{"id":"de78b1f8-923f-4278-a3a1-0000ece5d140","arxiv_id":"2507.16643","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Trion states in monolayer TMDs are asymmetric, and the measured exciton-trion splitting includes the bright-dark exciton energy difference, which dominates the signal in WSe2.","lead":"This paper shows that the measured energy gap between exciton and trion light peaks in monolayer semiconductors is not the trion binding energy, because trions form on a darker, more tightly bound exciton. It offers a corrected interpretation that resolves a long-standing mismatch between theory and experiment in WSe2 and MoSe2.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Quantitative WSe2 T3 agreement depends on a mass-ratio input tuned to 33% versus the independent DFT value of 27%, and this input directly controls the dominant bright-dark exciton contribution.","rationale":"The reader's weakest assumption identifies exactly the load-bearing input for the paper's most striking quantitative result. The paper is transparent about this: it states that the 35 meV T3 energy is reproduced by assuming a 33% mass difference, compared to the theoretically predicted 27%. Because the T3 decomposition is dominated by ΔE_BD = 30.7 meV, this assumed mass difference is not a minor parameter—it sets the magnitude of the claimed reinterpretation. The model has independent support in the form of standard variational calculations and use of experimentally fitted Rytova-Keldysh parameters, and the qualitative idea that dark-exciton binding differences enter the measured trion splitting is plausible and likely correct. However, the specific quantitative claim that the true T3 trion binding energy is only about 4 meV is not robust to the mass-input uncertainty. A targeted recalculation with the DFT value and experimentally plausible bounds would settle whether the headline numbers survive. Since the reader already conditioned the verdict on independent determination of these parameters, the appropriate final verdict remains CONDITIONAL, and no verdict change is needed.","tokens_in":8916,"tokens_out":4980,"duration_ms":55476,"concrete_test":"Recompute the WSe2 T3 (and T2) trion with the independent DFT mass ratio me,2/me,1 = 0.73 (27% mass difference) while keeping all other inputs fixed (κ = 4.4, r0 = 3.85 nm, and the Ref. [16] exchange energy), and repeat with the 15–30% experimental mass-uncertainty bounds. If the resulting ΔE_exp stays within experimental error of 35 meV and ΔE_bind remains near 4–6 meV, the concern is resolved. If ΔE_exp shifts by more than about 5 meV or ΔE_bind rises substantially, the quantitative agreement is parameter-driven rather than a robust prediction of the asymmetric-trion model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline decomposition for WSe2 T3—35 meV splitting into 30.7 meV of bright-dark exciton binding difference and only 4.2 meV of true trion binding—is computed with me,2/me,1 = 0.36/0.54 = 0.667, i.e., a 33% mass difference, explicitly chosen to reproduce the T3 energy. The independent DFT value is about 27%, and the paper itself notes that experimentally determined masses deviate 15–30% from theory. Since ΔE_BD constitutes roughly 88% of the T3 splitting, the central quantitative claim is directly proportional to this unmeasured input. If the actual mass ratio is closer to the DFT value, ΔE_BD shrinks and ΔE_bind grows, weakening the 'only 4 meV' and 'twice as extended' statements. The conceptual asymmetry mechanism is more robust, but the quantitative explanation for the headline WSe2 result is not yet independently constrained.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This Letter argues that the commonly measured exciton-trion splitting ΔE_exp in monolayer TMDs is not simply the trion binding energy, because dark excitons are more strongly bound than the bright exciton. The authors solve the three-body Schrödinger equation with Rytova-Keldysh potentials, a short-range exchange interaction, and a correlated 2D-hydrogenic basis, and decompose ΔE_exp into a bright-dark exciton binding-energy difference ΔE_BD and a trion binding energy ΔE_bind defined relative to the dark exciton. For WSe2 and MoSe2 they report agreement with measured trion energies, with the WSe2 T3 trion dominated by ΔE_BD (30.7 