{"id":"06ad6cb0-89f9-413f-9650-d513bfb09dd4","arxiv_id":"2512.22523","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Undoped Ta2NiS5 photoexcited by 800 nm light develops a momentum-localized in-gap feature that rate-equation analysis assigns to a fluence-tunable mixture of trions and excitons, with trions forming without a photoexcited hole.","lead":"This paper reports that after a short laser pulse, undoped Ta2NiS5 shows a bright, long-lived state inside its electronic gap, which the authors interpret as a mixture of trions and excitons. If right, it would mean ultrafast photoemission can watch and tune charged quasiparticles in bulk materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The trion/exciton decomposition is not independently measured; the single-particle trion channel is assumed in the rate-equation model and is not discriminated from conventional alternatives.","rationale":"The reader's weakest assumption identifies the same load-bearing issue: the trion/exciton decomposition is an output of a model that already contains the single-particle trion channel, not a measured property of the spectral feature. My stress-test agrees and sharpens it. The paper does provide real evidence—a clear fluence-dependent in-gap state, slow dynamics, a dispersion/mass consistent with prior equilibrium trion work (ref. 17), and a rate-equation fit that outperforms a trion-free model. However, those supports do not settle whether the QPF is actually a two-species trion+exciton mixture. The absence of a spectrally resolved signature, combined with the model's structural assumption, makes the central claim conditional rather than established. A direct discrimination against conventional trion formation would be the decisive test. No ad hominem or theatrical judgment is intended; this is a standard identifiability concern in kinetic modeling of unresolved spectral features.","tokens_in":12997,"tokens_out":7040,"duration_ms":82073,"concrete_test":"Refit the three normalized QPF decay traces with a competing rate-equation model that includes only conventional secondary trion formation (e.g., d n_T/dt = C'_T n_e n_X - R_T n_T n_h, with no C_T n_e single-particle source) and compare fits via reduced chi-square/BIC. If the competing model fits comparably, the data do not discriminate the single-particle channel. Additionally, generate synthetic decay curves from the conventional model and fit them with Eq. (1); if the fit returns a spurious nonzero C_T, the published decomposition is not identifiable.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—transient trions in undoped Ta2NiS5 formed via a single-particle pathway without a photoexcited hole—depends on converting an unresolved in-gap feature (QPF) into separate n_X(t) and n_T(t) populations. The only observable is the total QPF intensity, assumed proportional to n_X+n_T (Fig. 3C). The decomposition into trions and excitons comes from integrating Eq. (1), whose linear term -C_T n_e already asserts that trions form without consuming a hole; additionally, the rates are forced symmetric (C_T=R_X, C_X=R_T) based on qualitative overlap arguments in the SM. No spectrally resolved trion/exciton signature is presented; the expected peak separation is tens of meV, comparable to the 69 meV QPF width, so the two states are not resolved. The claim therefore rests on the assumed model structure. Alternatives—a phonon sideband, a surface/defect state, or hot-carrier trapping—could plausibly produce a bright, momentum-localized, slow-decaying feature with fluence-dependent dynamics. The paper's supplementary shows that an exciton-only model fails, but it does not test the key alternative: conventional trion formation via exciton-plus-electron capture (a two-particle process that does require a photoexcited hole). Because that alternative was not fit, the unconventional single-particle pathway is not established by the data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports time- and angle-resolved photoemission (trARPES) measurements on undoped Ta2NiS5 following 800 nm optical excitation. The authors observe a bright, momentum-localized in-gap feature (QPF) that appears on a ~400 fs timescale and decays over ~2 ps. Based on a comparison of its effective mass (~3 me) with equilibrium trions from a companion study (ref. 17), and on a fluence-dependent rate-equation analysis, they conclude that the QPF is a mixture of excitons and trions, with trions formed via an unconventional 'single-particle' pathway that does not require a photoexcited hole. The rate equations (Eq. 1) include a linear trion-creation term (C_T n_e), and the trion/exciton populations are extracted by fitting the total QPF intensity (A·σ) to n_X(t)+n_T(t) at three pump fluences. The paper further claims that the trion-to-exciton ratio is controlled by fluence, with trions dominating at late times, and that an exciton-only model fails to capture the dynamics.","tokens_in":13421,"tokens_out":3345,"duration_ms":39202,"significance":"If the identification is correct, the work is significant: it would demonstrate transient trions in an undoped, strongly correlated quasi-1D