{"id":"83a22675-d46e-4296-b2d4-f995623cafe5","arxiv_id":"2411.13695","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"In layered ReS2, exciton coherence times are sub-picosecond, homogeneously limited, and remain measurable up to room temperature, with thickness-independent population dynamics.","lead":"Researchers measured the quantum coherence lifetime of excitons in layered rhenium disulfide and found it survives at room temperature with little disorder. The result matters because exciton coherence is the resource for quantum information and nanophotonics in atomically thin materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"FWM deconvolution is pulse-limited: with T2 comparable to the 115 fs pulse, gamma and sigma<=1 meV are not uniquely determined, so the homogeneous-limit claim needs a full 2D refit.","rationale":"The reader's weakest assumption correctly identifies the FWM fitting and signal-interpretation premise as the softest load-bearing point in the paper. My pass sharpens that concern: the homogeneous/inhomogeneous decomposition is not merely a matter of absent echo fingerprints, but a parameter-identification problem in a pulse-limited regime. Since T2 is comparable to the 118 fs pulse, the model cannot cleanly separate gamma, sigma, and a possible nonresonant background without a full 2D fit and accurate pulse autocorrelation. This does not invalidate the experiment, but it means the headline values should be treated as conditional on model assumptions until the raw data are reanalyzed. The paper's independent support (PL polarization analysis, consistency of gamma across locations and thickness) is encouraging, but it does not resolve this specific identifiability issue. The reader already assigned CONDITIONAL, and my concern reinforces that condition rather than moving the verdict, so UNCHANGED is appropriate.","tokens_in":14074,"tokens_out":5275,"duration_ms":49771,"concrete_test":"Refit the raw 2D FWM maps for multilayer Flake-2 and monolayer Flake-6 at 10 K using the full response in (tau12,t): a resonant exciton component (gamma, sigma) convolved with the independently measured pulse intensity autocorrelation, plus an instantaneous nonresonant term A_nr delta(t)delta(tau). Compare model evidence (AIC or F-test) and check whether best-fit gamma shifts by more than 20% or sigma exceeds 1 meV when A_nr is included; also inspect the spectrally resolved FWM phase versus t at fixed tau12 for the t=tau echo twist that would be hidden in the integrated amplitude.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that ReS2 excitons are homogeneously limited (sigma<=1 meV, gamma 5-8 meV at 10 K) rests on fits of FWM amplitude versus tau12 (Fig. 2h-i; SI Sec. 3) to a model that convolves a Gaussian pulse (118 fs FWHM, Methods) with an exponential coherence decay and Gaussian inhomogeneous broadening. For gamma=5-8 meV, T2=160-260 fs, only 1.4-2.3 times the pulse duration. In this regime, the time-integrated FWM decay is dominated by the pulse autocorrelation, so the extracted gamma is a deconvolution that is highly sensitive to pulse shape, chirp, and any instantaneous nonresonant component (e.g., substrate or two-photon response). The absence of a photon echo in the 2D map only excludes sigma substantially larger than gamma; it does not, by itself, bound sigma to 1 meV when the echo is masked by the finite pulse and fast decay. An unmodeled coherent background would appear as an extra fast component and could bias gamma upward while making sigma appear negligible. Without raw 2D maps, independent pulse characterization, and fit residuals, the homogeneous-limit conclusion is not uniquely determined by the data shown. The room-temperature claim is even more vulnerable because T2 there likely approaches or falls below the pulse duration.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports resonant four-wave mixing (FWM) micro-spectroscopy on mechanically exfoliated ReS2 flakes of varying thickness, extracting exciton coherence times T2 in the sub-picosecond range and multiple population decay timescales T1. The authors claim that the excitonic linewidth is homogeneously limited, with an inhomogeneous broadening upper bound σ ≤ 1.0 meV and homogeneous linewidth γ ≈ 5–8 meV at 10 K, that this coherence is remarkably robust against temperature and optical power (discernible signatures even at room temperature), and that the measured T1 and T2 support a direct band gap in ReS2 regardless of layer thickness.","tokens_in":14544,"tokens_out":4281,"duration_ms":59038,"significance":"If the central claims hold, ReS2 would be an unusual 2D semiconductor that combines in-plane anisotropy, low disorder, and