{"id":"f37cba5f-d0d2-4c0d-bfd5-5f76937a85ae","arxiv_id":"2607.21298","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A comprehensive spectroscopic reference for platinum, correlating core-level satellites with electron energy-loss features and showing that spin-orbit coupling is needed to reproduce the measured valence band.","lead":"This paper measures platinum's full X-ray photoelectron spectrum and explains the extra satellite peaks by matching them to electron energy-loss features such as plasmons and interband transitions. The result is a labelled reference that anyone using X-ray photoelectron spectroscopy on platinum catalysts can use to avoid misreading those satellites.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"RHEELS-to-satellite matching uses flexible offsets and contains internal inconsistencies; the claimed internally consistent assignment is not yet quantitatively supported.","rationale":"The reader's weakest assumption—that RHEELS loss features correspond one-to-one in energy to photoemission satellites—is the right locus, and my concern sits within it. I partially agree because I focus not on the physical transferability of loss channels (which is a standard and defensible practice for metals, and which the paper addresses via screening arguments) but on the absence of a quantitative matching criterion and on specific internal contradictions that make the assignment chain overfit and underdetermined. The paper has real strengths: a well-characterized clean Pt surface, measurements at two photon energies, a prior W-metal methodology, literature comparisons, and open data on Zenodo. The line-width analysis and valence-band G0W0+SOC comparison are plausible and useful. However, the headline claims—'first complete description' and 'first internally consistent interpretation of satellite structure'—depend on the reliability of the satellite assignments. The 40.7 eV 'overtone' with no RHEELS counterpart, the 56.7 eV satellite assigned to a feature labelled differently in Table I, and the inconsistent surface-plasmon energy (24.5 vs 23.1 eV) are concrete signs that the matching is not yet disciplined. These are addressable with a re-fit and a statistical control, so the appropriate verdict remains CONDITIONAL as the reader concluded; the concerns reinforce the need for the conditions rather than overturning the paper's value.","tokens_in":26205,"tokens_out":7323,"duration_ms":78511,"concrete_test":"Build a prediction table from the RHEELS feature list in Table I: for each core level, predict satellite positions as BE_core + feature_energy for all features, and compare with the observed S_N separations using a tolerance of ±1.5 eV (the paper's stated offset range). Count the match rate. Then perform a Monte Carlo control (10^4 runs) in which the observed satellite positions are randomly permuted within the same energy window, keeping their multiplicities, to estimate the expected chance match rate. If the true match rate is not significantly above the random rate at p<0.05, the correlation-based assignments are not statistically supported. Separately, re-fit the Zenodo RHEELS spectrum to obtain unbiased peak positions for features d and g; if g is not within ±1 eV of 2d (using the same d value used for the overtone claim), then the 'plasmonic overtone' assignment for g should be rev","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central reference claim rests on matching RHEELS loss features to PES satellites, but the matching has no explicit decision rule. The paper permits variable offsets of 0.7–2.5 eV between RHEELS and HAXPES positions and invokes them post hoc, without modeling the shift or propagating uncertainty. Because there are ~11 RHEELS features and ~30 satellite positions across core levels, some coincidences are expected by chance; no statistical control is reported. More concretely, the assignments are not internally consistent: (1) Pt 4d S2 relative to 4d5/2 at 40.7 eV is labelled a 'plasmonic overtone' even though no RHEELS feature lies near 40.7 eV—the nearest are f=34.8 eV and g=46.7 eV, each >5 eV away. (2) In §III.B.1, Pt 4s S2 at 56.7 eV is said to correspond to '5p1/2 excitation at w = 59.5 eV (feature i)', but Table I assigns feature i at 59.5 eV to a plasmonic overtone and feature j at 68.2 eV to 5p1/2 ionisation; 56.7 eV is actually closer to h (53.4 eV, 5p3/2). (3) The surface-plasmon position is listed as d=24.5 eV in Table I, but §III.B.2 cites '≈23.1 eV' for the same loss in the Pt 3d analysis, and the overtone g=46.7 eV is justified as 'double that of the surface plasmon (d)', which would be 49 eV. These inconsistencies affect both the 'first complete description' of RHEELS features and the satellite assignments built on them, weakening the central claim of an internally consistent spectroscopic reference.