{"id":"8613c845-7c63-41c5-a098-42cea2f009f4","arxiv_id":"2412.05219","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Single-crystal neutron and X-ray diffraction establish noncentrosymmetric space group I4_1md for CeGaGe, and some samples show a transition to primitive P4_3/P4_1 symmetry on cooling.","lead":"Neutron and X-ray experiments on the candidate Weyl semimetal CeGaGe show its crystal structure lacks a mirror center, with the space group I4_1md rather than the centrosymmetric I4_1/amd proposed earlier. The work resolves a long-standing ambiguity that powder diffraction alone could not settle, clarifying the symmetry that future electronic-structure and magnetic studies must use.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 911/1236 weak-reflection statistic is unweighted and applied after a |Fo|^2/σ^2 ≥ 3 cut; under positive-intensity selection this can favor whichever model predicts larger |Fc|^2, so the I4_1md assignment may not be statistically established.","rationale":"The reader's weakest assumption correctly identifies the weak-reflection subset and the anomalous GoF as the fragile part of the argument. I agree that systematic effects such as background subtraction, absorption, extinction, and outlier rejection could bias weak intensities. My concern is more specific: the quantitative '911 of 1236' metric is an unweighted comparison applied after a positive-intensity selection, and the higher-prediction model has an intrinsic advantage under such selection. This is a concrete statistical mechanism, not merely a generic worry about data quality. Because the reported GoF values are far from 1, one cannot fall back on χ²-type significance arguments, and the paper provides no alternative statistical test. The SCXRD data are consistent with I4_1md but are nearly insensitive to Ga/Ge ordering, so they do not independently rescue the neutron analysis. The central assertion may well be correct, and there is strong qualitative evidence, but the decisive quantitative statistic is not yet demonstrated to be unbiased. This leaves the verdict CONDITIONAL: the conclusion is plausible and probably right, but the paper should either provide a bias-free statistical comparison or release the reduced data so an independent reduction and refinement can verify the result. I do not see grounds for outright rejection, since the physical picture and the qualitative agreement are compelling, but the claimed 'definitive' resolution is not fully supported by the statistics as presented.","tokens_in":21459,"tokens_out":6665,"duration_ms":69385,"concrete_test":"Generate synthetic data from the refined I4_1/amd model: for each of the 4430 hkl, sample observed intensities from the reported σ distribution around the I4_1/amd |Fc|^2, apply the same |Fo|^2/σ^2 ≥ 3 cut and the same GSAS-II refinement path, and compute the fraction of the 1236 weak reflections that are 'better' for I4_1md under the paper's agreement metric. If the synthetic I4_1/amd data yields a comparable fraction (roughly 70% or more), the metric is demonstrably biased and the assignment is not established. If the synthetic fraction is near 50%, the concern is weakened, and the real data should instead be reanalyzed with a weighted paired test using the reported σ values.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The load-bearing evidence for I4_1md over I4_1/amd is the set of 1236 weak reflections discussed in Sec. III around Eq. (1) and Fig. 5: 911 are said to agree better with I4_1md. This count is not corrected for selection bias. The data were pre-filtered with |Fo|^2/σ^2 ≥ 3, so every retained weak reflection has a positive observed intensity that is statistically significant relative to its reported uncertainty. For reflections whose true intensity is near zero, this selection systematically biases the observed values upward relative to the true mean. Since I4_1md predicts larger |Fc|^2 than I4_1/amd for 986 of the 1236 reflections, the larger-prediction model wins more of the unweighted absolute-difference comparisons even if I4_1/amd were the true structure. The reported GoF values of 19.13 and 22.40, far from 1, indicate that the σ estimates are not reliable enough to convert this asymmetry into a formal significance. No Hamilton R-factor test, paired residual test, or likelihood comparison using σ is reported. The SCXRD corroboration is weaker than it might appear because Ga and Ge have nearly equal X-ray form factors, so the two models are nearly degenerate in X-ray refinement; the Flack parameter near 0.5 is not independent evidence for noncentrosymmetry unless resonant effects are demonstrated at Mo Kα. Thus the central claim currently rests on a statistic that can be biased by the very selection used to define the weak-reflection subset.