robust exciton coherence at room temperature, with potential for nanophotonic and quantum applications. The work uses a state-of-the-art heterodyne FWM micro-spectroscopy setup and includes 2D FWM maps, power- and temperature-dependent measurements, and multi-exponential population fits with alternative fit checks (SI Sec. 5). These are real strengths. The main concerns are the pulse-limited deconvolution of the FWM decays, the lack of a quantitative statistical criterion for the inhomogeneous upper bound, the non-identifiability of the optical-phonon fit parameters, and the indirect nature of the direct-gap inference.","major_comments":[{"comment":"The central homogeneous-limit claim rests on deconvolving the FWM amplitude versus τ12 with a model that convolves a Gaussian pulse (FWHM 115 fs, fixed at 118 fs in the fitting) with an exponential coherence decay and Gaussian inhomogeneous broadening. For γ = 5–8 meV, T2 = 2ℏ/γ ≈ 160–260 fs, which is only 1.4–2.3 times the pulse duration. In this regime the integrated FWM decay is dominated by the pulse autocorrelation, making γ and σ highly sensitive to the assumed pulse shape, chirp, and any unmodeled instantaneous (nonresonant) background. The absence of a photon echo in the 2D maps (Fig. 2f–g) only excludes σ ≫ γ; it does not set a quantitative upper bound of σ ≤ 1 meV when the echo is masked by the finite pulse and fast decay. I request a full 2D fit over (τ12, t) with independent pulse characterization (amplitude and phase) and an explicit nonresonant component, together with fit residuals, to demonstrate that γ and σ are uniquely determined.","section":"Results – Coherence dynamics; Methods"},{"comment":"The bound σ ≤ 1.0 meV is stated to be 'determined by the consistent fit quality up to this limit,' but no statistical criterion is provided (e.g., χ² threshold, F-test, or residual analysis). Without a quantitative goodness-of-fit measure, the bound is not falsifiable. Please specify the fitting metric, the range of σ tested, and how the upper limit is formally defined.","section":"Results – Coherence dynamics"},{"comment":"The fit of γ(T) to γ(T) = γ0 + αT + β(exp(E0/kBT) − 1)^(−1) has multiple solutions for (E0, β); the authors constrain E0 using Ref. [39] to about 50 meV and then report β = 5.0 ± 2.6 meV (EX1) and 6.3 ± 1.1 meV (EX2). Because E0 and β are strongly correlated, the quoted uncertainties on β and the derived claim of 'approximately 30 times lower' optical-phonon coupling than MoSe2/WSe2 are conditional on an external literature value. Please report the joint confidence regions or present the fit across a range of fixed E0 values, and disclose the uncertainty of the literature constraint, rather than stating a single best-fit value.","section":"Results – Coherence dynamics (temperature dependence)"},{"comment":"The inference that ReS2 has a direct gap at all thicknesses is based on the similarity of measured T1 and T2 to those of direct-gap MX2 monolayers and the contrast with indirect-gap bilayer MoSe2. This is indirect evidence; the manuscript itself notes that the nature of the gap is under debate. Please soften this claim or provide a direct test (e.g., comparison with calculated radiative rates, or temperature-dependent PL/absorption linewidth analysis) to make the conclusion proportionate to the evidence.","section":"Conclusion"}],"minor_comments":[{"comment":"The sentence 'The exciton properties varies with the orientation' should be 'vary'.","section":"Introduction"},{"comment":"The x-axis 'Flake numbers' is not defined in the caption; please list the corresponding layer counts (bulk-like, multilayer, few-layer, etc.) for each flake number.","section":"Fig. 3h"},{"comment":"The pulse FWHM is given as 115 fs ± 10 fs and then fixed at 118 fs in the model; clarify whether 118 fs is the mean autocorrelation width, a deconvolved pulse duration, or a fitted value, and specify the chirp compensation procedure.","section":"Methods"},{"comment":"The fitting function for the FWM amplitude is referenced but not shown in the main text; include the full expression or at least explicitly define γ, σ, and the pulse convolution, so the reader can assess the deconvolution assumptions.","section":"SI Sec. 3"},{"comment":"Reference [39] is used to constrain E0, but the main text does not state the uncertainty of this constraint; please disclose the range or error bar adopted from that work.","section":"References"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":null,"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Resonant four-wave mixing in layered ReS2 measures sub-picosecond exciton coherence times, with homogeneously limited linewidths and coherence