meV) rather than ΔE_bind (4.2 meV); the associated wavefunction is strongly asymmetric, with the bright-exciton bond length stretched to about twice the commonly assumed trion size.","tokens_in":9164,"tokens_out":9724,"duration_ms":106706,"significance":"The conceptual message is important and likely correct as a matter of definition: the measured splitting contains the bright-dark exciton energy difference, and this contribution can be comparable to or larger than the trion binding contribution. The paper is also transparent: it separates the two contributions in Table I, explains the variational method in the main text and Supplemental Material, and explicitly identifies the mass-ratio assumption used to reproduce the WSe2 T3 energy. These are genuine strengths. The significance is nevertheless conditional, because the quantitative claims currently rely on parameters that are adjusted to match the same experimental values—most notably the WSe2 upper-conduction-band mass ratio and the MoSe2 exchange strength. The framework and mechanism are valuable, but the quantitative explanation is not yet independently constrained.","major_comments":[{"comment":"The headline WSe2 T3 decomposition is calibrated to the experiment rather than predicted from independent inputs. The text states that the 35 meV singlet energy is reproduced by 'assuming a 33% mass difference' between the upper and lower conduction bands, while the independent DFT estimate is about 27% and the manuscript notes that experimentally determined masses deviate 15–30% from theory. Because ΔE_BD = 30.7 meV is about 88% of the 35 meV splitting, the central statement that only about 4 meV is true trion binding—and the related claim that ⟨r1⟩ ≈ 4.11 nm is about twice the usual trion size—is directly proportional to this fitted input. Please report the T2 and T3 results at the DFT mass ratio and across the stated experimental uncertainty, giving ΔE_BD, ΔE_bind, and the wavefunction expectation values, and state explicitly which conclusions survive over that range.","section":"Results (Table I, Fig. 3b)"},{"comment":"The MoSe2 column is also fitted: the 9 meV exchange interaction is inferred from the experimental trion energy ('consistent with an exchange interaction on the order of 9 meV'), and the cited 9 meV estimate from Ref. [18] is for MoS2 rather than an independent MoSe2 input. With the exchange strength set to 9 meV, ΔE_BD = 9.0 meV and the total becomes 24.3 meV, so this agreement is a consistency check rather than a quantitative prediction. I ask that IX(MoSe2) be computed with the same DFT-based method as for WSe2, or that a sensitivity scan over IX be provided and the wording adjusted accordingly.","section":"Results (MoSe2 discussion, Table I)"}],"minor_comments":[{"comment":"The cross-reference 'As shown in Fig. 1b)' appears to be a typo; Fig. 1 has no panel b, and the surrounding text discusses the exchange-strength dependence shown in Fig. 3a.","section":"Results, p. 3"},{"comment":"Please state the units and normalization convention for IX and for the Gaussian exchange potential VX(r). As written, Eq. (3) uses IX as if it were an energy, while the contact interaction in 2D has units of energy times area; the calibration of cX relative to IX should be made explicit.","section":"Eq. (3) and Eq. (4)"},{"comment":"For comparison against experiment, please quote the experimental uncertainties from Refs. [7] and [12]; without them, the claimed agreement at the level of 0.2–3 meV has no quantitative benchmark.","section":"Table I"},{"comment":"The punctuation error 'Here. L^{2l}_i' should be corrected to 'Here, L^{2l}_i'.","section":"Supplemental Material, Eq. (10)"},{"comment":"It would be helpful to state explicitly in the caption that r1 is the bright-exciton bond length and r2 is the dark-exciton bond length, since this assignment is otherwise only given in the main text.","section":"Fig. 4 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper makes a valuable conceptual point and the variational machinery appears sound, but the quantitative headline is currently tied to parameters that are tuned to the data. I would ask the authors to add a sensitivity analysis for the mass ratio and the exchange strength, and to soften the 'quantitatively explains' claim until the WSe2 T3 decomposition is shown to be robust over the plausible range of these inputs. A referee with detailed knowledge of TMD conduction-band masses and DFT exchange parameters would be particularly useful."