{"id":"1a6d8855-b67b-4fa2-8344-d6062ee05c2b","arxiv_id":"2608.10842","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Cathodoluminescence from different lateral regions of an extended electron beam adds incoherently, while a single electron can coherently excite two nearby pillars; a quantum eraser geometry is proposed to restore interference.","lead":"Electron beams can generate light in materials, but the wave-like coherence of the electron does not simply pass to the emitted light. New experiments show this for the first time and propose a way to recover the lost coherence by erasing which-path information.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The TR experiment's discrimination between coherent and incoherent excitation depends entirely on the companion-paper value σ=150 nm for the defocused beam's lateral coherence; without an independent in-situ coherence measurement, Fig. 1 cannot exclude an incoherently broadened source.","rationale":"The theoretical result of Eq. (2) is correct and independently checkable, so the concern is not with the derivation. The two-pillar experiments provide a separate positive demonstration of coherent excitation, but the first half of the central claim (no lateral-coherence interference) rests on Fig. 1. The most vulnerable element is the value σ=150 nm used to construct the alternative coherent prediction. This value comes from the authors' own companion paper [30], is not re-measured in the same session and geometry, and no uncertainty is propagated into Fig. 1(d). The discarded low-k region and lack of error bars increase the sensitivity to this parameter. A concrete in-situ coherence measurement would settle the issue. I agree with the reader's weakest assumption, and the conditional verdict stands: the paper should be accepted only if the coherence value is independently verified and the quantitative comparison is strengthened with uncertainty analysis.","tokens_in":11317,"tokens_out":13913,"duration_ms":121454,"concrete_test":"Perform an in-situ measurement of the transverse coherence function of the SEM e-beam under the exact defocused condition used for Fig. 1 (e.g., electron diffraction from a calibrated grating or the method of ref. [30] at the sample plane with the 6-µm spot). Using the measured coherence function, compute the predicted angle-resolved TR intensity for a partially coherent e-beam via the cross-spectral density (Gaussian-Schell model) over the full 6-µm spot. If the corrected coherent prediction still differs from the data by more than the experimental noise, the claim stands; if the corrected coherent curve overlaps the data, Fig. 1 does not support the incoherent-sum conclusion.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The no-postselection derivation in SM Eq. (2) is mathematically sound: summing over final transverse electron momenta converts the coherent amplitude into an incoherent position integral via Plancherel, so no lateral interference should appear in angle-resolved CL without electron postselection. The load-bearing step is therefore experimental. The 'coherent sum' curve that the data are compared against is computed from Eqs. (7)-(8) using a single Gaussian source with σ=150 nm, taken as 5% of the 6-µm beam width from the authors' companion paper [30]. If the true transverse coherence length of the defocused e-beam at the sample is substantially smaller than 150 nm, the coherent and incoherent angular profiles become nearly indistinguishable, and the measured similarity of focused and defocused TR angular distributions provides no evidence for the claim. The paper gives no error bars, no statistical test, and discards the low-k region via an impurity explanation, so the visual agreement cannot quantitatively exclude this alternative. Moreover, the coherent prediction itself is not directly observable in a no-postselection measurement; a correct partially coherent treatment of a 6-µm beam with a 150-nm coherence length (e.g., Gaussian-Schell model) would be a more appropriate benchmark, and this is not provided. Thus the central experimental demonstration of the incoherent-sum claim rests on an unverified parameter rather than on an internal test.