persisting to room temperature.","keywords":["rhenium disulfide","exciton coherence","four-wave mixing","homogeneous linewidth","anisotropic excitons","two-dimensional semiconductors","population dynamics","direct band gap"],"falsifier":"Measure the same ReS2 flakes with substantially shorter pulses (below 50 fs) and check whether the extracted gamma remains 5 to 8 meV and the photon echo remains absent; if gamma changes or an echo appears, the homogeneous-limit claim is an artifact of pulse convolution. An independent check compares the FWM-derived gamma with the low-temperature photoluminescence or absorption linewidth, where a large discrepancy would reveal unaccounted inhomogeneous broadening.","tokens_in":13887,"feed_emoji":"🔬","tokens_out":4598,"duration_ms":44104,"temperature":0.7,"pith_summary":"This paper reports direct measurements of the coherence time T2 and population lifetimes T1 for anisotropic excitons in layered ReS2 using resonant four-wave mixing micro-spectroscopy. It finds T2 in the hundreds of femtoseconds, corresponding to homogeneous linewidths of about 5 to 8 meV at 10 K, with inhomogeneous broadening no larger than 1 meV. The coherence remains discernible up to room temperature and is unusually robust against optical power, while the population dynamics span from about 150 fs to nanoseconds. The authors argue these results indicate a low-disorder material and a direct band gap that does not depend on layer thickness.","feed_headline":"ReS2 excitons stay coherent up to room temperature","feed_subtitle":"Four-wave mixing finds homogeneous linewidths and sub-picosecond T2 in layered ReS2.","key_machinery":"The central object is heterodyne-detected four-wave mixing (FWM) micro-spectroscopy using roughly 115 fs pulses: a two-pulse sequence measures coherence decay as a function of inter-pulse delay tau_12, while a three-pulse sequence measures population decay as a function of tau_23. The FWM signal is fit with a model that convolves the pulse shape with a system response consisting of exponential homogeneous decay (gamma) and Gaussian inhomogeneous broadening (sigma). The absence of a photon echo, signaled by the FWM maximum appearing at real time t = 0 rather than t = tau_12, is the diagnostic that sigma is much smaller than gamma.","core_discovery":"The central claim is that the excitonic linewidth in pristine exfoliated ReS2 is homogeneously limited: the four-wave-mixing signal shows free induction decay with no photon echo, and fitting gives a homogeneous width gamma of about 5 to 8 meV with inhomogeneous broadening sigma <= 1 meV across mono- to bulk-like flakes. From gamma they derive T2 = 2*hbar/gamma, placing T2 in the hundreds of femtoseconds. They further claim T2 depends only weakly on layer thickness, T1 is quasi-independent of thickness, and coherent signatures persist to room temperature, which they attribute to weak optical-phonon coupling. These observations lead them to conclude that ReS2 has a direct band gap regardless of layer thickness.","pith_inferences":["If the homogeneous limit is genuine, bare exfoliated ReS2 flakes could serve as a room-temperature platform for anisotropic exciton quantum optics, including polaritonics and coherent control experiments, without needing encapsulation.","The near-linear temperature dependence of gamma with weak optical-phonon coupling suggests quasi-one-dimensional dephasing along the Re chains; this could be tested by measuring T2 in ReSe2 or in ReS2 under uniaxial strain along the b-axis.","The assignment of the roughly 150 fs T1 component to bright-to-dark scattering could be probed directly by resolving the dark-state population with time-resolved photoluminescence or by applying a magnetic field to change the bright-dark splitting."],"forward_implications":["ReS2 reaches the homogeneous limit without hexagonal boron nitride encapsulation, unlike MoSe2 and WSe2 monolayers.","Coherent exciton signatures remain measurable at room temperature, suggesting feasible room-temperature coherent optoelectronics.","The weak thickness dependence of T2 means multilayer and bulk-like flakes retain monolayer-like coherence properties.","The population dynamics with roughly 150 fs, 0.5 to 1 ps, and nanosecond components implicate bright-to-dark scattering as the dominant non-radiative channel.","The comparison with indirect-gap MoSe2 bilayers supports a direct band gap in ReS2 across all measured thicknesses."],"supporting_citations":[{"why":"Provides the four-wave-mixing methodology and the MoSe2 monolayer photon-echo baseline with large inhomogeneous broadening that ReS2 is contrasted against.","marker":"[29]"},{"why":"Establishes the homogeneous-limit