{"id":"f9efcd29-76f9-4e3c-95de-f04eb4e85483","arxiv_id":"2607.15411","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A new time-dependent GW method with full electron-phonon coupling across the Brillouin zone reproduces the phonon-mediated K-to-Q intervalley exciton transfer in monolayer WSe2 on a ~0.5 ps timescale.","lead":"This paper introduces a new first-principles method that simulates how excitons move across a material's momentum space while interacting with lattice vibrations in real time. Applied to monolayer WSe2, it reproduces the ultrafast (~0.5 ps) transfer of excitons from the bright K-valley to the dark Q-valley seen in experiments.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Classical fixed-bath phonon treatment (Eq. 10) and single-trajectory phases could shift the claimed 0.5 ps K-to-Q transfer timescale; needs quantum/back-reaction check.","rationale":"After reading the paper, the most load-bearing assumption for the central quantitative claim is indeed the treatment of phonons as a classical fixed bath in Eq. (10). The intervalley transfer is a phonon-driven rate process; the classical replacement erases the distinction between phonon absorption and emission, a distinction that is central at 100 K where the thermal occupation of the relevant modes can be suppressed. The paper's own Discussion admits overestimation of absorption at low temperature, and the neglect of back-reaction leaves the phonon distribution unfixed. Moreover, the reported dynamics is a single random-phase trajectory, so the 38/222 fs peak times are not accompanied by uncertainty. This is exactly the reader's weakest_assumption. The missing SI limits verification of derivation and q-grid convergence, but the physical model of the lattice bath is the deepest uncertainty. The proposed test—a quantum-kinetic Boltzmann comparison—would directly quantify the error introduced by the classical approximation. Therefore, the verdict remains conditional: the method and central result are plausible, but the phonon modeling must be validated or improved.","tokens_in":12909,"tokens_out":10285,"duration_ms":112033,"concrete_test":"Using the same GW band structure and e-ph matrix elements, compute the K-to-Q exciton transfer timescale with a quantum-kinetic Boltzmann equation that includes Bose-Einstein phonon occupations at 100 K and explicit absorption/emission asymmetry. Compare the time at which the Q-valley exciton population peaks and the I_K/I_Q ratio at 500 fs with the TD-aGW-ph results. If the quantum-kinetic transfer time differs from ~0.5 ps by more than ~30%, the classical fixed-bath approximation of Eq. (10) is load-bearing.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central simulation result—a ~0.5 ps phonon-mediated K-to-Q exciton transfer in WSe2—rests on Eq. (10), where lattice phonons are replaced by a classical, fixed-temperature reservoir: ⟨â+â†⟩ → √(4n+2) cos(ωt+φ), with no electronic back-reaction on the phonon equations of motion. The paper itself states this treatment 'does not distinguish the asymmetry between quantum absorption and emission processes' and may 'overestimate phonon-absorption channels ... at low temperature.' Since the K-to-Q transfer rate is governed by phonon absorption/emission, an overestimate of the absorption channel would accelerate the predicted transfer, potentially biasing the claimed ~0.5 ps timescale and the K/Q intensity ratio (Fig. 3). The back-reaction neglect further prevents the phonon bath from being modified as excitons scatter, which could affect the long-time spectral widths (Fig. 3h) and the eventual quasi-equilibrium ratio. Additionally, the reported dynamics is a single random-phase realization; without an ensemble average, the peak times (38 fs, 222 fs) and the sub-100 fs oscillations may not be robust. Thus the quantitative accuracy of the central claim is contingent on the least-secure assumption in the phonon model.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces TD-aGW-ph, a first-principles method that generalizes the time-dependent adiabatic GW approach to include coherent electron-hole and electron-phonon couplings across the full Brillouin zone. The central equation of motion, Eq. (8), propagates the interacting one-particle density matrix in a Bloch basis with finite-momentum components, where the interaction matrix element Eq. (9) combines the e-h kernel with phonon-modulated couplings. Phonons are treated classically through the replacement in Eq. (10), with a fixed thermal bath and random initial phases, and electronic back-reaction on the phonons is neglected. The method is implemented in BerkeleyGW and applied to monolayer WSe2 in a tr-ARPES pump-probe geometry. The simulations show that a bright K-valley exciton transfers to dark Q-valley excitons within approximately 0.5 ps, producing a K-to-Q in-gap photoemission intensity transfer that is absent when phonons are removed. The paper claims this is the first ab initio real-time treatment of coherent exciton-phonon dynamics with full finite-momentum coupling.","tokens_in":13266,"tokens_out":4258,"duration_ms":45046,"significance":"If the central