{"id":"ec768828-a8fd-447e-a83a-6cbb4cc948ce","arxiv_id":"2607.22089","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"By detecting recoil-projectile coincidences in 40 keV Xe transmission, carbon and hydrogen contamination on freestanding graphene is quantified, with PMMA-free transfer plus 400 °C annealing reaching nearly clean single-layer graphene.","lead":"A new ion-beam method counts carbon and hydrogen atoms on atomically thin materials by detecting both the knocked-out atom and the transmitted beam particle at the same moment. It ranks which graphene transfer process leaves the least contamination and shows how heating can nearly clean the surface.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Absolute coverages rest on an uncalibrated cross-section × detector-efficiency product; a 20% systematic error would overturn the 'atomically clean' claim.","rationale":"The reader's weakest assumption focused on leakages from false coincidences/support recoils. I identify the more specific issue of absolute calibration: the unvalidated product of SIMNRA cross-sections and MCP detection efficiency under a single 0.54 factor. This equally undermines the quantitative central claim, and a calibration test would settle it. Since the method is otherwise plausible and the relative comparisons are robust, the CONDITIONAL verdict remains appropriate; no change is needed.","tokens_in":13986,"tokens_out":7566,"duration_ms":87675,"concrete_test":"Measure the coincidence yield for a sample with known areal density, e.g., a self-supporting monolayer of hBN (areal densities of B and N are known to ~1 ML) or a NIST-traceable carbon foil of known thickness, using the same 40 keV Xe setup and analysis pipeline. Compare the extracted ML values to the known 1 ML. If the ratio deviates from unity by more than 2× the statistical uncertainty, the current calibration factor is wrong and the reported absolute coverages must be revised. A second, complementary check: repeat the graphene measurement at a different projectile energy (e.g., 60 keV Xe); the inferred coverages should reproduce those at 40 keV if the cross-section model and detection-efficiency corrections are correct.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative conclusion that annealed Flattened graphene reaches 1.04 ± 0.02 ML C and 0.20 ± 0.01 ML H is obtained by dividing measured coincidence yields by a single scale factor: σ_eff/(A_UC·0.54), where σ_eff comes from SIMNRA with the Universal potential and 0.54 is stated as the MCP open-area ratio. This conflates two distinct quantities. First, the Universal-potential cross sections for 40 keV Xe on C/H are not independently validated at these low energies; differences of tens of percent are plausible. Second, the 0.54 factor corrects only for geometric open area, not for the actual detection efficiency of MCPs for slow recoils (H recoils ≈ 1.2 keV, C recoils ≈ 12 keV). MCP efficiency for such low-energy ions is typically lower and velocity-dependent. The reported uncertainties are purely statistical; no systematic uncertainty is assigned to the calibration. A 10% change in the efficiency/cross-section product shifts the C coverage to 1.14 ML, outside the cited ±0.02 precision, and doubles the H coverage shift. The central claim that the surface is 'approximately atomically clean within experimental uncertainty' therefore does not yet withstand quantitative scrutiny.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript introduces a recoil-projectile coincidence detection scheme in time-of-flight medium-energy ion scattering in transmission geometry to quantify surface contamination of freestanding single-layer graphene. Using 40 keV Xe projectiles, the authors detect H, C, and O recoils in coincidence with transmitted Xe and convert the measured coincidence yields into monolayer-equivalent coverages using SIMNRA cross sections and an MCP open-area correction. They apply the method to three graphene transfer routes (PMMA-free 'Flattened', and two PMMA-assisted types), before and after UHV annealing. The principal results are that thermal annealing at 400 °C for 1 h reduces C/H coverages, with the PMMA-free sample reaching 1.04 ± 0.02 ML C and 0.20 ± 0.01 ML H and remaining nearly uncontaminated for at least 140 min, whereas PMMA-transferred samples recontaminate within ~1–2 h. The method is claimed to be quantitative, element-specific, isotopically resolved, and minimally destructive.","tokens_in":14266,"tokens_out":16029,"duration_ms":172565,"significance":"If the absolute calibration can be established, the method would be a valuable quantitative, large-area (mm-scale average) complement to local STEM characterisation, with the unique ability to quantify hydrogen coverage on freestanding 2D materials. The coincidence concept and kinematic separation are physically reasonable; the convergence of annealed Flattened-graphene carbon coverage to 1.04 ML provides a useful internal consistency check; the radiation damage estimate is careful; and the relative comparisons between samples are likely robust. The main weakness is the uncalibrated absolute scale, which currently prevents the quantitative claims from being fully supported.","major_comments":[{"comment":"The absolute