{"id":"747e8835-3085-4d12-931d-98477a8aab5a","arxiv_id":"1908.09132","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"The authors demonstrate that candidate EMCD signals can be extracted from zone-axis single-pass STEM measurements using an 8-blade patterned aperture, though the signals remain below the conventional detection threshold.","lead":"This paper tests a patterned aperture design for measuring magnetic circular dichroism in an electron microscope, reporting candidate signals that depend on sample tilt and momentum transfer. It also introduces a data processing pipeline for weak EELS difference signals, with data and code shared openly.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No null test means the heavily parameterized extraction pipeline could produce the reported 'candidate' EMCD signals from processing artifacts alone, so the central claim that candidates are extractable and orientation/qy-dependent is not yet supported.","rationale":"The reader's weakest assumption identifies exactly the load-bearing gap: the candidate signals are assumed to reflect EMCD rather than processing artifacts, and no null test rules out false positives. I agree. The paper is transparent about its limitations, openly shares data and code, and carefully avoids claiming detection above the Rose criterion; those are genuine strengths that justify a conditional rather than reject verdict. However, the extraction workflow is sufficiently flexible and the reported signals sufficiently weak that without a null control the central claim is not independently secured. The sign anomaly on Fe L3, the post-hoc selection of qy ranges, and the empirically thresholded orientation masks all increase the risk. The proposed null-control run is decisive: if the same pipeline produces comparable candidate SNR on an EMCD-free dataset, then the 'promising candidates' and their apparent dependencies are artifacts; if it does not, the paper's cautious interpretation is materially strengthened. Since the reader already reached CONDITIONAL and this concern is the same one, no verdict adjustment is needed.","tokens_in":16644,"tokens_out":4538,"duration_ms":56887,"concrete_test":"Download the Zenodo data and code and run the exact published fmincon pipeline on a null 4D-EELS datacube constructed from the measured Fe data by replacing each chiral-minus spectrum with the corresponding chiral-plus spectrum plus independent Poisson noise, preserving all ROIs, qy ranges, mask thresholding, and optimizer settings. If this null control yields L2/L3 SNR comparable to the reported ~2.2 dB candidate, or reproduces qy-dependent sign inversions like Fig. 9, then the pipeline manufactures candidates and the central claim is unsupported; if the null remains at or below the noise floor, the candidate signals gain real evidential weight.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is deliberately limited: it reports 'promising EMCD candidate signals' below the Rose criterion, not confirmed EMCD. What must be true for even that claim is that the extraction pipeline does not manufacture candidate signals in the absence of a magnetic signal. This condition is untested. The pipeline optimizes 17 free parameters with fmincon and applies non-standard corrections: an energy-dependent post-edge normalization (Eq. 8), a second-derivative profile-matching step (Eq. 7), a non-integer energy shift, and a two-pseudo-Voigt EMCD model with the L3 and L2 amplitudes constrained to opposite sign (Table 1). Each step can imprint spectral structure onto a difference spectrum. The authors themselves note that the Fe L3 sign inversion in Fig. 9 is not reciprocated on L2 and attribute it to misalignment and mixing of magnetic and non-magnetic signals, and they identify imperfect crystal/spectrometer alignment as among the strongest factors influencing extraction. Yet no null experiment is reported: no non-magnetic specimen, no label-swap, no synthetic noise-only dataset, and no permutation test for the orientation masks. Since the orientation masks are qualitatively thresholded and the outer qy range is selected post hoc, the claimed sensitivity to orientation and momentum transfer could arise from the same artifact without any EMCD. Thus the central claim rests on an untested assumption that the pipeline output is not a false positive.