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a structural determination of the candidate Weyl semimetal CeGaGe using single-crystal neutron diffraction (SCND) at 100 K, complemented by single-crystal X-ray diffraction (SCXRD), powder X-ray diffraction (PXRD), and energy-dispersive X-ray spectroscopy (EDX). The authors claim that the SCND data confirm the noncentrosymmetric space group I4_1md (No. 109) over the centrosymmetric I4_1/amd (No. 141), citing global R-factors of 0.161 vs 0.224 and a weak-reflection agreement count of 911/1236. They also report that some flux-grown crystals show additional weak reflections at 100 K consistent with a transition to primitive chiral space groups P4_3/P4_1, while a float-zone crystal retains I4_1md down to 100 K.","tokens_in":21857,"tokens_out":11563,"duration_ms":107695,"significance":"If the central claim holds, the paper would resolve a long-standing structural ambiguity in an equiatomic rare-earth intermetallic and would establish CeGaGe as a structurally valid noncentrosymmetric platform for studying Weyl physics in the presence of magnetic moments. The paper's demonstration that PXRD is insensitive to the space-group choice, and the careful SCND measurement with extensive high-angle coverage, are valuable contributions. The compositional consistency between EDX and the I4_1md refinement is a useful cross-check. However, the statistical evidence for the space-group assignment is presently incomplete: the key model comparison relies on unweighted counts and goodness-of-fit values far from unity, and the SCXRD Flack parameter does not provide the independent corroboration claimed. The conclusion may well be correct, but the paper's stated 'definitively resolves' is stronger than the evidence currently supports.","major_comments":[{"comment":"The 911/1236 statistic is an unweighted count of which model gives a smaller absolute residual, applied to reflections pre-selected with |Fo|^2/σ^2 ≥ 3. Because every retained weak reflection has a positive, statistically significant observed intensity, this selection biases |Fo|^2 upward relative to the true intensity. The I4_1md model predicts larger |Fc|^2 for 986 of the 1236 reflections, so the larger-prediction model can win more of the absolute-difference comparisons even when I4_1/amd is the true structure. The authors should present a bias-corrected comparison (e.g., include below-threshold reflections, use σ-weighted residuals, or simulate the selection under both models) or apply a formal hypothesis test such as the Hamilton R-factor test.","section":"Section III, weak-reflection analysis (Fig. 5, 1236-reflection subset)"},{"comment":"The paper reports GoF = 19.13 for I4_1md and 22.40 for I4_1/amd, values far from the ideal of 1, indicating that the σ estimates or the models are substantially imperfect. The R-factor difference (0.161 vs 0.224) is presented as decisive, but with GoF this large the difference may be dominated by systematic errors in TOF reduction, absorption, extinction, or background treatment rather than by space-group symmetry. The manuscript does not report the weighted R-factor, the number of parameters and degrees of freedom, or a formal likelihood/Hamilton test that uses the reported σ's. Without such a test, the statistical significance of the R-factor difference is not established.","section":"Section III, reliability factors and goodness-of-fit"},{"comment":"The claim that a Flack parameter of 0.49(8) 'indicates that the parameter is well defined and so the structure is noncentrosymmetric' is not logically valid. A centrosymmetric structure refined in a noncentrosymmetric space group with an inversion twin can also yield a well-defined Flack parameter near 0.5. Because Ga and Ge have nearly equal X-ray form factors at Mo Kα and the anomalous dispersion of Ce is weak at this wavelength, the SCXRD refinement cannot independently distinguish the two structural models; the slightly better R-factors (0.0249 vs 0.0263) are likely not significant. The SCXRD data therefore do not provide the corroboration claimed in the abstract and conclusions.","section":"Section IV, SCXRD and Flack parameter"}],"minor_comments":[{"comment":"The word 'complimentary' should be 'complementary' in the abstract, the introduction, and Section IV (Section IV, first paragraph).","section":"Abstract and Introduction"},{"comment":"The R-factor in Eq. (1) is defined with |Fo| and |Fc| (not their squares), but the