semiconductor and establish trARPES as a probe of charged many-body quasiparticles. The experimental data are of high quality: the time-mapping analysis (Fig. 2) and the fluence series (Fig. 3) are well presented, and the authors are explicit about the model assumptions. The paper also ships a quantitative rate-equation framework and compares against the exciton-only model, which is a useful negative control. The central claim, however, depends on model-based decomposition of a spectrally unresolved feature, and the single-particle trion channel is inserted by assumption rather than discriminated from alternatives. The significance would be substantially enhanced by a test that does not presuppose the existence of the single-particle pathway.","major_comments":[{"comment":"The central claim that trions form via a single-particle pathway without a photoexcited hole is not established by the data. The observable used for fitting is the total QPF intensity, assumed proportional to n_X+n_T, yet the model in Eq. (1) already includes the linear trion-creation term C_T n_e and the symmetric assignment C_T=R_X, C_X=R_T. The decomposition into n_T(t) and n_X(t) is therefore generated by a model that contains the conclusion. The exciton-only model (fig. S3) rules out a model with no trions at all, but it does not test the key alternative: conventional trion formation via exciton-plus-electron capture, which would require a photoexcited hole. A fit of that alternative model (e.g., with a term proportional to n_X n_e or n_e^2 n_h) to the same data is necessary to substantiate the 'no-hole pathway' claim.","section":"§3 (Eq. 1), Fig. 3C–D"},{"comment":"The fluence dependence of the rates undermines the claim that the model predicts composition from measured dynamics without fluence information. The extracted rates vary strongly with fluence: α_1p^{-1} goes from 446±74 fs (490 μJ/cm²) to 175±14 fs (50 μJ/cm²), and α_2p^{-1} from 350±60 fs to 833±69 fs. Because all fits use identical initial conditions (n_e(0)=n_h(0)=1 by normalization), the three fluences are not distinguished by the source term; the fluence dependence is absorbed entirely into per-fluence fitted rates. The statement 'the model is not supplied a priori with fluence information' is thus misleading. Moreover, the abstract's claim that 'the same model with unchanged rates reproduces the full dynamics' in surface-doped samples is inconsistent with Table S1, where rates change by more than a factor of two across the pristine-fluence series.","section":"§3, Table S1"},{"comment":"The identification of the QPF as a bound trion/exciton mixture is not fully compelling because alternative origins of the in-gap feature are not ruled out. The increased Gaussian width (69 meV vs 46 meV) is attributed to 'the unavoidable presence of excitons,' but a phonon sideband, a surface/defect state, or a hot-carrier relaxation artifact could produce a broad, momentum-localized, slowly decaying feature with fluence-dependent dynamics. The comparison to equilibrium trions via effective mass (fig. S1) is suggestive, but the peak separation expected between trions and excitons is tens of meV, comparable to the 69 meV width, so the feature is spectrally unresolved. The paper should explicitly discuss and, where possible, exclude these alternative interpretations, or present an independent discriminator.","section":"§4 (Fig. 4D), §2"}],"minor_comments":[{"comment":"The abbreviation 'QPF' (quasiparticle feature) is used, but the supplementary occasionally writes 'QFP'; please unify.","section":"Throughout"},{"comment":"In the derivation of C_T, C_X, R_X, R_T (Eqs. S12–S15), the paper states that 'geometric overlap factors are typically of order unity,' which justifies C_T=R_X and C_X=R_T. This is a qualitative estimate; the systematic uncertainty in this equality should be quantified or at least discussed as a limitation.","section":"SM, Rate Equations"},{"comment":"The caption has two items labeled '(A)'; the second should be labeled '(B)' or renumbered.","section":"SM, fig. S1 caption"},{"comment":"The phrase 'TaNiS5' should be 'Ta2NiS5'.","section":"SM, first paragraph"},{"comment":"The intensity curves are normalized to their peak and have background subtracted. This normalization removes absolute fluence information; please state explicitly in the text or caption that the fits therefore use only the shape of the normalized decay, not the absolute quasiparticle density.","section":"§3, Fig. 3B"}],"recommendation":"major_revision","confidential_remarks":"The reader's concern about circularity is well grounded in the manuscript: the single-particle trion channel is assumed in Eq. (1), and the data do not independently distinguish it from a conventional hole-requiring pathway. The paper would be suitable for publication in a strong journal only after the model comparison is extended to include alternative trion-formation channels and after the fluence dependence of the fitted rates is reconciled with the 'unchanged rates' claim. I would also suggest the editors consider whether the companion paper (ref. 17) is under review elsewhere, since a substantial part of the identification rests on its reported equilibrium trion signatures."