behavior in hBN-encapsulated MoSe2 and WSe2 monolayers, serving as the comparison for bare ReS2.","marker":"[32]"},{"why":"Describes the heterodyne detection with spectral interference that underpins the FWM micro-spectroscopy method.","marker":"[37]"},{"why":"Documents monolayer-like electronic and vibrational decoupling in bulk ReS2, the basis for the weak thickness dependence claim.","marker":"[9]"},{"why":"Identifies the linearly polarized anisotropic excitons EX1 and EX2 in ReS2, used for the polarization-resolved FWM analysis.","marker":"[10]"},{"why":"Provides the spin-dark state energies and exchange-split Rydberg series used to interpret the population dynamics and direct-gap conclusion.","marker":"[20]"},{"why":"Reports resonant internal quantum transitions and femtosecond radiative decay in WSe2 monolayers, giving the comparison for the ultrafast T1 component.","marker":"[44]"},{"why":"Shows order-of-magnitude shorter T1 and T2 in indirect-gap MoSe2 bilayers, the key comparison for the direct-gap argument.","marker":"[46]"}],"fun_headline_variants":["Sub-picosecond exciton coherence in ReS2 at room temperature","No photon echo in ReS2: homogeneous exciton lines","ReS2 layer thickness barely affects exciton coherence","Direct band gap in ReS2 confirmed by homogeneous exciton coherence"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The extracted homogeneous linewidths depend on fitting the FWM decays with a model of exponential coherence decay plus Gaussian inhomogeneous broadening convoluted with a roughly 115 fs Gaussian pulse; if the finite pulse duration (comparable to T2 in the monolayer) or an unmodeled nonresonant contribution distorts the decay, the homogeneous-linewidth interpretation would be compromised.","fun_headline_variants_meta":{"raw":{"variants":["Sub-picosecond exciton coherence in ReS2 at room temperature","No photon echo in ReS2: homogeneous exciton lines","ReS2 layer thickness barely affects exciton coherence","Direct band gap in ReS2 confirmed by homogeneous exciton coherence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001386,"raw_usage":{"total_tokens":5593,"prompt_tokens":913,"completion_tokens":4680,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":529,"completion_tokens_details":{"reasoning_tokens":4620}},"tokens_in":529,"tokens_out":4680,"duration_ms":31929,"temperature":1.0,"reasoning_tokens":4620,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:59:42.431196+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the same ReS2 flakes with substantially shorter pulses (below 50 fs) and check whether the extracted gamma remains 5 to 8 meV and the photon echo remains absent; if gamma changes or an echo appears, the homogeneous-limit claim is an artifact of pulse convolution. An independent check compares the FWM-derived gamma with the low-temperature photoluminescence or absorption linewidth, where a large discrepancy would reveal unaccounted inhomogeneous broadening.","supporting_citations":[{"cited_title":"& Kasprzak, J","cited_arxiv_id":null,"evidence_quote":"Provides the four-wave-mixing methodology and the MoSe2 monolayer photon-echo baseline with large inhomogeneous broadening that ReS2 is contrasted against."},{"cited_title":"& Kasprzak, J","cited_arxiv_id":null,"evidence_quote":"Establishes the homogeneous-limit behavior in hBN-encapsulated MoSe2 and WSe2 monolayers, serving as the comparison for bare ReS2."},{"cited_title":"& Kasprzak, J","cited_arxiv_id":null,"evidence_quote":"Describes the heterodyne detection with spectral interference that underpins the FWM micro-spectroscopy method."},{"cited_title":"S., Ho, C","cited_arxiv_id":null,"evidence_quote":"Documents monolayer-like electronic and vibrational decoupling in bulk ReS2, the basis for the weak thickness dependence claim."},{"cited_title":"A., Van Der Zande, A","cited_arxiv_id":null,"evidence_quote":"Identifies the linearly polarized anisotropic excitons EX1 and EX2 in ReS2, used for the polarization-resolved FWM analysis."},{"cited_title":"O., Rodríguez -Fernández, C., Jadczak, J., Bryja, L., Faugeras, C., Basko, D","cited_arxiv_id":null,"evidence_quote":"Provides the spin-dark state energies and exchange-split Rydberg series used to interpret the population dynamics and direct-gap conclusion."},{"cited_title":"& Huber, R","cited_arxiv_id":null,"evidence_quote":"Reports resonant internal quantum transitions and femtosecond radiative decay in WSe2 monolayers, giving the comparison for the ultrafast T1 component."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows order-of-magnitude shorter T1 and T2 in indirect-gap MoSe2 bilayers, the key comparison for the direct-gap argument."}],"review_version":1}