{"id":"09e9358d-a53a-4290-b775-83e38500402a","arxiv_id":"2509.04900","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Time-resolved reflectivity in NiPS3 reveals two relaxation channels, attributed to spin-orbit entangled exciton coherence (1-9 ps) and spin reordering (1-4 ns), with the slow channel showing critical slowing down near the Neel temperature.","lead":"Researchers used ultrafast laser pulses to watch how excitons in the 2D antiferromagnet NiPS3 lose coherence and how magnetic order recovers after photoexcitation, finding signatures of critical slowing down near the magnetic transition. The results suggest that magnetic fluctuations directly control exciton stability, which matters for designing optically controllable correlated 2D materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Slow-channel spin-reordering assignment rests on a free-TN power-law fit and lacks a magnetic control; a non-magnetic thermal recovery origin is not excluded.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing premise: the assignment of τ2 to spin reordering and its critical divergence rests on a three-parameter power-law fit and an under-constrained exponent comparison, with no magnetic control. I agree with that identification. My independent reading of the paper confirms that this is the most consequential weakness: the headline claim of dynamical coupling between SOEE and AFM order depends on τ2 being a magnetic observable. The abstract and main text state that critical slowing down near TN is the key evidence; if τ2 instead reflects thermal or phonon-mediated recovery that merely becomes slow near the structural/lattice anomaly at the same temperature, the central claim loses its magnetic specificity. I also noted an internal inconsistency the reader flagged: the main text says τ1 decreases monotonically from ~9 ps to ~4 ps between 75 K and 120 K and then gradually toward room temperature, while SM S7 says τ1 increases by 20–25% in the 135–150 K range and attributes this to magnetic stabilization of phase coherence. Fig. 3(b) is described in the main text as a gradual decrease, so the claimed stabilization near TN is asserted in the supplement but not visible in the main-text description; this directly affects the coupled-order-parameter interpretation. The Ginzburg-Landau section also says τ1 is associated with phase relaxation of the exciton order parameter, but the free energy depends only on |ψ|², and the supplement admits the phase degree of freedom does not appear in the relaxation time; this is a real gap between the phenomenological model and the claimed assignment. The power-law fit details in SM S8 are themselves mutually inconsistent (m = 0.44 ± 0.04 in the log-log text versus m = 0.44 ± 0.18 averaged from the split fits, and the upper-side uncertainty of ±0.28 is too large to confirm a universality class). The concrete test I propose is a magnetic-field dependence of τ2 near TN, which would directly discriminate spin-critical slowing down from thermal or phonon recovery; I also propose the cheaper check of fixing TN to the independently measured 153 K and comparing the split exponents. Agreement with the reader is 'agree' because we identify the same weakest premise, and the verdict should remain CONDITIONAL: the paper is publishable as a careful study of temperature-dependent transient reflectivity with a plausible spin-exciton coupling narrative, but the central claim requires the magnetic control experiment or a more rigorous fixed-TN scaling analysis before it can be regarded as established. I do not see a reason to move to REJECT: the data quality, the internal consistency of the two-pump-probe configurations, and the independent support from literature timescales for τ1 give real weight to the interpretation.","tokens_in":17967,"tokens_out":2587,"duration_ms":20781,"concrete_test":"Measure τ2(T) in the same crystal under an applied magnetic field of ~5–10 T along the easy axis, or in the isostructural non-magnetic counterpart (e.g., MnPS3 is still magnetic; better: use a non-magnetic analog such as FePSe3 is also magnetic—so instead use magnetic-field suppression of TN or a diamagnetic isostructural compound if available, or measure the same sample above TN with a field that cannot affect phonon/thermal recovery). If τ2(T) near TN is unchanged under a field that shifts TN, the magnetic assignment fails; if the divergence tracks the field-shifted TN, the spin origin is confirmed. Simpler analytical check: refit the SM S8 data with TN fixed to the independently measured 153 K from Fig.