{"id":"81ac066f-fb2a-48b1-b425-0105ba746dd9","arxiv_id":"2608.01210","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":10,"one_line_summary":"Template-matched EBSD distinguishes triple-layer interlayer spacings in Nb-Ni and Nb-Co mu-phases differing by ~0.02-0.06 Å, with trends in partial agreement with XRD/DFT and in conflict with the paper's own HR-STEM data for Nb-Ni.","lead":"Electron backscatter diffraction, a common scanning-electron-microscope technique, was used to measure tiny gaps between atomic layers in Nb-Ni and Nb-Co intermetallic phases. The method picks up trends that usually require expensive local or averaged probes, so it could map subtle structure changes over large sample areas.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper's own HR-STEM validation (Table 3: 0.436 Å -> 0.386 Å for 51Nb-49Ni -> 58Nb-42Ni) reverses the EBSD dt trend (Table 2: 0.380 Å -> 0.420 Å), contradicting the abstract's 'good agreement with HR-TEM'.","rationale":"I read the paper's central claim as the abstract's statement that EBSD pattern matching predicts composition-dependent dt in agreement with XRD and HR-TEM. The reader's weakest_assumption targets the dynamical-simulation/PC model risk, which is real but partially mitigated by the paper's own parameter tests and PC precision estimates. The more direct, checkable problem is internal: the one independent HR-TEM dataset presented in the paper (Table 3) has the opposite Nb-Ni ordering to the EBSD result (Table 2), while Section 4.2 and the abstract assert agreement. This is a manuscript-internal inconsistency, not an outside-consensus disagreement, so it should be resolved before the central claim is accepted. A blinded re-extraction of the STEM motifs will settle it. I agree with the reader's CONDITIONAL verdict — this does not change it — but I would put the HR-STEM discrepancy, rather than the Bethe/PC concern, at the top of the required revisions.","tokens_in":30513,"tokens_out":10814,"duration_ms":118149,"concrete_test":"Perform a blinded re-analysis of the raw HAADF-STEM images for the two Nb-Ni foils using the §2.7 motif-extraction pipeline, without sample labels, and compute dt for each. If the unlabeled result preserves Table 3's ordering (58Nb-42Ni < 51Nb-49Ni), the abstract's HR-TEM agreement is falsified. If it instead yields 58Nb-42Ni > 51Nb-49Ni, Table 3 is a labelling/transcription error and the validation can be restored after correction. Ideally also independently refine the 6c2 z coordinate from the existing XRD data to provide a third check.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Central claim: the abstract's statement that EBSD pattern matching 'successfully predicts the change of interlayer spacing ... in good agreement with XRD and HR-TEM'. The decisive validation is an independent measurement. Table 3 reports HR-STEM dt = 0.436 ± 0.035 Å for 51Nb-49Ni and 0.386 ± 0.035 Å for 58Nb-42Ni, i.e., dt decreases with Nb content. Table 2 reports EBSD dt = 0.380 ± 0.012 Å and 0.420 ± 0.008 Å, i.e., dt increases by 0.040 Å. Thus the paper's internal HR-STEM data have the opposite ordering to the EBSD result for the Nb-Ni pair. Section 4.2 nonetheless states 'HR-STEM investigations in the present study reveal an increase of the dt site by about 0.1Å', which is not what Table 3 shows. Even if the two HR-STEM values overlap within their 0.035 Å error bars, the claimed 0.02–0.06 Å-level agreement is not demonstrated by these data. Because this contradiction sits exactly on the validation sentence of the abstract, the manuscript as written cannot support 'good agreement with HR-TEM'. It must be resolved by verifying the raw STEM motif extraction, sample labelling in Fig. 11, and the dt-K comparison before the EBSD capability claim is accepted.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates whether EBSD with dynamical template matching can resolve subtle crystallographic differences in Nb-Ni and Nb-Co mu-phases, specifically the interlayer spacing dt between the triple layer and Kagome layer, and the 3a site occupancy. The authors build a library of EMSoft-simulated EBSD patterns in which the 6c2 z-position (i.e., dt) and 3a occupancy are varied, match experimental patterns from four arc-melted samples using global orientation refinement and PCA, and extract dt values from polynomial fits to NCC landscapes. They report dt = 0.361 Å (48Nb-52Co), 0.380 Å (51Nb-49Ni), 0.399 Å (56Nb-44Co), 0.420 Å (58Nb-42Ni), i.e., dt increases with Nb content, and interpret this as a successful prediction in agreement with XRD and HR-TEM. They also report a full Nb occupancy of the 3a site for all samples, while acknowledging that occupancy changes could not actually be resolved. The paper includes extensive simulation-parameter studies, projection-centre