{"id":"12a8a4f4-79be-4474-8ccd-c94735866498","arxiv_id":"1908.02045","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"For tens-of-keV helium beams, uncertainty from the scattering potential in backscattering ion-beam analysis of thin films is below 3%, and smaller in transmission.","lead":"This paper measures how the choice of atomic interaction model affects ion-beam analysis of thin films with tens-of-keV beams. It finds that for light helium ions the model uncertainty stays below a few percent, which helps make thin-film measurements more reliable.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The <3% uncertainty claim lacks error propagation: fitted screening corrections may be degenerate with electronic stopping and are not shown to be uniquely identified.","rationale":"The paper is a sensitivity study, and its central claim is a bound: for the investigated conditions, uncertainties in the scattering potential propagate to less than 3% in quantitative analysis. The experimental spectra and TRBS simulations provide real evidence that the potential choice changes the multiple-scattering background and that reduced screening improves the fit. I do not see an internal inconsistency in the main construction. The weakest point is not the existence of a potential effect but the quantitative credibility of the '<3%' number. That number is obtained by comparing a limited set of potential models and by selecting c_a by eye-like fitting, without any uncertainty budget, covariance analysis, or check against other simulation parameters such as energy-loss straggling, interface roughness, or TRBS's multiple-scattering cutoff. This matters because the MS background intensity is sensitive to both the potential and the energy-dependent cross sections along the trajectory, so c_a and S_e are not cleanly separable. The reader's weakest assumption, that the MS background is dominated by 10-20 degree collisions, is real but secondary: it mainly limits transferability to other geometries and energies, whereas the missing error propagation affects the validity of the central uncertainty claim even for the investigated conditions. Since the concern is about missing evidence rather than a demonstrated fatal flaw, the appropriate verdict remains CONDITIONAL, unchanged from the reader.","tokens_in":11961,"tokens_out":5229,"duration_ms":63305,"concrete_test":"Re-analyze the 30 keV HfN spectrum with TRBS in a two-parameter fit: scan S_e and c_a over a grid (e.g., S_e ±10%, c_a 0.8-1.0) and compute a chi-square or negative-log-likelihood surface over the full spectrum, instead of fitting width and background sequentially. Report the joint 1-sigma and 2-sigma confidence regions. If the region's projection onto S_e exceeds ±3%, or if the region contains the line c_a = 1.0 for some S_e, then the screening correction is not uniquely determined and both the <3% quantitative-uncertainty claim and the statement that the potential overestimates the MS background need revision. A useful cross-check is to fit the same spectrum with c_a fixed at 1.0 while adding an energy-loss straggling parameter or interface roughness; if that model matches the experimental spectrum within the noise, the evidence for a potential correction is not unique.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central <3% uncertainty figure rests on the difference between TFM and ZBL fits and on screening-length corrections c_a extracted by matching the multiple-scattering background (Sec. 3.1.1, Figs. 1 and 3). Two issues make this load-bearing. First, no statistical or systematic uncertainties are reported for the fitted c_a or for the corresponding electronic stopping S_e; the paper does not show a joint fit or confidence region. Because changing S_e by ±10% strongly alters the Hf-peak width (Fig. 3) and also changes the energy at which deeper-scattered particles contribute, c_a and S_e can be partially degenerate: a larger c_a (weaker potential) might be compensated by a different S_e and still reproduce the measured spectrum. Without quantifying this covariance, the statement that the potential overestimates the multiple-scattering background is an interpretation, not a demonstrated identification. Second, even if c_a is physically meaningful, the conversion to a potential uncertainty and then to <3% in quantitative analysis assumes the MS background is dominated by 10-20° collisions (Sec. 3.1.1, Fig. 4); the x-error bars given cover only the angle range, not the resulting uncertainty in S_e or thickness. The difference between TFM and ZBL (~3.4% in S_e) is a model-spread estimate, not a measurement uncertainty, and could be larger or smaller than the true potential uncertainty.