{"id":"28e674a9-f353-4a84-a130-32af416994b0","arxiv_id":"2505.19224","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper establishes the A1g/E2g Raman intensity ratio as a figure of merit for detecting nanoscale contamination in monolayer MoSe2 and identifies new peaks associated with such protuberances.","lead":"This paper shows that nanoscale bumps on a single layer of the semiconductor MoSe2 have a lower ratio of two Raman spectral peaks than clean areas, which could serve as a quick way to spot contamination. The finding gives researchers a non-destructive spectroscopic fingerprint for checking the quality of 2D material devices.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The A1g/E2g drop over nanoprotuberances is quantitatively consistent with tip–sample distance loss of near-field enhancement, not with contamination chemistry; a tip-approach curve on flat MoSe2 would test this.","rationale":"The reader identified the same weakest assumption, and the manuscript's own quantitative data sharpen it. Using the far-field value as a baseline, protuberance pixels have essentially the same A1g/E2g ratio as the far field, whereas flat near-field pixels show roughly 1.8 times higher ratios. This is exactly the signature expected from distance-dependent near-field enhancement: when the tip is effectively lifted by an 8–25 nm bump, the A1g-selective near-field enhancement is lost, leaving a ratio close to the far-field value. The paper lists this as a contributing factor but does not measure its magnitude, so the central claim that the ratio drop is a reliable contamination signature is not yet supported. A tip-approach curve on flat MoSe2 would settle the matter directly with the same instrument. The work is otherwise plausible and the chemical peaks are suggestive, but the figure of merit needs this control. The reader's CONDITIONAL verdict remains appropriate; no adjustment is needed.","tokens_in":7432,"tokens_out":3687,"duration_ms":39947,"concrete_test":"Record TERS spectra on a flat MoSe2/hBN region while sweeping tip–sample separation from the engaged setpoint to roughly 30 nm retracted, covering the 8–25 nm height range of the protuberances, using the same PTTP tip and laser. Plot A1g/E2g versus separation. If the ratio falls from about 2.8 to about 1.5 over 8–25 nm, the nanoprotuberance contrast is quantitatively explained by topography, and the claimed contamination signature would need residual, height-independent evidence to stand. If the ratio is flat over that range, the topographic artifact is excluded and the chemical interpretation is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires the A1g/E2g drop to report contaminant chemistry. The paper's own numbers point the other way: the far-field mean ratio is 1.47 ± 0.08, while near-field flat regions give 2.8 ± 0.8 and 2.5 ± 0.5, and nanoprotuberances give 1.5 ± 0.4 and 1.7 ± 0.3. Protuberance pixels therefore sit at the far-field value, and the apparent 'signature' is that near-field enhancement—which preferentially boosts A1g—is lost when the tip rides over an 8–25 nm bump. The authors list tip–sample separation as 'another contributing factor' in the Results section and cite ref 21, but never quantify it. Because TERS enhancement decays on a nanometer scale, an 8–25 nm height increase can plausibly reproduce the entire observed contrast without any contamination-specific dielectric or chemical effect. Without a distance-dependence calibration, the figure of merit is a topography marker, not a material signature.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports tip-enhanced Raman spectroscopy (TERS) hyperspectral maps of monolayer MoSe2 on hBN, correlating nanoscale protuberances (50–200 nm diameter, 8–25 nm height) with local changes in the A1g/E2g Raman intensity ratio and with new Raman peaks attributed to contamination and oxidation. The authors propose the A1g/E2g ratio reduction over protuberances as a figure of merit for surface contamination. The central evidence is two TERS maps (HS1, HS2) in which the mean ratio is roughly 1.5–1.7 on protuberances versus 2.5–2.8 on surrounding flat areas, while a far-field map of the same region gives 1.47 ± 0.08. Additional spectral peaks (e.g., 974, 998, 1168, 1226, 1423, and 1585 cm−1) are assigned to MoO3 and carbonaceous