{"id":"7ecc8911-93e2-4535-b17e-5586c1b6588b","arxiv_id":"1908.05734","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Convergent beam electron diffraction shows MBE-grown Cd3As2 thin films have the centrosymmetric tetragonal 4/mmm point group, consistent with the Dirac semimetal description.","lead":"This paper uses a specialized electron microscopy technique to determine the crystal symmetry of thin films of the topological semimetal Cd3As2. It finds the films are centrosymmetric, confirming they are Dirac semimetals rather than Weyl semimetals.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 4/mmm verdict rests on a visually assigned 2mm symmetry in a single [110] whole-pattern CBED; without a quantitative symmetry metric or raw data, a false second mirror would collapse the claim.","rationale":"The reader's weakest_assumption identifies the same load-bearing step: the visual assignment of 2mm to the [110] WP and the reliability of the simulation comparison. I agree that this is the critical point, and I sharpen it: because the bright-field symmetry is identical for 4/mmm and 4mm along [110], the entire point-group determination reduces to the presence of one extra mirror in the HOLZ-containing whole pattern. This is a binary visual judgment on a single experimental pattern, with no quantitative metric, no stated simulation parameters, and no raw data provided. If that judgment is wrong, the film would be assigned to 4mm, a Weyl semimetal, directly contradicting the paper's Dirac classification. The claim is logically sound given the reported pattern, and the paper is competent, but the evidence as presented is not independently verifiable. Therefore the appropriate verdict is CONDITIONAL: the central claim should be accepted only if the raw [110] WP and simulation inputs are made available and the 2mm vs m distinction is quantitatively confirmed. This is not an accusation of error; it is a reproducibility requirement for a claim that hinges on a single pattern's symmetry.","tokens_in":7486,"tokens_out":10754,"duration_ms":109220,"concrete_test":"From the raw data, extract the experimental [110] whole pattern (including the FOLZ ring) and compute a normalized cross-correlation or R-factor between the pattern and its reflection across the mirror that is present in the I41cd simulation, versus the reflection across the orthogonal mirror. Apply the same metric to simulated I41/acd and I41cd patterns at the same thickness, voltage, and absorption parameters. If the contrast for the second mirror in the experimental pattern is not significantly above the noise level and not significantly closer to the I41/acd simulation than to the I41cd simulation, the 2mm assignment is not established and the film could be non-centrosymmetric 4mm.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that the films are centrosymmetric 4/mmm rests entirely on the experimental [110] whole pattern (WP) showing 2mm rather than m. This is a whole-pattern/HOLZ effect: the bright-field disc along [110] is 2mm for both 4/mmm and 4mm, so the BP cannot discriminate. The only evidence for the second mirror is visual inspection of Fig. 3(b) and the FOLZ magnification. If the recorded WP actually has only one mirror (e.g., due to a small tilt, specimen bending, FIB damage, or noise in the weak HOLZ reflections), the pattern is fully consistent with the non-centrosymmetric 4mm point group and the central claim fails. The paper does not report the raw patterns, a quantitative symmetry metric, or the Bloch wave simulation inputs (thickness, absorption, Debye-Waller factors, accelerating voltage). The simulation for I41cd showing only m is a symmetry-required outcome for an untilted zone axis, so it is not independent confirmation; the experimental pattern is the only load-bearing evidence. Without a means to distinguish a true 2mm WP from an m WP with noise, the inversion-center conclusion is not independently checkable.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a convergent beam electron diffraction (CBED) study of molecular-beam-epitaxy-grown Cd3As2 thin films. From the symmetries of the bright-field discs and whole patterns recorded along [2-01], [021], [1-10] and [001], together with HAADF-STEM imaging and Bloch wave simulations, the authors conclude that the films belong to the centrosymmetric tetragonal point group 4/mmm (space group I41/acd), and therefore are Dirac semimetals. The key differentiating observation is the claimed 2mm symmetry in the [1-10] whole pattern, which is expected for 4/mmm but not for the noncentrosymmetric 4mm (I41cd) structure.","tokens_in":7705,"tokens_out":3679,"duration_ms":38227,"significance":"If the conclusion is correct, the paper resolves an important structural ambiguity for thin-film Cd3As2: it assigns the centrosymmetric structure, aligning the films with the Dirac-semimetal description and showing that MBE growth and strain relaxation do not alter the bulk vacancy ordering. The multiple-zone-axis strategy, use of the Buxton tables, and comparison with Bloch wave simulations of both candidate structures are