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a comparison of time-of-flight MEIS spectra for 25-100 keV He backscattered from a thin HfN film and for B+ backscattered from a TiN/W/Si stack with Monte-Carlo simulations using the TRBS code and either the Thomas-Fermi-Moliere (TFM) or the Universal (ZBL) screened Coulomb potential. The authors find that at these low energies both potentials overestimate the multiple-scattering background, with TFM giving a better overall description; at 30 keV a linear screening-length correction ca=0.87 is required to reproduce the measured background. They argue that this potential uncertainty translates into a systematic uncertainty of about 1-2% in spectrum-width determinations (thickness, electronic stopping) and about 3% in the extracted electronic stopping when comparing the two potentials, leading to the abstract's claim of \"<3% uncertainty in quantitative analysis.\" For transmission geometry they find negligible influence of the potential on peak positions for He projectiles, but more pronounced effects for Ne and for simulated implantation profiles.","tokens_in":12291,"tokens_out":5427,"duration_ms":58429,"significance":"The manuscript addresses a practical and timely question for ion-beam analysis: how strongly do uncertainties in the interatomic scattering potential affect quantitative MEIS/LEIS results? Its strengths include the use of independently characterized thin films, a series of primary energies for He, layer-resolved TRBS simulations, and explicit cross-checks through cutoff-angle variation and collision-number decomposition. The finding that backscattering geometries effectively select trajectories that suppress the influence of potential uncertainties is a useful and nontrivial message for practitioners. However, the headline uncertainty figure (<3%) is derived from model spread and sequential fits rather than a formal error propagation, and the paper would benefit substantially from a quantitative sensitivity analysis to support the central quantitative claim.","major_comments":[{"comment":"The screening correction ca=0.87 is obtained by a sequential fitting procedure: first the electronic stopping Se is adjusted to match the width of the Hf peak, then ca is varied to reduce the multiple-scattering background. No uncertainties are reported for either parameter, and no joint confidence region is shown. Because changing Se also alters the energy at which deeper-scattered particles contribute to the background, ca and Se may be partially degenerate; the statement that the measured background is lower than predicted by the uncorrected potential depends on this identifiability. Please quantify the covariance, for example by a contour plot of the fit residuals as a function of both ca and Se, or otherwise demonstrate that the extracted ca is uniquely constrained.","section":"Section 3.1.1, Fig. 3"},{"comment":"The conversion of fitted screening corrections into interaction distances assumes that the multiple-scattering background is dominated by collisions with deflection angles around 15 degrees (range 10-20 degrees). The supporting evidence (cutoff-angle tests and the collision-number decomposition in the inset) is plausible but indirect. The x-error bars show only the resulting uncertainty in the distance of closest approach; no uncertainty is propagated to the derived Se or thickness values. Consequently, the stated <3% uncertainty is an estimate based on the spread between TFM and ZBL, not a measurement uncertainty. The paper should either propagate the angle-range and fitting uncertainties through the quantification or explicitly label the <3% figure as a model-spread estimate rather than a measured uncertainty.","section":"Section 3.1.1, Fig. 4"},{"comment":"The abstract claims 'resulting in an uncertainty of <3% in quantitative analysis', but the quantitative basis in the paper is the ~3.4% difference in electronic stopping between TFM and ZBL at the lowest energy, together with a 1-2% ambiguity in defining the spectrum width. These contributions are not combined through any error propagation, and the relationship between 'difference between two potentials' and 'uncertainty' is not defined. Please provide the missing propagation, or alternatively rephrase the claim as 'on the order of a few percent' with the caveat that this is an estimate based on the spread between two screened potentials.","section":"Abstract and Section 4"}],"minor_comments":[{"comment":"The section numbering jumps from 3.2 (Transmission simulations) to 3.4 (Ion implantation); there is no Section 3.3.","section":"Section 3.2/3.4"},{"comment":"The caption contains the placeholder text 'Fig. Error! Unknown switch argument.' and should be corrected.","section":"Figure 6 caption"},{"comment":"The phrase 'These measurements were performed in double transmission geometry' is unclear for the B+ marker experiment; the beam