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports a multi-technique spectroscopic study of polycrystalline Pt using RHEELS, SXPS (1.7 keV), and HAXPES (5.9 keV), supported by DFT and G0W0 calculations with and without spin-orbit coupling. Core-level binding energies, spin-orbit splittings, line widths, and satellite separations are tabulated; RHEELS loss features a–k are correlated with photoelectron satellites S1–S6 across the Pt 3s–5p core levels; and the valence band is compared with orbital-projected DOS using four photoionisation cross-section weighting schemes. The central claims are that this provides the first complete assignment of Pt characteristic energy-loss features and the first internally consistent interpretation of satellites across accessible Pt core levels, and that G0W0+SOC with a Pb-corrected 6p cross-section best reproduces the valence band.","tokens_in":26619,"tokens_out":4191,"duration_ms":43530,"significance":"If the satellite assignments are correct, the paper would give XPS and HAXPES users a valuable energy-loss-based reference for distinguishing loss satellites from chemical states in Pt, and the measured binding energies, spin-orbit splittings, and FWHM values are broadly consistent with prior literature. The data are made available on Zenodo, and the RHEELS/PES correlation is in principle non-circular because the two measurements are independent. However, the paper's central novelty—the unified satellite assignment—currently rests on qualitative matching with unquantified tolerances and contains internal inconsistencies. These issues must be resolved before the proposed reference can be relied upon for routine spectral interpretation.","major_comments":[{"comment":"In Table III, Pt 4d S2 relative to 4d5/2 is placed at 40.7 eV and assigned to a 'plasmonic overtone.' No RHEELS feature lies near 40.7 eV; the nearest features are f = 34.8 eV and g = 46.7 eV, each more than 5 eV away. Since overtones are defined as multiples or sums of the surface (24.5 eV) and bulk (34.8 eV) plasmons, the expected overtone positions would be around 49, 59, and 70 eV. The assignment is therefore not supported by the presented loss spectrum and needs to be either re-assigned or justified with a quantitative criterion.","section":"Section III.B.1, Table III"},{"comment":"The text assigns the Pt 4s S2 satellite at 56.7 eV to '5p1/2 excitation at w = 59.5 eV (feature i),' but Table I assigns feature i at 59.5 eV to a 'plasmonic overtone' and feature j at 68.2 eV to 5p1/2 ionisation. The text thus contradicts its own Table I. In addition, 56.7 eV is closer to feature h at 53.4 eV (5p3/2) than to feature i at 59.5 eV. This inconsistency affects the analogous Pt 3s S2 assignment and weakens the claimed internal consistency.","section":"Section III.B.1 (Pt 4s S2)"},{"comment":"The surface-plasmon position is given as d = 24.5 eV in Table I, but Section III.B.2 cites '≈23.1 eV' for the same loss when assigning the Pt 3d S1 satellite at 25.0 eV. Furthermore, feature g at 46.7 eV is justified as 'double that of the surface plasmon (d),' which would be 49.0 eV rather than 46.7 eV. The manuscript should use a single consistent loss-position table with uncertainties, and if probe-dependent shifts are expected, they should be modelled rather than invoked case-by-case.","section":"Section III.B.2 and Table I"},{"comment":"The matching of RHEELS loss features to PES satellites has no explicit decision rule. The paper accepts offsets of 0.7–2.5 eV for features h, j, and k, and similar flexible offsets appear in other assignments, without propagating the stated ±0.3–0.5 eV uncertainties or providing any statistical control. With 11 RHEELS features and roughly 30 satellite positions across core levels, coincidental matches within a few eV are expected. A quantitative matching criterion—for example, a tolerance tied to combined uncertainties, an energy-shift model, or a false-discovery assessment—is required to support the claim of an internally consistent satellite reference.","section":"Section III.B (general matching procedure)"},{"comment":"Approach (3) determines the 6p photoionisation cross-section by least-mean-square optimisation against the experimental HAXPES valence band, and this optimised value is then used as one of the four approaches in the ranking that concludes the Pb-correction approach is best. Since the LMS value is fit to the same data used for comparison, the comparison is partly circular. The authors should report the L2 residual, use cross-validation, or treat the LMS result only as a consistency check rather than as an independent ab initio alternative.","section":"Section II.C and Table IV"}],"minor_comments":[{"comment":"The Pt 4d rows are duplicated: the satellite list appears once under the combined 4d5/2/4d3/2 entries and again with separate rows. Merge these rows or label them clearly to avoid confusion.","section":"Table II"},{"comment":"The text reads 'FWHM ... were 8.2 and 11.5 eV for Pt 3p3/2 and 3p3/2.' The second entry should be 3p1/2.","section":"Section III.B.2"},{"comment":"The last row 'S1 352.9 55.3' appears to belong to Pt 3s but is separated from the 3s block; check the table formatting and column alignment.","section":"Table II, Pt 3s row"},{"comment":"The statement 'first complete description of the characteristic energy loss features from Pt metal' is strong given the extensive prior EELS literature summarised in Table S2. Consider softening the claim or explicitly stating what new assignments or completeness criteria are being introduced compared with Schröder et al. and Seignac and Robin.","section":"Section III.A"},{"comment":"The caption refers to SXPS data 'in orange' and HAXPES 'in grey,' but the printed greyscale version may not distinguish these