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports experimental progress toward single-pass STEM-EMCD on a bcc Fe [001] zone axis using an 8-blade patterned aperture with a mirror-symmetry plane. The authors describe a complete data acquisition workflow, including custom hardware, synced 4D EELS and 4D STEM-diffraction acquisition, and a signal extraction pipeline. The extraction pipeline uses a 17-parameter fmincon optimization that models pre-edge backgrounds, performs second-derivative profile matching between the chiral spectra (Eq. 7), applies a non-integer energy shift and an energy-dependent post-edge normalization slope/intercept (Eq. 8), and fits the difference signal as two pseudo-Voigt peaks with opposite-sign amplitudes (Table 1). The authors report candidate EMCD signals whose strongest Fe L2 SNR is 2.2 dB, explicitly below the Rose criterion of 5 (7 dB), and observe dependence of these candidates on orientation masks and on the qy integration range, including a sign inversion on Fe L3. Data and code are made available under open licenses.","tokens_in":17046,"tokens_out":3354,"duration_ms":37659,"significance":"If the candidate signals could be established as genuine EMCD, the paper would provide a useful methodological template for dose-efficient, single-pass zone-axis EMCD with patterned apertures, including a detailed treatment of data synchronization, spectral alignment, and artifact correction. The authors are commendably transparent about the limitations: they frame the results as 'candidate' signals, state that none meet the Rose criterion, and discuss possible artifact sources such as crystal/spectrometer misalignment. The open release of data and code is a strength. However, the central empirical claim that 'promising EMCD candidate signals can be extracted' is not yet supported because the pipeline that produces these candidates has not been shown not to manufacture apparent signals from processing artifacts alone. The significance of the paper therefore rests on the outcome of a null test that is currently absent.","major_comments":[{"comment":"The central claim that candidate EMCD signals are extracted is not yet supported by a null test. The extraction pipeline optimizes 17 free parameters—including pre-edge amplitudes and slopes, the profile-matching scalar ks in Eq. (7), the chiral-minus energy shift, the post-edge normalization slope and intercept, and the EMCD peak amplitudes a1 and a2 whose signs are constrained to be opposite (Table 1)—and applies these steps to the same dataset from which the signal is extracted. With no non-magnetic specimen, no label-swapped chiral pair, and no synthetic noise-only dataset, one cannot exclude that the reported difference features arise from overfitting or from the parameterization itself. Since the authors themselves attribute the Fe L3 sign inversion to misalignment and mixing of magnetic and non-magnetic signals, a null experiment or a permutation test is essential to establish that the pipeline does not produce false positives in the absence of a magnetic signal. Please add such a control and report whether any 'candidate' survives it.","section":"Methods: EMCD signal extraction; Results: Candidate EMCD spectra (Figs. 5–8)"},{"comment":"The reported sensitivity to sample orientation and to momentum transfer is based on selections that are post hoc and qualitative. The qy range 17.6–24.2 mrad is motivated by the observation that it yields a 'more convincing' Fe L2 signal, and the orientation masks 'Orient 01' and 'Orient 02' are generated by an 'empirically-determined' qualitative VDF threshold. Unless the threshold choices and qy window are shown to be stable under variation or were fixed before inspecting the extracted signals, the claimed dependence of the candidate signal on orientation and qy could be a product of the selection procedure rather than of the underlying physics. Please quantify the sensitivity of the results to the mask threshold and qy range, or use a cross-validation or pre-registration scheme.","section":"Results: Influence of qy; Fig. 9; Fig. 4c,d"},{"comment":"The reported SNR values (e.g., 'SNR L2 = 2.22 dB' in Fig. 5) are not defined operationally. It is unclear whether the noise is estimated from the residual between the data and the pseudo-Voigt fit, from adjacent energy channels, from frame-to-frame variation, or from some other source. Because the SNR is the basis for the 'candidate' versus 'confirmed' distinction and for the comparison to the Rose criterion, the noise model must be specified explicitly. Without this, the SNR values cannot be independently interpreted or reproduced.","section":"Methods: EMCD signal extraction; Figs. 5–8"}],"minor_comments":[{"comment":"The text states that the post-edge normalization line is 'fit to the post-edge ratios' with slope m and intercept d, but then lists m and d among the 17 parameters passed to fmincon. Please clarify whether the linear regression is performed inside the objective function at each iteration or whether m and d are free parameters; the current wording is contradictory.","section":"Methods: EMCD signal extraction, paragraph following Eq. (8)"},{"comment":"The caption sentence 'Note that the EMCD amplitudes were not constrained to be positive and negative as above; rather, they were constrained to have opposite sign from each other' is confusing because the table already shows a1 ≥ 0 and a2 ≤ 0. Consider simplifying the caption to state that a1 and a2 are constrained to opposite signs.","section":"Table 1 caption"},{"comment":"The quantity labeled 'mL/mS' in the cumulative-sum panels is not defined in the text or