calibration, and an HR-STEM motif-extraction comparison for two Nb-Ni samples.","tokens_in":30961,"tokens_out":4016,"duration_ms":43092,"significance":"If the central claim is correct, the paper would demonstrate a genuinely new capability: extracting sub-angstrom interlayer spacings from EBSD patterns at statistically meaningful sampling scales, complementing local HR-STEM and bulk XRD. The study is unusually careful in several respects: it provides open data and code, tests the influence of EMSoft Bethe parameters, dmin, acceleration voltage, and projection-centre misalignment on pattern distinctiveness, and uses a large-area projection-centre calibration plane. The internal consistency of the EBSD trend across four samples and its qualitative match to literature trends for Nb-Co are encouraging. However, the validation is substantially weakened by a direct internal contradiction with the paper's own HR-STEM data for Nb-Ni, and by the use of template libraries built from the same literature structures used as benchmarks. The significance of the result as a stand-alone measurement is therefore not established; at present it is best viewed as a promising but incomplete methodology study.","major_comments":[{"comment":"The abstract's claim of 'good agreement with XRD and HR-TEM' is contradicted by the paper's own HR-STEM data. Table 3 reports dt = 0.436 ± 0.035 Å for 51Nb-49Ni and 0.386 ± 0.035 Å for 58Nb-42Ni, i.e., a decrease of ~0.05 Å with increasing Nb content. Table 2 reports EBSD dt values of 0.380 ± 0.012 Å and 0.420 ± 0.008 Å, i.e., an increase of 0.040 Å. Section 4.2 states that 'HR-STEM investigations in the present study reveal an increase of the dt site by about 0.1Å', which is the opposite of what Table 3 shows. This is a load-bearing contradiction: the abstract explicitly advertises HR-TEM agreement, and Section 4.2 uses it to validate the EBSD trend. The authors must correct the STEM analysis or its interpretation, verify sample labelling in Fig. 11, and then revise the abstract and Section 4.2 accordingly. If the HR-STEM data remain as reported, the claim of HR-TEM agreement must be re","section":"Abstract and §4.2 vs. Table 3"},{"comment":"The validation is partly circular. The template library for Nb-Ni is generated from the structures of Joubert et al. (2004), and the Nb-Co starting structures are DFT-relaxed cells from Luo et al. (2023) — the same literature sources used later as benchmarks for the 'agreement'. The dt search range is explicitly 'according to the expected range of IL parameters from [5]'. The fitted dt values are therefore obtained by interpolation within a library that is centred on the expected literature values. This does not constitute an independent measurement of dt. The paper should either use templates derived from crystallographically independent sources (e.g., structures relaxed with different methods or bare experimental XRD-derived coordinates without literature z-values) or clearly state that the method can only rank candidate structures within a pre-specified structural model, not measure d","section":"§2.5.2, §3.6, §4.2"},{"comment":"The authors correctly state that they 'cannot fully exclude residual systematic errors in the absolute alignment' and cite Alkorta et al. showing that projection-centre offsets can mimic lattice strain. The reported sample-to-sample differences in dt are 0.02–0.06 Å, whereas a 0.5% lattice-parameter change (which the authors estimate could be present in the XRD data) corresponds to a phantom dt shift of 0.015 Å. A systematic PC error of ~1.7 px could therefore contribute a large fraction of the observed compositional trend. The precision analysis is rigorous (scatter around the plane is ~0.001), but precision does not constrain accuracy. The paper should provide a quantitative uncertainty budget that includes systematic PC errors, or explicitly restrict all claims to relative trends within the same calibration setup. As written, the absolute dt values and the compositional gradients are","section":"§4.1 and Fig. 8"},{"comment":"The authors admit that they 'cannot at present identify those or any crystallographic features which are decisive in the pattern matching.' While the empirical NCC trend is internally consistent, the claim that EBSD 'resolves' interlayer spacings is weakened if the physical origin of the NCC changes is unidentified. The paper would be strengthened by either a more detailed band-by-band or zone-axis analysis linking specific intensity changes to the 6c2 displacement, or by a clear statement in the abstract and conclusions that the method currently provides an empirical, library-based correlation rather than a physically interpretable measurement.","section":"§4.3"}],"minor_comments":[{"comment":"The definition of the