-dependent: smaller crystals should show quantitatively different diffuse scattering than larger ones.","The four distinct pore-channel types, differing in ligand orientation, break the local point symmetry and should have different host–guest interactions, offering a route to bias specific disordered or ordered daughter configurations.","Chemical substitution controls the disorder: the aliphatic hca2− ligand gives an ordered analogue (Zn(hca)) while the aromatic hba2− favours disorder, and Co-substituted variants exist, so the configurational landscape appears navigable by synthesis."],"fun_headline_variants":["MOF's disorder hides a Truchet tiling","Zinc MOF's chaos follows ice-like rules","Truchet tiles explain MOF's diffuse scattering","Hidden Truchet pattern in a disordered MOF"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire model rests on transferring the local coordination rules of the ordered lithium analogue Li(inox) to disordered Zn(hba): every Zn2+ must form two short and two long Zn–O bonds, and every ligand chain must have uniform polarity; if either local rule is violated, the correlated-disorder model collapses.","fun_headline_variants_meta":{"raw":{"variants":["MOF's disorder hides a Truchet tiling","Zinc MOF's chaos follows ice-like rules","Truchet tiles explain MOF's diffuse scattering","Hidden Truchet pattern in a disordered MOF"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000604,"raw_usage":{"total_tokens":2632,"prompt_tokens":698,"completion_tokens":1934,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":442,"completion_tokens_details":{"reasoning_tokens":1884}},"tokens_in":442,"tokens_out":1934,"duration_ms":11425,"temperature":1.0,"reasoning_tokens":1884,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T05:08:40.774871+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Decisive experiment: a high-dynamic-range single-crystal diffraction measurement at l = odd in the (h0l) plane. If the features at l = (2n+1)/2 remain broad with a ~50 Å correlation length, the Truchet-tile model is supported; if they sharpen to instrument-limited Bragg peaks as the crystal grows, they are true superstructure reflections and the c-doubled model stands. Independently, Zn K-edge EXAFS or total-scattering PDF would show whether Zn–O bonds are two-short, two-long or four nearly equal.","supporting_citations":[],"review_version":1}