captions. Define mL and mS.","section":"Figs. 5–8"},{"comment":"The oxygen K edge at 532 eV and the Fe L2,3 edges at 709/723 eV are correctly identified, but the text says the oxygen signal is 'primarily dominant in the background regions'; the caption would be clearer if it noted that the oxygen edge appears because the summation includes frames without Fe.","section":"Results: Fig. 2b"},{"comment":"In Eq. (4), the ratio bk,2/bk,1 equals ΔEk only if the Taylor expansion coefficients are defined consistently with the amplitude Ak; this is correct, but the step would benefit from a one-sentence note that the mean spectrum is normalized to avoid a scaling ambiguity.","section":"Methods: 4D EELS pretreatment"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a transparent methods paper with open data/code, and its claims are deliberately modest ('candidate' signals below the Rose criterion). The missing null test is the load-bearing issue: it can be fixed within the scope of a revision by adding a non-magnetic control, a label-swap test, or synthetic-noise injections. I therefore recommend major revision rather than rejection, provided the authors can supply a null control or otherwise convincingly rule out pipeline-induced false positives."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is an honest, carefully written methods paper, not an overclaiming detection paper. The authors implement the patterned-aperture idea from Negi et al. in a zone-axis STEM geometry, build a custom aperture, acquire 4D EELS and 4D CBED data, and release data and code on Zenodo. They explicitly report candidate signals below the Rose criterion and describe a long list of experimental and instrumental limitations. That transparency is real, and the paper is useful for anyone trying to reproduce or extend single-pass STEM-EMCD.\n\nThe genuinely new part is the combination of a single-pass patterned-aperture acquisition with a fairly elaborate data treatment pipeline: energy-dependent post-edge normalization, second-derivative profile matching, and a 17-parameter optimization that fits a two-pseudo-Voigt EMCD model with opposite-sign L3/L2 amplitudes. The paper also maps how the candidate signal varies with orientation mask and qy, which is the right kind of experiment to run. The data sharing is a genuine strength.\n\nWhere it gets soft: the central claim is not that EMCD has been detected, but that promising candidate signals can be extracted and that these depend on orientation and momentum transfer. The stress-test concern lands on that claim. The pipeline has many free parameters and several nonstandard corrections, and there is no null test. No non-magnetic sample, no label swap, no noise-only synthetic dataset, no permutation test on the orientation masks. The qy range is selected post hoc, and the orientation masks are thresholded qualitatively. Under those conditions, an apparent orientation/qy dependence could be manufactured by the pipeline or by residual spectral aberrations rather than by real EMCD. The authors themselves note the Fe L3 sign inversion is not reciprocated on L2 and blame misalignment and mixing of magnetic and non-magnetic signals, which reinforces that artifacts are plausible.\n\nThat said, the paper does not pretend to have confirmed detection. It defines candidate signals as below Rose criterion and says so repeatedly. So the missing null test is not fatal to the paper's value as a methods-development report, but it is a real gap in support for the specific claim of extractability and sensitivity. Adding even one null experiment, or a synthetic-data injection at comparable noise levels, would substantially strengthen the paper.\n\nThe citation pattern looks fair; prior work by Negi et al. and double-aperture experiments are credited. The math and code are presented in enough detail to reproduce, which is more than many papers in this area offer.\n\nBottom line: this deserves serious peer review, not desk rejection. The right referee report would ask for a null test and a less post hoc choice of qy range, but the paper is a solid, readable contribution to a difficult experimental problem. I would bring it to a reading group focused on EELS or EMCD methods, and I would cite it if working in that area.","headline":"A transparent, well-documented methods paper whose EMCD candidate signals are plausible but under-validated: the heavily parameterized extraction pipeline is never tested against a non-magnetic or synthetic null, so the central claim rests on an untested assumption.","tokens_in":17570,"tokens_out":1532,"would_cite":true,"duration_ms":18931,"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":"A mirror-symmetric patterned aperture in a scanning transmission electron microscope can extract candidate electron magnetic circular dichroism signals in a single pass on the [001] zone axis of bcc iron, with strength and sign that…","keywords":["electron magnetic circular dichroism","EMCD","patterned aperture","scanning