{"id":"eaa10db2-2e0c-4aeb-bd92-a02a3b1a0a88","arxiv_id":"2607.17265","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A dissipative real-time TDDFT framework with first-principles electron-phonon collision integrals captures carrier relaxation and valley depolarization in Si and WS2 on femtosecond timescales.","lead":"This paper builds a way to add electron-phonon scattering into real-time simulations of electrons in crystals, so excited carriers can relax and lose coherence instead of evolving forever in a perfect quantum loop. It is useful because it connects first-principles scattering calculations to measurable ultrafast signals such as time-resolved photoemission.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Markovian assumption is strained in the very benchmarks used to validate the method, so the quantitative agreement with experiment is not a clean test.","rationale":"The reader's weakest_assumption correctly identifies the load-bearing concern: the Born–Markov approximation with an equilibrium phonon bath is used exactly in the regimes where its validity is questionable. This is also the most critical issue because the central claim rests on the quantitative benchmarks; if the benchmarks are not clean tests, the claim is under-supported. The paper itself explicitly acknowledges the Markovian limit violation for the Si momentum lifetimes, which strengthens the concern. The proposed test directly assesses whether memory effects alter the extracted lifetimes, which would settle whether the approximation is benign or limiting. I agree with the reader's CONDITIONAL verdict: the method is a credible extension, but the validation needs stronger evidence. No adjustment to the verdict is necessary.","tokens_in":17794,"tokens_out":6812,"duration_ms":65846,"concrete_test":"Re-run the bulk Si relaxation using the full non-Markovian collision integral (Eq. S13) instead of the Markovian limit, keeping all other settings (laser, active space, coarse-graining) identical. Compare the extracted energy and momentum lifetimes with the Markovian results. If the momentum lifetimes change by more than a factor of two, or the energy lifetimes by more than 30%, the Markovian approximation is load-bearing and the experimental agreement is not a clean validation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the framework provides a practical first-principles route to simulate relaxation and decoherence is validated primarily by quantitative agreement with experiment: Si energy lifetimes of 40/106 fs versus 30/120 fs, and WS2 intervalley lifetime of 17 fs versus 16 fs. However, the paper itself notes that the extracted Si momentum lifetimes (8 and 14 fs) approach the Markovian validity limit set by the maximum Si phonon energy (~65 meV, bath correlation time ~10 fs). The WS2 lifetime of 17 fs is similarly short compared to typical phonon correlation times. In both cases the extracted relaxation times are only marginally longer than the memory time, meaning the Born–Markov approximation is stretched exactly in the regime used for validation. Moreover, the transition rates W are computed from static equilibrium electron-phonon matrix elements and applied during strong-field driven dynamics, where the adiabatic basis may not remain continuously connected to the reference band states. The paper's own Discussion acknowledges these limitations but still presents the numerical agreement as verification. If non-Markovian effects or field-modified rates are significant on these timescales, the reported lifetimes could be shifted, and the validation would not cleanly test the framework.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a dissipative real-time TDDFT framework in which the unitary Kohn-Sham propagation in real space is supplemented by a Born-Markov electron-phonon collision integral acting on a reduced single-particle density matrix in an adiabatic active space. The collision integral is built from first-principles electron-phonon matrix elements, coarse-grained onto a computationally tractable k-grid, and applied to the diagonal (Boltzmann-like) and off-diagonal (dephasing) sectors. The method is benchmarked on hot-carrier relaxation in bulk Si and intervalley scattering in monolayer WS2, and it is combined with a tSURFF-based tr-ARPES simulation to connect the microscopic dynamics to a photoemission observable. The reported quantitative agreements are Si energy lifetimes of 40/106 fs versus experimental 30/120 fs, and a WS2 intervalley lifetime of 17 fs versus an experimental exciton depolarization time of 16 fs.","tokens_in":18101,"tokens_out":7557,"duration_ms":85316,"significance":"If the central claims hold, the framework would be a valuable practical tool: it extends rt-TDDFT beyond unitary dynamics while retaining real-space nonlinear and spectroscopic capabilities, and it derives the dissipative kernel from first-principles electron-phonon interactions rather than from phenomenological parameters. The paper ships open code and input files, which is a notable strength for reproducibility. The method's main value lies in providing a single framework for population relaxation, coherence decay, and time-resolved observables. However, the validation is not yet as clean as the text suggests: the benchmark lifetimes are comparable to the phonon bath correlation time, the WS2 comparison is against an excitonic observable rather than free-carrier scattering, and the strong-field applicability is