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I'll cut to the chase: the experiment looks solid and the observation is real, but the paper's headline claim—that the in-gap state is a trion population formed by a single-particle pathway—is not established by the data as presented. What we have is a bright, momentum-localized, fluence-dependent feature in undoped Ta2NiS5 that appears after photoexcitation and decays on a ~2 ps timescale. That is new and worth reporting. The trARPES data is careful: good time resolution, clean background subtraction, and the temporal mapping in Fig. 2 is well done. The comparison of the effective mass with the equilibrium trion in ref 17 is suggestive, and the authors honestly show that an exciton-only model fails to capture the decay dynamics.\n\nThe weakness is in the decomposition. The QPF is one unresolved peak; the expected exciton-trion energy difference is tens of meV, comparable to the measured 69 meV width. The populations n_X and n_T are never measured—they come from integrating Eq. (1), which already contains the single-particle trion channel C_T n_e and explicitly neglects the conventional two-particle trion channel (exciton + electron capture). When you fit that model to the same integrated intensity you used to define the feature, the 'trion-dominated at late times' result is partly baked in. The rate symmetry C_T=R_X, C_X=R_T is plausible but rests on qualitative overlap arguments, not a detailed calculation.\n\nI also want to flag a discrepancy: the abstract states that in surface-doped samples the same model with unchanged rates reproduces the full dynamics, but I found no such data in the main text or supplementary. That claim needs to be either shown or removed.\n\nAlternatives—phonon sidebands, surface or defect states, hot-carrier relaxation artifacts—are not ruled out, and the paper doesn't attempt to fit a model with conventional trion formation. So the evidence for the unconventional mechanism is not compelling yet.\n\nThat said, this is a serious experimental paper. The observation of a fluence-dependent in-gap quasiparticle with slow dynamics in a bulk undoped quasi-1D material is a useful contribution to the trARPES and exciton/trion community. It should go to peer review, but a careful referee should push on the model-dependence of the composition and the missing surface-doped data. If I were the editor, I'd send it out and ask for either stronger evidence or a scaled-back interpretation.","headline":"Real fluence-dependent in-gap state in undoped Ta2NiS5, but the trion/exciton decomposition and single-particle pathway are model-generated, not directly measured.","tokens_in":13906,"tokens_out":4103,"would_cite":false,"duration_ms":41154,"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":"In undoped Ta2NiS5, optical excitation alone creates a mixed population of trions and excitons, with trions forming via a single-particle channel that needs no photoexcited hole.","keywords":["trions","excitons","trARPES","Ta2NiS5","quasi-one-dimensional","ultrafast dynamics","rate equations","Coulomb correlations"],"falsifier":"Measure the trion and exciton contributions independently, for example by tuning the probe energy or using a momentum microscope to resolve the few-meV difference in photoemission kinetic energy between the two species, or by varying the pump wavelength to suppress the single-particle channel: if the late-time decay becomes fluence-independent or the feature splits into two resolvable peaks, the single-peak, two-species interpretation would be falsified.","tokens_in":12928,"feed_emoji":"⚛️","tokens_out":2637,"duration_ms":32081,"temperature":0.7,"pith_summary":"The paper reports that pumping pristine, undoped Ta2NiS5 with an ultrafast near-infrared pulse produces a bright, momentum-localized state inside the band gap, which the authors identify as a mixture of excitons and trions (charged excitons). The key claim is that trions—conventionally thought to require charge doping—are generated here purely optically, through a direct single-particle pathway in which a conduction electron binds to a virtual exciton. A rate-equation model with just two rates captures the fluence-dependent decay, showing that trions dominate the late-time relaxation while excitons dominate early at high fluence. If correct, this establishes that charged many-body bound states can be created and controlled by light alone in a correlated quasi-one-dimensional semiconductor, opening a new route to manipulate neutral and charged quasiparticles on ultrafast timescales.","feed_headline":"Trions form without doping in Ta2NiS5","feed_subtitle":"Ultrafast laser pulses alone create a mixed trion/exciton population; fluence controls the ratio.","key_machinery":"The central mechanism is a pair of coupled rate equations for conduction electrons, holes, excitons, and trions, reduced to two effective rates: alpha_1p for all single-particle processes (trion creation and exciton decay) and alpha_2p for two-particle processes (exciton creation and trion decay). The model assumes the measured in-gap