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper addresses whether lateral coherence of an electron beam imprints itself on cathodoluminescence (CL) angular patterns when no post-selection of the transmitted electron is performed. The theoretical claim, derived in the end matter as Eq. (2), is that the total CL excitation probability is an incoherent sum over lateral electron positions weighted by the transverse electron density, because summing over final transverse electron momenta invokes the Plancherel theorem. Experimental support is presented in two geometries: (i) transition radiation from a Si3N4 film excited by focused versus defocused beams, where the similar angular distributions are interpreted as an incoherent sum over the beam spot; and (ii) two Au nanopillars excited by a single narrow beam, where a time-of-flight component in the Fourier-transformed interferogram is interpreted as evidence for coherent excitation of both pillars. A third configuration shows that a 500-nm beam overlapping two pillars 400 nm apart produces no interference, attributed to the limited lateral coherence of the beam. A quantum-erasure scheme using a bilayer graphene grating is proposed to recover interference via electron-photon correlation.","tokens_in":11550,"tokens_out":6056,"duration_ms":75204,"significance":"If the central claim holds, it provides a clear and counterintuitive criterion: without electron post-selection, angle-resolved CL from an extended beam is blind to lateral electron coherence, whereas a single electron coupling within its evanescent field to multiple scatterers produces coherent far-field interference. The Plancherel derivation in the end matter is elegant, parameter-free, and machine-checkable in its mathematical steps. The two-pillar time-of-flight signature in Fig. 2(g) is a genuinely clean observable, and the proposal for a which-path eraser experiment is a useful forward-looking idea. The main weakness is that the experimental demonstration of the incoherent-sum claim in Fig. 1 is qualitative and depends on a single coherence parameter taken from a companion paper, rather than on an internal, error-quantified test.","major_comments":[{"comment":"The central experimental evidence for the incoherent-sum claim is a qualitative comparison. The text states that the measured focused and defocused TR curves \"agree well with the incoherent sum,\" but no incoherent-sum curve is actually plotted, no error bars are shown, and no statistical test (e.g., residual analysis or chi-square) is reported. The low-k region is excluded by attributing it to an impurity, but the impurity contribution is not modeled or subtracted. Because this figure carries the main burden of the paper's first experimental conclusion, the manuscript needs a quantitative comparison between data and both the incoherent-sum and coherent-sum predictions, including uncertainty estimates and a stated criterion for distinguishing the two scenarios.","section":"Fig. 1(d) and Eqs. (7)-(8)"},{"comment":"The discriminating power of the TR experiment depends almost entirely on the assumed lateral coherence length of the defocused beam, modeled as a Gaussian with sigma = 150 nm (5% of the 6-micron beam width) and taken from the companion paper ref. [30]. If the true coherence length at the sample were substantially smaller than 150 nm, the coherent-sum angular profile would approach the point-dipole/incoherent profile, making the focused and defocused measurements nearly indistinguishable. The manuscript provides no in-situ measurement of the coherence at the sample position and no demonstration that the angular data themselves bound sigma. Without this, the observed similarity of focused and defocused TR profiles cannot exclude an incoherently broadened source.","section":"Fig. 1(d) and ref. [30]"},{"comment":"Equation (8) as printed, Gamma_CL proportional to |f|^2 exp[(k_parallel^0 / sigma)^2], is dimensionally inconsistent and predicts a growth of intensity with emission angle, which is opposite to the coherent narrowing described in the text. The correct Gaussian Fourier transform from Eq. (7) should produce an exponent of order -(sigma k_parallel)^2 (up to a factor of two). The authors must correct this equation and confirm that the calculated \"coherent sum\" curve in Fig. 1(d) was generated with the corrected expression.","section":"End matter, Eq. (8)"},{"comment":"The interpretation of the t = 25 fs band as the electron time of flight between pillars is a load-bearing step for the coherent-excitation claim, but the manuscript does not quantify the uncertainty of the extracted time delay or compare the predicted dispersion slopes from SM Eqs. (9)-(10) with the measured band shapes. I recommend overlaying the predicted bands on the Fourier-transformed data and reporting the uncertainty in the height difference; this would make the two-pillar coherence evidence quantitative rather than visual.","section":"Fig. 2(g) and SM Eqs. (9)-(10)"}],"minor_comments":[{"comment":"The description \"integrated over an angular range of 0.05 rad along the kx/k0 = 0 direction\" is ambiguous; please specify whether this is a slice of width 0.05 in kx/k0, and clarify how the plotted curves are normalized.","section":"Fig. 1(d)"},{"comment":"The notation r is used both for the far-field position vector and its magnitude; please use a unit vector or explicit notation, e.g., k = k0 rhat, to avoid confusion in the phase factor.","section":"End matter, Eq. (7)"},{"comment":"The