coverages are obtained from Eq. (3), which divides measured coincidence yields by σ_eff·A_SLG·0.54. The scale factor depends on unquantified inputs: SIMNRA/Universal σ_eff for 40 keV Xe on C,H, and the MCP 'effective open area ratio' 0.54 used as recoil detection efficiency. The latter is a geometric open-area fraction, not a measured efficiency; for slow recoils (H≈1.2 keV, C≈12 keV) MCP efficiency is typically lower and velocity-dependent, and the manuscript is ambiguous whether 0.54 should enter once or twice. Since the expected 1-ML yield uses the same σ_eff, the observed 1.04 ML for annealed Flattened graphene is an internal consistency check, not an independent calibration. A 10% error in σ_eff×ε shifts C coverage to ≈1.14 ML and has a larger relative effect on the 0.20 ML H coverage. Cross-section uncertainty is acknowledged but not propagated, so the 'within experime","section":"Section 5.3, Eq. (3), Table 1, Discussion"},{"comment":"The quantification assumes that the selected TOF windows contain only single-scattering events from C and H in freestanding graphene. However, Fig. 2b shows a feature near 32 keV attributed to H recoils from the Quantifoil support, and Table 1 lists 'Quantifoil overlap after annealing' for Graphenea. This suggests support recoils are not fully separated by the coincidence gate, despite Section 5.2 stating that Xe on Quantifoil is effectively stopped. False coincidences are also only described as 'low intensity' (Section 2.1) without a quantitative bound. If support or false coincidences leak into the H/C windows, the absolute coverages—especially the small post-annealing H coverage—are biased. Please show the gated spectra/windows, quantify leakage bounds, and explain how the 'Quantifoil overlap' was excluded from the reported values.","section":"Sections 2.2, 5.3, Fig. 2b, Table 1"}],"minor_comments":[{"comment":"The claim of 'isotopic resolution' is not demonstrated. No isotope-resolved features (e.g., D vs H, 13C vs 12C) are shown; with ~9 ns pulse length, the expected 12C/13C recoil TOF separation is only a few ns. Please either demonstrate isotopic separation or temper the claim.","section":"Abstract; Section 4"},{"comment":"Each transfer route is represented by a single sample (n=1). The transfer-route comparison would be strengthened by replicate samples or an explicit statement that the comparison is anecdotal rather than statistically validated.","section":"Section 2.2, Table 1"},{"comment":"The evaluated detector region is limited to angles <7.85° and Ω=0.059 sr; the effective cross-section is taken as dσ/dΩ × Ω under the assumption of a flat angular distribution. Please quantify the error from this approximation, or justify the constancy more explicitly.","section":"Section 5.3"},{"comment":"The temperature was monitored with a thermocouple attached to the sample holder, not on the graphene membrane. The quoted annealing temperatures may therefore differ from the actual sample temperature; this should be stated as an uncertainty.","section":"Section 5.4"},{"comment":"The TEM clean-area fractions (43%, 42%, 72%) are based on a single ImageJ thresholding protocol. Please provide details of the threshold selection and report the sensitivity of the fractions to the chosen parameters.","section":"Section 5.5, Figure 4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope and the method is promising, but the central quantitative claim rests on an uncalibrated absolute scale. The relative comparisons and internal consistency are convincing; the required fix is a systematic uncertainty budget and, ideally, an independent calibration. I do not see a need to question the authors' integrity; the gaps are technical and addressable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nQuick take on arXiv:2607.22089. The core is a recoil-projectile coincidence filter in transmission ToF-MEIS, which cleanly separates H and C recoils from the graphene itself from Quantifoil support recoils and the direct beam. That is genuinely new relative to refs [30] and [32]; I don't see this coincidence condition in prior transmission MEIS/recoil work. The demonstrated convergence to 1 ML carbon after annealing for the PMMA-free sample is a nice internal check and gives me some confidence that the absolute carbon calibration isn't wildly off. The transfer-method comparison (PMMA-free vs PMMA-based) shows large, consistent differences in both initial contamination and recontamination kinetics; those are probably robust.\n\nThe paper is honest about its main limitation: it states in Section 3 that absolute accuracy depends on scattering cross sections and detector acceptance, and that this contributes systematically to coverage but not to relative comparisons. Good. But the quantitative claim in the abstract and conclusion—'approaches atomically clean single-layer graphene within experimental uncertainty'—rests on an uncertainty budget that only includes counting statistics. The calibration is effectively a single factor: SIMNRA Universal-potential cross section times a 0.54 MCP open-area correction. The 0.54 ignores the actual detection efficiency for slow H (~1.2 keV) and C (~12 keV) recoils, which is likely velocity-dependent and could be noticeably lower. The stress-test note is right that a 10% shift in that factor moves C coverage to 1.14 ML, outside the quoted ±0.02, and roughly doubles the H coverage shift. That doesn't destroy the qualitative story, but it means 'atomically clean within experimental uncertainty' is too strong as written unless they assign a systematic uncertainty to the calibration.