clearly; add distinct line styles or labels in the figure panels.","section":"Figure 2 and caption"}],"recommendation":"major_revision","confidential_remarks":"The experimental dataset is valuable and the measured core-level parameters appear reliable, but the satellite assignment methodology is the core contribution and is not yet quantitatively supported. The internal inconsistencies listed in the major comments should be addressed in a revised version before the manuscript can be accepted as a reference work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a genuinely useful experimental reference for Pt: core-level binding energies, FWHMs, and spin-orbit splittings across SXPS and HAXPES, plus a multi-energy RHEELS dataset and deposited data on Zenodo. The first complete 5p spin-orbit splitting is a real, citable result, and the valence-band comparison with G0W0 + SOC makes a convincing case that relativistic effects are required. If you do XPS or HAXPES on Pt catalysts, the tables alone are worth keeping on your desk.\n\nThe weak part is the satellite-assignment narrative. The paper matches RHEELS loss features to photoelectron satellites with no explicit decision rule, tolerating offsets of 0.7–2.5 eV, and with roughly 11 loss features and 30 satellite positions, some coincidences are inevitable. No statistical control is reported. That alone would make me cautious, but there are also concrete internal contradictions in the text. Pt 4s S2 at 56.7 eV is assigned to “5p1/2 excitation at w = 59.5 eV (feature i)”, yet Table I labels feature i as a plasmonic overtone and assigns the 5p1/2 ionisation to feature j at 68.2 eV. Pt 4d S2 at 40.7 eV is called a plasmonic overtone even though the nearest RHEELS features are f = 34.8 and g = 46.7 eV. The surface plasmon is listed as d = 24.5 eV in Table I, but the 3d analysis cites ≈23.1 eV, and overtone g = 46.7 eV is justified as “double” the surface plasmon, which would be 49 eV. Those are not cosmetic issues; they are load-bearing for the central claim of an internally consistent interpretation.\n\nThe LMS cross-section weighting (Approach 3) is fitted to the same spectrum it is later compared with, so the ranking of approaches is partly circular. It is a minor issue because the main conclusion favors the Pb-correction approach, but it should be disclosed as a fitting exercise rather than a prediction.\n\nWhat is solid: the core-level positions match prior literature well, the line-width trends are sensible, and the RHEELS assignments to interband transitions and plasmons are mostly consistent with older EELS work. The satellite correlation is plausible in outline but needs either a systematic matching procedure with uncertainty propagation, or an explicit downgrade to “provisional”. The internal inconsistencies should be corrected before publication.\n\nThis paper deserves a serious referee. The dataset is valuable, the problems are fixable, and the intended audience—XPS/HAXPES users studying Pt—will cite it. I would accept peer review, require the satellite assignments to be tightened or relabeled, and recommend citing the reference tables while treating the satellite scheme as suggestive until revised.","headline":"Solid Pt reference dataset worth having, but the satellite-assignment framework is too loose and contains internal inconsistencies that need fixing before the 'internally consistent' claim can stand.","tokens_in":27172,"tokens_out":2299,"would_cite":true,"duration_ms":23717,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["79.60.-i","79.20.Uv","71.45.Gm","71.15.Mb"],"model":"deepseek-v4-flash","headline":"This paper establishes a unified, internally consistent reference for the photoelectron spectrum of metallic platinum by assigning every core-level satellite to a specific electron energy-loss event.","keywords":["platinum metal","photoelectron spectroscopy","HAXPES","RHEELS","plasmon satellites","spin-orbit coupling","density of states","energy-loss assignment"],"falsifier":"Take the same Pt foil and record the 4f satellites with HAXPES at several photon energies between 2 and 8 keV while also collecting a momentum-resolved loss spectrum on a Pt single crystal; if the 7.1 eV satellite has no counterpart in the interband loss region at any momentum value, or if the RHEELS-to-HAXPES offsets change systematically with photon energy, the one-to-one loss dictionary underlying the assignment is wrong.","tokens_in":1450,"feed_emoji":"⚛️","tokens_out":4331,"duration_ms":106677,"temperature":0.7,"pith_summary":"Metallic platinum emits photoelectron spectra whose main peaks carry tails of extra 'satellite' features; these are often mistaken for chemical states in catalyst studies. This paper argues that every such satellite has a mechanical origin: a specific way the outgoing electron loses energy, such as kicking a bulk or surface plasmon, driving an interband transition, or ionising a 5p or 4f shell. The authors measure those loss energies independently with high-energy electron scattering, then use them as a dictionary to label the satellites seen in soft- and hard-X-ray photoemission across all accessible Pt core levels. They also show that reproducing the platinum valence band requires spin-orbit coupling in the calculations and a realistic, small 6p orbital weight. If the