{"id":"8ea1c854-e963-4f5a-a29a-ff421f55db00","arxiv_id":"2504.15939","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Phase-resolved sum-frequency microscopy with phonon enhancement images monolayer hBN and maps its absolute crystal orientation, showing that CVD-grown triangular islands have nitrogen-terminated zigzag edges.","lead":"This paper demonstrates a new microscopy technique that makes invisible single layers of hexagonal boron nitride visible by using laser pulses tuned to the material's atomic vibrations, and reads out the crystal direction from the emitted light. The approach images large areas in under a second, which could speed up characterization of 2D materials and devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The N-termination claim rests on a DFT sign that the paper never shows; until that sign is checked, the central edge-polarity conclusion is unsupported.","rationale":"The reader's weakest_assumption is exactly the same load-bearing point: the absolute polarity assignment rests on the DFT sign that is deferred to the SI. My own review of the full text confirms the main text contains no derivation or independent calibration of that sign, and the sentence in the main text is the only bridge between the measured phase and the N-termination conclusion. The experimental measurement (SFG amplitude sign as a function of azimuthal angle) is clearly presented, and the 180-degree distinction is demonstrated internally, so the experimental technique itself is not in question. The concern is not that DFT is 'outside consensus'; rather, it is that the specific sign convention is a single point of failure with no redundancy in the paper. This is internally consistent with the reader's verdict: the appropriate action is to require the DFT sign analysis to be made explicit and checked, plus an error-bar and data-availability statement. An unconditional accept would be premature; a rejection would mischaracterize a paper that does present a reproducible phase-resolved technique. Hence CONDITIONAL is the right call, and agreement_with_reader is 'agree'.","tokens_in":10566,"tokens_out":1602,"duration_ms":13616,"concrete_test":"Independently reproduce the SI Section S2 DFT result for the sign of the resonant SFG susceptibility of monolayer hBN, including the phonon displacement coordinate convention: obtain the sign of the Raman tensor and Born effective charge such that the sign of chi_eff^{(2)} at the TO phonon is fixed relative to an experimentally defined +x axis. Then independently calibrate the z-cut quartz reference phase by measuring a known sample (e.g., a reference with known susceptibility sign from literature). If both checks reproduce the assumed positive sign, the edge-termination claim holds; if either sign flips, the claim flips to B-terminated.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central crystallographic claim has two parts: (i) threefold phase imaging gives the lattice orientation, and (ii) the absolute B-N vs N-B polarity is fixed by DFT (SI Section S2), so the observed positive SFG amplitude implies N-terminated zigzag edges. Part (i) is well supported by the azimuthal Fourier analysis and the 180-degree discrimination demonstrated in Figs. 3a-b. Part (ii) is the load-bearing step: the main text only states \"Through comparison with density functional theory calculations (SI Section S2), we expect a positive SFG amplitude\" when the B-N armchair direction points along +x. The actual DFT calculation, including all sign conventions (lattice coordinates, Cartesian axes, Raman tensor, Born effective charge, field polarization, and the z-cut quartz reference phase), is relegated to the SI. If the calculation is internally consistent that is fine, but the main text gives no check against an independent datum, and the existing literature (e.g., Li et al. 2013, Nano Lett. 13, 3329) does not report a phase-resolved hBN susceptibility. A single sign error in the DFT or in the quartz reference would flip every \"N-terminated\" conclusion to B-terminated, because the B-N and N-B directions differ only by a sign of the 3fold Fourier component. This is exactly what the authors use the phase to distinguish. The claim's empirical scope (\"vast majority ... N-terminated\") is therefore only as secure as the hidden DFT sign. The relative-orientation conclusions (zigzag vs armchair edges, island-to-island orientations) do not depend on this sign and are robust; the absolute polarity claim does.