traverses the TiN film twice, so the wording should be clarified (e.g., 'with the beam traversing the TiN film on the way in and out').","section":"Section 4"},{"comment":"The clause 'as B electronic stopping is still larger' should read 'For B, electronic stopping is still larger'.","section":"Section 4"},{"comment":"The abstract states 'uncertainty of <3%' while the conclusion states '~3% for the lowest investigated energy'; these numbers should be made consistent or the difference explained.","section":"Abstract vs. Section 4"}],"recommendation":"major_revision","confidential_remarks":"The experimental and simulation content appears sound, and the core observation that the screened potentials overestimate the multiple-scattering background is well supported. The main gap is that the central <3% uncertainty claim is presented more strongly than the analysis justifies; a revision with a proper two-parameter sensitivity analysis and a more careful wording of the uncertainty estimate would bring the paper to an acceptable level. No issues with novelty or citation practice were noted; the section numbering and figure caption glitches are minor and easily corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a useful, workmanlike sensitivity study that quantifies something the MEIS/LEIS community has long waved at: the choice of screened potential changes the multiple-scattering background and, at the few-percent level, the extracted thickness and electronic stopping. It doesn't reshape the field, but it gives practitioners a concrete number to cite.\n\nWhat's actually new: a systematic experimental plus simulation comparison of TFM and ZBL potentials in both backscattering and transmission geometries, for He and Ne, with the notable result that light ions in transmission are nearly insensitive to the potential while heavier ions are not. The mapping of screening corrections onto closest-approach distances and the connection to established LEIS correction factors is a nice touch and gives the paper a practical handle beyond the two specific samples.\n\nThe central claim holds up: both potentials overestimate the multiple-scattering background, and the resulting systematic effect on quantitative analysis is on the order of a few percent for the investigated conditions. The data support that; the slab decomposition in Fig. 2 and the cutoff-angle checks in Fig. 4 are good cross-validation.\n\nThe soft spots are real but not load-bearing. The headline <3% figure is an order-of-magnitude estimate, not a propagated uncertainty. The fitted screening corrections and electronic stopping values come without error bars, and there is no joint covariance analysis. The stress-test worry that ca and Se could be degenerate is partly answered by Fig. 3, where stopping mainly shifts the peak width and screening mainly changes the background, but a joint confidence region would have settled it cleanly. Also, only two samples, and no code or data release. None of this contradicts the main message.\n\nWho's this for? Ion-beam analysts who use MEIS or LEIS for thin-film metrology and want a defensible systematic-uncertainty term. It deserves a serious referee; the niche is small but the paper is careful within its scope. I'd ask the authors to add a paragraph acknowledging the unquantified covariance and, if easy, a simple joint-fit plot, but I wouldn't demand new experiments.","headline":"A useful sensitivity study that quantifies how screened-potential choice shifts MEIS/LEIS results by a few percent, with the caveat that the headline uncertainty is an estimate without formal error propagation.","tokens_in":12739,"tokens_out":1737,"would_cite":true,"duration_ms":21659,"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":"For backscattered He ions at tens of keV, standard screened Coulomb potentials overestimate the multiple-scattering background, yet the resulting systematic uncertainty in quantitative analysis stays below 3 percent.","keywords":["medium-energy ion scattering","screened Coulomb potential","multiple scattering","screening length correction","electronic stopping","quantitative ion beam analysis","Monte Carlo simulation","trajectory selection"],"falsifier":"Take a thin film of known thickness and record its backscattered-ion spectrum at several primary energies from 20 keV up to 100 keV and at several detector angles. Fit each spectrum with the same screening-length correction factor ca: if the ca values that reproduce the multiple-scattering background drift by more than the quoted 3 percent with energy or angle, the single-factor potential description fails; conversely, if one ca reproduces all spectra, the transferability claim holds.","tokens_in":11797,"feed_emoji":"⚛️","tokens_out":7115,"duration_ms":65716,"temperature":0.7,"pith_summary":"This