claims hold, this is a substantial methodological advance: it enables real-time first-principles simulation of coupled exciton-phonon dynamics with finite-momentum coherence, a capability previously inaccessible to GW-BSE-based approaches. The qualitative physics is compelling: the K-to-Q intervalley transfer emerges naturally in the TD-aGW-ph calculation and disappears in the clamped-nucleus TD-aGW limit, establishing that phonons are the essential driver. The implementation in a widely used code, the primitive-cell formulation, and the explicit attribution of most K-valley dephasing to intervalley exciton-phonon scattering are all valuable. However, the quantitative accuracy of the results is currently contingent on two approximations: the classical fixed-bath phonon treatment with a single random-phase trajectory, and the phenomenological dephasing time fitted to the same experiment used for validation. These approximations do not invalidate the qualitative mechanism, but they do limit the strength of the quantitative claims.","major_comments":[{"comment":"The central quantitative predictions — the ~0.5 ps K-to-Q transfer timescale, the IK/IQ ratio in Fig. 3e, and the spectral widths dK, dQ in Fig. 3h — rest on replacing the phonon operators by a classical, fixed-temperature cosine with random phases, Eq. (10), and on a single realization of those phases. The manuscript itself concedes that this treatment 'does not distinguish the asymmetry between quantum absorption and emission processes' and 'may overestimate phonon-absorption channels... at low temperature.' Because the K-to-Q transfer rate is governed by the phonon absorption/emission balance, this caveat can bias the transfer timescale and the valley intensity ratio, not just the long-time line shape. The authors should (i) compare against a quantum Bose-Einstein-resolved treatment or a phonon back-reaction calculation to quantify the error, and (ii) provide at least a small ensemble","section":"Method, Eq. (10); Results, Fig. 3"},{"comment":"The dephasing time tau_deph = 2.2 ps is fitted to the experimental decay of IK + IQ and then inserted into the simulated spectra that are compared with that same experiment. This makes the aggregate intensity decay agreement partly circular. The valley-resolved K-to-Q transfer and the IK/IQ ratio are not fitted and constitute the more informative signals, but the paper should (i) state explicitly that the total-intensity comparison is not an independent validation, (ii) show the sensitivity of IK(t), IQ(t), and IK/IQ to tau_deph (e.g., by varying tau_deph around 2.2 ps, or showing the no-dephasing curve alongside a larger dephasing), and (iii) where possible, determine tau_deph from a separate observable. The qualitative intervalley transfer claim does not depend on tau_deph, but the claim of 'excellent agreement with experiment' in the abstract and conclusion does.","section":"Results, Fig. 3b and text near 'Dephasing time'"},{"comment":"The linearization of the e-h interaction with respect to phonon displacements invokes the approximation ∂K_e-h/∂u = 0. This is a standard ansatz in exciton-phonon work, but here it is used for a nonequilibrium, finite-momentum transfer calculation in which phonon-induced changes in screening or in the e-h kernel could alter the relative energies and couplings of K- and Q-valley excitons. Because the method is advertised as first-principles and the quantitative transfer rate is sensitive to such details, the authors should provide a frozen-phonon test of this assumption — for example, recomputing the lowest K and Q exciton energies and e-h kernels at representative displaced geometries — to show that the neglected terms are numerically small. Currently this assumption is untested and is part of the quantitative claim.","section":"Method, Eq. (11)"}],"minor_comments":[{"comment":"There is a typo 'tau_deth' in the text near the dephasing discussion; it should be 'tau_deph'. Also, 'e-hexcitations' should read 'e-h excitations'.","section":"Introduction and Fig. 3b caption"},{"comment":"The derivation of the central equation of motion and the explicit forms of M, the TDA truncation, and the tr-ARPES expression are relegated to the SI. The main text would benefit from a compact summary of the key derivation steps and equation numbers in the SI, since the formal structure is the paper's main contribution.","section":"Method, Eq. (8)"},{"comment":"The experimental data in Fig. 3b are normalized to the theoretical IK+IQ at 1000 fs. This normalization choice affects the apparent agreement in panels c–e. Please state the normalization protocol more prominently and, if possible, show an unnormalized or independently normalized comparison.","section":"Results, Fig. 3b"},{"comment":"The pump-pulse description (1.85 eV, 50 fs cosine-squared envelope) should include the fluence or peak field amplitude and the polarization convention used in the simulation, as these affect the initial exciton density and therefore the subsequent dynamics.","section":"Results, 'We use the new TD-aGW-ph method...'"