text and figures discuss |Fo|^2 vs |Fc|^2. Please clarify the convention used in the reported R-factors and the axis labels of Figs. 4 and 7.","section":"Section III, around Eq. (1)"},{"comment":"The sentence 'Each of the structure factors had |Fo|^2/σ_o^2 ≥ 3, which is larger than the forced signal to noise cutoff of |Fo|^2/σ_o^2 ≥ 1' is confusing; specify which cutoff is applied by the TOPAZ/MANTID pipeline and which is imposed by the authors.","section":"Section III, data reduction paragraph"},{"comment":"The composition entries for the I4_1/amd refinement (Ga 0.96(2) and Ge 1.00(8)) are not explained; if these are site occupancies on a shared 8e site, they should sum to 1. Please clarify the normalization convention.","section":"Table I and Section II"},{"comment":"The transition to P4_3/P4_1 is inferred from weak extra reflections in SCXRD, and the authors speculate about impurity or stoichiometry effects. A brief discussion of possible multiple domains or multiple crystals would help readers assess the single-crystal evidence.","section":"Section V, structural transition"},{"comment":"Reference [73] points to Supplemental Material via a DOI link that may not be stable; consider providing an archive identifier or a note on how to access the data.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper's central claim is plausible and the SCND data set is valuable, but the evidence as presented does not meet the standard for a 'definitive' space-group assignment. The selection-bias issue in the weak-reflection count is serious and should be addressed through simulations or formal statistical tests. If the authors cannot provide such an analysis, the conclusions should be softened to report a strong preference rather than confirmation. The SCXRD Flack parameter argument should be revised. I recommend major revision rather than rejection because the underlying data may support the I4_1md assignment after proper analysis."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The central structural claim, that CeGaGe is I4_1md, is probably correct, but the main statistical comparison is biased and the goodness-of-fit values are unexplained. The paper is worth a careful referee.\n\nWhat it does well: this is the first side-by-side single-crystal neutron and X-ray refinement comparison for CeGaGe, and the first observation of a possible temperature-driven I4_1md to P4_3/P4_1 transition in some samples. The TOPAZ experiment is competently done; 4430 reflections, 19 rotations, good merging. The physical insight about high-angle Bragg planes containing only Ga or only Ge in I4_1md, compared to mixed planes in I4_1/amd, is clear and explains why the weak reflections carry the discrimination. The PXRD demonstration that powder data cannot distinguish the two models is a useful caution.\n\nThe soft spots are real. The 911/1236 statistic is unweighted and applied after a |Fo|^2/σ^2 ≥ 3 cut. That cut biases observed intensities upward for weak reflections, and since I4_1md predicts larger |Fc|^2 for 986 of the 1236 reflections, the larger-prediction model will win more absolute-difference comparisons even if I4_1/amd were true. So this count is not the decisive evidence the paper makes it out to be. The R-factor difference (0.161 vs 0.224) is suggestive but is driven by the same weak reflections, so it shares the same problem. The reported goodness-of-fit values of 19.13 and 22.40 are far from 1; the paper does not discuss what that implies about σ estimation, and it matters for any formal significance claim. The SCXRD Flack parameter of 0.49(8) at Mo Kα is weak independent support because Ga and Ge have nearly equal X-ray form factors; the Flack value mostly reflects the inversion twin ratio. The structural transition is based on one flux-grown crystal at two temperatures, and the speculation about impurities is reasonable, but it is not a headline result.\n\nThe fix is straightforward: report a Hamilton R-factor ratio test or a paired residual test using the σ's, and say plainly what the GoF values mean. The conclusion will probably survive, but the current argument oversells a biased statistic.