paper asks whether uncertainties in the screened Coulomb potential that governs keV ion-atom collisions corrupt quantitative analysis by ion scattering. For backscattered He ions with tens of keV primary energy, the answer is no in practice: the standard TFM and ZBL potentials both overestimate the multiple-scattering background, yet the resulting systematic uncertainty in extracted thickness, composition, and electronic stopping stays below 3 percent for the conditions investigated. The robustness comes from trajectory selection, since the analytical high-energy edge is set by single-scattering collisions at small impact parameters, where the potential is well known, while the uncertain part of the potential affects mainly the background. Transmitted light ions are even less sensitive when electronic stopping dominates, and heavier ions such as Ne become the case where the potential choice matters.","feed_headline":"Scattering-potential errors keep keV-ion analysis under 3%","feed_subtitle":"Even when screened potentials overshoot the background, thickness and stopping values stay accurate to about 3 percent.","key_machinery":"The mechanism carrying the analysis is the decomposition of a simulated backscattering spectrum by depth slab and by number of collisions, compared with the experimental spectrum in two regions: the high-energy edge, where a single backscattering collision dominates and normalization is safe, and the low-energy multiple-scattering background, where the potential at large impact parameters is tested. A simulation with a 15-degree cutoff shows that the background is mostly built from projectiles that suffered at least three collisions, and an angular window of 10-20 degrees defines the relevant distances of closest approach. A single linear screening-length correction factor ca, inserted into the screened Coulomb potential, is varied until the background intensity matches; the fitted ca values are then mapped onto distances of closest approach to compare different energies and geometries.","core_discovery":"The central claim is that for He projectiles at 25-100 keV, Monte-Carlo simulations using either the TFM or the ZBL screened Coulomb potential reproduce single scattering at the high-energy edge but overestimate the multiple-scattering background, meaning the real potential is weaker at large impact parameters than both models. Fitting the background fixes a linear screening-length correction of ca = 0.87 at 30 keV, and the series of spectra shows the correction factor decreasing as the probed distance of closest approach grows. Nevertheless, the quantities that ion-beam analysis wants, thickness, composition, and electronic stopping, change by only 1-3 percent across the plausible potential range, because they are extracted from spectral features that single scattering dominates. The same trajectory-selection argument explains why transmitted He is nearly unaffected by potential changes while transmitted Ne, for which nuclear stopping dominates, responds at a level comparable to a 10 percent change in electronic stopping.","pith_inferences":["Beyond the paper, the trajectory-selection principle implies that any ion-beam method that deliberately rejects multiply scattered events inherits this robustness, while methods such as implantation or radiation-damage modeling that integrate all trajectories do not benefit from it.","The decreasing correction factors toward lower energies suggest that at low-energy ion scattering energies a linear ca may saturate or break down; a direct angular-resolved measurement at few-keV energies would test whether the potential correction is non-linear in distance.","A testable extension is to apply the same background-fitting scheme to films of different thicknesses at a fixed energy: the paper's analysis predicts that the composite background fraction and the fitted ca should be essentially thickness-independent, separating potential error from straggling and interface roughness."],"forward_implications":["Thicknesses, compositions, and electronic stopping values extracted from backscattered He spectra at tens of keV carry a systematic uncertainty below 3 percent from the choice of potential.","For light ions in transmission, screening uncertainties are negligible when electronic stopping dominates, so energy-loss values from transmission experiments need no potential correction.","For heavier projectiles such as Ne in transmission, nuclear stopping dominates and screening corrections shift the peak position by roughly 0.6 keV per 0.1 change in ca, comparable to a 10 percent change in electronic stopping, so the potential uncertainty cannot be ignored there.","In marker experiments using B+ on a TiN/W stack, screening corrections do not shift the W peak position, so energy-loss determinations