\n\nOther soft spots, in proportion: the transfer-method comparison relies on single samples per condition, so sample-to-sample variability is not assessed. The 'isotopic resolution' claim is plausible given the flight-time separation, but there's no explicit demonstration of resolving, say, 12C vs 13C or 1H vs 2H. The false-coincidence contribution is described as low-intensity but not quantified; a referee should ask for a bound or a control measurement. The Quantifoil support recoils are separated by flight time, which seems fine.\n\nOverall: this is a useful, well-described method paper with a clear internal consistency check and honest discussion of its own limitations. It deserves a serious referee. I'd send it to review, and the main request would be: quantify the systematic uncertainty in the cross-section/efficiency product, ideally with an independent calibration sample (e.g., known areal density of organic film or a self-supporting carbon foil), and add replicate samples or repeated measurements for at least one transfer condition. With that, the central claim becomes much stronger.\n\nMy verdict: conditional, leaning positive. I'd bring it to the reading group for a methods discussion. I'd cite it if I work on ion-beam characterization of 2D materials.\n\nRecommendation for you: if this lands on your desk, treat it as a solid experimental methods paper worth refereeing; don't desk-reject; ask for the systematic uncertainty work rather than challenging the core method.","headline":"A genuinely new coincidence-based transmission ToF-MEIS scheme for quantifying C and H contamination on freestanding 2D materials; the relative comparisons are probably solid, but the absolute 'atomically clean' claim needs a systematic calibration before it fully lands.","tokens_in":14757,"tokens_out":3149,"would_cite":true,"duration_ms":32316,"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 introduces a recoil-projectile coincidence method in keV ion transmission that identifies and quantifies individual surface contaminants—including hydrogen—on freestanding ultrathin materials with isotopic resolution and monolayer","keywords":["graphene","surface contamination","ion beam analysis","time-of-flight medium energy ion scattering","recoil-projectile coincidence","monolayer coverage","hydrogen quantification","ultrathin materials"],"falsifier":"Measure the coincidence yield from a bare holey-carbon grid with no graphene under identical 40 keV Xe conditions and compare it with the H/C window counts from a graphene sample; if the bare-grid yield accounts for more than the stated statistical uncertainty of the graphene coverage, the absolute monolayer values are not solely due to graphene contaminants. Alternatively, repeat the measurement on isotopically labelled 13C graphene and check that the carbon recoil peak shifts to the 13C flight time and that the inferred areal density remains 1 monolayer; failure to observe the predicted isot","tokens_in":13898,"feed_emoji":"⚛️","tokens_out":7590,"duration_ms":73839,"temperature":0.7,"pith_summary":"The authors take on a measurement problem: surface contamination controls the properties of 2D materials, but standard microscopy only sees nanometer-scale regions and cannot quantify hydrogen. They claim that requiring coincidence between a recoil atom and the projectile that scattered it suppresses background enough to turn 40 keV Xe ion transmission through freestanding graphene into an element-specific, area-averaged contamination measurement, with precision set by counting statistics. Applied to three differently transferred graphene samples, the method shows that thermal annealing at 400 °C for 1 h drives PMMA-free transferred graphene to 1.04 ± 0.02 monolayers of carbon and 0.20 ± 0.01 monolayers of hydrogen, approaching atomically clean single-layer graphene. The same measurements expose different recontamination kinetics: PMMA-transferred samples re-adsorb carbon and hydrogen within hours at 2×10⁻⁸ mbar, while PMMA-free graphene stays nearly clean for at least 140 minutes. If correct, this gives the ultrathin-materials field a quantitative, large-area cleanliness metric that resolves the very elements—especially hydrogen—that local imaging misses.","feed_headline":"Coincidence counting puts a number on graphene contamination","feed_subtitle":"Element-specific coverage shows PMMA-free graphene reaches 1.04 carbon monolayers after annealing","key_machinery":"The load-bearing mechanism is the recoil-projectile coincidence condition. In a pulsed 40 keV Xe beam transmitted through a freestanding target, the time-of-flight spectrum has distinct peaks for H recoils (~619 ns), C recoils (~671 ns), and transmitted Xe projectiles (~1197 ns); selecting only events where one particle falls in the recoil window and a second falls in the projectile window suppresses uncorrelated background and assigns each event to a