asserted rather than quantitatively bounded.","major_comments":[{"comment":"The paper itself states that the extracted Si momentum lifetimes (8 and 14 fs) approach the Markovian validity limit set by the maximum Si phonon energy (~65 meV, ~10 fs bath correlation time). The WS2 intervalley lifetime of 17 fs is similarly only marginally longer than typical optical-phonon periods. Since the central validation rests on quantitative agreement with experiment for these short lifetimes, the agreement is not a clean test of the Born-Markov approximation. Please provide a quantitative estimate of the error incurred by the Markov and equilibrium-bath assumptions, e.g., by comparing the full memory integral in Eq. (S13) with the Markov result for the same rates, or by showing convergence of the extracted lifetimes with respect to the memory-time cutoff τ_E. Without such a check, the claim that the numerical agreement verifies the framework is overstated.","section":"Ultrafast carrier relaxation dynamics in silicon (Fig. 2d) and Ultrafast Valley Dynamics in WS2 (Fig. 3d)"},{"comment":"The experiment quoted for the 16 fs lifetime probes exciton valley depolarization, while the simulation describes single-particle free-carrier intervalley scattering and explicitly omits long-range electron-hole interactions. The paper acknowledges this simplification but still presents the 17 fs versus 16 fs agreement as validation. This comparison is not apples-to-apples: the agreement could be coincidental or arise from cancellation of errors. Please either provide a theoretical argument for why the single-particle intervalley rate should match the exciton depolarization time, or reframe the WS2 result as a qualitative/semi-quantitative benchmark and temper the validation claim accordingly.","section":"Ultrafast Valley Dynamics in WS2, comparison with Ref. [36]"},{"comment":"The main text states that Eq. (9) is the single-particle reduction of an underlying CPTP open-system evolution. The SI derives the general nonlinear quantum kinetic equation (S17) and then applies the secular approximation to obtain Eq. (S20) and the diagonal Boltzmann form. It is not demonstrated that the resulting diagonal + dephasing structure, with the state-dependent γ_n of Eq. (10), preserves complete positivity. The citation to Ref. [26] is not a proof. Since complete positivity is a central formal claim, either provide a direct proof for Eq. (9), state that positivity is only approximate and must be monitored, or explain how Ref. [26] covers this specific nonlinear secular form.","section":"Eq. (9) and SI Sec. S.1"},{"comment":"The transition rates W are computed from static equilibrium electron-phonon matrix elements, while the collision integral is applied in the instantaneous adiabatic basis of the driven system. The paper asserts that the adiabatic states remain continuously connected to the reference band states, but this is not quantified. For strong-field driving, the adiabatic Houston states are shifted in crystal momentum by the vector potential, so the e-ph matrix elements in the active space should in principle be evaluated at the instantaneous shifted momenta. The present benchmarks use moderate intensities, but the abstract and Discussion advertise the method for strong-field phenomena. Please specify the maximum field strength / vector-potential shift for which the static-rate approximation is valid, or describe how the rates are updated during the propagation.","section":"Coupling active-space dissipation to real-space propagation; Discussion"}],"minor_comments":[{"comment":"The heading 'A veraged Transition Rate' contains a typo; it should be 'Averaged Transition Rate'.","section":"Methods heading"},{"comment":"The residual states |d_μ(t)⟩ are constructed from the finite set of TDKS orbitals |ψ_μ(t)⟩, so Eq. (17) is a decomposition of the subspace spanned by the TDKS orbitals, not of the full single-particle Hilbert space. The statement 'we strictly preserve the Hilbert space completeness' is too strong and should be rephrased to refer to the completeness of the propagated orbital manifold.","section":"Eq. (16) and Methods 'Basis Construction'"},{"comment":"The text reports a CP-light lifetime of 20 fs and an LP-light lifetime of 17 fs, but the fits are not shown in Fig. 3(c,d). Please indicate the fit windows and, ideally, the fit uncertainties, since the extracted lifetimes are central to the validation.","section":"Fig. 3 and text"},{"comment":"The convolution model for the tr-ARPES intensity includes parameters A1, A2, and τ0, but no details are given about how these are constrained or how sensitive τ0 is to the choice of pump/probe envelopes. A brief sensitivity statement would be useful.","section":"Eq. (14)"},{"comment":"The manuscript contains placeholders 'State individual contributions' and 'Declare competing interests or state none.' These should be completed before publication.","section":"Author contributions / Competing interests"}],"recommendation":"major_revision","confidential_remarks":"This is a potentially useful methodological contribution with open code and data, but the validation story needs substantial strengthening. The most important issues are (i) demonstrating that the Markovian approximation is quantitatively reliable in the exact regime used for validation, and (ii) reconciling the WS2 comparison with the fact that the experiment measures exciton depolarization while the simulation is single-particle. The CPTP claim for Eq. (9) also needs proof or qualification. These are fixable within the manuscript's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth your time: the paper delivers a real implementation of Born-Markov electron-phonon collision integrals inside real-space rt-TDDFT (Octopus), with EPW-computed rates and an active-space/real-space split that is both practical and honest. The benchmarks are the first of their kind, but they don't cleanly validate the framework, because the Markovian assumption is stretched on the very timescales used for the headline agreement.