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript demonstrates phase-resolved, phonon-enhanced sum-frequency generation (SFG) microscopy for imaging CVD-grown monolayer hexagonal boron nitride (hBN) islands on fused silica. The authors use heterodyne detection with balanced paired-pixel imaging and azimuthal sample rotation to record the amplitude and phase of the threefold SFG component. They report an ~800-fold intensity enhancement from the E' transverse optical phonon, enabling acquisition over ~150x200 µm^2 fields in about a second. The orientation maps are correlated with AFM topography, and the measured phase, combined with DFT calculations, leads to the conclusion that the triangular islands have nitrogen-terminated zigzag edges. The paper also discusses future applications to twist angles, stacking, strain, and infrared upconversion devices.","tokens_in":10804,"tokens_out":8278,"duration_ms":80223,"significance":"If it holds, the central result offers a fast, label-free, large-area technique for full crystallographic characterization of hBN monolayers, a material that is essentially invisible to linear optical microscopy. The experimental execution is careful: phase-resolved heterodyne detection isolates the sign of the threefold response, quartz normalization provides an amplitude and phase reference, rotational Fourier analysis cleanly separates the desired 3-fold component, and AFM images independently corroborate the topography. The huge phonon enhancement places SFG on par with excitonic SHG in TMDs, which argues for wide applicability. The absolute edge-termination claim, however, is only as secure as the DFT sign of the nonlinear susceptibility, which is not presented in the main text; this is the key caveat to the paper's headline conclusion.","major_comments":[{"comment":"The absolute assignment of N-terminated versus B-terminated zigzag edges rests on the sign of the DFT-calculated nonlinear susceptibility and on the phase of the z-cut quartz reference. The main text states only that 'Through comparison with density functional theory calculations (SI Section S2), we expect a positive SFG amplitude when the B-N arm-chair crystal direction points along the positive x axis'. Since the threefold Fourier component is sign-sensitive, a sign error in the DFT calculation or an uncontrolled sign flip in the quartz reference would reverse the edge-termination conclusion for every island. This issue is load-bearing for the headline claim. Please present the calculated sign of the susceptibility and all sign conventions (lattice coordinates, Cartesian axes, field polarizations, quartz reference phase) in the main text or in a dedicated, clearly explained SI section, and provide at least one independent check of the sign (for example, a phase measurement on an hBN sample whose termination is known from STM/TEM, or a cross-check of the quartz reference sign against an independent nonlinear measurement).","section":"Results, 'Full Crystallographic Imaging of hBN Monolayers' (last paragraph)"},{"comment":"The sentence 'we do not observe B-terminated edges' is not an independent empirical finding; it is the same DFT-sign assignment just described. The measurement determines a threefold phase, and the identification of that phase with N-termination is contingent on the DFT sign. Please state this contingency explicitly and consider softening the claim unless an independent experimental calibration is provided.","section":"Results, Fig. 3f and discussion of 'B-terminated edges'"}],"minor_comments":[{"comment":"The abstract quotes '100x100 µm2' while the Results text and Figure 1 caption mention '150x200 μm' and '200x150 µm2'; please harmonize the stated field of view.","section":"Abstract vs. Results"},{"comment":"The term 'IR-subdiffractional spatial resolution' should be defined; the resolution is limited by the visible SFG wavelength, not by the IR wavelength, so a brief clarification would help avoid confusion.","section":"Discussion"},{"comment":"The paper states that 'for almost all islands' the armchair directions point toward the triangle corners and that the 'vast majority' have N-terminated edges, but it does not report the number of islands analyzed or the criteria for classifying the few irregular flakes; please provide the island count and the fraction of each orientation.","section":"Results, edge-termination statistics"},{"comment":"The black lines indicating armchair directions may be difficult to distinguish from the island edges; a zoomed panel or a legend showing a single flake would improve clarity.","section":"Figure 3e"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a strong experimental contribution with careful control measurements and a clear presentation of the imaging method. The central risk is the absolute polarity assignment, which depends on a DFT sign that is not shown in the main text. If the authors can make that sign verifiable and provide a cross-check, the paper would be suitable for publication. I have no concerns about novelty or fit with the journal's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things before reading. This paper introduces the first phase-resolved sum-frequency microscope applied to monolayer hBN, and it is a genuinely useful tool. It solves a real problem: hBN monolayers are nearly invisible optically, and SHG intensity cannot distinguish B-N from N-B directions. The ~800-fold phonon enhancement, the sub-diffraction resolution, and the ability to map crystal orientation over 100x100 µm in under a second are all real.\n\nWhat is new and solid: the heterodyne balanced imaging, the quartz normalization, and the rotational Fourier analysis are carefully executed. The relative orientation maps, which show which armchair direction points where, are robust and directly cross-checked against AFM topography. The geometry-based conclusion that the islands have zigzag edges, rather than armchair, is well supported. I buy that part.