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The one to know: this paper changes how we should read trion spectra in TMDs. The measured exciton-trion splitting contains a large contribution from the bright-dark exciton binding difference, so the naive 'trion binding energy' is wrong. That point is new and, I think, correct. The decomposition is an identity; the variational trion calculation is standard and the discussion of asymmetric bond lengths is convincing.\n\nWhat is genuinely new: the identification that trions in WSe2 and MoSe2 are spatially asymmetric because dark excitons bind more strongly than bright ones, and that the measured splitting therefore includes the bright-dark binding difference. The paper does a good thing by showing this resolves the old problem of needing an ad hoc dielectric constant to reproduce trion energies. The T1 result for WSe2 and the dark trion prediction matching the hexciton onset are meaningful successes.\n\nThe soft spot is the quantitative claim for WSe2 T3. The 35 meV splitting is decomposed into 30.7 meV of bright-dark binding difference and 4.2 meV of true binding, but the mass ratio between the upper and lower conduction bands is set to 33% to reproduce the 35 meV. The independent DFT value is around 27%; the paper concedes experimental masses deviate 15–30% from theory. Since the dominant piece scales with this input, the 'only 4 meV' and 'twice as extended' statements are not independently constrained. If the mass ratio is closer to the DFT value, the bright-dark contribution shrinks and the picture moderates. The T2 result (26 vs. 29 meV) is a more honest prediction, and it supports the framework.\n\nThe MoSe2 comparison is softer too: the 9 meV exchange energy is effectively fitted, though consistent with estimates in MoS2.\n\nOverall, the paper deserves a serious referee. The conceptual mechanism is likely right and will be cited; the numerical values should be treated as parameter-dependent illustrations, not independent validations. The authors are honest about what is fitted and what is predicted. This belongs in a good journal after a careful review of the mass sensitivity.\n\nMy recommendation: send it to peer review.","headline":"A genuinely new conceptual point about trion spectra—the measured splitting includes the bright-dark exciton binding difference—backed by careful numerics, but the headline WSe2 T3 number leans on a fitted mass ratio.","tokens_in":9679,"tokens_out":1623,"would_cite":true,"duration_ms":16630,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Dark-exciton offset dominates the measured trion splitting in WSe2","keywords":["trions","monolayer transition metal dichalcogenides","bright-dark exciton splitting","trion binding energy","exchange interaction","Rytova-Keldysh potential","WSe2","MoSe2"],"falsifier":"A direct measurement of the upper conduction-band electron mass in WSe2, for example from Landau-level spectroscopy or quantum transport, could pin the mass ratio near the DFT value of 27% rather than the assumed 33%; combined with a measurement of the dark-trion binding energy relative to the dark exciton, this would reveal whether the true T3 binding is really about 4 meV or substantially larger.","tokens_in":2041,"feed_emoji":"🔬","tokens_out":4815,"duration_ms":113519,"temperature":0.7,"pith_summary":"The paper argues that the energy difference between the trion peak and the bright exciton peak, routinely read as the trion binding energy, actually contains a second contribution: the binding-energy difference between the dark and bright excitons. Because dark excitons are more strongly bound than bright excitons in monolayer TMDs, the trion wave function becomes asymmetric, with the bright-exciton bond stretched relative to the dark-exciton bond. In WSe2 the model assigns roughly 31 meV of the measured 35 meV splitting to the bright\\u2013dark difference and only about 4 meV to true trion binding, making that trion about twice as extended as usually assumed. The same calculation reproduces the measured trion energies in WSe2 and MoSe2 using dielectric screening parameters