appropriate and give the argument a coherent internal logic. The paper's usefulness would be enhanced by making the symmetry evaluation quantitative and by reporting simulation parameters; in its present form the decisive step rests on visual inspection of a small number of patterns.","major_comments":[{"comment":"The central claim that the films have 4/mmm rather than 4mm rests on the assignment of 2mm symmetry to the experimental [1-10] whole pattern (WP), but this assignment is made by eye from a single false-colored pattern and no quantitative symmetry metric is reported. A pattern with one mirror plus noise can easily be read as having two mirrors, and the second mirror is the only feature that distinguishes 4/mmm from 4mm in this zone axis; as the authors themselves note, the bright-field disc along [1-10] cannot discriminate between the two point groups. Please provide a quantitative symmetry measure (for example, intensity residuals after imposing the candidate mirror operations, or a correlation coefficient against symmetrized versions), and show the unprocessed or minimally processed patterns, so that the 2mm-versus-m decision can be checked independently.","section":"Results and Discussion, Fig. 3"},{"comment":"The Bloch wave simulations are used as confirmation of the Buxton-table analysis, but the manuscript gives no simulation parameters (thickness, absorption parameters, Debye-Waller factors, accelerating voltage, orientation or beam tilt) and no quantitative comparison between simulated and experimental patterns. Without these, the reader cannot assess whether the simulated I41cd [1-10] WP is representative of the experimental scattering conditions or whether the absence of the second mirror in that simulation is robust. Please include the input parameters and a direct side-by-side comparison with the experimental pattern, ideally using the same symmetry metric applied to the data.","section":"Results and Discussion, Fig. 4"},{"comment":"The final discrimination between 4/mmm and m3m in Table I relies on the HAADF-STEM evidence for ordered Cd vacancies and on the observed differences in HOLZ ring diameters between zones A, B and C. This reasoning is reasonable, but the paper does not show the supplementary <100> WP that is cited in the text as additional confirmation; please show that pattern or explicitly describe how it enters the point-group intersection. Without it, the argument that the additional zone axis rules out m3m is not fully documented in the main text.","section":"Results and Discussion, Table I and Fig. 5"}],"minor_comments":[{"comment":"The phrase 'reported room temperature crystal structures of Cd3As2 reported differ' is redundant; please rephrase for clarity.","section":"Abstract and Introduction"},{"comment":"Figure 3(b) is described as a magnification of parts of the FOLZ in the [1-10] WP, but the caption does not state the exact region or the magnification; a dashed box in Fig. 3(a) would help the reader locate the magnified area.","section":"Fig. 3 caption"},{"comment":"Reference [26] gives 'A. Pietrasz' and 'K. Lukaszew', but the same authors appear as 'Pietraszko' and 'Lukaszewicz' in Reference [17]; please check and correct these names.","section":"References"},{"comment":"Table I is difficult to read because all possible point groups are concatenated in a single column; a properly formatted table with separate rows for each zone axis and each candidate point group would improve clarity.","section":"Table I"},{"comment":"The statement that 'additional confirmation' comes from a <100> WP refers to the Supplemental Material but does not cite a specific supplementary figure; please add the figure number.","section":"Results and Discussion"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Honggyu and colleagues have produced a clean, well-exposed determination of the point group of MBE-grown Cd3As2 thin films. The result—4/mmm, centrosymmetric, space group I41/acd—settles an ambiguity that matters for topological classification: these are Dirac, not Weyl, semimetals. This is the first CBED-based point group measurement on this material, and the dynamical nature of CBED is exactly what is needed to break Friedel's law and distinguish the two candidate structures.\n\nThe experimental logic is sound. The authors record CBED from four zone axes, use the Buxton tables to list the compatible diffraction groups, and show that the observed m, m, 2mm, and 4mm symmetries in BP and WP reduce the candidates to 4/mmm and m3m, with HAADF-STEM ruling out the cubic phase. The Bloch wave simulation of the I41cd structure reproduces the expected single mirror in the [110] WP, which is a useful check that the table reading is correct, though not an independent confirmation of the experimental observation.