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that the slow component τ2 (1–4 ns) specifically tracks antiferromagnetic spin reordering and its critical slowing down near TN. The support is a power-law fit τ2 = [Δ(1-T/TN)^m]^-1 with three free parameters (m, Δ, TN) in the main-text inset of Fig. 3(d) and SM S8, plus the coincidence of m ≈ 0.44 with the 3D Heisenberg exponent. The main text quotes m = 0.44 ± 0.18, while SM S8 gives m = 0.44 ± 0.04 from only one side and reports asymmetric split fits (below TN: m = 0.44 ± 0.08; above TN: m = 0.43 ± 0.28). The upper-side uncertainty is so large that the claimed universality-class confirmation is not established. Moreover, TN is a free parameter (fit value 157 ± 4.5 K, while susceptibility gives 153 K and the text elsewhere uses 155 K), so a divergence enforced near a freely chosen TN can mimic critical slowing down even for a non-magnetic process such as thermal diffusion or phonon-bottleneck recovery. Because the same τ2 is observed in the anisotropic (Kerr) configuration, the spin origin is plausible but still not uniquely established: any spin-coupled lattice or carrier recovery shows the same temperature trend. The paper explicitly notes (SM S5) that the Kerr signal contains a phonon-modulated component, so a purely phonon/thermal contribution to τ2 has not been isolated. No control measurement (e.g., applied magnetic field, comparison with a non-magnetic isostructural compound, or a direct measurement of the spin correlation time) is presented.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports time-resolved non-degenerate isotropic and anisotropic pump-probe reflectivity measurements on the van der Waals antiferromagnet NiPS3 over 5-294 K. The authors observe a biexponential decay following 3.1 eV excitation, with a fast component (1-9 ps) assigned to spin-orbit entangled exciton (SOEE) coherence and a slow component (1-4 ns) assigned to spin reordering dynamics. They also observe ~27 GHz coherent acoustic phonons. The central claims are that the slow component shows critical slowing down near the Neel temperature with an exponent consistent with the 3D Heisenberg universality class, that the fast component loses coherence near the exciton dissociation temperature, and that the coupled temperature dependence demonstrates dynamical exciton-spin coupling. A Ginzburg-Landau free energy with coupled order parameters is proposed to interpret the results, and fluence-dependent measurements are used to argue for a competing many-body scenario near 135 K.","tokens_in":18360,"tokens_out":4128,"duration_ms":38402,"significance":"If the assignments are correct, the paper would provide direct time-domain evidence for dynamical coupling between spin-orbit entangled excitons and antiferromagnetic order in a 2D van der Waals magnet, with implications for nonequilibrium critical phenomena and optical control of correlated excitations. The experimental work includes useful strengths: repeatability checks at several temperatures, a linear fluence-dependence check of the peak signal, temperature- and fluence-dependent characterization of the acoustic phonon mode, and consistency between isotropic and anisotropic (Kerr) measurements. However, the interpretive claims are currently undermined by internal inconsistencies and by the absence of control measurements that would rule out non-magnetic origins for the slow relaxation component. The significance of the paper as a definitive demonstration of exciton-spin dynamical coupling is therefore not yet established, although the data set is valuable and likely to be of interest to the community.","major_comments":[{"comment":"The power-law fit of τ2 that underlies the claim of magnetic critical slowing down uses the Neel temperature as a free parameter, with the fit returning TN = 157 ± 4.5 K, while the susceptibility measurement in SM S1.3 gives TN = 153 K and the main text elsewhere adopts 155 K. With three free parameters (m, Δ, TN), a divergence can be accommodated even for a non-magnetic process such as thermal diffusion. Moreover, SM S8 reports separate fits above and below TN with m = 0.43 ± 0.28 on the upper side; this uncertainty is large enough that the claimed agreement with the 3D Heisenberg exponent is not established. The main-text value of m = 0.44 ± 0.18 also differs materially from the SM S8 value of m = 0.44 ± 0.04 obtained from the log-log plot. Please fix TN to a value determined independently, report the fit with fixed TN, quantify the sensitivity of m to the choice of TN, and state the uncertainty honestly; the current presentation overstates