species, and AFM phase contrast is used to support the contamination interpretation.","tokens_in":7660,"tokens_out":4273,"duration_ms":38783,"significance":"If the ratio reduction truly tracks contamination chemistry, the paper would offer a practical nano-Raman fingerprint for detecting contamination in 2D heterostructures. The main strengths are that the ratio contrast is reproduced in two independent TERS maps, the authors explicitly compare near-field and far-field behavior, and the chemical-peak observations, while tentative, are a valuable starting point for identifying trapped contaminants. The paper is also honest in listing tip–sample separation as a contributing factor. However, the proposed figure of merit is not yet established, because the observed contrast is quantitatively consistent with a topographic artifact from reduced near-field enhancement when the tip rides over the bumps, and the paper does not provide a distance-dependence calibration to separate that effect from a contamination-specific chemical response.","major_comments":[{"comment":"The mean ratios over nanoprotuberances (1.5 ± 0.4 in HS1 and 1.7 ± 0.3 in HS2) are statistically indistinguishable from the far-field ratio (1.47 ± 0.08) reported for the same region. This means the data are fully consistent with a loss of near-field enhancement when the tip rides over 8–25 nm high features, rather than with a contamination-specific effect on the A1g mode. The authors mention tip–sample separation as \"another contributing factor\" but do not quantify it. Because TERS enhancement decays on a nanometer length scale, a tip-approach curve on flat MoSe2 (or an equivalent distance-dependence model) is required to separate the topographic contribution from any material-specific contribution. Without such a calibration, the ratio contrast is a topography marker, not a material signature of contamination.","section":"Results, TERS figure of merit; Table 1"},{"comment":"No pixel counts or numbers of spectra are given for the inside/outside protuberance classes, and no statistical test is reported. Given the large standard deviations (e.g., 2.8 ± 0.8 for HS1 surroundings), the separation between the two populations needs to be quantified. Please report the number of spectra in each class and the distribution (e.g., histograms or box plots), and provide a significance test (e.g., Welch's t-test or Mann-Whitney U). This is necessary to support the claim that the ratio reliably distinguishes the regions.","section":"Results, Table 1"},{"comment":"The conclusion that a reduced A1g/E2g ratio \"serves as a reliable nano-Raman signature of contamination in monolayer MoSe2 samples\" is stronger than what the data support. The measured contrast is a correlation with protuberance topography; the attribution to contamination rests on the auxiliary spectral peaks and on AFM phase contrast, not on the ratio itself. Either add a calibration experiment that rules out the topographic mechanism, or rephrase the claim so that the ratio is presented as a nano-topography marker that accompanies contamination-related spectral signatures rather than as a standalone contamination signature.","section":"Conclusions"}],"minor_comments":[{"comment":"The assignments of the 974 cm−1 and 998 cm−1 peaks to MoO3-related modes are plausible, but they are not confirmed by complementary techniques such as XPS or EDS; the text uses \"likely\" and \"may indicate,\" which is appropriate, but the limitations should be restated in the Conclusions.","section":"Results, Chemical analysis"},{"comment":"The PCA preprocessing is described only briefly; please state how many principal components were retained and discuss whether the five-component reconstruction could distort the amplitudes of weak Lorentzian peaks used for the intensity maps.","section":"Methods"},{"comment":"The phase maps and ratio maps lack explicit axis scales and quantified color-bar units; adding scale bars, axis ticks, and unambiguous color-bar labels would improve reproducibility and readability.","section":"Figures 2 and 3"},{"comment":"There are several typographical errors with missing spaces, for example \"correspondingrespectivelytoregions\" and \"The resultsreinforce that theintensity ratiois lower\"; the manuscript needs a careful proofreading pass.","section":"Results"},{"comment":"The sentence about Figure 6, \"The map shown in Figure 6 was acquired at the edge of the MoSe2 flake, which accounts for the presence of the observed peak in regions that would otherwise be expected to be flat,\" is awkward and should be clarified regarding which regions are inside versus outside the protuberance mask.","section":"Results, Chemical analysis"},{"comment":"Reference 21 is a general textbook; citing a specific TERS tip-sample distance-dependence study would strengthen the discussion of the topographic factor and help the reader assess its magnitude.