\n\nThe soft spot, and it is a genuine one, is the load-bearing symmetry assignment in the experimental [110] whole pattern. The claim that the WP shows 2mm and not just m is made by visual inspection of Fig. 3(b); there is no quantitative symmetry metric (e.g., residual after mirror averaging), no raw data in the supplement, and the simulation inputs are not specified in enough detail to reproduce. If the recorded pattern actually has only one mirror because of a slight beam tilt or specimen bending, the conclusion would collapse to 4mm. That is a real risk, although the fact that two independent zone axes and the [001] pattern all line up with the same assignment mitigates it.\n\nA minor overreach: the conclusion that the film structure is 'not modified by any of the parameters of thin film growth' is broader than the evidence—only one film thickness, one substrate orientation, and one growth condition were examined.\n\nOverall, the central claim is convincing and important for the field. The paper deserves peer review; the referees should ask for a quantitative WP symmetry analysis and more complete simulation metadata. I would cite this for the structural determination, with the caveat that the key WP pattern needs independent confirmation.\n\nRecommendation: accept with minor revisions.","headline":"Careful CBED work that convincingly places MBE-grown Cd3As2 in the centrosymmetric 4/mmm point group, with the caveat that the crucial symmetry call is visual.","tokens_in":8253,"tokens_out":2463,"would_cite":true,"duration_ms":23531,"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":"Convergent beam electron diffraction shows that molecular-beam-epitaxy-grown Cd3As2 thin films belong to the centrosymmetric tetragonal point group 4/mmm, making them Dirac semimetals.","keywords":["cadmium arsenide","Cd3As2","convergent beam electron diffraction","point group determination","Dirac semimetal","inversion symmetry","molecular beam epitaxy","CBED symmetry analysis"],"falsifier":"Re-record the $[\\bar{1}10]$ whole-pattern CBED under controlled thickness and tilt and compute a quantitative mirror-symmetry score, for example the correlation between the pattern and its reflection across each candidate mirror. If either mirror score is not clearly above the noise level, or if a Bloch wave calculation for I41cd with the same absorption and thickness parameters as the experiment also shows 2mm whole-pattern symmetry, the paper's central claim is falsified.","tokens_in":7316,"feed_emoji":"🔬","tokens_out":7448,"duration_ms":68097,"temperature":0.7,"pith_summary":"This paper establishes that molecular-beam-epitaxy-grown cadmium arsenide thin films are centrosymmetric, and therefore Dirac semimetals, by determining their point group with convergent beam electron diffraction. The central result is that the whole-pattern CBED along $[\\bar{1}10]$ shows 2mm symmetry, which is compatible with the centrosymmetric 4/mmm point group but not with the non-centrosymmetric 4mm group proposed for the same material. This resolves a long-standing ambiguity in the room-temperature structure of Cd3As2, where the arrangement of Cd vacancies determines whether an inversion center exists. The result matters because inversion symmetry is what separates a Dirac semimetal from a Weyl semimetal, and it shows that CBED can resolve such subtle structural differences in thin films.","feed_headline":"Electron beams show Cd3As2 films keep their inversion center","feed_subtitle":"The 2mm whole-pattern symmetry settles the Dirac-versus-Weyl structure debate for MBE-grown films.","key_machinery":"The central object is whole-pattern CBED symmetry, specifically the presence of two perpendicular mirror planes in the $[\\bar{1}10]$ whole pattern including the higher-order Laue zone rings. Because dynamical electron diffraction breaks the usual centrosymmetric intensity relation, the whole-pattern symmetry directly reports the three-dimensional point group. The paper pairs this with the diffraction-group classification tables and with Bloch wave simulations of both candidate structures; the simulation shows that the non-centrosymmetric I41cd model gives only one mirror in the $[\\bar{1}10]$ whole pattern, while the centrosymmetric I41/acd model gives 2mm.","core_discovery":"The paper reports that MBE-grown Cd3As2 films are centrosymmetric with point group 4/mmm, the same point group as the bulk I41/acd structure. The decisive evidence is the $[\\bar{1}10]$ zone-axis CBED pattern, where both the bright-field disc and the whole pattern show 2mm symmetry. According to the diffraction-group tables, that combination is compatible with 4/mmm and incompatible with the non-centrosymmetric 4mm point group, for which the whole pattern would show only one mirror. Bloch wave simulations for the two reported structures reproduce this difference, and HAADF-STEM images show the systematically ordered Cd vacancies of the I41/acd structure. Along $[001]$, the pattern shows 4mm symmetry; combined with the other zone axes this narrows the point group to 4/mmm rather than m3m.","pith_inferences":["A quantitative check would be to apply a pixel-level mirror-symmetry statistic to the recorded $[\\bar{1}10]$ whole pattern; the paper reports no such metric and relies on visual inspection.","If the assignment is correct, angle-resolved photoemission and magnetotransport on these films should show one Dirac node pair and no chiral