{"id":"02da7bb8-ba01-43d0-adad-91f2f571c621","arxiv_id":"2607.19698","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"An analytical inversion of the modified Kratzer model extracts bandgap, screening length, exciton mass, and dielectric constant of monolayer TMDCs from three exciton energies plus magnetoexciton shifts.","lead":"Measured exciton line positions and their magnetic-field shifts can be converted, via closed-form formulas, into the bandgap, screening length, exciton mass, and dielectric constant of monolayer TMDCs. The method is demonstrated on published spectra of WSe2, WS2, MoS2, MoSe2, and MoTe2 in different dielectric environments.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Table I's Chen et al. row shows E4 predicted at 1.893 eV versus the tabulated experimental 1.918 eV, 25 meV below experiment and also below the retrieved bandgap Eg=1.906 eV; this unaddressed discrepancy contradicts the claimed parameter-free prediction of higher Rydberg states.","rationale":"The reader’s weakest assumption targets the calibrated constants g^2 and c. I agree those matter, but I found a more direct, unaddressed discrepancy in the paper’s own validation table. The inversion algebra in Eqs. (7)–(12) is internally self-consistent: the printed formulas reproduce the table entries (e.g., Liu E1–E3 yields r0≈4.09 nm), so the Chen E4 mismatch is not a trivial arithmetic typo. It is a concrete counterexample to the “predict higher Rydberg states” part of the central claim. The severity is bounded: this is one dataset and the experimental assignment could in principle be disputed, so the appropriate final verdict remains conditional rather than rejection. Hence the reader’s CONDITIONAL verdict is retained; no change is needed. The reader and I partially agree because both concern validation, but the specific load-bearing weakness I identify is different from the g^2/c calibration issue.","tokens_in":947,"tokens_out":1122,"duration_ms":216308,"concrete_test":"Check the original Chen et al. (2019) paper to confirm whether the 1.918 eV line is the 4s state. If it is, attempt to fit all four zero-field energies E1–E4 with the modified Kratzer formula; if no (Eg, η, ξ) set reproduces them within experimental accuracy, the model’s Rydberg series is too compressed and the predictive claim fails for this dataset. If it is not the 4s state, the table’s “E4” label is misleading and should be corrected.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Table I lists for the Chen et al. hBN-encapsulated WSe2 dataset four measured s-state exciton energies (1.727, 1.858, 1.884, 1.918 eV). The retrieval uses E1–E3 and then “predicts” E4=1.893 eV and E5=1.898 eV. The E4 prediction is 25 meV below the experimental 1.918 eV, and the retrieved bandgap Eg=1.906 eV is actually below the measured 4s energy, so the model would regard the tabulated 4s state as unbound. The text discusses only the Molas dataset as having the “largest deviations” and never mentions this Chen mismatch. This is not a calibration subtlety: it is a direct failure of the central claim that higher Rydberg states can be predicted without additional fitting. It also indicates that the one-parameter Kratzer inversion fitted to E1–E3 compresses the upper Rydberg series for a nominally optimal hBN/hBN sample. The paper should either correct the table, revise the prediction claim, or explain why the 1.918 eV line is not the 4s state. I also note that the prose definition of λ before Eq. (7) appears inverted relative to Eq. (8), though the displayed formulas are internally self-consistent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an analytical two-stage procedure to retrieve quasiparticle bandgap, screening length, reduced exciton mass, and dielectric constant of monolayer TMDCs from the three lowest measured s-state exciton energies and from magnetoexciton energy shifts. The first stage inverts a modified Kratzer model (Molas et al.) to obtain ξ, η, and Eg directly, then converts ξ to the screening length r0 using a universal coefficient g^2 = 0.205. The second stage uses an analytical magnetoexciton interpolation (Eq. 14) with a calibrated crossover constant c = 0.905 to retrieve μ from magnetic-field data, and then κ from η. The framework is applied to a wide range of experimental datasets for WSe2, WS2, MoS2, MoSe2, and MoTe2 in different dielectric environments. The retrieved parameters are compared with literature RK values and with the authors' earlier RK-based analytical retrieval, and the predicted diamagnetic coefficients, radii, and magnetoexciton spectra are compared with experiment and numerical solutions.","tokens_in":23852,"tokens_out":3164,"duration_ms":37413,"significance":"If the central claim were fully substantiated, the paper would offer a valuable, computationally inexpensive alternative to numerical RK fitting for extracting material parameters from excitonic spectra. The analytical inversion formulas are explicit and the algebra is internally coherent. The authors also provide independent-looking tests of predictive quantities (diamagnetic coefficients, radii, magnetoexciton spectra) that do not enter the retrieval. However, the strength of the claim is moderated by two calibrated constants that directly enter the retrieved r0 and μ, by the absence