intensity is proportional to the sum of trion and exciton populations, n_T + n_X, and exploits the symmetry that the two channels have equal rates. The energetics of Ta2NiS5—total trion binding energy exceeding the band gap—is what enables the single-particle trion-creation pathway and is taken from prior work, not recalculated here.","core_discovery":"The paper shows that in Ta2NiS5, a quasi-one-dimensional correlated semiconductor, photoexcitation of undoped samples generates a transient in-gap quasiparticle feature that reaches maximum intensity about 400 fs after the pump pulse and decays over ~2 ps. By combining time- and angle-resolved photoemission (trARPES) with coupled rate equations, the authors identify this feature as a superposition of excitons and trions. The trions are formed through an unconventional single-particle mechanism: a conduction-band electron binds directly to a virtual exciton, without requiring an intermediate electron-hole binding step. This is energetically possible because the total trion binding energy exce","pith_inferences":["Beyond the paper: if this single-particle trion channel is generic to materials with total trion binding energy exceeding the band gap, it would predict that optical pumping alone can drive trion formation in other quasi-one-dimensional and strongly correlated semiconductors, a testable extension to materials like Ta2NiSe5 or TaSe3.","Beyond the paper: the fluence-dependent trion/exciton ratio implies an optical switch between charge-neutral and charged quasiparticle populations; one could probe this with time-resolved, momentum-resolved electron energy-loss or photoemission of the two-exciton decay products.","Beyond the paper: the decomposition into trions and excitons is purely derived from the rate-equation structure; a direct spectral measurement at higher probe energy or lower temperature might resolve the small binding-energy difference and independently verify the fractions.","Beyond the paper: the slow downward drift of the feature's center energy (attributed to screening changes) suggests that the trion/exciton mixture itself renormalizes the electronic structure, which could be observed as a fluence-dependent shift in the valence band or in the equilibrium trion dispersion."],"forward_implications":["Trions can be created purely optically in an undoped bulk semiconductor, eliminating the need for external charge doping and suggesting a general route to light-controlled charged excitations.","Pump fluence provides a tunable knob: low fluence produces a trion-dominated nonequilibrium state, while high fluence initially favors excitons, allowing controlled studies of mixed charged/neutral quasiparticle populations.","The trion decay channel necessarily generates two excitons per trion, implying that trion recombination cascades into exciton populations and shapes the late-time photophysics.","The success of the two-rate model in both undoped and surface-doped samples unifies transient and equilibrium trions, suggesting that light-induced and doping-induced trions are the same quasiparticle species.","Time-resolved photoemission can resolve charged many-body bound states, not just neutral excitons, in bulk correlated materials."],"fun_headline_variants":["Fluence steers trion-exciton competition in Ta2NiS5","Laser light alone creates trions in pristine Ta2NiS5","Optical pulses make trions without any doping","Trions on demand: pump light forms them in Ta2NiS5","Light-created trions compete with excitons in Ta2NiS5"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The in-gap photoemission feature is a bound quasiparticle whose total measured intensity is proportional to the sum of trion and exciton populations, and the trion/exciton decomposition is not directly measured but instead emerges from rate equations that already contain the single-particle trion channel.","fun_headline_variants_meta":{"raw":{"variants":["Fluence steers trion-exciton competition in Ta2NiS5","Laser light alone creates trions in pristine Ta2NiS5","Optical pulses make trions without any doping","Trions on demand: pump light forms them in Ta2NiS5","Light-created trions compete with excitons in Ta2NiS5"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001126,"raw_usage":{"total_tokens":4530,"prompt_tokens":767,"completion_tokens":3763,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":511,"completion_tokens_details":{"reasoning_tokens":3668}},"tokens_in":511,"tokens_out":3763,"duration_ms":25422,"temperature":1.0,"reasoning_tokens":3668,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T13:49:21.785282+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the trion and exciton contributions independently, for example by tuning the probe energy or using a momentum microscope to resolve the few-meV difference in photoemission kinetic energy between the two species, or by varying the pump wavelength to suppress the single-particle channel: if the late-time decay becomes fluence-independent or the feature splits into two resolvable peaks, the single-peak, two-species interpretation would be falsified.","supporting_citations":[],"review_version":1}