normalised cross-correlation is incomplete: the denominator is written as (||A−Abar|| · ||B−Bbar||) but the norm is defined as 'qP i,j (Ai,j − Abar)' without a square-root symbol or square inside. Please correct the notation.","section":"Eq. (1)"},{"comment":"The text refers to '48Nb-42Co' in the misorientation-angle discussion; this should presumably be '48Nb-52Co' for consistency with the sample list.","section":"Figure 3"},{"comment":"The caption says 'No noise and offset applied.' without a panel label; several panels share this description. Please clarify which panel is which and use consistent notation for the SNR/offset values.","section":"Figure S4 caption"},{"comment":"Reference [55], cited for the influence of Bethe parameters on EBSD intensities, is a paper on Cu/W interface cohesion properties. It appears unrelated to the statement; please verify the citation and replace it with the appropriate source.","section":"Reference [55]"},{"comment":"The NCC threshold of 0.65 is introduced without a sensitivity analysis. Since this threshold filters patterns before averaging orientations, a brief comment on how the results depend on this choice would be helpful.","section":"§2.6"},{"comment":"The HR-STEM dt and dt−K values are given with standard deviations of 0.035 Å and 0.017–0.067 Å. Given the small sample sizes, please state the number of unit cells/motif triplets analysed and whether the scatter is dominated by local strain variations or measurement uncertainty.","section":"Table 3"}],"recommendation":"major_revision","confidential_remarks":"The internal contradiction between the abstract/Section 4.2 and Table 3 is the most serious issue. If the HR-STEM data cannot be re-analysed to match the EBSD trend, the abstract must be rewritten to remove the HR-TEM agreement claim. The circularity of the template library also needs to be addressed head-on; if the authors can show that the fitted dt values are insensitive to the choice of starting structures (e.g., by using a library built from independent DFT or by widening the search range), the paper could become a solid methodological contribution. The manuscript is otherwise well within scope for the journal and the data/code availability is a strong positive."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The EBSD trend is credible, but the abstract's 'good agreement with HR-TEM' is contradicted by the paper's own Table 3 for the Nb-Ni pair: the HR-STEM values (0.436 Å for 51Nb-49Ni, 0.386 Å for 58Nb-42Ni) decrease with Nb content, while the EBSD values (0.380 Å and 0.420 Å) increase. Section 4.2 even says HR-STEM 'reveal[s] an increase of the dt site by about 0.1Å,' which is not what their table shows. This internal contradiction sits squarely on the headline claim, so the validation sentence in the abstract cannot stand as written.\n\nWhat is genuinely new: using EBSD template matching against a library of dynamical simulations to read out triple-layer interlayer spacings in Nb-Ni and Nb-Co mu-phases, and getting a consistent trend across four compositions that matches the direction of XRD/DFT literature. The work is reproducible—EBSD data and template simulations are on Zenodo, code in AstroEBSD—and the sensitivity tests for projection-centre offsets, noise, Bethe parameters, and minimum d-spacing are careful. The authors are candid about not being able to rule out systematic PC errors or identify the crystallographic features driving the NCC changes. That honesty earns credit.\n\nSoft spots:\n\n- The HR-STEM discrepancy is not minor. It could be a sample-labelling issue in Fig. 11 or a motif extraction problem, but as printed it overturns the 'good agreement' claim. This must be resolved.\n- The 3a-site result is muddy: Section 3.6 reports full Nb occupancy of the 3a site for all samples, Section 4.2 says it disagrees with Nb-Co HR-STEM observations, and the conclusion says it could not be resolved. The status needs a clear statement.\n- Mild circularity: the template library is seeded from the same literature structures (Joubert et al., group DFT) used as benchmarks, and the search range is centered on expected values. That does not disqualify the EBSD results—the patterns are independent measurements—but the method's accuracy is anchored to input structures. A blind test or external calibration would strengthen it a lot.\n\nWho this is for: EBSD method developers and anyone studying Frank-Kasper phases who wants a scanning, area-averaged probe for interlayer spacings. Not a field-reshaping paper, but a practical capability.