from delta-layer peak positions remain robust.","Implantation depth profiles are more sensitive: a 10 percent change in screening or electronic stopping moves the He profile maximum by about 5 percent of the range, and for Ne the screening change alone moves it by roughly 10 percent."],"supporting_citations":[{"why":"defines the Thomas-Fermi-Moliere screened potential used as one of the two interaction models in the simulations.","marker":"[17]"},{"why":"provides the ZBL universal potential and the SRIM electronic stopping values that anchor the simulations.","marker":"[18]"},{"why":"supplies the Monte-Carlo simulation code whose spectra are compared with experiment throughout the study.","marker":"[32]"},{"why":"gives previously published screening-length corrections from LEIS, against which the extracted corrections are compared.","marker":"[22]"},{"why":"justifies normalizing simulations at the high-energy edge, where single scattering dominates.","marker":"[33]"},{"why":"establishes the 1-2 percent systematic uncertainty baseline for quantification at low keV energies.","marker":"[34]"},{"why":"shows how trajectory selection via impact parameter affects energy-loss measurements, underpinning the transmission analysis.","marker":"[29]"}],"fun_headline_variants":["He ion beams: potential uncertainties cause only <3% error in analysis","Potential uncertainties limit keV-ion analysis error to <3% for He backscatter","Scattering potential uncertainty: keV ion analysis still <3% accurate","Light ions not bothered by potential errors: keV analysis under 3%","For He at tens of keV, potential uncertainties yield <3% analysis error"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire uncertainty estimate assumes that the multiple-scattering background is dominated by collisions with deflection angles around 10 to 20 degrees, so one fitted screening factor represents the potential at the interaction distances that matter; if a different angular range dominates elsewhere, the correction factors need not transfer.","fun_headline_variants_meta":{"raw":{"variants":["He ion beams: potential uncertainties cause only <3% error in analysis","Potential uncertainties limit keV-ion analysis error to <3% for He backscatter","Scattering potential uncertainty: keV ion analysis still <3% accurate","Light ions not bothered by potential errors: keV analysis under 3%","For He at tens of keV, potential uncertainties yield <3% analysis error"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000806,"raw_usage":{"total_tokens":3488,"prompt_tokens":842,"completion_tokens":2646,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":2548}},"tokens_in":458,"tokens_out":2646,"duration_ms":18792,"temperature":1.0,"reasoning_tokens":2548,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:54:56.775118+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a thin film of known thickness and record its backscattered-ion spectrum at several primary energies from 20 keV up to 100 keV and at several detector angles. Fit each spectrum with the same screening-length correction factor ca: if the ca values that reproduce the multiple-scattering background drift by more than the quoted 3 percent with energy or angle, the single-factor potential description fails; conversely, if one ca reproduces all spectra, the transferability claim holds.","supporting_citations":[{"cited_title":"Molière, Theorie der {Streuung} schneller geladener {Teilchen} {I}","cited_arxiv_id":null,"evidence_quote":"defines the Thomas-Fermi-Moliere screened potential used as one of the two interaction models in the simulations."},{"cited_title":"Ziegler, J.P","cited_arxiv_id":null,"evidence_quote":"provides the ZBL universal potential and the SRIM electronic stopping values that anchor the simulations."},{"cited_title":"Primetzhofer, E","cited_arxiv_id":null,"evidence_quote":"supplies the Monte-Carlo simulation code whose spectra are compared with experiment throughout the study."},{"cited_title":"Takeuchi, Evaluation of scree ning length corrections for interaction potentials in impact - collision ion scattering spectroscopy, Nucl","cited_arxiv_id":null,"evidence_quote":"gives previously published screening-length corrections from LEIS, against which the extracted corrections are compared."},{"cited_title":"Biersack, E","cited_arxiv_id":null,"evidence_quote":"justifies normalizing simulations at the high-energy edge, where single scattering dominates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"establishes the 1-2 percent systematic uncertainty baseline for quantification at low keV energies."},{"cited_title":"Sigmund, A","cited_arxiv_id":null,"evidence_quote":"shows how trajectory selection via impact parameter affects energy-loss measurements, underpinning the transmission analysis."}],"review_version":1}