of propagated uncertainties, and by at least one clear predictive failure (Chen et al. E4 in Table I). The validation is also largely against the authors' own earlier RK retrieval rather than fully independent benchmarks. These issues are load-bearing for the paper's central 'parameter-free prediction' claim, but they are addressable in revision.","major_comments":[{"comment":"The Chen et al. hBN-encapsulated WSe2 dataset lists measured E4 = 1.918 eV, but the retrieval using E1–E3 predicts E4 = 1.893 eV, a 25 meV discrepancy, and also yields Eg = 1.906 eV, which is below the measured E4. Thus the model would treat the measured 4s state as unbound. The text only discusses the Molas dataset as having the largest deviations and does not address this failure. This directly contradicts the claim that higher Rydberg states are predicted without additional fitting. The authors should either correct the table, explain why 1.918 eV is not the 4s state, or explicitly restrict the prediction claim.","section":"Table I, Chen et al. row"},{"comment":"The retrieval of r0 uses the universal coefficient g^2 = 0.205, and the retrieval of μ uses the calibrated crossover factor c = 0.905 in the cubic Eq. (24). These constants are fitted to the same class of materials/experiments that the retrieval is applied to, so the 'without additional fitting parameters' claim in the abstract and §II D is overstated: the model itself has two fitted parameters, and any inaccuracy in their universality (as the authors acknowledge for g^2 in §III C) directly biases r0 and μ. The paper should provide a sensitivity analysis (e.g., how r0 and μ change when g^2 ranges over 0.195–0.216 and c over ±5%) and should clearly distinguish model calibration from parameter-free prediction.","section":"§II B and §II C, Eqs. (12) and (24)"},{"comment":"The validation is largely against the authors' own earlier RK-based retrieval (rows labeled 'Retrieval [32]') rather than fully independent benchmarks. For several datasets the agreement with 'Retrieval [32]' is excellent, but this is not an independent check because the same experimental inputs and similar analytical approximations are used. The paper should more clearly separate independent experimental/theoretical benchmarks from the authors' own previous results, and should include propagated experimental uncertainties for the retrieved quantities (currently no error bars are given for Eg, r0, μ, κ in the 'This work' rows).","section":"Tables I–III and §III A"},{"comment":"Even outside the Chen case, the paper's predictive claim for E4 and E5 rests on the assumption that the modified Kratzer model with a single universal g^2 reproduces the Rydberg series for every material and environment. The paper does not quantify the accuracy of this assumption across the full table; for example, in the Chernikov WS2 row the predicted E4 (2.329 eV) is 15 meV below the experimental 2.344 eV. A systematic table of residuals (predicted minus measured) for all available higher states would allow the reader to assess the actual predictive power, and would likely require a more cautious statement in the conclusion.","section":"§III A, Chen et al. and other Rydberg predictions"}],"minor_comments":[{"comment":"The prose defines λ = (2ξ+1)δ/(2ξ+3), but solving that relation for ξ gives (δ−3λ)/(2(λ−δ)), not the displayed Eq. (8) ξ = (3δ−λ)/(2(λ−δ)). The displayed formulas appear self-consistent, but the textual definition should be corrected or reconciled with Eq. (8).","section":"Before Eq. (7)"},{"comment":"Several typos and unfinished cross-references remain: 'T ransition' in the title/abstract, 'Tabe' in the text, 'detailled', 'konstant', 'reponsible', 'root-meam-square', and 'Subsecs. III A and ??'. Figure captions 3 and 4 refer to 'Eq. (??)' instead of Eq. (14).","section":"General"},{"comment":"The data availability statement says data are available 'upon reasonable request.' For a paper whose central contribution is a quantitative retrieval procedure, providing the numerical values of all retrieved parameters (with uncertainties) and the per-dataset residuals in a machine-readable table or repository would strengthen reproducibility.","section":"Data availability"},{"comment":"The symbol ξ is called 'effective screening parameter' in Eq. (1), but the relation to the physical screening length r0 in Eq. (3) is model-dependent. The paper would benefit from an explicit statement that ξ is not directly the RK screening length, particularly for readers applying the formulas to new materials.","section":"Notation"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of cond-mat.mtrl-sci and offers a useful analytical tool. However, the central claim of parameter-free prediction is weakened by the fitted constants g^2 and c, and the Chen et al. E4 discrepancy is a concrete predictive failure that should be addressed head-on. The validation would be much more convincing if it included independent benchmarks and error propagation. I recommend major revision rather than rejection because the inversion algebra is sound and the framework may become publishable with an honest discussion of its limitations."