\n\nRecommendation: send it to peer review. The flaws are fixable with clearer writing, a resolution of the HR-STEM numbers, and a more restrained abstract. The core idea and the reproducible workflow deserve referee time, even though I would not cite the current version as validation.","headline":"The EBSD trend is credible, but the abstract's 'good agreement with HR-TEM' is contradicted by the paper's own Table 3—worth a serious referee, but the validation claim needs fixing.","tokens_in":31491,"tokens_out":4513,"would_cite":false,"duration_ms":41586,"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":"EBSD pattern matching can measure atomic-layer spacings in mu-phase intermetallics, resolving differences as small as 0.02–0.06 Å between compositions.","keywords":["EBSD","mu-phase","interlayer spacing","dynamical template matching","electron backscatter diffraction","intermetallics","projection centre calibration","Nb-Ni"],"falsifier":"Measure a sample with a known uniform interlayer spacing (e.g., a high-purity mu-phase single crystal whose structure has been refined by high-resolution XRD or neutron diffraction), run the same template-matching routine, and check whether the inferred dt matches the refined value within the reported scatter of about 0.01 Å; a systematic offset larger than the sample-to-sample differences would falsify the claim that EBSD resolves these spacings.","tokens_in":30418,"feed_emoji":"🔬","tokens_out":4110,"duration_ms":39951,"temperature":0.7,"pith_summary":"This paper asks whether electron backscatter diffraction (EBSD) can detect crystallographic features that change only the intensity of diffraction patterns, not the apparent orientation. It shows that by matching experimental patterns against a library of dynamically simulated templates that differ only in the spacing between atomic layers in the triple layer of mu-phase Nb-Ni and Nb-Co intermetallics, the method recovers interlayer spacings that track composition. The inferred spacings for four alloys agree with published XRD and HR-STEM values, so EBSD could become a way to map subtle structural variations across large sample areas. The same approach failed to resolve site-lattice occupancy of the 3a site, and the authors cannot yet say which pattern features drive the match.","feed_headline":"EBSD reads atomic-layer spacings in mu-phase alloys","feed_subtitle":"Pattern matching resolves 0.02–0.06 Å differences between Nb-Ni and Nb-Co mu-phase compositions, matching XRD and HR-STEM trends.","key_machinery":"The load-bearing mechanism is template matching between experimental EBSD patterns and dynamical diffraction simulations (EMSoft) of crystal structures whose 6c2 z-position (and hence triple-layer spacing dt) is systematically varied. A normalised cross-correlation metric, refined through an interior-point optimisation of orientation and projection centre, selects the best-matching structure; a fourth-degree polynomial fit to the NCC scores across the template grid locates the optimal dt. The distinction works because changes in dt alter pattern intensities in specific regions, even when the overall patterns look nearly identical.","core_discovery":"The central claim is that EBSD, combined with dynamical template matching, can resolve the interlayer spacing dt between the Kagome and triple layers in Nb-Ni and Nb-Co mu-phases. By generating template libraries in which the z-position of the 6c2 site is varied in steps of 0.025 Å and comparing experimental patterns from four alloy compositions using normalised cross-correlation, the authors infer dt values of 0.361 (48Nb-52Co), 0.380 (51Nb-49Ni), 0.399 (56Nb-44Co), and 0.420 Å (58Nb-42Ni). These values increase with Nb content in agreement with literature trends from XRD and HR-STEM, and the sample-to-sample differences are larger than the scatter within each sample. The authors are carefu","pith_inferences":["A natural extension is to apply the same template-matching workflow to other Frank-Kasper or ordered intermetallic phases where interlayer spacing is known to couple to slip behaviour, provided the identity of the variable atomic site is known in advance.","Because the authors find an orientation dependence of the inferred dt and cannot pin down the crystallographic features behind the NCC changes, part of the signal may be a compensation effect between simulation inaccuracies and projection-centre offsets; testing this would require a ground-truth sample with a known, uniform dt.","The failure to resolve 3a occupancy suggests that occupancy-related intensity changes are below the current detection floor; combining the method with excess/deficiency-corrected simulations or energy filtering may push sensitivity lower.","If the trends hold, EBSD could serve as a screening tool to identify local variations in spacing within a single grain, producing maps of 'structural phase' that are invisible to standard orientation maps."],"forward_implications":["EBSD can track composition-dependent interlayer spacings across areas