surface Fermi arcs; looking for the absence of Fermi arcs is a testable consequence the paper does not pursue.","The same diffraction-group strategy could settle inversion-center debates in other vacancy-ordered topological materials where x-ray diffraction is ambiguous.","A caveat implicit in the paper is that the intermediate-temperature P42/nmc phase also belongs to 4/mmm, so CBED alone cannot distinguish it from I41/acd; the HAADF-STEM vacancy pattern is what breaks that tie."],"forward_implications":["Because the films belong to 4/mmm, electronic-structure predictions based on the centrosymmetric bulk structure apply directly to MBE-grown films.","The film structure is not altered by residual strain or low growth temperature, so the symmetry-protected Dirac nodes should survive in the thin-film geometry.","Whole-pattern CBED symmetry, particularly the $[\\bar{1}10]$ 2mm pattern, discriminates between I41/acd and I41cd in cases where x-ray diffraction cannot.","The intermediate-temperature P42/nmc phase is excluded for these films; HAADF-STEM identifies the vacancy arrangement as that of I41/acd."],"supporting_citations":[{"why":"Supplies the centrosymmetric I41/acd structure that the films are matched to.","marker":"[15]"},{"why":"Proposed the non-centrosymmetric I41cd/4mm structure and the Weyl-semimetal scenario that the paper rules out.","marker":"[5]"},{"why":"One of the reported I41cd structures used as the competing model in the CBED simulations.","marker":"[12]"},{"why":"Provides the diffraction-group tables that convert observed bright-field and whole-pattern symmetries into possible point groups.","marker":"[19]"},{"why":"Supplies the CBED theory and microdiffraction analysis methods used to assign symmetries.","marker":"[21]"},{"why":"Implements the Bloch wave simulation code used to compute CBED patterns for both candidate structures.","marker":"[24]"},{"why":"Describes the MBE growth and film properties, including the (112) epitaxial orientation used to index the zone axes.","marker":"[22]"}],"fun_headline_variants":["CBED settles Dirac vs Weyl for Cd3As2 films","Electron diffraction pins Cd3As2 to 4/mmm point group","Cd3As2 thin films are centrosymmetric, CBED confirms","Inversion center confirmed in Cd3As2 films via CBED","Diffraction shows Cd3As2 keeps its inversion symmetry"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion depends on the visual judgment that the experimental $[\\bar{1}10]$ whole-pattern CBED has two perpendicular mirrors, and on Bloch wave simulations that reproduce the difference between the two structures; if beam tilt, specimen bending, thickness, or absorption effects changed the apparent symmetry, or if the simulation parameters favored one structure, the 4/mmm assignment would fail.","fun_headline_variants_meta":{"raw":{"variants":["CBED settles Dirac vs Weyl for Cd3As2 films","Electron diffraction pins Cd3As2 to 4/mmm point group","Cd3As2 thin films are centrosymmetric, CBED confirms","Inversion center confirmed in Cd3As2 films via CBED","Diffraction shows Cd3As2 keeps its inversion symmetry"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000409,"raw_usage":{"total_tokens":2107,"prompt_tokens":917,"completion_tokens":1190,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":533,"completion_tokens_details":{"reasoning_tokens":1109}},"tokens_in":533,"tokens_out":1190,"duration_ms":10046,"temperature":1.0,"reasoning_tokens":1109,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:04:45.734388+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-record the $[\\bar{1}10]$ whole-pattern CBED under controlled thickness and tilt and compute a quantitative mirror-symmetry score, for example the correlation between the pattern and its reflection across each candidate mirror. If either mirror score is not clearly above the noise level, or if a Bloch wave calculation for I41cd with the same absorption and thickness parameters as the experiment also shows 2mm whole-pattern symmetry, the paper's central claim is falsified.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the centrosymmetric I41/acd structure that the films are matched to."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proposed the non-centrosymmetric I41cd/4mm structure and the Weyl-semimetal scenario that the paper rules out."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"One of the reported I41cd structures used as the competing model in the CBED simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the diffraction-group tables that convert observed bright-field and whole-pattern symmetries into possible point groups."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the CBED theory and microdiffraction analysis methods used to assign symmetries."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Implements the Bloch wave simulation code used to compute CBED patterns for both candidate structures."},{"cited_title":"Schumann, M","cited_arxiv_id":null,"evidence_quote":"Describes the MBE growth and film properties, including the (112) epitaxial orientation used to index the zone axes."}],"review_version":1}