transmission electron microscopy","EELS","zone axis","bcc iron","magnetic moment mapping"],"falsifier":"Apply the exact extraction routine to 4D EELS data from a non-magnetic sample, or to synthetic spectra containing only Poisson noise plus the same background, and check whether signed Fe $L_3$/$L_2$ difference features at SNR near 2 dB still emerge; if they do, the candidate signal is a processing artifact. A separate check is to intentionally rotate the aperture or the diffraction pattern by a known small angle and verify that the predicted $q_y$ sign inversion of the Fe $L_3$ difference appears.","tokens_in":16454,"feed_emoji":"🧲","tokens_out":9862,"duration_ms":96267,"temperature":0.7,"pith_summary":"This paper is a progress report toward measuring magnetic moments with atomic-column resolution in the electron microscope. It tries to establish that an eight-blade, mirror-symmetric patterned aperture, mounted in front of the EELS spectrometer, can extract electron magnetic circular dichroism (EMCD) candidate signals from a ferromagnetic sample in a single STEM pass on a zone-axis orientation. On a 10-nm bcc iron film at low spatial resolution, the authors recover difference spectra on the Fe L3 and L2 edges; the strongest L2 candidate has a signal-to-noise ratio of 2.2 dB, below the conventional Rose-criterion threshold of 5, so the paper labels them candidates rather than confirmed detections. The candidate signal depends on crystal orientation relative to the zone axis and on the momentum-transfer window selected by the aperture, which supports the paper's two design hypotheses. If validated further, the approach would remove the need for repeated scans with sub-atomic registration and would scale toward atomic-resolution magnetic mapping.","feed_headline":"Single-pass zone-axis EMCD yields candidate magnetic signals","feed_subtitle":"The candidate Fe L-edge dichroism tracks crystal orientation and momentum transfer, a step toward atomic-scale magnetic mapping.","key_machinery":"The load-bearing element is the patterned aperture itself: an eight-blade 'mirrored ventilator' design with a mirror symmetry plane, installed so that the mirror plane is parallel to the spectrometer's energy dispersion axis. It divides the diffraction plane into chiral plus and chiral minus collection regions, integrates over the $q_x$ dimension, and leaves the non-dispersive momentum transfer $q_y$ resolved in a two-dimensional $q$--$E$ spectrum. The argument is carried by combining this aperture with a 4D workflow: probe-position time stamps are assigned to every EELS frame, missing frames are interpolated, spectra are aligned along energy with a Taylor-coefficient shift correction, peak profiles are matched by subtracting a scaled second derivative, and a 17-parameter least-squares optimization extracts the difference signal as two pseudo-Voigt peaks with opposite signs.","core_discovery":"On its own terms, the central discovery is that single-pass STEM-EMCD on a zone axis is experimentally accessible with a patterned aperture, and that the recovered difference signal is structured rather than arbitrary: its strength tracks the local crystal orientation, and its Fe $L_3$ sign can flip with the non-dispersive momentum transfer $q_y$. The most favorable extraction, using pixels closest to the Fe [001] zone axis and the outer $q_y$ window from 17.6 to 24.2 mrad, yields a candidate $L_2$ signal with SNR 2.2 dB together with the expected signs on $L_3$ and $L_2$. When pixels with larger zone-axis mistilt are included, the $L_2$ signal weakens while a feature remains on $L_3$; the authors interpret this as mixing of magnetic and non-magnetic contributions due to imperfect symmetry. They also observe a sign inversion of the $L_3$ difference at intermediate $q_y$ values, which they attribute to a slight rotation of the diffraction pattern relative to the aperture's mirror plane and to small aperture misalignments.","pith_inferences":["A decisive next test would be a false-positive audit: run the same 17-parameter extraction on a non-magnetic specimen or on synthetic noise with matched count statistics and check whether signed $L_3$/$L_2$ difference features at SNR near 2 dB still appear.","The observed $q_y$ sign inversion, if reproducible, could be turned into a calibration tool: a deliberate small rotation of the aperture or diffraction pattern should shift the inversion position in a predictable way, allowing non-magnetic mixing to be estimated and subtracted.","If the orientation sensitivity persists at atomic scale, the usable field of view will be limited by beam-tilt-induced orientation drift, so descan or smaller scan areas may be prerequisites for atomic-column EMCD mapping.","The single-pass design implies all information for quantification is in one pass; comparing per-pixel candidate signals against a bulk or micromagnetic reference would test whether the extracted moment ratio is quantitatively meaningful."],"forward_implications":["A