\n\nThe genuinely new part is the integration. The collision integral itself derives from standard open-quantum-system theory (Simoni et al. and others), but combining it with a full real-space TDKS propagation, an adiabatic active space with residual-space completion, natural-orbital reconstruction, and tSURFF-based tr-ARPES is not something I've seen done. The Si results give a clean separation between fast momentum redistribution and slower energy relaxation, which is physically appealing. The authors also deserve credit for spelling out their own limitations: they note that the 8-14 fs momentum lifetimes approach the ~10 fs phonon memory time, that the WS2 comparison is free-carrier vs. exciton, and that equilibrium rates are applied during driven dynamics.\n\nThe soft spots are proportionally serious. The 'quantitative agreement' claims rest on the 40/106 fs energy lifetimes and the 17/16 fs WS2 comparison. But if the momentum lifetimes are only marginally longer than the bath correlation time, then the collision integral is being applied where its central assumption is questionable. The energy lifetimes are longer, but they inherit whatever the early-time dynamics did, so agreement with experiment is not decisive. Similarly, the WS2 17 vs 16 fs match may be fortuitous given the exciton-vs-free-carrier mismatch. Finally, there are no convergence checks on active-space size, coarse-graining patches, or exponential-fit windows, and the public repository contains the EPW interface and input files but not the Octopus-side dissipative code, so full reproducibility is not yet possible.\n\nThis is a credible method paper with a sound core. It belongs in the review process, but it should come back with a validation case where the Markovian limit is comfortably satisfied and with the Octopus code released. I'd be happy to referee it.","headline":"A credible and genuinely implemented e-ph collision-integral extension of rt-TDDFT, but the benchmark agreement is not a clean test because the Markovian limit is stretched on the very timescales used for validation.","tokens_in":18623,"tokens_out":3423,"would_cite":true,"duration_ms":35045,"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":"Adding first-principles electron-phonon collision integrals to real-time TDDFT yields dissipative dynamics that reproduce measured lifetimes in silicon and WS2.","keywords":["electron-phonon scattering","real-time TDDFT","Born-Markov approximation","collision integral","density matrix","carrier relaxation","intervalley scattering","tr-ARPES"],"falsifier":"A clean falsifier: measure the momentum-resolved population decay in photoexcited silicon under identical pump conditions but at a different lattice temperature; if the extracted sub-20 fs momentum lifetimes change substantially with bath temperature in a way the static-equilibrium-rate prediction cannot reproduce, the Markovian thermal-bath assumption is violated. Alternatively, an exact non-Markovian memory-kernel calculation on the same system that yields energy lifetimes more than a factor of two from the 40 fs and 106 fs values would undercut the validation.","tokens_in":17711,"feed_emoji":"⚛️","tokens_out":6363,"duration_ms":52153,"temperature":0.7,"pith_summary":"The paper sets out to remove a standing limitation of real-time time-dependent density functional theory: its unitary evolution cannot describe irreversible carrier relaxation, thermalization, or decoherence. It proposes to treat the electronic system as an open quantum system weakly coupled to a thermal phonon bath, deriving a collision integral for the reduced one-body density matrix from first-principles electron-phonon matrix elements within the Born-Markov approximation. The resulting equations keep the coherent real-space Kohn-Sham propagation intact while adding phonon-mediated transitions that redistribute carriers in energy and crystal momentum and damp coherences. If the framework works as claimed, it gives a practical first-principles route to simulate carrier cooling, intervalley scattering, and time-resolved spectroscopic signals in realistic materials, with validation shown for bulk silicon and monolayer WS2.","feed_headline":"Dissipative TDDFT matches hot-carrier lifetimes in silicon and WS2","feed_subtitle":"First-principles collision integrals add irreversible phonon scattering to the density matrix, matching experiments in Si and WS2.","key_machinery":"The central object is the reduced single-particle density matrix rho(1) of the Kohn-Sham system, evolved by d rho(1)/dt = -i[H_KS, rho(1)] + I_e-ph[rho(1)]. The collision integral I_e-ph is the load-bearing piece: its diagonal entries are first-principles electron-phonon transition rates W(m,k)<-(n,k') that transfer carriers between bands and momenta with Pauli blocking, and its off-diagonal entries decay coherences with a rate gamma_n(k,t) built from the same W. In the