the universality-class confirmation.","section":"§3(d) inset and SM S8"},{"comment":"There is a direct contradiction about the behavior of the fast relaxation time τ1 in the range 135-150 K. The main text states that τ1 decreases from about 9 ps at low temperature to about 4 ps near 120 K and then gradually decreases toward room temperature, with no mention of an increase. But SM S7 states that τ1 increases by about 20-25% in the range 135 < T < 150 K and that this increase is the signature of magnetic stabilization of excitonic coherence, explicitly referring to Fig. 3(b) of the main text. This inconsistency is load-bearing because the increase near TN is used as evidence for a dynamical coupling between SOEE coherence and the antiferromagnetic order. Please reconcile the two descriptions and present the actual temperature dependence of τ1 with appropriate uncertainties; if the increase is present, it must be shown in the main-text figure and analysis, and if it is not reproducible, the SM S7 statement should be corrected.","section":"Main text, §3(b), vs. SM S7"},{"comment":"The assignment of the slow component τ2 to spin reordering dynamics is not uniquely established because no control measurement is provided that would exclude a thermal or phonon origin. The only supporting line of argument is the temperature dependence of τ2 and its similarity to a power-law divergence, but a phonon-bottleneck or thermal-diffusion recovery can also produce a growing relaxation time near TN. The anisotropy measurement in SM S5 is presented as supporting the spin origin, yet the same section explicitly notes that the Kerr signal contains a phonon-modulated component, so the two channels are not cleanly separated. A measurement under applied magnetic field, a comparison with a non-magnetic isostructural compound, or a direct measurement of the spin correlation time (e.g., via time-resolved Faraday/Kerr rotation with a magnetic pump) would be needed to support the central claim. At minimum, the possible non-magnetic contributions to τ2 should be quantitatively estimated and discussed.","section":"§3(d), SM S5, general assignment of τ2"}],"minor_comments":[{"comment":"There are several typographical errors: 'pronouns temperature dependence' should be 'pronounced temperature dependence', and 'exhibhits' should be 'exhibits'. These occur in the abstract and in §3(d) (or Fig. 4 discussion), respectively.","section":"Abstract and main text"},{"comment":"The text cites 'Ho et al. [31]' for micro-thermoreflectance studies of A1 band-edge excitons, but reference [31] is actually Chu et al., Nat. Mater. 16, 200 (2017), while the Ho, Hsu, and Muhimmah paper appears as reference [33]. Please correct the citation to match the intended source.","section":"Reference [31] in §3(b)"},{"comment":"The equations in SM S7 are presented with unusual formatting, such as 'τζ ——— 1/(aζΓζ)' and 'τψ ——— 1/(aψΓψ)', which appears garbled. These should be written as standard equations for clarity.","section":"SM S7"},{"comment":"The statement that at 135 K 'the pump excitation disrupts the spin fluctuations' and thereby prolongs τ2 is presented as an interpretation, but no microscopic calculation or independent measurement is given. Consider softening this claim or adding a more quantitative analysis of exciton-spin coupling in the fluence-dependent regime.","section":"Main text, §4 fluence dependence"}],"recommendation":"major_revision","confidential_remarks":"The central experimental dataset appears carefully acquired and is presented with useful reproducibility checks. The main concern is that the paper's headline claim of dynamical exciton-spin coupling rests on a small number of temperature-dependent parameters that are internally inconsistent between the main text and the Supplementary Material, and on a critical-exponent fit with a free TN and large uncertainties. These issues are fixable in revision if the authors can present a consistent τ1(T) curve and a more robust treatment of the τ2 divergence. If the inconsistency in τ1 cannot be resolved, the claim of magnetic stabilization of excitonic coherence should be withdrawn. A control experiment would substantially strengthen the paper but may be outside the present manuscript's scope; at minimum, the limitations of the current evidence should be stated explicitly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is a serious experimental study: a systematic temperature- and fluence-dependent transient reflectivity dataset on NiPS3 covering 5–294 K, with two clearly separated relaxation channels and a 27 GHz acoustic phonon. That dataset, and the two-channel decomposition, is a genuine step beyond earlier single-temperature measurements, and the observation of critical slowing down in the slow component near TN is worth taking seriously. The methods section is careful, and the analysis is reproducible in principle.