specific scatterer. Coverage is computed as n_C/H = (A_UC · A_C/H) / (2 · σ_eff · A_SLG · 0.54), where A_UC is the graphene unit-cell area, A_C/H is the coincidence count in the recoil window, σ_eff is the effective scattering cross section com","core_discovery":"The central claim is that a coincidence condition—one detector hit in the time-of-flight window of H or C recoils paired with a second, coincident hit in the window of transmitted Xe projectiles—converts keV ion transmission through freestanding graphene into an element-specific, area-averaged contamination measurement. The authors demonstrate that carbon and hydrogen dominate surface contamination, that PMMA-free 'Flattened' graphene has the lowest native contamination, and that after annealing at 400 °C for 1 h its measured carbon areal density (1.04 ± 0.02 monolayers) and hydrogen coverage (0.20 ± 0.01 monolayers) are consistent with nearly atomically clean single-layer graphene. They fur","pith_inferences":["Independently calibrating the microchannel-plate detection efficiency, rather than folding it into a single 0.54 open-area factor, would let different laboratories compare absolute monolayer values and could turn this method into a standard for reporting 2D-material cleanliness.","The strong contrast between PMMA and PMMA-free recontamination suggests that transfer-induced residues or defect sites—not the intrinsic graphene lattice—set the adsorption kinetics; repeating the measurement on h-BN or MoS2 transferred by both routes could test this directly.","Widening the coincidence analysis to include recoil energy or angle information could yield not just total coverage but details of the in-plane distribution or binding configuration of contaminants on the 2D layer.","Combining the method with controlled gas dosing could deliver element-resolved sticking coefficients, a direction the authors mention as future work but do not yet demonstrate."],"forward_implications":["Element-specific quantification of both carbon and hydrogen on freestanding graphene, including isotopic resolution, becomes possible over macroscopic sample areas—closing a gap left by atomic-resolution imaging, which struggles to quantify hydrogen.","PMMA-free transferred graphene reaches 1.04 ± 0.02 ML C and 0.20 ± 0.01 ML H after 400 °C for 1 h, providing a concrete numeric benchmark for an atomically clean, large-area graphene surface.","PMMA-transferred graphene recontaminates within roughly 70–130 minutes at 2×10⁻⁸ mbar, while PMMA-free graphene remains clean for at least 140 minutes, showing that transfer route—not just initial cleanliness—controls the post-annealing surface state.","Because the method is minimally destructive (estimated at most ~0.26 ppm of lattice atoms damaged per typical measurement), it enables time-resolved in-situ studies of adsorption kinetics, annealing protocols, and cleaning efficiency.","Since the approach only requires ion transmission through sufficiently thin samples, the same recoil-projectile coincidence scheme transfers directly to other 2D materials and ultrathin films for studies of adsorption, implantation, and sticking coefficients."],"fun_headline_variants":["Ion coincidence counts carbon and hydrogen on graphene","Recoil-projectile coincidence quantifies 2D contamination","PMMA-free graphene stays clean hours after anneal","Element-specific ion counts gauge graphene atomic purity"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The quantitative coverage formula assumes that every coincidence event selected in the H/C time-of-flight windows is a single-scattering event from an atom in the freestanding graphene layer, with false coincidences, support-film recoils, and multiple scattering making negligible contributions; if any of those leak into the windows, the absolute areal densities are biased even if relative comparisons between samples survive.","fun_headline_variants_meta":{"raw":{"variants":["Ion coincidence counts carbon and hydrogen on graphene","Recoil-projectile coincidence quantifies 2D contamination","PMMA-free graphene stays clean hours after anneal","Element-specific ion counts gauge graphene atomic purity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00043,"raw_usage":{"total_tokens":2058,"prompt_tokens":795,"completion_tokens":1263,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":539,"completion_tokens_details":{"reasoning_tokens":1202}},"tokens_in":539,"tokens_out":1263,"duration_ms":14044,"temperature":1.0,"reasoning_tokens":1202,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T05:48:30.184807+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the coincidence yield from a bare holey-carbon grid with no graphene under identical 40 keV Xe conditions and compare it with the H/C window counts from a graphene sample; if the bare-grid yield accounts for more than the stated statistical uncertainty of the graphene coverage, the absolute monolayer values are not solely due to graphene contaminants. Alternatively, repeat the measurement on isotopically labelled 13C graphene and check that the carbon recoil peak shifts to the 13C flight time and that the inferred areal density remains 1 monolayer; failure to observe the predicted isot","supporting_citations":[],"review_version":1}