single STEM pass can record both the chiral spectra and the local diffraction pattern from the same region, so repeated scans with sub-atomic registration are no longer required for this class of measurement.","Because the patterned aperture accepts a larger fraction of inelastically scattered electrons than point-like or slit geometries, it should improve the dose efficiency of EMCD, directly attacking the signal-to-noise bottleneck.","Quantitative zone-axis EMCD will demand near-perfect alignment of the crystal symmetry directions with the spectrometer dispersion axis and minimal beam tilt during scanning, since orientation drift measurably weakens the extracted signal.","Selecting the momentum-transfer window in the non-dispersive direction changes the strength and even the sign of the candidate signal; outer $q_y$ windows with fewer Bragg contributions gave the strongest $L_2$ candidates.","The processing chain developed here is transferable to other EMCD geometries and provides a template for the atomic-resolution version of the experiment."],"supporting_citations":[{"why":"Proposes the patterned-aperture geometry and its zone-axis simulations; the eight-blade design tested here comes from this proposal.","marker":"[20]"},{"why":"Supplies the SNR and Rose-criterion methodology used to classify the recovered signals as candidates rather than confirmed detections.","marker":"[13]"},{"why":"First experimental demonstration of EMCD in the transmission electron microscope; defines the effect this experiment adapts to single-pass STEM.","marker":"[6]"},{"why":"Shows EMCD detection under zone-axis diffraction geometry, providing the direct comparison for the zone-axis signal reported here.","marker":"[15]"},{"why":"Explains how asymmetry in the diffraction geometry mixes magnetic and non-magnetic signals; used to interpret the loss of Fe L2 on mistilt and the L3 sign inversion.","marker":"[27]"},{"why":"Quantifies the effect of dynamical diffraction asymmetry on EMCD intensity, supporting the orientation-dependence interpretation.","marker":"[28]"}],"fun_headline_variants":["Single-pass zone-axis EMCD yields candidate magnetic signal","Patterned aperture enables single-pass zone-axis EMCD","Zone-axis EMCD: candidate signal sensitive to orientation and q","Single-pass EMCD on zone axis: progress in signal extraction","Candidate magnetic dichroism from single-pass zone-axis EMCD"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the difference spectra produced by the processing chain are genuine EMCD and not artifacts of that chain; the manuscript reports no null test on a non-magnetic sample or on synthetic noise to support this assumption.","fun_headline_variants_meta":{"raw":{"variants":["Single-pass zone-axis EMCD yields candidate magnetic signal","Patterned aperture enables single-pass zone-axis EMCD","Zone-axis EMCD: candidate signal sensitive to orientation and q","Single-pass EMCD on zone axis: progress in signal extraction","Candidate magnetic dichroism from single-pass zone-axis EMCD"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001046,"raw_usage":{"total_tokens":4407,"prompt_tokens":967,"completion_tokens":3440,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":583,"completion_tokens_details":{"reasoning_tokens":3358}},"tokens_in":583,"tokens_out":3440,"duration_ms":25466,"temperature":1.0,"reasoning_tokens":3358,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:20:26.524656+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Apply the exact extraction routine to 4D EELS data from a non-magnetic sample, or to synthetic spectra containing only Poisson noise plus the same background, and check whether signed Fe $L_3$/$L_2$ difference features at SNR near 2 dB still emerge; if they do, the candidate signal is a processing artifact. A separate check is to intentionally rotate the aperture or the diffraction pattern by a known small angle and verify that the predicted $q_y$ sign inversion of the Fe $L_3$ difference appears.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proposes the patterned-aperture geometry and its zone-axis simulations; the eight-blade design tested here comes from this proposal."},{"cited_title":"& Leifer, K","cited_arxiv_id":null,"evidence_quote":"Supplies the SNR and Rose-criterion methodology used to classify the recovered signals as candidates rather than confirmed detections."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"First experimental demonstration of EMCD in the transmission electron microscope; defines the effect this experiment adapts to single-pass STEM."},{"cited_title":"& Zhu, J","cited_arxiv_id":null,"evidence_quote":"Shows EMCD detection under zone-axis diffraction geometry, providing the direct comparison for the zone-axis signal reported here."},{"cited_title":"& Zhu, J","cited_arxiv_id":null,"evidence_quote":"Quantifies the effect of dynamical diffraction asymmetry on EMCD intensity, supporting the orientation-dependence interpretation."}],"review_version":1}