\n\nThe soft spots are real, though. The most serious is an internal inconsistency: the main text describes tau1 as decreasing monotonically from ~9 ps at low T to ~3 ps at high T, but SM S7 claims an increase of 20–25% in the 135–150 K range and uses that increase as evidence for spin-mediated stabilization of the exciton. That range is exactly what the central coupling claim leans on. The reader cannot tell which description matches the data, and the authors need to resolve this before the interpretation can be assessed.\n\nThe second issue is the assignment of tau2 to spin reordering. The power-law fit has TN as a free parameter (fit value 157±4.5 K vs. 153 K from susceptibility), and the exponent is quoted as m=0.44±0.18 in the main text but split into m=0.44±0.08 below TN and m=0.43±0.28 above TN in the supplement. The upper-side uncertainty is so large that claiming confirmation of the 3D Heisenberg universality class is overreach. More importantly, a divergence near a freely chosen TN could also describe thermal diffusion or phonon-related recovery. The Kerr data help, but the paper itself notes the Kerr signal includes a phonon-modulated component, so the spin origin of tau2 is plausible, not proven. A control measurement—magnetic field, a non-magnetic isostructural compound, or a direct spin-correlation time—would have made the case.\n\nCitation mismatches are minor but annoying: refs [31]/[33] appear swapped, and the power-law formula citation points to a paper that does not obviously contain it. That kind of sloppiness should be cleaned up.\n\nAll that said, the paper is not a casual product. The GL theory is a reasonable interpretative framework, though post-hoc, and the claims are clearly stated. The paper deserves peer review because the dataset is valuable and the questions are important; it just needs major revision on the tau1 inconsistency and a much more cautious treatment of the exponent and the spin origin of tau2. I would send it out rather than desk-reject, and I would bring it to the reading group to discuss what controls would be needed to nail the interpretation.","headline":"Careful systematic transient reflectivity data on NiPS3 with two relaxation channels, but the central spin-coupling interpretation is undermined by an internal tau1 inconsistency and an overfit critical exponent.","tokens_in":18902,"tokens_out":3727,"would_cite":true,"duration_ms":30816,"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":"Spin-orbit entangled excitons in NiPS3 are dynamically coupled to the antiferromagnetic order: the exciton coherence collapses exactly as spin fluctuations slow down critically near the Néel temperature.","keywords":["spin-orbit entangled excitons","NiPS3","two-dimensional antiferromagnet","transient reflectivity","pump-probe spectroscopy","critical slowing down","exciton-spin coupling","van der Waals magnets"],"falsifier":"Perform the same pump-probe measurement while sweeping an external magnetic field through the Néel point, or on a non-magnetic isostructural compound such as ZnPS$_3$: if the divergence of $\\tau_2$ and the associated shortening of $\\tau_1$ survive unchanged, the attribution to magnetic critical slowing down is wrong. A direct time-resolved measurement of the spin correlation time (for instance by magnetic X-ray or neutron scattering) compared point-by-point with $\\tau_2(T)$ would settle the identification as well.","tokens_in":17696,"feed_emoji":"🧲","tokens_out":11291,"duration_ms":84633,"temperature":0.7,"pith_summary":"This paper claims that spin-orbit entangled excitons (SOEEs) in the layered antiferromagnet NiPS$_3$ are dynamically coupled to the magnetic order, not merely influenced by it. Temperature-dependent pump-probe reflectivity shows two relaxation channels after photoexcitation: a fast decay ($\\tau_1 = 1$–$9$ ps) assigned to SOEE coherence and a slow decay ($\\tau_2 = 1$–$4$ ns) assigned to spin reordering. The slow channel grows and diverges near the Néel temperature ($T_N = 155$ K) with critical slowing down, while the fast channel simultaneously collapses near the exciton dissociation temperature ($T_\\mathrm{ED} = 120$ K) and stays short above it. The authors read the coincidence of these two behaviours as the exciton coherence being destroyed by the same spin