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the journal's scope and the experiments are interesting. The central claim, however, needs to be backed by control measurements that decouple topography from chemistry. A tip-approach curve on flat MoSe2 (or a distance-dependent TERS model) would be the decisive experiment; alternatively, the authors could reframe the ratio as a topographic marker and combine it with the chemical-peak maps to form a multivariate contamination signature. I also recommend that the editor ask for a clear statistical statement (number of pixels/spectra, significance test) in the revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Jane Elisa et al. report a correspondence between AFM-visible nanoprotuberances on monolayer MoSe2/hBN and a drop in the TERS A1g/E2g intensity ratio, along with several spatially correlated Raman peaks (974, 1168, 1226, 1423, 1579/1587/1600 cm–1) that plausibly arise from MoO3 and carbonaceous contamination. That is a genuinely new set of observations, and the two hyperspectral maps give the result a degree of reproducibility. The paper is clearly written, and the authors are transparent about the interpretive uncertainties.\n\nThe load-bearing soft spot is the topography artifact, and the stress-test note is right: the protuberance ratios (1.5±0.4 and 1.7±0.3) essentially coincide with the far-field value (1.47±0.08), while flat near-field regions give 2.5–2.8. That is precisely the pattern you would expect if the near-field enhancement—which preferentially boosts A1g—is simply lost when the tip rides up over an 8–25 nm bump. The authors mention tip–sample separation as a contributing factor and cite Novotny and Hecht, but they never quantify it. Without a tip-approach curve on flat MoSe2, the claimed 'reliable signature of contamination' is just as plausibly a topographic height marker. That distinction matters if someone tries to use the ratio to diagnose contamination independent of topography.\n\nTwo smaller points: the paper gives no pixel counts or number of spectra per region, so the statistical weight of those means is unclear, and the chemical assignments are plausible but not confirmed by any complementary measurement. The HS1/HS2 maps come from a single sample, so flake-to-flake variability is untested.\n\nAll that said, the paper is honest and the raw data are likely usable by others. The right fix is not rejection but a major revision that includes a distance-dependence calibration and proper statistics. If the ratio drop survives that control, the claim becomes solid; if it does not, the paper still has value as a demonstration of a TERS artifact mechanism. I'd send it to review.","headline":"A useful empirical observation—A1g/E2g suppression over MoSe2 nanoprotuberances—but the data are consistent with a topographic loss of near-field enhancement, so the 'contamination signature' claim needs a distance-dependence control before it can be called chemical.","tokens_in":8208,"tokens_out":3872,"would_cite":true,"duration_ms":36585,"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":"The paper shows that a reduction in the A1g/E2g intensity ratio of monolayer MoSe2 in tip-enhanced Raman maps is a reliable nano-Raman signature of nanoscale contamination, and that contaminant-related peaks reveal oxidation and…","keywords":["tip-enhanced Raman spectroscopy","MoSe2 monolayer","nanoprotuberances","A1g/E2g ratio","surface contamination","van der Waals heterostructures","MoO3 oxidation","nano-Raman hyperspectral imaging"],"falsifier":"Take the same TERS setup over clean MoSe2 on stepped hBN or lithographically defined mounds with the same 8–25 nm heights as the protuberances but no contamination; if the $A_{1g}/E_{2g}$ ratio falls over these clean raised features as strongly as over the nanoprotuberances, the proposed figure of merit tracks topography rather than contamination.","tokens_in":7295,"feed_emoji":"🔬","tokens_out":7311,"duration_ms":58349,"temperature":0.7,"pith_summary":"Nano-Raman (tip-enhanced Raman) measurements on monolayer MoSe2 on hBN show that regions containing 50–200 nm nanoprotuberances have a consistently lower intensity ratio of the $A_{1g}$ (240 cm$^{-1}$) to $E_{2g}$ (287 cm$^{-1}$) Raman bands than the surrounding flat areas. The paper proposes this ratio reduction as a reliable nano-Raman signature of surface contamination, giving experimenters a figure of merit that works where confocal Raman cannot resolve individual protuberances. The same hyperspectral data also reveal contaminant- and oxidation-related peaks (near 974, 998, 1168, 1226, 1423, and 1585 cm$^{-1}$) that appear or change across the protuberances, linking the topographic features to carbon-based contamination and partial oxidation of MoSe2. If the claim holds, TERS ratio maps become a practical way to detect and localize nanoscale impurities in two-dimensional heterostructures.","feed_headline":"Raman ratio drop exposes nanoscale contamination on MoSe2","feed_subtitle":"Tip-enhanced Raman maps of the A1g/E2g band ratio localize 50–200 nm protuberances in monolayer MoSe2.","key_machinery":"The central object is the hyperspectral TERS intensity-ratio map of the $A_{1g}$ to $E_{2g}$ Raman modes of monolayer MoSe2 ($A_{1g}$ at 240 cm$^{-1}$, $E_{2g}$ at 287 cm$^{-1}$). The maps are built from nano-Raman hyperspectra recorded with 15.6 nm steps using AFM-based TERS with plasmon-tunable tip pyramids, then processed by principal component analysis and Lorentzian curve fitting. The ratio map itself is the figure of merit: it separates protuberance pixels from flat-area pixels in two independent sample regions, and it is complemented by amplitude maps of individual contaminant-related Raman peaks that show complementary spatial distributions.","core_discovery":"On its own terms, the paper claims that the intensity ratio $I_{A_{1g}}/I_{E_{2g}}$ of monolayer MoSe2, measured by tip-enhanced Raman spectroscopy, drops markedly over nanoprotuberances and therefore serves as a nano-Raman figure of merit for contamination. In the two near-field maps, the mean ratio on protuberances was 1.5 ± 0.4 and 1.7 ± 0.3, versus 2.8 ± 0.8 and 2.5 ± 0.5 in the surrounding flat regions, while a far-field average over the same area gave 1.47 ± 0.08 without spatial discrimination. The authors attribute the lower on-protuberance ratio to a weaker near-field enhancement of the $A_{1g}$ mode, related to loss of light coherence in the near field, and note that a slight increase in tip–sample separation over the raised features may also contribute. Complementary spectral peaks assigned to MoO3 and to carbon/organic species localize preferentially on or around the protuberances, supporting the interpretation that the protuberances are contamination with associated oxidation.","pith_inferences":["If the ratio suppression reflects contamination chemistry rather than the raised topography, the same figure of merit is likely to work for other monolayer TMDCs such as MoS2 and WS2, whose $A_{1g}/E_{2g}$ pairs respond similarly to near-field coherence.","A control experiment on clean sample steps of comparable height would separate the tip–sample distance contribution from the chemical signature; the paper identifies this distance effect as a possible alternative explanation.","The complementary spatial behavior of lower-frequency vs higher-frequency Lorentzian components on and around protuberances suggests local charge doping or strain gradients, which could be tested with nanoscale photoluminescence mapping."],"forward_implications":["TERS ratio maps can localize individual nanoprotuberances that conventional far-field Raman averages cannot resolve.","The $A_{1g}/E_{2g}$ ratio can serve as a surface-quality metric when assembling MoSe2-based van der Waals heterostructures.","The new contaminant-related peaks (MoO3 near 998 cm$^{-1}$, C=C near 1585 cm$^{-1}$, organic modes at 1168, 1226, and 1423 cm$^{-1}$) identify oxidation and carbon contamination within the protuberances.","The ratio contrast appears consistently across two