\n\nSend it to a serious referee. It is a careful experiment on a material of real interest in the magnetic Weyl family, and the structural question matters for subsequent band-structure and transport work.","headline":"The CeGaGe I4_1md assignment is likely right but the headline weak-reflection statistic is biased and the GoF values are unexplained; still deserves a genuine referee.","tokens_in":22422,"tokens_out":4655,"would_cite":true,"duration_ms":46028,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["61.05.fm"],"model":"deepseek-v4-flash","headline":"Single-crystal neutron diffraction shows that CeGaGe is noncentrosymmetric, with space group I4_1md (109) rather than the centrosymmetric I4_1/amd (141) that powder X-ray data suggested.","keywords":["CeGaGe","Weyl semimetal","noncentrosymmetric crystal structure","space group I4_1md","single-crystal neutron diffraction","inversion symmetry","Ga/Ge site ordering","structural phase transition"],"falsifier":"On the same float-zone crystal, re-collect the weak high-angle reflections with longer counting time or a different neutron wavelength and refine both space-group models with unbiased sigma estimates; if the 911-of-1236 preference for $I4_1md$ is not reproduced, or if the two models become statistically indistinguishable, the central claim is falsified.","tokens_in":21260,"feed_emoji":"🔬","tokens_out":10108,"duration_ms":93127,"temperature":0.7,"pith_summary":"Single-crystal neutron diffraction shows that the candidate Weyl semimetal CeGaGe is noncentrosymmetric, with space group $I4_1md$ (109) rather than the centrosymmetric $I4_1/amd$ (141) that powder X-ray data had suggested. The distinction matters because broken inversion symmetry is one of the two routes to Weyl points in a band structure, so the answer decides whether CeGaGe is structurally eligible for inversion-breaking Weyl physics. The neutron experiment exploits the roughly 12% difference in the neutron scattering lengths of Ga and Ge, which X-rays cannot tell apart. The decisive evidence is a set of weak high-angle reflections whose Bragg planes contain only Ga or only Ge in $I4_1md$ but a mix of both in $I4_1/amd$. Some flux-grown crystals additionally show a subtle transition to primitive chiral $P4_3$/$P4_1$ symmetry on cooling.","feed_headline":"Neutron study: CeGaGe breaks inversion symmetry","feed_subtitle":"Weak high-angle reflections settle Ga/Ge order X-rays cannot see, clearing a path for Weyl semimetal studies.","key_machinery":"The load-bearing contrast is the difference between the bound coherent neutron scattering lengths of Ga and Ge, which differ by about 12%, while their X-ray scattering factors are nearly identical across scattering angles. In the noncentrosymmetric $I4_1md$ structure, Ga and Ge occupy distinct $4a$ sites; in the centrosymmetric $I4_1/amd$ structure, they would randomly share an $8e$ site. The experiment's discriminating power comes from weak, high-angle reflections whose Bragg planes can contain Ga without Ge or vice versa in $I4_1md$ but would contain both elements in $I4_1/amd$. The comparison of refined structure factors for the two models, including the $R$-factor gap and the 911-of-1236 count, is what carries the assignment.","core_discovery":"CeGaGe forms with noncentrosymmetric $I4_1md$ symmetry at room temperature and, in float-zone crystals, remains so at 100 K. This is established by single-crystal time-of-flight neutron diffraction with 4430 indexed reflections: refinements give $R = 0.161$ for $I4_1md$ versus $0.224$ for $I4_1/amd$, and among 1236 weak reflections with $|F_c|^2$ between $10^{-4}$ and $1.25$, 911 put the observed intensities in better agreement with $I4_1md$. The physically intuitive content is that high-angle reflections such as $(-2,-2,-16)$ correspond to Bragg planes that can be occupied by Ga alone or Ge alone in the noncentrosymmetric structure, whereas the centrosymmetric alternative would force both elements onto the same site and nearly extinguish those reflections. Single-crystal X-ray diffraction corroborates the noncentrosymmetric assignment through a Flack parameter of $0.49(8)$, indicating near-perfect inversion twinning. The paper also reports that some flux-grown crystals show a subtle structural transition from body-centered $I4_1md$ to primitive chiral $P4_3$ (or $P4_1$) symmetry between room temperature and 100 K.","pith_inferences":["The reported sample-to-sample variation suggests that transport and magnetic measurements of CeGaGe may have been performed on crystals with different low-temperature crystallographic symmetries; re-characterizing the very crystals used in those experiments would show whether this is so.","The same high-angle, weak-reflection logic could be used to screen other RXZ candidates (rare earth, Al/Ga, Si/Ge) where X-ray contrast between the two p-element atoms is poor, especially for detecting hidden symmetry-lowering transitions.","If the $I4_1md$-to-$P4_3$/$P4_1$ transition is continuous or occurs near the magnetic ordering temperature, its thermodynamic signature would be difficult to see in specific