assumption that the lateral coherence length scales as a fixed 5% of the beam diameter is stated without justification; if this is an empirical finding from ref. [30], it should be stated explicitly and its applicability to the different beam currents and defocus distances in the present experiments should be discussed.","section":"Figs. 1 and 3"},{"comment":"The main text reports a beam current of 1.6 nA for the AR CL measurements, while the methods section states 0.8 nA for the Fig. 3 configuration; please reconcile these values.","section":"Experimental setup and methods"},{"comment":"The phrase \"conclusive experimental evidence\" is stronger than what the current quantitative analysis supports; I suggest tempering this to \"evidence consistent with\" until the error analysis and in-situ coherence validation requested above are provided.","section":"Abstract and main text"}],"recommendation":"major_revision","confidential_remarks":"The key experimental claim relies on a coherence parameter from companion paper ref. [30] that is not yet published as far as the manuscript indicates. If that paper is not available or its measurement does not apply at the sample position in the present SEM, the TR conclusion is substantially weakened. I recommend the editor ask for an in-situ measurement of the transverse coherence or an error-bounded analysis that shows the TR data themselves constrain sigma."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know: this is a solid experimental paper with a correct theoretical core, but the headline claim is a notch stronger than the evidence. The central physics—that without post-selecting the electron, CL excitation is an incoherent sum over lateral positions—is mathematically airtight via the Plancherel reduction in the SM, and the two-pillar experiment convincingly shows coherent excitation of two separated scatterers by a single electron. That part is genuinely new and should be cited.\n\nThe soft spots are concentrated in the TR comparison. In Fig. 1, the discrimination between coherent and incoherent excitation rests entirely on the claim that the defocused 6-µm beam still has a 150-nm coherence length, taken from the authors' companion paper. If that value is wrong, the two curves become nearly indistinguishable, and the measured similarity of focused and defocused patterns proves nothing. The paper gives no error bars, no statistical test, and discards the low-k region via an impurity explanation. That is a fair and serious concern. It does not sink the paper, because the two-pillar results stand on their own, but it means the TR demonstration is not as \"conclusive\" as the abstract says.\n\nI also want to flag that the central theoretical result is not new—it is drawn from refs. 19 and 32—and the paper is honest about that. What is new is the experimental test of it, plus the which-path eraser proposal. The eraser is a reasonable idea but explicitly untested, so it should be framed as a suggestion, not a result.\n\nWho is this for? Anyone working on electron-beam spectroscopy, quantum optics with free electrons, or cathodoluminescence. The experimental technique of AR-CL interferometry with sub-nanometer electron beams is well executed, and the two-pillar coherent-excitation result is likely correct. The paper deserves a serious referee. I would send it to review, but I would ask for a quantitative comparison of the TR angular profiles (e.g., chi-squared or residual analysis), an explicit discussion of the sensitivity to the assumed coherence parameter, and ideally an in-situ measurement of that parameter. With those additions it would be a solid publication; without them, the TR section remains too hand-wavy.","headline":"A correct and useful experimental paper, but the 'conclusive' wording outruns the TR data, which hinge on an externally measured coherence parameter.","tokens_in":12104,"tokens_out":1276,"would_cite":true,"duration_ms":871733,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper shows experimentally that the angular cathodoluminescence pattern from an electron beam is an incoherent sum over the beam's lateral positions when no electron post-selection is performed, while a single electron passing…","keywords":["cathodoluminescence","electron beam coherence","plasmon interference","transition radiation","which-path information","quantum eraser","angle-resolved CL","evanescent field excitation"],"falsifier":"In the same SEM, measure the 6-micrometer defocused beam's lateral coherence directly by electron diffraction while simultaneously recording the angular TR pattern; if the coherence length turns out to be much smaller than the assumed 150 nm, the observed insensitivity of the TR pattern to spot size would no longer discriminate between coherent and incoherent excitation, and the paper's first conclusion