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the explicit inversion formulas in Eqs. (8)–(11) and (24) are a legitimate new tool, and the paper is mostly careful about comparing with RK calculations and experiments. But the headline claim—higher Rydberg states predicted without extra fitting—has a clear counterexample in Table I. For the Chen et al. hBN/WSe2 dataset, E4 is predicted at 1.893 eV against a measured 1.918 eV, 25 meV off; worse, the retrieved Eg is 1.906 eV, which is above the predicted E4/E5 and below the measured 4s line. A model that puts the 4s state above its own continuum is not predicting that series, and the paper never mentions this row when discussing deviations. That needs to be addressed, not buried.\n\nWhat is actually new: previous work used the Kratzer model only for the forward spectrum; here the algebra is inverted. The two-stage retrieval—zero-field energies give Eg, ξ, η; magneto data give μ via the cubic—is clean and internally consistent. The survey across WSe2, WS2, MoS2, MoSe2, MoTe2 in different dielectric environments is useful, and the RK comparisons give a reasonable sanity check.\n\nThe soft spots beyond the Chen issue: the two constants g^2=0.205 and c=0.905 are fitted, not derived, and they directly enter r0 and μ. That does not kill the method—the paper is upfront about the ranges—but it does mean 'no additional fitting' is an overstatement; the same class of experimental spectra was used to calibrate the constants. There are no propagated error bars on the retrieved parameters, which is a problem for a paper whose business is retrieval from noisy data. No code or data is released, and some 'independent' checks are partly in-sample. The lambda definition before Eq. (7) also looks inverted relative to Eq. (8), though the displayed formulas are self-consistent.\n\nIf the Chen E4 issue turns out to be a misassignment of the experimental line, or if the authors can explain the model's failure there, the paper is a solid contribution to the characterization toolbox. As it stands, the central claim is overstated and the manuscript needs a serious revision, not a desk reject. I would send it to referees, and I would specifically ask for a careful check of every predicted E4/E5 against experimental tables. The target reader is an experimentalist measuring TMDC exciton spectra who wants a fast analytical estimate instead of numerical fitting.","headline":"Useful Kratzer inversion formulas, but the 'no additional fitting' claim has a visible counterexample in Table I.","tokens_in":24272,"tokens_out":4606,"would_cite":false,"duration_ms":45984,"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":"This paper claims that just three measured exciton energies plus magnetic-field shifts uniquely determine the bandgap, screening length, reduced mass, and dielectric constant of monolayer TMDCs, then analytically predict all higher exciton","keywords":["transition-metal dichalcogenides","exciton spectroscopy","magnetoexcitons","modified Kratzer model","analytical inversion","material parameter retrieval","Rytova–Keldysh potential","screening length"],"falsifier":"Measure exciton energies and magnetoexciton shifts for a monolayer TMDC in a deliberately different dielectric environment (for example, on a high-κ substrate or with an added encapsulating layer), retrieve r₀ and μ using the paper's formulas, and compare against independently measured values (e.g., from scanning tunneling spectroscopy or cyclotron resonance). If the retrieved parameters deviate from the independent measurements by more than the quoted uncertainties, the universality of g² and c is falsified.","tokens_in":23309,"feed_emoji":"⚛️","tokens_out":2834,"duration_ms":38475,"temperature":0.7,"pith_summary":"The authors aim to turn optical and magneto-optical exciton spectra of monolayer transition-metal dichalcogenides into direct, closed-form retrieval of four fundamental material parameters: the quasiparticle bandgap, screening length, reduced exciton mass, and surrounding dielectric constant. Using the exactly solvable modified Kratzer model, they derive inversion formulas that take the three lowest zero-field exciton energies as input and output the bandgap, screening parameter, and energy scale, while a separate analytical magnetoexciton expression recovers the reduced mass from the magnetic-field dependence of exciton energies. Once these parameters are known, the framework predicts higher Rydberg states, diamagnetic coefficients, exciton radii, and complete magnetoexciton spectra without any additional fitting or matrix diagonalization. A sympathetic reader would care because this replaces expensive iterative numerical procedures with transparent, nearly instantaneous formulas, making rapid characterization of 2D semiconductors from spectroscopy feasible.","feed_headline":"Three exciton lines reveal bandgap, mass, and screening","feed_subtitle":"A solvable exciton model turns three measured energies and magneto-optical shifts into full material parameters — no numerical fitting.","key_machinery":"The load-bearing object is the modified Kratzer exciton spectrum, εₙ = −η/(n−1/2+ξ)²Ry, where ξ captures dielectric screening and is tied to the screening length r₀ via ξ = g√η r₀/a₀. The inversion hinges on the ratio Δ(ξ) = (ε₃−ε₂)/(ε₃−ε₁), which is independent of both the bandgap and the energy scale; substituting the measured energies turns it into the cubic 3λ−4λ³=δ, solved trigonometrically