fluctuations that freeze at the transition. If correct, the work turns optical pump-probe into a tool for tracking and controlling correlated exciton-spin dynamics in two-dimensional antiferromagnets.","feed_headline":"Exciton coherence collapses as NiPS3 spins slow near 155 K","feed_subtitle":"Pump-probe data tie fast exciton decay to slow spin reordering across the Néel transition in this 2D antiferromagnet.","key_machinery":"The argument is carried by the decomposition of the measured $\\Delta R/R(t)$ into a bi-exponential decay with a common rise time, convolved with the Gaussian instrument response, plus a damped harmonic oscillator representing a 27 GHz longitudinal coherent acoustic phonon. The fast exponent is identified with SOEE coherence by benchmarking $\\tau_1$ against known exciton lifetimes and by the Rothwarf–Taylor fit; the slow exponent is identified with spin reordering by the power-law divergence of $\\tau_2$ near $T_N$ and by the matching energy scale and critical exponent. The conceptual engine is the Ginzburg–Landau free energy with two coupled order parameters, $\\psi$ for exciton coherence and $\\zeta$ for antiferromagnetic order, whose coupling term $\\lambda|\\psi|^2\\zeta^2$ lets the growth of spin order stabilise the exciton phase, and whose linearised time-dependent equation produces the critical slowing down observed in $\\tau_2$.","core_discovery":"On the paper's own terms, the central discovery is that the transient reflectivity of NiPS$_3$ pumped at 3.14 eV and probed at 1.57 eV separates into two physical channels whose temperature dependence pins them to distinct degrees of freedom. The fast component $\\tau_1$ matches the lifetime (~10 ps) of spin-orbit entangled excitons established by ultranarrow photoluminescence and THz studies, shortens from roughly 8–9 ps below $T_\\mathrm{ED} = 120$ K to about 3 ps above it with a tail persisting past $T_N$, and fits a Rothwarf–Taylor bottleneck expression with $\\Delta E = 66 \\pm 12$ meV, about half the reported 132 meV exciton binding energy. The slow component $\\tau_2$ grows near 120 K and follows $\\tau_2 = [\\Delta(1 - T/T_N)^m]^{-1}$ with $m = 0.44 \\pm 0.18$, an exponent the authors identify with the three-dimensional Heisenberg universality class, and $\\Delta \\approx 1.07$ meV, matching the spin-wave gap measured by neutron scattering and electron spin resonance. The paper's central evidence for dynamical coupling is that the SOEE coherence is lost precisely as spin fluctuations undergo critical slowing down near $T_N$, a feedback captured in a Ginzburg–Landau free energy of two coupled order parameters, the exciton coherence $\\psi$ and the AFM order $\\zeta$, interacting through $\\lambda|\\psi|^2\\zeta^2$.","pith_inferences":["A magnetic-field sweep through $T_N$ would test the coupling claim directly: the $\\tau_2$ divergence and the $\\tau_1$ collapse should both shift or smear with field if both channels are magnetic in origin, whereas a field-insensitive $\\tau_1$ would point to a separate decoherence source.","The framework invites an exponent for the coupling itself: measuring the exciton coherence time as a function of reduced temperature $|T - T_N|/T_N$ could yield its own critical exponent for the exciton-spin interaction, beyond the qualitative feedback the paper describes.","If the singlet-polaron picture is right, the fast channel may not be pure exciton dephasing but the decoherence of a magnetically dressed quasiparticle; a two-pulse (pump-pump) experiment could distinguish local dephasing from transport of the dressed state.","Repeating the same protocol on isostructural $M$PS$_3$ compounds with different magnetic orders (or on the non-magnetic member) would map how the sign and strength of the coupling term $\\lambda$ depend on the spin lattice, turning a single-material study into a design rule for exciton-magnet coupling."],"forward_implications":["Time-resolved reflectivity can separate excitonic coherence from spin-reordering dynamics in a 2D antiferromagnet, giving each degree of freedom its own optical readout in one measurement.","The slow-channel exponent $m \\approx 0.44$ places the spin fluctuations of NiPS$_3$ near $T_N$ in the three-dimensional Heisenberg universality class, connecting nonequilibrium optical data to equilibrium critical phenomena.","Because spin fluctuations govern the exciton decoherence time, external control of magnetic correlations (strain, fields, heterostructure stacking) should translate directly into control of exciton coherence.","The fluence experiments near 135 K show mirror-image behaviour — $\\tau_1$ shortens while $\\tau_2$ lengthens with increasing pump fluence — indicating competing many-body channels that optical excitation can steer.","The 27 GHz acoustic phonon mode softens with temperature but is inert to fluence, providing an internal lattice response that stays independent of the spin channel."],"supporting_citations":[{"why":"Reports the ultranarrow ~400 µeV photoluminescence linewidth that sets the ~10 ps SOEE lifetime benchmark anchoring the $\\tau_1$ assignment.","marker":"[1]"},{"why":"Optical pump–terahertz probe study giving a ~17 ps SOEE lifetime that corroborates the $\\tau_1$ timescale.","marker":"[16]"},{"why":"Provides the ~132 meV exciton binding energy against which the Rothwarf–Taylor $\\Delta E$ from the $\\tau_1$ fit is compared.","marker":"[19]"},{"why":"Supplies the Rothwarf–Taylor bottleneck model used to fit the temperature dependence of the fast relaxation time $\\tau_1$.","marker":"[36]"},{"why":"Gives the power-law form $\\tau_2 = [\\Delta(1 - T/T_N)^m]^{-1}$ used to extract the critical slowing down near $T_N$.","marker":"[31]"},{"why":"Provides the 3D Heisenberg critical exponent value used to identify the universality class of the $\\tau_2$ divergence.","marker":"[39]"},{"why":"Source of the time-dependent Ginzburg–Landau relaxation framework used in the coupled-order-parameter theory.","marker":"[40]"},{"why":"Neutron-scattering spin-wave gap value used to validate the energy scale $\\Delta$ extracted from the $\\tau_2$ power-law fit.","marker":"[41]"},{"why":"Theory of excitons dressed by the antiferromagnetic background (singlet polarons) that motivates the dynamical coupling picture.","marker":"[22]"}],"fun_headline_variants":["Excitons lose coherence as NiPS3 spins slow","Spin slowdown matches exciton collapse in NiPS3","Critical spin slowing kills exciton coherence in NiPS3","Two timescales tie exciton and spin order in NiPS3"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the slow nanosecond relaxation component is genuinely the antiferromagnetic spins reordering and that its growth near $T_N$ is magnetic critical slowing down; this identification rests on a three-parameter power-law fit whose exponent resembles the 3D Heisenberg value, without a control measurement such as an applied magnetic field or a non-magnetic isostructural crystal to rule out a non-magnetic origin such as thermal diffusion or phonon recovery.","fun_headline_variants_meta":{"raw":{"variants":["Excitons lose coherence as NiPS3 spins slow","Spin slowdown matches exciton collapse in NiPS3","Critical spin slowing kills exciton coherence in NiPS3","Two timescales tie exciton and spin order in NiPS3"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001296,"raw_usage":{"total_tokens":5391,"prompt_tokens":1148,"completion_tokens":4243,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":764,"completion_tokens_details":{"reasoning_tokens":4174}},"tokens_in":764,"tokens_out":4243,"duration_ms":24841,"temperature":1.0,"reasoning_tokens":4174,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:25:38.048302+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform the same pump-probe measurement while sweeping an external magnetic field through the Néel point, or on a non-magnetic isostructural compound such as ZnPS$_3$: if the divergence of $\\tau_2$ and the associated shortening of $\\tau_1$ survive unchanged, the attribution to magnetic critical slowing down is wrong. A direct time-resolved measurement of the spin correlation time (for instance by magnetic X-ray or neutron scattering) compared point-by-point with $\\tau_2(T)$ would settle the identification as well.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the ultranarrow ~400 µeV photoluminescence linewidth that sets the ~10 ps SOEE lifetime benchmark anchoring the $\\tau_1$ assignment."},{"cited_title":"Warshauer, H","cited_arxiv_id":null,"evidence_quote":"Optical pump–terahertz probe study giving a ~17 ps SOEE lifetime that corroborates the $\\tau_1$ timescale."},{"cited_title":"Warshauer, H","cited_arxiv_id":null,"evidence_quote":"Provides the ~132 meV exciton binding energy against which the Rothwarf–Taylor $\\Delta E$ from the $\\tau_1$ fit is compared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the power-law form $\\tau_2 = [\\Delta(1 - T/T_N)^m]^{-1}$ used to extract the critical slowing down near $T_N$."},{"cited_title":"Hamad, C","cited_arxiv_id":null,"evidence_quote":"Theory of excitons dressed by the antiferromagnetic background (singlet polarons) that motivates the dynamical coupling picture."}],"review_version":2}