fixed by independent Rydberg-exciton measurements rather than adjusted to fit the trions. If this is right, optical spectra must be read with the dark exciton as the reference state for trion binding.","feed_headline":"Dark-exciton offset dominates the measured trion splitting in WSe2","feed_subtitle":"Only about 4 of the 35 meV is true trion binding; the rest is bright\\u2013dark exciton difference.","key_machinery":"The working object is the variational trion wave function $\\psi_T(r_1,r_2,\\theta,\\alpha)=u(r_1,r_2,\\theta)\\exp(im\\alpha)/\\sqrt{2\\pi r_1 r_2}$ with $m=0$, where $r_1$ and $r_2$ are the electron\\u2013hole distances associated with the bright and dark excitons, $\\theta$ is the angle between them, and $u$ is expanded in two-dimensional hydrogenic basis functions. Coulomb interactions are treated with the Rytova\\u2013Keldysh potential, and exchange is added as a short-range Gaussian $V_X(r)=c_X\\exp(-r^2/\\sigma_X^2)$ whose strength follows from density-functional estimates. The mechanism that carries the argument is the additive offset $\\Delta E_{\\rm BD}$: once exchange and the conduction-band mass difference are included, the two branches of the wave function no longer contribute equally, the dark-exciton branch dominates, and the measured trion\\u2013bright-exciton splitting inherits the bright\\u2013dark exciton binding difference rather than measuring pure trion binding.","core_discovery":"The central claim is that the measured quantity $\\Delta E_{\\rm exp}=\\omega_{\\rm BX}-\\omega_{\\rm T}$ decomposes as $\\Delta E_{\\rm exp}=\\Delta E_{\\rm bind}+\\Delta E_{\\rm BD}$, where $\\Delta E_{\\rm bind}$ is the trion binding energy referenced to the dark exciton and $\\Delta E_{\\rm BD}=E_{\\rm DX}-E_{\\rm BX}$ is the bright\\u2013dark exciton binding-energy difference. Including a short-range exchange interaction on the bright exciton and a 33% mass difference between the upper and lower conduction bands, the authors compute trion energies in WSe2 of $-21.2$, $-26.0$, and $-35.0$ meV for T1, T2, and T3, close to the measured $-21$, $-29$, and $-35$ meV, with the T3 splitting broken into $-30.7$ meV of bright\\u2013dark offset and only $-4.2$ meV of actual binding. The resulting wave function for T3 has $\\langle r_1\\rangle \\approx 4.1$ nm for the bright-exciton bond length, roughly twice the size conventionally assumed. In MoSe2 the same framework works with a comparable exchange energy of about 9 meV, but there the genuine trion binding remains the dominant part of the splitting because of the heavier carrier masses.","pith_inferences":["If the T3 trion is really bound by only about 4 meV, its large bright-exciton bond length makes it unusually sensitive to the dielectric environment, so few-meV shifts between samples could be a natural consequence rather than an inconsistency.","The same asymmetry should govern dark trions and the hexciton feature: the paper's computed dark-trion position 41.6 meV below the bright exciton offers a direct, testable prediction for density-dependent photoluminescence.","A Landau-level or transport measurement of the WSe2 upper conduction-band mass near the DFT value of 27% would shift the T3 decomposition toward a larger true binding energy, providing a sharp experimental check of the 4 meV number.","The framework invites applying the same bright\\u2013dark decomposition to charged excitons in other two-dimensional semiconductors and to trions in twisted heterobilayers, where the dark-exciton offset may be comparable to the binding."],"forward_implications":["Commonly quoted trion binding energies for WSe2 should be revised downward: the T3 trion is bound by about 4 meV relative to the dark exciton, not 35 meV relative to the bright exciton.","The spatial size of the WSe2 T3 trion is roughly twice the conventional estimate, so overlap integrals and density estimates that assume a compact symmetric trion will need rescaling.","Optical spectra should be analyzed with the dark exciton as the reference state, making $\\Delta E_{\\rm BD}$ a required input alongside the trion binding energy.","The model restores consistency between trion energies and the independently measured dielectric screening constant $\\kappa=4.4$, eliminating the need for a downward adjustment of $\\kappa$ to fit trions.","The