\n\nThe soft spot is the absolute polarity. The paper concludes nitrogen-terminated zigzag edges from the sign of the SFG amplitude. That sign is assigned by comparing with DFT calculations described in SI Section S2, but the main text only says \"we expect a positive SFG amplitude.\" If the DFT sign convention is off, or if the z-cut quartz reference introduces a sign flip, every nitrogen-terminated label flips to boron-terminated. The stress-test note lands on this correctly. It is not fatal to the paper's main value: the imaging capability, the orientation maps, and even the zigzag-vs-armchair conclusion all stand independent of that sign. But the edge-termination headline is only as secure as the hidden sign. The authors should bring the DFT sign analysis into the main text or verify it against an independent experimental reference.\n\nMinor issues: the quantitative values (1.9x10^-10 m^2/V, 580 pm/V, ~790x enhancement) come without error bars, and no raw data or analysis code are shared. For a techniques paper, that is thinner than ideal, but it does not undermine the central demonstration.\n\nWho this is for: anyone working on hBN or van der Waals heterostructures who needs fast, label-free screening of monolayer location and orientation. It is a strong methods paper and deserves a serious referee. My recommendation: send it to review, and ask the authors to make the DFT sign and the error budget visible before it is final.","headline":"New phase-resolved SFG microscopy gives hBN orientation maps that SHG cannot; the edge-polarity claim needs the DFT sign shown.","tokens_in":11488,"tokens_out":3039,"would_cite":true,"duration_ms":27846,"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":"Phase-resolved SFG microscopy images the full crystal orientation of hBN monolayers with an ~800-fold phonon boost.","keywords":["sum-frequency generation microscopy","hexagonal boron nitride","phonon enhancement","heterodyne detection","crystallographic orientation","zigzag edge termination","two-dimensional materials","wide-field nonlinear imaging"],"falsifier":"Perform atomically resolved scanning tunnelling microscopy or transmission electron microscopy on the same type of CVD-grown triangular hBN islands and record which element terminates the zigzag edges; if the terminating atoms are boron rather than nitrogen, the DFT sign assignment used to interpret the SFG phase is inverted.","tokens_in":10356,"feed_emoji":"🔬","tokens_out":5611,"duration_ms":53183,"temperature":0.7,"pith_summary":"Hexagonal boron nitride monolayers are nearly invisible to ordinary optics, yet their crystal orientation matters for almost every van der Waals device application. This paper establishes that phase-resolved sum-frequency generation (SFG) microscopy can image these monolayers and recover their full crystallographic orientation in a wide-field measurement. The key is tuning the mid-infrared drive onto hBN's transverse optical phonon, which gives an ~800-fold resonant enhancement of the SFG signal and makes the material as easy to image as excitonic transition metal dichalcogenides. With heterodyne phase detection and azimuthal rotation, the 3-fold lattice symmetry is resolved including the B-N versus N-B direction, and the authors find that CVD-grown triangular islands have nitrogen-terminated zigzag edges. If true, this gives a fast, label-free way to locate and orient hBN monolayers and other non-centrosymmetric van der Waals materials.","feed_headline":"Phonon-boosted SFG maps hBN monolayers' crystal orientation","feed_subtitle":"Resonant mid-infrared drive boosts the signal ~800-fold, revealing nitrogen-terminated zigzag edges over large areas.","key_machinery":"The central object is the effective second-order nonlinear susceptibility $\\chi^{(2)}(\\omega)$ of monolayer hBN, whose resonant form couples the IR-active transverse optical phonon at $\\omega_{TO}=1368\\ \\mathrm{cm^{-1}}$ to the Raman polarizability; because the monolayer lacks inversion symmetry, the phonon is both IR- and Raman-active, giving the ~800-fold signal boost. The measurement apparatus is a wide-field, phase-resolved SFG microscope using balanced paired-pixel heterodyne imaging, and the crystallographic readout comes from a rotational Fourier analysis that isolates the $\\cos(3\\varphi)$ threefold component of the SFG amplitude, where $\\varphi$ is the angle between the armchair crystal direction and the in-plane laser polarization. The sign of that component distinguishes B-N from N-B directions.","core_discovery":"Phase-resolved, phonon-enhanced SFG microscopy can serve as a full crystallographic imaging tool for monolayer hBN: it visualizes the monolayer over ~100x150 micron fields in under a second, measures the resonant nonlinear susceptibility through interference with a local oscillator, and, by recording the threefold azimuthal component of the SFG amplitude, maps the in-plane crystal orientation pixel by pixel at an infrared-subdiffractional resolution set by the visible