implementation, coherent dynamics run on a real-space grid while the collision integral acts in a time-dependent adiabatic active space; the active-space density matrix is mapped back to natural orbitals with a phase- and rot","core_discovery":"The central claim is that equations (8) and (9) — coherent real-time Kohn-Sham propagation supplemented by a Born-Markov electron-phonon collision integral built from first-principles transition rates — correctly capture irreversible relaxation and decoherence in solids. The collision integral's diagonal part reduces to a Boltzmann gain-loss equation with Pauli blocking, while its off-diagonal part damps interband coherences using the same microscopic scattering channels. The paper validates this by matching simulated Si hot-carrier energy lifetimes (40 fs at 2.6 eV, 106 fs at 1.6 eV) to experimental values (30 fs and 120 fs), and simulated WS2 intervalley scattering (17 fs under linear pola","pith_inferences":["If the Born-Markov limit is genuinely pushed to its edge at about 10 fs, the cleanest next test is to extend the collision integral with a memory kernel and see whether the shortest momentum lifetimes change; until then, the sub-20 fs numbers should be read as trends rather than precise rates.","The same framework could be applied to solid-state high-harmonic generation, where decoherence competes with coherent driving; one testable prediction is that including electron-phonon scattering suppresses high harmonic yields at long pulse durations in a computable way.","The reported separation of fast momentum redistribution and slower energy relaxation in silicon suggests a two-stage hot-carrier picture that could be probed by transient absorption or two-photon photoemission with femtosecond resolution.","Because only equilibrium phonon rates enter, the method currently cannot describe hot-phonon feedback; a straightforward extension would couple the phonon occupations self-consistently, which should slow relaxation under strong excitation — a prediction testable against the pump-intensity dependence of valley lifetimes."],"forward_implications":["Conventional rt-TDDFT can be extended to open systems without abandoning real-space propagation, so relaxation and decoherence become first-principles outputs rather than phenomenological add-ons.","Because population relaxation and coherence decay arise from the same microscopic transition rates, pump-induced dephasing, damping of coherent oscillations, and carrier cooling can be described consistently in one simulation.","The validated energy lifetimes in silicon and intervalley lifetime in WS2 imply the method can predict hot-carrier cooling timescales relevant to photovoltaics and valleytronics.","Combined with a photoemission simulator, the framework turns internal carrier-population dynamics into computed time- and angle-resolved photoemission intensities, letting experiments and theory be compared directly.","The active-space construction keeps extra cost limited, making dissipation feasible for realistic materials and extendable to other reservoirs such as photons, magnons, or substrates."],"fun_headline_variants":["Dissipative TDDFT adds electron-phonon scattering, matches silicon and WS2 lifetimes","First-principles electron-phonon scattering in real-time TDDFT for hot-carrier relaxation","New rt-TDDFT method captures irreversible phonon scattering in solids","Real-time TDDFT with phonon collision integrals predicts Si and WS2 relaxation","Dissipative TDDFT models hot-carrier lifetimes with first-principles phonon scattering"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the randomizing effect of lattice vibrations can be treated as instantaneous and memory-free, with the phonons staying in thermal equilibrium at 300 K, on the few-tens-of-femtoseconds timescales used to validate the method — a premise the paper itself flags as strained for the shortest momentum lifetimes.","fun_headline_variants_meta":{"raw":{"variants":["Dissipative TDDFT adds electron-phonon scattering, matches silicon and WS2 lifetimes","First-principles electron-phonon scattering in real-time TDDFT for hot-carrier relaxation","New rt-TDDFT method captures irreversible phonon scattering in solids","Real-time TDDFT with phonon collision integrals predicts Si and WS2 relaxation","Dissipative TDDFT models hot-carrier lifetimes with first-principles phonon scattering"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000628,"raw_usage":{"total_tokens":2709,"prompt_tokens":678,"completion_tokens":2031,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":422,"completion_tokens_details":{"reasoning_tokens":1919}},"tokens_in":422,"tokens_out":2031,"duration_ms":11847,"temperature":1.0,"reasoning_tokens":1919,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T18:31:38.591168+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A clean falsifier: measure the momentum-resolved population decay in photoexcited silicon under identical pump conditions but at a different lattice temperature; if the extracted sub-20 fs momentum lifetimes change substantially with bath temperature in a way the static-equilibrium-rate prediction cannot reproduce, the Markovian thermal-bath assumption is violated. Alternatively, an exact non-Markovian memory-kernel calculation on the same system that yields energy lifetimes more than a factor of two from the 40 fs and 106 fs values would undercut the validation.","supporting_citations":[],"review_version":1}