correction is material dependent: in MoSe2 the true trion binding still dominates the measured splitting, so the reinterpretation matters most in materials with light carriers such as WSe2."],"supporting_citations":[{"why":"supplies the experimental WSe2 trion energies (T1, T2, T3) that the model reproduces.","marker":"[12]"},{"why":"supplies the MoSe2 trion energy and the experimental context for the spectra.","marker":"[7]"},{"why":"provides the experimentally fitted Rytova-Keldysh screening parameters and reduced masses used as input.","marker":"[15]"},{"why":"provides the DFT exchange energies and the theoretical conduction-band mass difference behind the bright-dark splitting.","marker":"[16]"},{"why":"provides the Rydberg exciton series that constrains the dielectric screening parameters.","marker":"[14]"},{"why":"supplies the variational few-body method and trion Hamiltonian used for the numerical calculation.","marker":"[20]"},{"why":"documents the bright-dark exciton splitting in WSe2 that motivates the reference-state shift.","marker":"[17]"},{"why":"supplies the band-structure masses for the conduction bands that set the mass-difference input.","marker":"[19]"},{"why":"supports the ordering of bright and dark exciton binding and the magnitude of exchange in related TMDs.","marker":"[18]"}],"fun_headline_variants":["Trion splitting: mostly bright-dark gap, not binding","Dark-exciton offset dwarfs true trion binding in WSe2","Most trion splitting is dark-exciton offset, not binding","Measured trion splitting in WSe2 is mostly dark-exciton gap"],"cache_read_input_tokens":11776,"weakest_assumption_plain":"The load-bearing input is the assumed 33% mass difference between the upper and lower conduction-band electrons in WSe2, chosen to match the T3 trion energy, while the independently predicted value is about 27% and measured masses deviate by 15\\u201330% from theory.","fun_headline_variants_meta":{"raw":{"variants":["Trion splitting: mostly bright-dark gap, not binding","Dark-exciton offset dwarfs true trion binding in WSe2","Most trion splitting is dark-exciton offset, not binding","Measured trion splitting in WSe2 is mostly dark-exciton gap"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001656,"raw_usage":{"total_tokens":6588,"prompt_tokens":971,"completion_tokens":5617,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":5539}},"tokens_in":587,"tokens_out":5617,"duration_ms":40888,"temperature":1.0,"reasoning_tokens":5539,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T15:05:13.646015+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct measurement of the upper conduction-band electron mass in WSe2, for example from Landau-level spectroscopy or quantum transport, could pin the mass ratio near the DFT value of 27% rather than the assumed 33%; combined with a measurement of the dark-trion binding energy relative to the dark exciton, this would reveal whether the true T3 binding is really about 4 meV or substantially larger.","supporting_citations":[{"cited_title":"Courtade, M","cited_arxiv_id":null,"evidence_quote":"supplies the experimental WSe2 trion energies (T1, T2, T3) that the model reproduces."},{"cited_title":"Kiper, H","cited_arxiv_id":null,"evidence_quote":"supplies the MoSe2 trion energy and the experimental context for the spectra."},{"cited_title":"Goryca, J","cited_arxiv_id":null,"evidence_quote":"provides the experimentally fitted Rytova-Keldysh screening parameters and reduced masses used as input."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the DFT exchange energies and the theoretical conduction-band mass difference behind the bright-dark splitting."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the Rydberg exciton series that constrains the dielectric screening parameters."},{"cited_title":"Zhang, T","cited_arxiv_id":null,"evidence_quote":"documents the bright-dark exciton splitting in WSe2 that motivates the reference-state shift."},{"cited_title":"Korm´ anyos, G","cited_arxiv_id":null,"evidence_quote":"supplies the band-structure masses for the conduction bands that set the mass-difference input."},{"cited_title":"Robert, B","cited_arxiv_id":null,"evidence_quote":"supports the ordering of bright and dark exciton binding and the magnitude of exchange in related TMDs."}],"review_version":1}