would lose its experimental support.","tokens_in":11067,"feed_emoji":"🔬","tokens_out":7369,"duration_ms":71646,"temperature":0.7,"pith_summary":"The paper aims to settle when the wave nature of an electron beam shows up in cathodoluminescence (CL). Its central claim is that without post-selecting the electron, the total CL excitation probability is an incoherent sum over the electron's transverse positions, weighted by the transverse electron density; different lateral parts of an extended beam do not interfere. By contrast, one electron whose evanescent field reaches two separated scatterers excites them coherently, producing interference in the far field. The authors test this with angle-resolved CL from a thin silicon nitride film (focused versus defocused beam) and with two gold nanopillars, and they propose a post-selection geometry to recover coherence. If right, it means the electron's lateral coherence does not directly transfer to the emitted light unless the electron's which-path information is erased.","feed_headline":"Single electrons keep light coherent; broad beams don't","feed_subtitle":"Cathodoluminescence from a defocused electron spot adds up incoherently; one electron couples two pillars into interference.","key_machinery":"The central object is the nonrecoil, no-post-selection excitation probability, Eq. (2): $P \\propto \\int d^2R \\, |\\langle n|\\hat{H}(R)|0\\rangle|^2 \\, |\\psi_\\perp^i(R)|^2$. Because the final transverse electron states are integrated out, cross terms between different lateral positions cancel under the Plancherel theorem, leaving an incoherent sum weighted by the transverse electron density. The complementary mechanism is the evanescent-field coupling of a single electron to two scatterers, which generates a phase difference set by the electron's time of flight between the pillars; this phase appears in the Fourier-transformed CL interferogram and is the signature of coherent excitation. For transition radiation, the coherently broadened source would shift the angular pattern toward the surface normal, a shift the experiments do not show. Together these pieces define the condition for observable CL interference: either the electron's evanescent field reaches the scatterers, or the electron's which-path information is erased.","core_discovery":"The paper establishes that, in a cathodoluminescence measurement without electron post-selection, the total excitation probability factors into an incoherent sum over the electron's lateral positions, weighted by the transverse electron density. This is a consequence of the Plancherel theorem applied to the nonrecoil interaction Hamiltonian. Hence an extended, even laterally coherent electron beam produces no CL interference between its different parts: the angular pattern from a 6-micrometer defocused beam on a thin silicon nitride film matches the incoherent sum of point-like transition-radiation sources, not the coherent superposition. The same electron, however, when passing within the evanescent field of two gold nanopillars, excites both coherently, as shown by far-field CL interferograms whose Fourier transform contains a band at the electron time of flight between the pillars. When the pillars lie outside the evanescent field and the beam is spread over both, the signal is again an incoherent sum. The paper argues that the difference is which-path information carried by the electron, and proposes a post-selection geometry (electron diffraction from a twisted bilayer of graphene) to erase it and recover CL interference.","pith_inferences":["The same Plancherel-theorem argument should apply to electron energy-loss spectroscopy: without post-selection, EELS spectra from an extended beam are an incoherent sum over lateral positions, so lateral coherence should not produce interference in EELS either.","A continuous sweep of the beam defocus in the thin-film transition-radiation experiment would provide a sharper test: the incoherent-sum prediction keeps the angular pattern fixed, while the coherent-sum prediction would progressively shift the pattern with spot size.","The proposed graphene-bilayer quantum eraser could be implemented with existing SEM and single-photon correlation equipment; the twist angle of the bilayer tunes the overlap of the diffraction disks and hence the strength of the which-path erasure.","A more general implication is that any attempt to image with CL using lateral electron coherence must either place multiple structures inside one electron's evanescent field or post-select the electron; lateral coherence alone never appears in non-post-selected far-field CL."],"forward_implications":["For standard, non-post-selected CL, defocusing the electron beam does not change the angular emission pattern's shape; the pattern is the point-source pattern incoherently summed over the