much larger than HR-STEM fields of view, enabling correlative studies of structure and mechanical properties.","The reported dt values increase with Nb content in both systems, matching published XRD and HR-STEM trends, so the method captures real structural differences rather than noise.","The method does not currently resolve 3a-site occupancy, meaning the technique's sensitivity is limited to features with a sufficient scattering contribution.","The approach is computationally heavy: the full mu-phase parameter space is far too large to sample exhaustively, so the method is only practical when the problem is constrained to a few parameters.","For simpler crystal structures or with faster simulation methods, the same workflow could make subtle spacing measurements routine in the scanning electron microscope."],"supporting_citations":[{"why":"Supplies the initial crystal structures and lattice parameters for the Nb-Ni mu-phase, which are modified to build the template libraries.","marker":"[5]"},{"why":"Provides DFT-relaxed Nb-Co structures and HR-STEM measurements of interlayer spacings used as initial coordinates and as comparison values for the EBSD results.","marker":"[1]"},{"why":"The EMSoft software used to generate all dynamical template patterns.","marker":"[35]"},{"why":"Provides the PCA-based noise and dimension reduction approach applied to the experimental EBSD patterns.","marker":"[24]"},{"why":"The interior-point global optimisation routine used for orientation and projection-centre refinement.","marker":"[51]"},{"why":"Constrains the effective energy spectrum of backscattered electrons, justifying the single-voltage template simulations.","marker":"[53]"},{"why":"Gives the Bethe parameter recommendations for heavier elements used in the dynamical simulations.","marker":"[54]"},{"why":"Demonstrates that projection-centre offsets can mimic lattice strain, motivating the need for careful PC calibration in this study.","marker":"[70]"}],"fun_headline_variants":["EBSD resolves interlayer spacings in mu-phase alloys","EBSD sees 0.02–0.06 Å layer shifts in Nb-Ni and Nb-Co","Pattern-matched EBSD reads atomic layer distances in mu-phases","EBSD distinguishes Nb-Ni and Nb-Co mu-phase spacings","EBSD benchmarks subtle interlayer spacing differences"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The inferred interlayer spacings are only as good as the simulated pattern intensities, which assume a single accelerating voltage, fixed Bethe parameters, and an affine projection-centre model under conditions where the authors cannot rule out systematic projection-centre errors.","fun_headline_variants_meta":{"raw":{"variants":["EBSD resolves interlayer spacings in mu-phase alloys","EBSD sees 0.02–0.06 Å layer shifts in Nb-Ni and Nb-Co","Pattern-matched EBSD reads atomic layer distances in mu-phases","EBSD distinguishes Nb-Ni and Nb-Co mu-phase spacings","EBSD benchmarks subtle interlayer spacing differences"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000296,"raw_usage":{"total_tokens":1602,"prompt_tokens":838,"completion_tokens":764,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":582,"completion_tokens_details":{"reasoning_tokens":683}},"tokens_in":582,"tokens_out":764,"duration_ms":7708,"temperature":1.0,"reasoning_tokens":683,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T00:26:50.226010+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure a sample with a known uniform interlayer spacing (e.g., a high-purity mu-phase single crystal whose structure has been refined by high-resolution XRD or neutron diffraction), run the same template-matching routine, and check whether the inferred dt matches the refined value within the reported scatter of about 0.01 Å; a systematic offset larger than the sample-to-sample differences would falsify the claim that EBSD resolves these spacings.","supporting_citations":[{"cited_title":"Joubert, B","cited_arxiv_id":null,"evidence_quote":"Supplies the initial crystal structures and lattice parameters for the Nb-Ni mu-phase, which are modified to build the template libraries."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides DFT-relaxed Nb-Co structures and HR-STEM measurements of interlayer spacings used as initial coordinates and as comparison values for the EBSD results."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The EMSoft software used to generate all dynamical template patterns."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The interior-point global optimisation routine used for orientation and projection-centre refinement."},{"cited_title":"Winkelmann, T","cited_arxiv_id":null,"evidence_quote":"Constrains the effective energy spectrum of backscattered electrons, justifying the single-voltage template simulations."}],"review_version":1}