to yield ξ. The magnetoexciton side uses an interpolation formula Eₙ(B) = E_g + εₙ + σₙB²/(1+σₙβₙ⁻¹B), with diamagnetic coefficients computed from the model's Laguerre-based wavefunctions, and the equality between measured and theoretical field-induced shifts closes a cubic equation for μ. Two empiri","core_discovery":"The central claim is that the modified Kratzer model—a hydrogen-like spectrum with an effective screening parameter ξ added to the principal quantum number—admits an exact inverse problem. From the measured energies of the three lowest s-state excitons, the ratio Δ=(E₃−E₂)/(E₃−E₁) isolates ξ through a trigonometric solution of a cubic equation, after which explicit formulas give the energy scaling factor η, the quasiparticle bandgap E_g, and the screening length r₀. Separately, an analytical interpolation for magnetoexciton energies, calibrated so that the strong-field Landau limit is reproduced, yields a cubic equation whose solution gives the reduced exciton mass μ, and hence the dielectri","pith_inferences":["A natural testable extension would be to calibrate g² per material family (W-based vs. Mo-based) or per dielectric environment; the paper's own comparison suggests g² may drift slightly, and a material-dependent g² could remove the systematic bias it reports in retrieved r₀.","The cubic equation for the reduced mass (Eq. 24) may admit multiple real roots; the paper does not discuss uniqueness criteria, so an editor-level inference is that physical selection rules (e.g., positivity, comparison with known mass ranges) become a practical requirement for automated use.","The analytical bridge between Kratzer and RK established here hints that a fully analytical inversion directly from the Rytova–Keldysh potential might be achievable—something the authors themselves flag as future work—which would eliminate the universal-g² approximation entirely.","For samples on very high-κ substrates or with strong environmental inhomogeneity, the universal crossover constant c=0.905, calibrated up to ~90 T, may need re-calibration; extending the formula to higher fields could either confirm or falsify the universality of this parameter."],"forward_implications":["If the inversion is reliable, material parameters can be extracted from a single optical spectrum plus one magneto-optical measurement, enabling high-throughput screening of 2D semiconductors without specialized fitting software.","The framework predicts higher Rydberg-state energies, exciton radii, and diamagnetic coefficients that have not yet been measured, giving concrete targets for future magneto-optical experiments.","Because the retrieval uses only energy differences, it is insensitive to absolute energy-calibration offsets, making it robust across different experimental setups.","The close agreement between the modified Kratzer and Rytova–Keldysh descriptions suggests that the simpler model can serve as a fast surrogate for the standard RK model in the experimentally relevant parameter range.","The same analytical machinery could be applied to other quasi-2D excitonic systems, such as monolayers of other semiconducting compounds, as long as the universal constants remain valid."],"fun_headline_variants":["Three exciton energies reveal bandgap, mass, screening","Solvable exciton model extracts material parameters from three lines","No fitting needed: three exciton lines yield TMD parameters","Analytical inversion: three exciton energies give full TMD constants","Rapid TMD characterization via three exciton energies"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole procedure stands on the assumption that the modified Kratzer model with the fixed universal constants g²=0.205 and c=0.905 faithfully represents the true Rytova–Keldysh exciton for every TMDC material and dielectric environment; if either constant varies, the retrieved screening length or reduced mass will be systematically biased.","fun_headline_variants_meta":{"raw":{"variants":["Three exciton energies reveal bandgap, mass, screening","Solvable exciton model extracts material parameters from three lines","No fitting needed: three exciton lines yield TMD parameters","Analytical inversion: three exciton energies give full TMD constants","Rapid TMD characterization via three exciton energies"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000195,"raw_usage":{"total_tokens":1234,"prompt_tokens":821,"completion_tokens":413,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":565,"completion_tokens_details":{"reasoning_tokens":330}},"tokens_in":565,"tokens_out":413,"duration_ms":5512,"temperature":1.0,"reasoning_tokens":330,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T11:57:32.859551+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure exciton energies and magnetoexciton shifts for a monolayer TMDC in a deliberately different dielectric environment (for example, on a high-κ substrate or with an added encapsulating layer), retrieve r₀ and μ using the paper's formulas, and compare against independently measured values (e.g., from scanning tunneling spectroscopy or cyclotron resonance). If the retrieved parameters deviate from the independent measurements by more than the quoted uncertainties, the universality of g² and c is falsified.","supporting_citations":[],"review_version":1}