}],"recommendation":"major_revision","confidential_remarks":"The manuscript presents a genuinely new methodological capability with a plausible qualitative mechanism. The main risk is not the conceptual framework but the validation strategy: the quantitative comparison relies on a classical phonon bath and a fitted dephasing time, both of which can bias the very quantities used to claim agreement. A revision that adds sensitivity studies, ensemble averaging, and a clearer separation between fitted and predicted observables would substantially strengthen the paper. I would not support rejection at this stage."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is not a routine application. The authors extend TD-aGW to include finite-momentum e-ph coupling via a linear-response expansion against phonon displacements, and the resulting EOM (Eq. 8) couples all momentum sectors. The WSe2 simulation shows a K-to-Q transfer that disappears when phonons are removed, so the mechanism is genuinely phonon-mediated. The core physics is probably right.\n\nWhat is new and good: the finite-momentum density-matrix coherence ρ_KQ is a new object in real-time first-principles dynamics. The primitive-cell formulation keeps the cost manageable (albeit O(N_q^2) on top), and the implementation sits in BerkeleyGW, so the method is concrete and in principle checkable. The qualitative agreement with tr-ARPES intensity transfer is real, and the ~0.5 ps transfer timescale emerges without being fit.\n\nThe soft spots are real but not fatal. The 2.2 ps dephasing is fitted from the very experiment used for comparison, so the aggregate decay comparison is partially circular; the valley-resolved transfer and the K/Q ratio are less affected. The classical phonon bath with no back-reaction, Eq. (10), is the load-bearing assumption for the quantitative timescale. The paper itself notes this may overestimate phonon absorption at low temperature, which would bias the transfer rate and the long-time spectral widths. A sensitivity check with quantum statistics or phonon feedback is missing. The single-trajectory dynamics means the quoted 38 fs and 222 fs peak times are one realization; the authors acknowledge ensemble averaging would smooth oscillations, but that also means those peak times are not averaged quantities. Finally, the derivation and convergence details are all in an SI that does not accompany the arXiv posting; referees cannot verify the central EOM without it.\n\nBottom line: this is a serious method paper with a plausible demonstration. It deserves a full referee process, but referees should ask for the SI, a sensitivity analysis of the phonon model, and an ensemble average (or at least a discussion of its effect) before the quantitative claims can be accepted. I would bring it to a reading group if the SI becomes available, and I would cite it as the first full-BZ real-time exciton-phonon method.","headline":"Genuinely new full-BZ real-time exciton-phonon method with a plausible WSe2 demonstration; quantitative claims are softer than the text suggests due to a fitted dephasing and a classical phonon bath.","tokens_in":13759,"tokens_out":2664,"would_cite":true,"duration_ms":27190,"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 a first-principles density-matrix method can simulate coherent exciton-phonon dynamics across the full Brillouin zone in real time, and demonstrates a ~0.5 ps intervalley exciton transfer in monolayer WSe2.","keywords":["time-dependent GW","exciton-phonon coupling","intervalley dynamics","tr-ARPES","monolayer WSe2","density matrix","nonequilibrium Green's function","finite-momentum coherence"],"falsifier":"Direct tr-ARPES measurement of monolayer WSe2 at low temperature (e.g., 10-20 K) tracking K- and Q-valley in-gap intensities: if the K-to-Q transfer takes longer than ~1 ps or the Q-valley signal fails to build up within 1 ps, the classical-phonon approximation underlying the claimed timescale would be contradicted. Alternatively, a quantum-kinetic calculation that accounts for phonon absorption/emission asymmetry at low T and yields a qualitatively different transfer time would fault the central claim.","tokens_in":12801,"feed_emoji":"⚛️","tokens_out":4114,"duration_ms":36780,"temperature":0.7,"pith_summary":"The paper introduces a first-principles method, TD-aGW-ph, that combines time-dependent adiabatic GW theory with electron-phonon coupling, allowing coherent exciton and phonon dynamics to be simulated in real time with full momentum resolution across the Brillouin zone. The authors use it to show that in monolayer WSe2, a pump-created bright K-valley exciton transfers to dark Q-valley excitons within roughly half a picosecond, mediated by thermal phonons. The calculated time-resolved photoemission intensities match published experiments, and the method attributes most of the bright-exciton dephasing to this intervalley transfer rather than to ad hoc dephasing parameters. If the approach holds, it gives a practical unified framework for simulating coupled electron-hole and electron-phonon nonequilibrium dynamics from first principles.","feed_headline":"WSe2 excitons switch valleys in under 0.5 ps","feed_subtitle":"New