heat; measuring the temperature dependence of elastic constants is a concrete testable extension that would locate the transition and reveal any coupling to magnetism.","Because the 911-of-1236 statistic is paired with a goodness-of-fit far from the ideal value of 1, a more rigorous statistical comparison of the two refinements would make the space-group assignment quantitative rather than indicative."],"forward_implications":["CeGaGe satisfies the structural prerequisite for inversion-breaking Weyl physics in float-zone crystals at least down to 100 K, so future band-structure calculations and transport experiments can be interpreted against a noncentrosymmetric model.","Powder X-ray diffraction cannot reliably assign the space group in this material family when the two p-element atoms are nearly isoelectronic; a complementary single-crystal neutron experiment is needed for a definitive answer.","Studies of CeGaGe must account for sample history: some flux-grown crystals transform to primitive chiral $P4_3$/$P4_1$ symmetry on cooling, which changes the symmetry conditions for Weyl nodes and magnetic order.","Earlier physical-property reports on CeGaGe were based on samples whose symmetry was not definitively known, and the new structural determination puts those measurements on firmer footing."],"supporting_citations":[{"why":"Early powder X-ray report assigning CeGaGe to the centrosymmetric I4_1/amd structure that the neutron experiment overturns.","marker":"[66]"},{"why":"Follow-up review of CeXX' intermetallics that repeats the centrosymmetric assignment from powder data.","marker":"[67]"},{"why":"Prior single-crystal study of flux-grown CeGaGe that fit powder data to I4_1md and reported the magnetic and transport baseline.","marker":"[69]"},{"why":"NdAlSi single-crystal neutron diffraction precedent showing the method resolves site ordering between elements with similar X-ray scattering factors.","marker":"[11]"},{"why":"Supplies the computed X-ray atomic scattering factors for Ga and Ge showing they are nearly indistinguishable.","marker":"[65]"},{"why":"Source of the bound coherent neutron scattering lengths whose 12% Ga/Ge difference makes the neutron measurement discriminating.","marker":"[71, 72]"}],"fun_headline_variants":["Neutrons settle CeGaGe: noncentrosymmetric Weyl candidate","High-angle reflections crack CeGaGe's symmetry","CeGaGe's inversion symmetry broken, neutrons confirm","Neutron diffraction: CeGaGe is not centrosymmetric"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The $I4_1md$ conclusion depends on the weak high-angle reflection intensities surviving time-of-flight data reduction (background subtraction, absorption, outlier rejection, and extinction correction) without systematic bias, and on the reported difference in goodness of fit (19.13 versus 22.40) being statistically meaningful even though both values are far from the ideal of 1.","fun_headline_variants_meta":{"raw":{"variants":["Neutrons settle CeGaGe: noncentrosymmetric Weyl candidate","High-angle reflections crack CeGaGe's symmetry","CeGaGe's inversion symmetry broken, neutrons confirm","Neutron diffraction: CeGaGe is not centrosymmetric"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000553,"raw_usage":{"total_tokens":2740,"prompt_tokens":1155,"completion_tokens":1585,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":771,"completion_tokens_details":{"reasoning_tokens":1516}},"tokens_in":771,"tokens_out":1585,"duration_ms":14053,"temperature":1.0,"reasoning_tokens":1516,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T20:50:22.264696+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"On the same float-zone crystal, re-collect the weak high-angle reflections with longer counting time or a different neutron wavelength and refine both space-group models with unbiased sigma estimates; if the 911-of-1236 preference for $I4_1md$ is not reproduced, or if the two models become statistically indistinguishable, the central claim is falsified.","supporting_citations":[{"cited_title":"doi:10.1088/1361-648x/ac987a","cited_arxiv_id":null,"evidence_quote":"Early powder X-ray report assigning CeGaGe to the centrosymmetric I4_1/amd structure that the neutron experiment overturns."},{"cited_title":"Equiatomic cerium intermetallics CeXX ′ with two p elements","cited_arxiv_id":null,"evidence_quote":"Follow-up review of CeXX' intermetallics that repeats the centrosymmetric assignment from powder data."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the computed X-ray atomic scattering factors for Ga and Ge showing they are nearly indistinguishable."}],"review_version":1}