mapping is correct, platinum's photoelectron spectra become a readable reference rather than a source of false chemical states.","feed_headline":"Platinum's XPS satellites traced to one energy-loss map","feed_subtitle":"A new reference separates loss satellites from chemical shifts, so Pt catalysts can be read correctly in XPS.","key_machinery":"The carrying mechanism is the RHEELS loss spectrum conceived as a dictionary: eleven labelled loss features (a-k) at 4.8, 13.4, 16.4, 24.5, 31.0, 34.8, 46.7, 53.4, 59.5, 68.2 and 71.3 eV, assigned to interband transitions, surface and bulk plasmons, plasmonic overtones, and 5p/4f ionisation, are used one-to-one to interpret the S1-S6 satellites riding on each core-level photoemission line. The same dictionary, combined with spin-orbit-resolved densities of states, anchors the valence-band comparison, with orbital photoionisation cross-sections adjusted so the 6p contribution is realistic.","core_discovery":"Using RHEELS, soft and hard X-ray photoelectron spectroscopy, and DFT/G0W0 calculations, the paper establishes what it calls the first complete description of Pt's energy-loss features and the first internally consistent interpretation of satellites across its accessible core levels. Eleven RHEELS features are assigned to interband transitions, surface and bulk plasmons, plasmonic overtones, and 5p/4f ionisation losses; these loss energies are then matched to the S1-S6 satellites in the 4s, 4p, 4d, 4f/5s/5p, 3s, 3p and 3d core lines. It also reports the previously missing Pt 5p1/2 binding energy and 5p spin-orbit splitting and shows that spin-orbit coupling is necessary for either DFT or G0W","pith_inferences":["If the loss dictionary is transferable, then in Pt alloys the satellite pattern should shift rigidly with the main photoemission line, so a constant offset between pure Pt and alloy Pt would confirm the assignment while a changing pattern would signal new chemical or final-state effects.","The 0.7-2.5 eV offsets between RHEELS and HAXPES positions for the semi-core ionisation features could encode the core-hole screening energy; compiling similar offsets for other metals might yield a systematic screening correction for comparing electron-loss and photoemission energies.","Satellite visibility appears to depend on the core level's angular momentum (surface plasmon losses are seen for p and d levels but not s levels); this selection-rule-like effect could be tested in other 5d metals, though the paper does not establish a mechanism."],"forward_implications":["Researchers measuring Pt-containing catalysts can now distinguish loss-induced satellites from chemically shifted peaks, reducing the chance of misassigning Pt oxidation states in XPS and HAXPES.","The first reported Pt 5p1/2 binding energy and the 5p spin-orbit splitting of 15.6 eV fill a gap in reference data for platinum.","Pt 3d is proposed as a practical HAXPES core level for chemical-state analysis because of its manageable lifetime width and higher photoionisation cross-section than 4f.","Valence-band modelling of platinum requires spin-orbit coupling, and the 6p contribution is much smaller than raw atomic cross-sections suggest.","The RHEELS-to-satellite correlation procedure, previously applied to tungsten, is transferable to other 5d transition metals."],"fun_headline_variants":["Pt XPS satellites traced to a single loss spectrum","One energy-loss map decodes platinum's XPS satellites","Platinum's missing 5p spin-orbit splitting resolved","Unified Pt reference: losses explain every satellite","Pt photoelectron satellites mapped to plasmon losses"],"cache_read_input_tokens":28416,"weakest_assumption_plain":"The load-bearing premise is that the energy losses seen when an incoming electron scatters off platinum in RHEELS are the same losses that appear as satellites in photoemission, where a photoelectron and a core hole are present; the paper itself notes 0.7-2.5 eV offsets, so if any loss channel shifts or is absent between the two probes, the corresponding satellite assignment fails.","fun_headline_variants_meta":{"raw":{"variants":["Pt XPS satellites traced to a single loss spectrum","One energy-loss map decodes platinum's XPS satellites","Platinum's missing 5p spin-orbit splitting resolved","Unified Pt reference: losses explain every satellite","Pt photoelectron satellites mapped to plasmon losses"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00014,"raw_usage":{"total_tokens":1013,"prompt_tokens":778,"completion_tokens":235,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":522,"completion_tokens_details":{"reasoning_tokens":159}},"tokens_in":522,"tokens_out":235,"duration_ms":3390,"temperature":1.0,"reasoning_tokens":159,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T07:51:06.807036+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the same Pt foil and record the 4f satellites with HAXPES at several photon energies between 2 and 8 keV while also collecting a momentum-resolved loss spectrum on a Pt single crystal; if the 7.1 eV satellite has no counterpart in the interband loss region at any momentum value, or if the RHEELS-to-HAXPES offsets change systematically with photon energy, the one-to-one loss dictionary underlying the assignment is wrong.","supporting_citations":[],"review_version":1}