independent hyperspectral maps, indicating reproducibility of the signature.","Far-field Raman measurements alone are insufficient to detect the contamination signature because they average over protuberance and surrounding regions."],"supporting_citations":[{"why":"Supplies the plasmon-tunable tip pyramid probes used for all TERS measurements and their near-field enhancement.","marker":"[13]"},{"why":"Provides the near-field Raman coherence framework the authors use to explain the weaker A1g enhancement on protuberances.","marker":"[11]"},{"why":"Theory of spatial coherence in near-field Raman scattering that underlies the coherence-loss interpretation of the ratio change.","marker":"[19]"},{"why":"Applies TERS coherence-length analysis to 2D materials, supporting the mode-dependent near-field enhancement argument.","marker":"[20]"},{"why":"Cited for the known effect of tip-sample separation on near-field optical signals, the alternative explanation the paper must discount.","marker":"[21]"},{"why":"Reference for MoO3 Raman bands used to assign the ~998 cm$^{-1}$ peak to oxidation of MoSe2.","marker":"[22]"},{"why":"Reference for the C=C stretching mode used to assign the ~1585 cm$^{-1}$ feature to carbon contamination.","marker":"[24]"}],"fun_headline_variants":["MoSe2 Raman ratio dips mark contamination spots","Raman ratio reveals 50-200 nm contamination in MoSe2","A1g/E2g ratio drop spots nanoscale impurities on MoSe2","Nanoscale protuberances on MoSe2 tagged by Raman ratio"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes the observed drop in the $A_{1g}/E_{2g}$ ratio over the bumps is caused by the contamination's effect on the near-field Raman signal, not simply by the tip being farther away on top of the raised features.","fun_headline_variants_meta":{"raw":{"variants":["MoSe2 Raman ratio dips mark contamination spots","Raman ratio reveals 50-200 nm contamination in MoSe2","A1g/E2g ratio drop spots nanoscale impurities on MoSe2","Nanoscale protuberances on MoSe2 tagged by Raman ratio"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000911,"raw_usage":{"total_tokens":3899,"prompt_tokens":912,"completion_tokens":2987,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":528,"completion_tokens_details":{"reasoning_tokens":2912}},"tokens_in":528,"tokens_out":2987,"duration_ms":21399,"temperature":1.0,"reasoning_tokens":2912,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:17:42.532023+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the same TERS setup over clean MoSe2 on stepped hBN or lithographically defined mounds with the same 8–25 nm heights as the protuberances but no contamination; if the $A_{1g}/E_{2g}$ ratio falls over these clean raised features as strongly as over the nanoprotuberances, the proposed figure of merit tracks topography rather than contamination.","supporting_citations":[{"cited_title":"G.; Ravikumar, R.; Kumar, T","cited_arxiv_id":null,"evidence_quote":"Reference for MoO3 Raman bands used to assign the ~998 cm$^{-1}$ peak to oxidation of MoSe2."},{"cited_title":"L.; Archanjo, B","cited_arxiv_id":null,"evidence_quote":"Supplies the plasmon-tunable tip pyramid probes used for all TERS measurements and their near-field enhancement."},{"cited_title":"G.; Rabelo, C.; Gadelha, A","cited_arxiv_id":null,"evidence_quote":"Provides the near-field Raman coherence framework the authors use to explain the weaker A1g enhancement on protuberances."},{"cited_title":"G.; Beams, R.; Jorio, A.; Novotny, L","cited_arxiv_id":null,"evidence_quote":"Theory of spatial coherence in near-field Raman scattering that underlies the coherence-loss interpretation of the ratio change."},{"cited_title":"G.; Jorio, A","cited_arxiv_id":null,"evidence_quote":"Applies TERS coherence-length analysis to 2D materials, supporting the mode-dependent near-field enhancement argument."},{"cited_title":"Principles of Nano-Optics, 2nd ed.; Cambridge University Press, 2012","cited_arxiv_id":null,"evidence_quote":"Cited for the known effect of tip-sample separation on near-field optical signals, the alternative explanation the paper must discount."},{"cited_title":"S.; Saito, R.; Dresselhaus, G","cited_arxiv_id":null,"evidence_quote":"Reference for the C=C stretching mode used to assign the ~1585 cm$^{-1}$ feature to carbon contamination."}],"review_version":1}