first-principles method couples excitons and phonons across momentum space, matching ultrafast photoemission.","key_machinery":"The central object is the interacting single-particle density matrix rho(t) in the Bloch band basis, propagated by a Liouville-von Neumann equation with a GW-level Hamiltonian. The new ingredient is the inclusion of finite-momentum (q != 0) density-matrix elements rho_{mk+q,nk}(t), driven by electron-phonon coupling matrix elements g_{mn nu}(k,q) and by the time-dependent electron-hole interaction kernel delta V^{e-h}; phonons enter classically as oscillating potentials with random initial phases. This allows the full Brillouin zone to be described within a primitive unit cell.","core_discovery":"The central claim is that coherent electron-hole (excitonic) and electron-phonon couplings can be incorporated into a real-time density-matrix equation of motion without needing supercells, by treating phonon perturbations in linear response and using Bloch states at equilibrium positions as the basis. This yields finite-momentum density-matrix elements that connect valleys (e.g., K to Q), and for monolayer WSe2 the resulting simulation shows a phonon-mediated direct-to-indirect exciton transition in about 0.5 ps, with in-gap photoemission intensity transferring from K to Q valleys. The paper further claims that most of the dephasing of the bright K exciton is due to this intervalley channel","pith_inferences":["The same linear-response density-matrix formalism could be extended to heterostructures and moiré systems, where finite-momentum exciton transfer between layers is the central process.","Including phonon back-reaction (allowing the lattice to respond to the electronic excitation) would likely alter the long-time spectral line shape and could reveal self-trapping regimes; the paper leaves this as future work.","The classical-phonon replacement may be the main source of quantitative error at low temperature; a fully quantum treatment should suppress spurious phonon-absorption channels and could slow the predicted transfer timescale.","Because the computational cost scales as N_q^2 relative to TD-aGW, the method is most practical for materials with moderate phonon-momentum grids; exploratory calculations will need to benchmark grid convergence."],"forward_implications":["If correct, TD-aGW-ph makes intervalley and finite-momentum exciton transfer directly simulable from first principles for a wide class of semiconductors, not just WSe2.","The method assigns most of the bright-exciton dephasing in WSe2 (over 80 percent in inverse-dephasing terms) to phonon-mediated K-to-Q transfer, reducing the need for fitted dephasing parameters in TD-aGW simulations.","The sub-100 fs oscillations seen in calculated valley intensities are predicted to be a single-trajectory coherent effect that would smooth out in ensemble-averaged experimental samples.","The classical-phonon treatment predicts a time-dependent spectral broadening from instantaneous level fluctuations, giving a testable line-shape signature beyond integrated intensities."],"fun_headline_variants":["Phonons tip WSe2 excitons from K to Q in half a picosecond","New GW method simulates exciton-phonon valley hops in WSe2","Simulation: Phonons drive WSe2 excitons across valleys in ~0.5 ps","Exciton-phonon coupling in WSe2: valley switch in 0.5 ps","Real-time GW shows phonons move WSe2 excitons between valleys"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The phonons are treated as a classical, fixed-temperature bath with random initial phases, ignoring quantum phonon statistics and any back-reaction of electrons on the lattice; if those neglected effects are significant, the predicted 0.5 ps transfer time and spectral broadening could shift.","fun_headline_variants_meta":{"raw":{"variants":["Phonons tip WSe2 excitons from K to Q in half a picosecond","New GW method simulates exciton-phonon valley hops in WSe2","Simulation: Phonons drive WSe2 excitons across valleys in ~0.5 ps","Exciton-phonon coupling in WSe2: valley switch in 0.5 ps","Real-time GW shows phonons move WSe2 excitons between valleys"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001188,"raw_usage":{"total_tokens":4777,"prompt_tokens":818,"completion_tokens":3959,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":562,"completion_tokens_details":{"reasoning_tokens":3857}},"tokens_in":562,"tokens_out":3959,"duration_ms":26186,"temperature":1.0,"reasoning_tokens":3857,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T23:25:57.381571+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Direct tr-ARPES measurement of monolayer WSe2 at low temperature (e.g., 10-20 K) tracking K- and Q-valley in-gap intensities: if the K-to-Q transfer takes longer than ~1 ps or the Q-valley signal fails to build up within 1 ps, the classical-phonon approximation underlying the claimed timescale would be contradicted. Alternatively, a quantum-kinetic calculation that accounts for phonon absorption/emission asymmetry at low T and yields a qualitatively different transfer time would fault the central claim.","supporting_citations":[],"review_version":1}