SFG wavelength. The paper's concrete material-science finding is that the vast majority of CVD-grown triangular hBN islands are single crystals with zigzag edges, and that these edges are nitrogen-terminated; the B-N versus N-B distinction is drawn from the sign of the phase-resolved SFG signal referenced against z-cut quartz and interpreted with density functional theory. The paper also quantifies a ~790-fold intensity enhancement of the resonant phonon contribution relative to the off-resonant electronic second-order response.","pith_inferences":["If the DFT sign convention were reversed, the same microscopy data would assign boron-terminated zigzag edges instead of nitrogen-terminated; the imaging method itself would survive, but the chemical label would flip.","The rotational Fourier analysis should transfer directly to any 3-fold-symmetric non-centrosymmetric 2D material with an IR-active, Raman-active phonon, giving phonon-selective twist-angle maps in heterostructures without needing lattice-resolved probes.","Because the spatial resolution is currently set by the visible SFG wavelength, using a shorter upconversion wavelength or structured illumination could push phonon-selective nonlinear imaging to the nanoscale, a testable extension the paper does not demonstrate."],"forward_implications":["Monolayer hBN islands can be located, oriented, and distinguished from contamination on transparent substrates without AFM, Raman, or destructive probes.","Phase-resolved SFG resolves 180-degree rotated domains that ordinary SHG intensity measurements cannot distinguish, enabling B-N versus N-B polarity imaging over macroscopic areas.","Because the response is resonant with a phonon and spectrally resolved, local strain, layer number, stacking, and twist-angle variations should show up as shifts and changes in the SFG spectrum across the field of view.","The strong phonon-enhanced nonlinearity makes monolayer hBN an efficient mid-IR-to-visible converter, with further quadratic gains expected for 3R-stacked multilayers.","The same microscope concept should apply to other van der Waals materials, interfaces, and molecular assemblies with broken inversion symmetry and IR- plus Raman-active phonons."],"supporting_citations":[{"why":"Supplies the paired-pixel balanced-imaging wide-field microscope geometry on which the SFG imaging is built.","marker":"[21]"},{"why":"Previous phase-resolved SFG microscopy work that this paper extends to crystallographic imaging of 2D materials.","marker":"[22]"},{"why":"Provides the off-resonant hBN SHG susceptibility baseline and the sixfold symmetry limitation that phase-resolved SFG overcomes.","marker":"[11]"},{"why":"Gives the sum-frequency phonon resonance model used to fit the measured hBN spectrum.","marker":"[32]"},{"why":"Ab initio nonlinear-optics formalism supporting the phonon-enhanced susceptibility expression used in the analysis.","marker":"[33]"},{"why":"Calculations of BN edge energetics that motivate the dominance of zigzag edges corroborated by the SFG images.","marker":"[38]"},{"why":"CVD growth and transfer procedures used to prepare the monolayer hBN islands on fused silica.","marker":"[24]"}],"fun_headline_variants":["Full crystallographic map of hBN monolayers via phonon-enhanced SFG","Phonon resonance gives 800x boost to map hBN crystal orientation","SFG microscope images hBN monolayers' crystal lattice in under a second","Phase-resolved SFG reveals nitrogen-terminated zigzag edges in hBN","Phonon-enhanced SFG maps hBN monolayers' crystal orientation in 1 s"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The B-N versus N-B edge-termination result rests on a density functional theory calculation, described only in the supplementary information, that fixes which sign of the SFG amplitude corresponds to a nitrogen-terminated zigzag edge; if that calculated sign is wrong, the conclusion flips to boron-terminated edges.","fun_headline_variants_meta":{"raw":{"variants":["Full crystallographic map of hBN monolayers via phonon-enhanced SFG","Phonon resonance gives 800x boost to map hBN crystal orientation","SFG microscope images hBN monolayers' crystal lattice in under a second","Phase-resolved SFG reveals nitrogen-terminated zigzag edges in hBN","Phonon-enhanced SFG maps hBN monolayers' crystal orientation in 1 s"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000831,"raw_usage":{"total_tokens":3659,"prompt_tokens":1009,"completion_tokens":2650,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":625,"completion_tokens_details":{"reasoning_tokens":2546}},"tokens_in":625,"tokens_out":2650,"duration_ms":16912,"temperature":1.0,"reasoning_tokens":2546,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:13:53.692611+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform atomically resolved scanning tunnelling microscopy or transmission electron microscopy on the same type of CVD-grown triangular hBN islands and record which element terminates the zigzag edges; if the terminating atoms are boron rather than nitrogen, the DFT sign assignment used to interpret the SFG phase is inverted.","supporting_citations":[],"review_version":1}