spot.","One electron passing between two pillars separated by less than the evanescent-field range yields a CL interferogram with a band at the electron time of flight, directly evidencing coherent excitation of both.","When two pillars are separated beyond the evanescent field and the beam is spread over both, the CL is an incoherent sum even if the beam is laterally coherent, because the electron retains which-path information.","Coherent CL from different lateral parts of an extended beam can be recovered by post-selecting electrons in the overlapping diffraction region of a diffractive element such as a twisted bilayer of graphene, correlating those electrons with the emitted photons."],"supporting_citations":[{"why":"Supplies the quantum formalism for electron-beam optical excitation without electron post-selection, from which Eqs. (1) and (2) are derived.","marker":"[19]"},{"why":"Reports the which-path erasure experiment in electron-photon entanglement that motivates the proposed post-selection geometry.","marker":"[21]"},{"why":"Earlier observation that Smith-Purcell radiation is insensitive to transverse beam broadening, the contrast case for the present coherence claim.","marker":"[28]"},{"why":"Companion measurement of the SEM's lateral coherence, giving the Gaussian standard deviation of 150 nm used to model the defocused beam.","marker":"[30]"},{"why":"Previous angle-resolved CL interferometry of plasmonic and dielectric scatterers that provides the analytical model and the coherent-excitation interpretation along the electron trajectory.","marker":"[31]"},{"why":"Provides the transition-radiation field expressions and the nonrecoil background used to compute the TR angular distributions.","marker":"[32]"}],"fun_headline_variants":["Broad electron beams: no interference; single electron: yes","Electron beam focus determines light coherence","Coherence dies with beam spreading, lives for point electrons","Spreading an electron beam erases its light coherence","Interference from one electron, not from a broad beam"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The transition-radiation conclusion hinges on the companion measurement that the defocused 6-micrometer beam still has a lateral coherence length of about 150 nm; if the actual coherence were much smaller, the focused and defocused TR patterns would be indistinguishable for both coherent and incoherent models, and the experiment would not discriminate them.","fun_headline_variants_meta":{"raw":{"variants":["Broad electron beams: no interference; single electron: yes","Electron beam focus determines light coherence","Coherence dies with beam spreading, lives for point electrons","Spreading an electron beam erases its light coherence","Interference from one electron, not from a broad beam"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000216,"raw_usage":{"total_tokens":1430,"prompt_tokens":941,"completion_tokens":489,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":557,"completion_tokens_details":{"reasoning_tokens":413}},"tokens_in":557,"tokens_out":489,"duration_ms":6182,"temperature":1.0,"reasoning_tokens":413,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:59:26.363275+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In the same SEM, measure the 6-micrometer defocused beam's lateral coherence directly by electron diffraction while simultaneously recording the angular TR pattern; if the coherence length turns out to be much smaller than the assumed 150 nm, the observed insensitivity of the TR pattern to spot size would no longer discriminate between coherent and incoherent excitation, and the paper's first conclusion would lose its experimental support.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the quantum formalism for electron-beam optical excitation without electron post-selection, from which Eqs. (1) and (2) are derived."},{"cited_title":"Ruimy, O","cited_arxiv_id":null,"evidence_quote":"Earlier observation that Smith-Purcell radiation is insensitive to transverse beam broadening, the contrast case for the present coherence claim."},{"cited_title":"Remez, A","cited_arxiv_id":null,"evidence_quote":"Companion measurement of the SEM's lateral coherence, giving the Gaussian standard deviation of 150 nm used to model the defocused beam."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Previous angle-resolved CL interferometry of plasmonic and dielectric scatterers that provides the analytical model and the coherent-excitation interpretation along the electron trajectory."},{"cited_title":"Determining Electron Beam Lateral Coherence in a Scanning Electron Microscope Using Electron Diffraction","cited_arxiv_id":"2606.28056","evidence_quote":"Provides the transition-radiation field expressions and the nonrecoil background used to compute the TR angular distributions."}],"review_version":1}